Forests

Old-growth Forest Study Methods

4.1 Introduction
As it would have been impracticable to assess all of the characteristics of old-growth forest described in Section 2.2, priority was given to assessing the primary attributes and influences that could be measured using remote sensing techniques, field survey and archival research.

The survey methodology essentially followed that developed by Woodgate et al (1994) for East Gippsland, using techniques and technologies which had proved to be reliable, objective and efficient and which were accepted as a standard procedure to be adopted in other regions of Victoria.

4.2 Study Team
The study itself was undertaken by a team of scientists and technicians with expertise in flora and fauna, interpretation of aerial photographs, historic places assessment and geographic information systems (see Section 6.2).

General oversight of the project was provided by a Steering Committee set up for the regional assessment of National Estate values in the Central Highlands by joint agreement between the then CNR and the Australian Heritage Commission. Scientific and technical advice was provided by members of the East Gippsland Old Growth Project Technical Advisory Committee. The project commenced on 1 January 1993 and field work and analysis was completed by the end of December 1993.

4.3 Study Objectives
The primary purpose of the study was to delineate the extent and describe the characteristics of old-growth forests in the Central Highlands. Major objectives were to:

• undertake a field survey, using aerial photographs, satellite imagery and data obtained from other sources, particularly existing stand class and vegetation mapping, to determine the nature and extent of older forests on all tenures of land in the study area. Gaps in the existing data bases were to be identified as a precursor to this study.
• undertake research into, and analysis of, historical data located within archival records
• assess and map the nature and degree of disturbance since European settlement
• create a digital database to support all information collected during the life of the project, and integrate with other relevant data layers in the Department's geographic information system including logging and fire history, forest type, floristic vegetation, land systems, land tenure and land use
• identify and describe the natural and cultural values of all forests in the study area, using digital spatial analysis techniques.

Figure 2. Relationship between the typical eucalypt growth stages described by Jacobs (1955) and those considered by the Victorian studies

The Victorian growth stages were then refined for interpretation from aerial photographs as 'crown forms', in effect serving as a surrogate for growth stages. Crown form (size, shape and composition) was mapped for the entire study area and growth stage for a given tree species was then inferred from this crown-form mapping. Crown cover density was also mapped simultaneously from the aerial photographs. The scale of aerial photography for approximately half of the study area was 1:25 000 (colour and black and white), and for the other half it was 1:40 000 (black and white). Polygon size averaged about 40 hectares for the study area, and the mapping was reliable down to about 10 hectares in size.

In the majority of vegetation classes identified in the Central Highlands, and particularly on high environmental quality sites (fertile sites), typical crown forms were consistent with the growth stages described by Jacobs (1955) and could be readily mapped from aerial photographs. However, for eucalypt-dominated vegetation on poorer-quality sites, and for non-eucalypt vegetation types, field inspection showed that senescing trees senesce and die without exhibiting the classic morphology of the Jacobs overmature growth form. Further, tree crowns in some cases were too small to permit the detection of senescing features from aerial photographs. Trees exhibiting these crown-form characteristics were considered to be atypical Jacobs trees in morphology. However, during the course of the study, it became apparent that other features that could be interpreted from aerial photographs could be used for the delineation of old-growth forest. Relative crown size, regularity of crown gaps, tree heights and regularity of crown-form distribution could be used in conjunction with forest type and vegetation mapping to develop more sophisticated assessments of forest growth stage and disturbance history.

Aerial photograph interpretation was used to assess the crown-cover projection and relative proportion of growth stages in the upper stratum in each forest stand. Crown-cover percent for each stand was estimated as follows: Class 1 - less than 10% cover; Class 2 - 10% to 29%; Class 3 - 30% to 49%; Class 4 - 50% to 69%; and Class 5 - equal to or greater than 70% crown cover.

Many forest stands contained more than one growth stage in the upper stratum or canopy and, in these cases, the relative proportion of each growth stage (senescing, mature and regrowth) was assessed. For this purpose five categories of relative abundance were used: absent, sparse, sub-dominant, co-dominant and dominant and, when applied to each of the three growth stages, a total of 38 growth-stage combinations were mapped.

4.5 Vegetation Classes
Vegetation in the study area was classified using the system of 'ecological vegetation classes' ('vegetation classes' in this report) which represent the highest level in the hierarchy of vegetation classification used across Victoria by NRE. The vegetation was originally mapped as part of the LCC descriptive studies for the Melbourne Area District 2 and was further refined for the old-growth forest study. These vegetation classes are defined by integrating biophysical datasets (geology, soils, climate and land system), floristic quadrat data and forest type maps. Twenty forest vegetation classes were described on public land in the study area, ranging in total extent from 30 ha for Box Woodland to 162 020 ha for Damp Forest.

By describing old-growth forests in terms of their constituent vegetation classes, the particular environmental conditions which produce and perpetuate the old-growth forest state can also be delineated for that class. Because the classification system is based on environmental processes, the impact of disturbance regimes which may effect the old-growth forest status of a stand can be consistently assessed for each vegetation class. Descriptions of vegetation classes in the study area, and within the definition of 'forest', are provided in Appendix 1.

4.6 Disturbance Influences
Historical and contemporary records were researched to identify and map the extent and severity of the major disturbance events known to have altered the primary attributes (floristics, structure or growth stages) of the forests in the study area. These disturbances included agricultural clearing, selective and clear felling harvesting, historical settlements, urban development, sawmill and tramway sites, mining, grazing, wildfires and fuel-reduction burns. Chapter 5 provides further detail about these disturbance types.

Areas for which there were no records of disturbance were described as 'undisturbed', although it was acknowledged that fire, at least, has probably affected all forests at some time in the past. It was also recognised that some unnatural disturbances may not have been recorded. The various mapped disturbances were described in terms of year or period of occurrence, intensity and origin - either natural or anthropogenic.

It is worth noting that, in the context of the Central Highlands study, unnatural disturbances are intended to refer to disturbances that are European in origin. Although Aboriginal influences were apparent prior to and during the early years of European settlement, this influence was considered to be part of the 'natural' environment in Australia. Limited understanding of Aboriginal disturbances has prevented their meaningful delineation in these studies.

4.7 Tree Ageing
Limited investigations (Banks 1993 unpubl.) into the ages of forest growth stages were undertaken through dendrochronological studies. Two Mountain Ash (Eucalyptus regnans) trees from the O'Shannassy and Watts River water supply catchments were selected for study. The trees were killed in the 1939 bushfires and had only recently collapsed. Radial cross sections were taken high in the bole where complete radii were available and tree age estimates of 309 +/- 6 years and 459 +/- 8 years were recorded. The O'Shannassy tree was in the late mature growth phase and had grown in an environment relatively free of disturbance, while the Watts catchment tree was older and exhibited greater variability in its growth rate. In addition to estimating tree age, the study provided evidence of the timing of previous wildfires and data on the magnitude and duration of the growth stages of trees.

Interpretation of this data suggests an immature growth stage for Mountain Ash lasting about 80 years, a mature phase for about 270 years, a senescing phase lasting for about 100 years and finally a stag phase for about 50 years.