Analysis of Data and Results of Old-growth Study
6.1 Introduction
Following completion of the field survey and assessment, the data were analysed on the assumption that every discrete forest stand, irrespective of age or disturbance history, contained or potentially contained some of the characteristics used to assess old-growth forests. Using the core characteristics described in Section 2.2, each of the forest stands was assessed and ranked for old-growth status.
6.2 Construction of Old-growth Database
Crown form, crown cover projection, vegetation classes and disturbances were all compiled in map form. Each map was entered into a Geographic Information System (GIS). A GIS is a computer system capable of storing maps, text and data pertaining to any point, or area, on the map. The GIS can be used to analyse data, permitting any combination of information to be analysed efficiently and in a repeatable manner. For example, by combining land tenure, vegetation class and timber harvesting information in the GIS, it is possible to determine the location and extent in hectares of harvested areas for each vegetation class on public land and to display this combination in map form.
Data for each mapsheet were entered into the GIS by various techniques depending mainly on the type and format of source data. These techniques included hand digitising or scanning of line work, or file transfer from databases outside the GIS.
The separate layers of spatial data, equivalent to maps, were overlayed on each other to produce composite maps for inquiry and analysis. Apart from storing data from disparate sources in a standard format so it could be easily retrieved, the GIS facilitates the updating of the datasets as new information becomes available.
6.3 Identification of Forested Vegetation Classes
A stepwise process was used to assess, rank and classify forest stands according to their level of old-growth characteristics. The first step was to delineate all forested vegetation classes, which are essentially a combination of structural and floristic vegetation data refined according to environmental attributes. Of the total of 29 vegetation classes identified and described for the study area, 20 were considered to fall within the definition of 'forest' (Section 2.3.2(b)) and 17 were classified as eucalypt-dominated forest. (See Table 1 and Appendix 1.)
6.4 Identification of Growth Stages
Prior to analysing the data according to the working definition of old-growth forest, it was necessary to use the primary data layers to identify the two characteristics required by the definition, that is, forest growth stage and level of disturbance.
On the basis of field checking across the study area, rules were developed to convert the crown form layer to a forest growth stage layer. In forest on the higher 'environmental quality' sites (those conforming to the classic Jacobs growth habit), the crown-form mapping accurately depicted the equivalent forest growth stages as described in Section 4.4.
For forest not conforming to the Jacobs model, a series of additional indicators were used to infer growth stage. These additional indicators were developed after extensive field observation which showed that, for certain vegetation classes with particular dominant species occurring on poorer-quality sites (usually with stand top-height less than about 28 metres), different yet distinct crown forms were consistently displayed. This permitted the senescing growth stages to be distinguished from younger growth stages.
In some cases the mature and senescing growth stages could not be differentiated. In these situations a conservative approach was followed and the mature component of the stand was considered to be equivalent to the senescing stage.
6.5 Stratification of Disturbance Level
The mapped disturbances were stratified according to their reliability (or probability of occurrence) and their likely severity or level of significance. Some disturbance records which had been mapped broadly were refined using additional evidence of that disturbance having occurred. For example, forest crown-cover mapping and inferred growth-stage mapping were used to refine both selective harvesting records and areas that were allocated (from historical records) for agricultural selections. Moreover, the effect of disturbance varied from one vegetation class to another, necessitating the construction of explicit rules to determine the impact, duration and area of occurrence. For example, grazing was unlikely to have an appreciable effect on vegetation classes unpalatable to domestic stock, even if records indicated that the licensed grazing area included these classes.
The response of a vegetation class to a given disturbance was conceptualised in terms of level of disturbance and recovery time. The parameters of the recovery curve were considered to vary for each vegetation class according to the resilience of that class to the particular disturbance, the rate of recovery to the pre-disturbed state, and the pre-existing level of stress in the ecosystem due to prior or contemporaneous disturbances. Thus, each recorded disturbance regime was used to assign a significance level based on the influence it was believed to have exerted on the present structural and floristic characteristics of a given forest stand.
The significance of each disturbance was then classified according to three levels. Areas with no record or evidence of disturbance history were described as undisturbed (disturbance level 1), even though some environmental weed species may be present and wildfires had almost certainly occurred at some time in the past. A negligible disturbance (level 2) was defined as a disturbance for which there is an authentic record but which is unlikely to have altered the structure (growth-stage ratios or crown-cover projection) or the usual floristic composition of species for that vegetation class; or if a disturbance did occur in the past but its effect is no longer significant. A significant disturbance (level 3) was defined as a disturbance regime for which there is an authentic record and which is likely to have altered the growth-stage ratios, crown-cover projection and the usual floristic composition of that vegetation class and which is detectable at the time of the survey.
