42 
IAN D. LUNT 
Fig. 1. Distribution of vegetation types in 61 quadrats 
in Derrimut Grassland Reserve. Open, semi-closed 
and closed circles and squares denote species-poor, 
moderate and species-rich Vtilpia grassland and 
Themeda grassland respectively: open stars denote 
Eleocharis sedgcland; closed stars denote Amphibro- 
mus-Agrostis grassland; triangles denote mixed- 
species herbland. Shaded areas were ploughed last cen¬ 
tury. Contour interval is 1 m (Melbourne & Metro¬ 
politan Board of Works 1978). Horizontal dotted line 
denotes an old fenceline. 
SITE HISTORY 
The property was owned by the Myers family 
from 1912 until the 1950s when it was bought by 
the Victorian Government, but it was grazed by 
the family’s stock until 1985 (D. Myers, personal 
communication; source also of the following 
agricultural information). Three to four hun¬ 
dred head of sheep (3 to 4 per hectare) were run 
until the late 1960s when they were replaced by 
80 to 90 head of cattle. On two occasions in the 
early 1960s the site was heavily grazed for a fort¬ 
night by 6,000 to 7,000 sheep. Although all of the 
vegetation was consumed, Themeda triandra 
and other grasses recovered after the first rains. 
The property was divided in two by an east-west 
fence (Fig. 1). Grazing pressure was greater in the 
southern paddock and where cattle congregated 
at the western end of the fence. Three areas of 
the reserve were ploughed, and presumably 
cropped, in the late 1800s (Fig. 1) but there has 
been no ploughing or cropping since 1912 and 
probably none since 1900. The site was rarely 
burnt and superphosphate was never applied. 
METHODS 
Sampling 
Minimal quadrat area (defined as the point at 
which a 10% increase in area yields only a 
5% increase in species; Mueller-Dombois & 
Ellenberg 1974) was determined from two sites 
dominated by T. triandra , and was found to be 
between 9 m 2 and 11 m 2 . A quadrat size of 15 m 2 
(5 x 3 m) was employed. 
Quadrats were sampled in November and 
December of 1986 and 1987. Fifty-one were 
located at 100 m intervals on six transects set 
200 m apart and ten were selectively placed in 
vegetations that were inadequately sampled on 
transects (Fig. 1). Quadrats were orientated to 
include uniform vegetation. In each quadrat, the 
cover of all species of vascular plants was 
recorded using the Braun-Blanquet scale 
(Mueller-Dombois & Ellenberg 1974), with the 
Braun-Blanquet category "r” being included in 
the category “T”. The reserve was exhaustively 
traversed between November 1986 and Januao' 
1988 and a comprehensive list made of all 
species not found in quadrats. Plant taxonomy 
follows Forbes & Ross (1988), and exotic species 
are marked with an asterisk. Specimens of many 
species are lodged with the National Herbarium 
of Victoria (MEL). 
Classification 
The program MAGIC — a polythetic, agglom- 
erative, non-hicrarchical cluster analysis utiliz¬ 
ing presence/absence data (Gullan 1978) — was 
used to classify quadrats according to fioristic 
composition. Quadrat and species groups from 
the classification were hand-sorted for clearer 
resolution of the fioristic table. Data are perma¬ 
nently stored on the database of the Flora and 
Fauna Survey Group of the Department of Con¬ 
servation, Forests and Lands of Victoria, as 
quadrats B18181 to B18234 (quadrats 1 to 54) 
and E01401 to E01407 (quadrats 55 to 61). 
The Tukey-Kramer test, suitable for un¬ 
planned, multiple comparisons among pairs of 
means based on unequal sample sizes (Sokal & 
Rohlf 1981), was used to test differences in spe¬ 
cies-richness between vegetation groups. 
