VEGETATION OF DERRIMUT GRASSLAND RESERVE 
47 
but may reflect low levels of soil moisture from 
late spring to early autumn (Lunt 1987); insuffi¬ 
cient soil moisture may possibly restrict the 
growth of T. triandra and prevent it from out- 
competing herbs. 
Vulpia grassland was dominated by the exotic 
annual * Vulpia bromoides , often with sub¬ 
dominant Stipa bigeniculata and occasionally 
with the exotic *Stipa neesiana (Fig. 8). This 
grassland occurred along drainage lines and in 
areas that were ploughed last century or grazed 
heavily this century, such as the western end of 
the fence that once divided the reserve (Fig. 1). 
The cover of T. triandra in Vulpia grassland was 
slightly under-estimated as quadrats were orien¬ 
tated to include uniform vegetation, and con¬ 
sequently to avoid small patches of T, triandra. 
With the possible exception of some sites along 
the drainage lines, areas of Vulpia grassland 
Fig. 8. Species-rich Vulpia grassland with forbs includ¬ 
ing Convolvulus erubescens, Helichrysum apiculatum 
and Eryngium rostratum; Themeda grassland occurs 
in background. Photographed in December 1986. 
were probably dominated by T. triandra prior to 
ploughing and grazing. The abundance in Vulpia 
grassland of exotics such as *Briza minor, *Lol- 
ium rigidum, *Trifolium species and *Romulea 
rosea probably reflected the relatively open 
cover. Most of these species were common in 
Themeda grassland as soil-stored seed (Lunt 
1990a) but germination was presumably sup¬ 
pressed by the closed sward. These species 
regenerated profusely when the Themada grass¬ 
land was burnt (Lunt 1990b). Vulpia grassland 
contained significantly more native species than 
Themeda grassland, and significantly more 
exotics than Themeda grassland, Amphibromus- 
Agrostis grassland and Eleocharis sedgeland 
(p < 0.05). Areas ploughed last century were 
amongst the most diverse in the reserve. On 
average, 17 natives and 13 exotics occurred in 
each quadrat (Fig. 3). 
Original vegetation 
Due to the almost complete destruction of the 
native vegetation of the basalt plains and the 
marked modification of relict patches, es¬ 
pecially by the establishment of exotic species, it 
is almost impossible to reconstruct accurately 
the vegetation of the Derrimut reserve as it 
existed at the time of European settlement. 
However, the following points may be made. 
The area was a natural grassland, devoid of 
trees (Stuwe 1986, McDougall 1987), with T. 
triandra dominating in all well-drained areas. 
The density of T\ triandra was probably lower 
and native herbs were presumably more com¬ 
mon in the abundant inter-tussock spaces. 
Patton (1935) illustrated a quadrat in Themeda 
grassland with 16 native species in 0.8 m 2 , a spe¬ 
cies richness far greater than that recorded from 
Derrimut. Many species were presumably de¬ 
pleted or eliminated by continual grazing, a lack 
of burning and isolation from seed sources in 
undisturbed grassland remnants. Native 
orchids, legumes, chenopods and lilies are now 
particularly rare. Circumstantial evidence of 
species losses is provided by the ratio of native 
grasses to native forbs, since grasses are gener¬ 
ally more tolerant than forbs of grazing and 
trampling (Crawley 1983). Whereas grasses 
comprise only 10% of the native species on the 
Keilor Plains, and 13% of those in an ungrazed 
rail reserve at St Albans (Willis 1964, Groves 
1965), they comprise 27% of native species in 
the Derrimut reserve and, similarly, 26% of 
those in the Laverton North Grassland Reserve 
(Platt 1983 and unpublished data). This high 