Whilst the classification recognises the origin or frequency of the disturbance (natural or anthropogenic) it was considered that, from an ecological perspective, there should be no arbitrary distinction between natural or anthropogenic disturbances. It was also considered that an assessment of disturbance level should be based on the physical impact of the disturbance and the time taken for its effects to abate.
6.6 Assignment of Old-growth Forest Status
Using the forest growth-stage layer derived from crown-form mapping, and the disturbances assigned to stands for each disturbance regime, old-growth forest status was identified according to the old-growth forest definition. That is, old-growth forest was identified as forest which contained senescing trees present in dominant, co-dominant and sub-dominant proportions of the upper stratum (at least 10% of total crown cover for the stand) and where regrowth was present in sparse (less than 10% of crown cover) proportion or non-existent; and where there were only negligible or no disturbances identified (disturbance levels 1 and 2).
The significance of younger forests dominated by mature growth stages with a lower proportion of old-growth forest structural characteristics, but which had negligible evidence of disturbance, was recognised as a separate class - 'negligibly disturbed' forest. By definition, these stands contained a regrowth component of less than 10% of the crown cover of the upper stratum.
6.7 Summary of Results
The extent of disturbance on forested public land, according to the above criteria, is indicated on Map 2 (found in back cover pocket).
The undisturbed category (28% of public land) covers a band of forest through the centre of the study area representing (from west to east) Wallaby Creek, Kinglake and the Watts, O'Shannassy and Upper Yarra River catchments.
The negligibly disturbed forest (20% of public land) generally embraces another swathe of country to the north-east of these undisturbed catchments. It comprises mainly the headwaters of the Thomson and Aberfeldy Rivers in the south and extends north over the Divide and through the headwaters of the Goulburn and Big Rivers.
Old-growth forest occurs throughout the study area, usually in small and scattered patches (Map 3). On public land it was identified in 14 vegetation classes (Table 1) and totalled about 28 820 hectares, representing 4.4% of the eucalypt-dominated forest in the study area.
The most extensive areas are found in the dry vegetation classes, located predominantly in the Aberfeldy, Upper Thomson, Mt Matlock, Terrible, Red Jacket and Enochs Point geographic units. An aggregation of relatively small stands occurs in the Kinglake unit and some is also found in the Latrobe (southern part of study area) and Black Snake units.
The Wet Forest old-growth stands are largely aggregated in Melbourne's water supply catchments within the Yarra Ranges National Park, primarily in the O'Shannassy and Watts River catchments. Damp forest old-growth occurs in small stands with the largest found in the O'Shannassy, Upper Yarra and Big River units.
The analysis indicates that about 91% of the total area of old-growth forest in the Central Highlands is found in four vegetation classes - Heathy Dry Forest (41.8%), Wet Forest (17.6%), Montane Dry Woodland (19.5%) and Heathy Woodland (11.9%).
Table 1 Area of old-growth forest by eucalypt-dominated vegetation class on public land
|
|
|
old-growth (hectares) |
|
| ||||
|
|
|
|
|
|
|
|
|
|
| ||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| ||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| ||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| ||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Notes: Data relates to eucalypt-dominated forest on public land only. Old-growth has not been assessed for non-eucalypt forest classes which comprise a further 16 500 ha. Some smaller patches of indicated old-growth forest included in this table still require field verification to confirm old-growth status and areas of less than 5 ha have generally not been included [conforming with the AHC/CNR study].
In addition to the primary old-growth results summarised above, other study outcomes include a series of maps and area statements for vegetation, disturbance and growth stage classes and, importantly, a database which can be readily updated as new and improved information becomes available.
6.8 Limitations of the Study
With almost three-quarters of a million hectares of native vegetation in the study area, it was not possible to field check all of these areas on the ground and on-going verification of old-growth forest will be required. The area mapped as old-growth forest is only as accurate and reliable as the information used to define it.
Small-scale (1 : 40 000) aerial photographs were used for approximately half of the area. Growth stages and vegetation classes mapped from these scales will reveal some discrepancies, particularly when used for on-ground operational planning.
Historical information on forest disturbance is not totally reliable in some areas and the quality of the data, in particular records of fuel reduction burning, selective harvesting and grazing, could be improved over time with field verification. Assumptions regarding the impact of, and recovery times following, disturbance could also be refined through further research and application of improved remote-sensing techniques.