56 
Cunninghamia 
Vol. 1 (1): 1981 
serrata-Angophora bakeri low open-forest (Site 6) at the well drained end, as expected. 
Site 3, the only site in the Eucalyptus sclerophylla-E.parramattensis-Banksia serratifolia 
woodland, had been recently burnt and was grouped with intermediate dry sites 
of the E. sderophylla-A. bakeri-B. serrata woodland most of which had been similarly 
burnt. In the species classification Banksia serratifolia is grouped with species of 
poorly drained situations (species group 5) whereas Banksia serrata is grouped with 
those of the best drained situations (species group 3). There is no evidence that the 
sites in the Eucalyptus sclerophylla-E. parramattensis-Banksia serratifolia woodland 
behave differently from sites in the Eucalyptus sclerophylla-Angopliora bakeri- 
Banksia serrata woodland though Banksia serrata evidently prefers better drained 
sites than Banksia serratifolia. There were no sites in the Lepidosperma longitudinale 
sedgeland, but its floristic composition and topographic position indicate that it 
would be located at the extreme wet end of the moisture/drainage gradient. 
Other factors, particularly fire, and location on supposedly older areas, were also 
considered important. The analyses indicate that the moisture/drainage factor is 
by far the dominating influence on plant species distribution, though the quantitative 
sampling was not detailed enough to show how important it was in determining the 
abundance of individual species. 
The results suggest that projective canopy cover is also an important factor in 
the distribution of the understorey species. However, the behaviour of two species, 
should be noted; on nearby areas of different geological substrate (Wianamatta 
Shale and Hawkesbury Sandstone) Hardenbergia violacea and Grevillea ntucronulata 
are colonizers following disturbance, which is often associated with canopy removal. 
On the Agnes Banks sand they are restricted to sites that have the highest cover and 
have remained unburnt for a long period. 
An estimate of the relative importance for plant distribution of the two factors, 
moisture/drainage and canopy cover, is indicated by the percentage variation these 
account for in the species ordination. The axis attributed to moisture/drainage 
accounts for 26.1 % of the total variance and that attributed to canopy cover 19.3%. 
The next three axes together account for only 16.9% of the total variance. 
A third factor, fire, though of much less importance, is indicated from general 
observations and in some of the analyses. Although eight of the 24 sites had been 
burnt within the previous 2 years (probably in Dec. 1974), no particular site or species 
groups are clearly indicative ol recent fire, though it is indicated in some axis com¬ 
ponents. Purdie & Slatyer (1976) have reported that, for dry sclerophyll forest 
near Canberra, most of the regrowth from surviving organs, from residual seed and 
from seed released after fire, commences in the first 12 months after burning. This 
seems to be the case for the vegetation at Agnes Banks, and the recording of species 
presence alone is inadequate to detect differences between areas burnt at different 
times in the past. 
ACKNOWLEDGMENTS 
I wish to thank Helen Bryant for her help in collecting the data and preparing 
the diagrams, and my colleagues at the National Herbarium of New South Wales, 
particularly R. G. Coveny, S. W. L. Jacobs, M. D. Fox and J. M. Powell for help 
with the identification of specimens and comments on the manuscript. 
REFERENCES 
Bureau of Meteorology (1979). Climatic survey Sydney, region 5, New South Wales. 
Department of Science and the Environment, Canberra. 
Carolin, R. C. (1970). Myall Lakes—an ancient and modern monument. Proc. 
Ecol. Soc. Austral. 5, 123-129. 
