36 
Curtninghamia 
Vol. 1 (1): 1981 
east. To the west are the Recent alluvial flats of the Hawkesbury-Nepean River, 
and beyond them the Triassic Hawkesbury Sandstone of the lower Blue Mountains 
is exposed along the Lapstone Monocline (New South Wales Dept, of Mines, 1966). 
The vegetation on the sand deposit is mainly woodland to low woodland with 
some smaller areas of open-scrub (structural forms'follow Specht, 1970). The 
dominant tree species, Eucalyptus sclerophylla, E. parramattensis and Angophora 
bakeri, are associated with a species-rich sclerophyllous shrub understorey, common 
species being Banksia serrata, B. serratifolia, Ricinocarpos pinifolius, Dillwynia 
glaberrima, Leptospermum attenuatum, Conospermum taxifolium, Pbilotheca salsolifolia 
and Acacia brownii. On the surrounding areas of clay the dominant tree species 
is Eucalyptus fibrosa subsp. fibrosa, though E. sclerophylla, E. parramattensis and 
Angophora bakeri occur occasionally. The rich shrub understorey is replaced by a 
more open one, predominantly of Melaleuca decora and Bursaria spinosa with a dense 
ground cover of grasses. 
Phillips (1947) described the vegetation on both the sand deposit and the Tertiary 
alluvial deposits as belonging to her Eucalyptus crebra [E. racemosa]-Angophora 
bakeri Association but distinguished the Agnes Banks vegetation in a separate 
shrub-stratum group characterized by Banksia serrata. Forster et at. (1977) listed 
the main tree and shrub species on the sand deposit and included it in a list of areas 
of botanical conservation interest. 
VEGETATION DESCRIPTION 
Methods 
Black and white aerial photography (Cumberland 1970 series, scale 1:15 000, 
date 9.7.1970 and Penrith series, scale 1:40 000, date 27.6.1966) and field traverses 
were used to prepare a vegetation map (Figure 1) showing the vegetation types extant 
in 1970. Tree canopy cover and topographic position were the main characteristics 
used to distinguish the different communities on the aerial photography. Field 
observations confirmed that these were related to species groups, and indicated 
particularly that drainage was an important factor in plant distribution. 
In September 1976 the vegetation was sampled at 250 m intervals along north/ 
south transects, the transects being chosen subjectively in order to sample the five 
plant communities. As considerable clearing had been carried out since 1970, 
sample points falling on cleared or mined areas were ignored, leaving 24 in essentially 
undisturbed native vegetation, though 8 of these had been burnt within the previous 
two years. 
Mining and consequently destruction of the plant communities has been uneven. 
The highest dunes have generally been mined first, so that only small areas of their 
vegetation remained whereas there were still large areas of vegetation on shallower 
sand. Since sand depth appears to influence plant community composition, the 
sampling of the five communities was uneven and the majority of sampling points 
were located in the two communities occupying the greatest area of remaining vege¬ 
tation, the Eucalyptus sclerophylla-Angophora bakeri-Banksia serrata woodland 
and the Eucalyptus parramattensis low open-woodland. Two communities had only 
one sampling point in them and one community was not sampled. Several points 
had been sampled previously in 1975, similar data being recorded. These data 
were not used in the later numerical analyses but were included in the basal area 
and stem density descriptions. 
At each sample point all vascular plant species present within a 20 x 20 m square 
quadrat, hereafter referred to as a sample site, were listed or collected for later 
identification, and stem diameters of all individuals greater than 5 cm (d.b.h.) were 
recorded. Projected canopy foliage cover was measured at 50 points within the 
site, using a vertical sighting tube with cross-wires. Notes were made on important 
shrub species, position on dune, drainage, aspect, slope and evidence of fire or other 
disturbance. 
