Predictive vegetation mapping of forest overstorey species using terrain

Johnson, Derek (2004). Predictive vegetation mapping of forest overstorey species using terrain MPhil Thesis, School of Biological Sciences, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
THE18018.pdf Full text application/pdf 11.16MB 2
Author Johnson, Derek
Thesis Title Predictive vegetation mapping of forest overstorey species using terrain
School, Centre or Institute School of Biological Sciences
Institution The University of Queensland
Publication date 2004
Thesis type MPhil Thesis
Supervisor Dr. David Lamb
Dr. John Neldner
Total pages v.p.
Language eng
Subjects 0602 Ecology
Formatted abstract
The effect of terrain on forest overstorey species composition was studied using a terrain class map consisting of nine terrain classes classified on parameters of aspect, slope position and steepness. The study area was approximately 10 km of eucalypt forest in State Forest 809, in Brisbane Forest Park, north-west of Brisbane, south-east Queensland. The terrain class map was produced by the Queensland Department of Primary Industries Forest Planning and Environment Branch (now within the Forestry and Wildlife Division of Queensland Environmental Protection Agency). A total of 12 transects each 30 m wide and ranging in length from 150 m to 500 m were placed perpendicularly across one or more terrain class boundaries to assess changes in overstorey species composition. Correlations between overstorey species and terrain classes were found, indicating in some cases that terrain correlated with overstorey species composition. Marked changes in both overstorey and understorey species composition were usually observed close to terrain class boundaries. Sensitivity testing, by shifting transects by the maximum estimated error of 20 m in either direction, was not found to markedly change the overall species composition on each terrain class. An area of uniform geological type was selected, but little was known about soils. 


The analysis package PATN for Windows was used to identify correlations between and within species and terrain class. Dendrograms, ordination and Two-way Tables were used. Of the fifteen overstorey species encountered, grey gum (Eucalyptus propinqua) favoured upper N to NW slopes and ridges. Brushbox (Lophostemon confertus) favoured mid and lower SE slopes and stream lines. Western slopes favoured forest oak (Allocasuarina torulosa) and white mahogany (Eucalyptus acmenoides). Steep upper NE slopes were unexpectedly found to be similar to western slopes, and favoured spotted gum (Corymbia citriodora), and also white mahogany and forest oak. Tallowwood (Eucalyptus microcorys) showed a weak association with flat areas. Grey ironbark (Eucalyptus siderophloia) appeared to be a generalist and was relatively evenly distributed across all terrain classes. Other species were too uncommon to show trends. 


It was found that the classes with the most similar species compositions also had similar characters of aspect and slope. The exception were the upper NE slopes, which had a markedly different species composition to other terrain classes with similar topographic characters. This indicated that while most terrain classes were able to indicate direct environmental influences like radiation, drainage and exposure, this terrain class indicated a more complex influence such as possibly fire, soil formation or some other influence less directly related to terrain - but still probably influenced by terrain. Alternatively, this terrain class received more radiation since the whole study area was on the NE side of a range. 


The study showed that terrain classes were potentially useful as a primary factor in spatial modelling of vegetation communities, and in the location of particular overstorey species. However, sampling was insufficient to reliably correlate species with terrain, and knowledge of soils and local geology were poor. Further work would be necessary to evaluate the reliability of these patterns.
Keyword Vegetation mapping

Document type: Thesis
Collection: UQ Theses (RHD) - UQ staff and students only
Citation counts: Google Scholar Search Google Scholar
Created: Wed, 23 Nov 2011, 16:40:56 EST by Melissa Lamb on behalf of Scholarly Communication and Digitisation Service