Journal of the Royal Society of Western Australia, 86(3), September 2003 
were relatively low, it could take several years before 
any detrimental impacts become obvious. We believe the 
latter is an important consideration and that the fencing 
of remnant vegetation should include a strategy (e.g. 
1080-baiting; see Williams et al. 1995) to reduce rabbit 
numbers prior to totally enclosing a bush remnant. 
The short-term response of remnant vegetation 
following the rabbit-proof fencing appeared to be mixed. 
The positives mainly resulted from keeping rabbits out, 
which helped to maintain the confined rabbit population 
at low levels. However, there did appear to be some 
deleterious effects due to the confinement of rabbits, such 
as a decrease in the percentage cover of sedges and 
grasses. 
Longer term effects of confining rabbits 
Although there was a significant effect between years 
(the two years were markedly different in rainfall during 
the growing season), there is no doubt that low numbers 
of confined rabbits (e.g. ~ 5 ha' 1 ) had a significant impact 
on fenced remnant vegetation. In particular, the 
abundance of seedlings and reshoots was reduced, and 
sedge and grass cover was diminished, relative to the 
exclosures in the corresponding 1-year and 5 to 6-year 
fenced remnant vegetation. Given that this effect 
occurred within 3-7 years from when the fences were 
originally erected (e.g. for 1 year plus 2 years of 
experimentation), then the long-term implication is that a 
substantial change in the biomass and composition of 
bush remnants is a likely consequence of confining even 
low numbers of rabbits within these remnants. It is 
possible that some of the deleterious effects of rabbit 
grazing may be exacerbated by environmental stress. 
That is, such effects may not become obvious until the 
fenced remnant vegetation undergoes an additional 
source of abiotic stress, such as below average rainfall. 
It is also noteworthy that our findings are consistent 
with those of other studies, and some examples of this 
are presented below. 
• Rabbits will often selectively browse seedlings of 
certain shrubs and trees. In fact, there may be no 
'safe' rabbit density for some tree and shrub 
seedlings (Morris 1939; Lange & Graham 1983). For 
example, with free-ranging, unconfined 
populations, even rabbit densities of around 4 ha* 1 
can prevent the regeneration/replacement of some 
plant species, particularly in arid Australia, and 
this can lead to significant soil erosion (Cooke 
1981,1987; Foran et al. 1985). 
• Rabbit-grazing can also impact on native grasses, 
and when rabbits are excluded, native perennial 
grasses will regenerate and rapidly replace many 
of the introduced annual grass species (Mallet & 
Cooke 1986). 
• In some sub-alpine areas, the effects of rabbit¬ 
grazing resulted in the loss of nine palatable forbs 
within seven years. However, where rabbits were 
excluded there was a net overall gain of two 
species (Leigh et al. 1987). The presence of rabbits 
led to a substantial reduction in the cover, biomass 
and species diversity of the forbs in this habitat. 
• In the Victorian mallee district, seventeen native 
species of ground-layer plants were recorded 
where rabbits had been excluded for two years but 
none of these plant species were found where 
rabbits had ready access to such areas (Cochrane & 
McDonald 1966). 
These findings, and the results of our study, strongly 
support the need for a strategy for reducing the numbers 
of rabbits present in bush remnants prior to the remnant 
vegetation being totally enclosed with rabbit-proof 
fencing. This could be achieved by a well conducted 
baiting program, preferably with 1080, which may or 
may not need to be integrated with a shooting program 
to mop up any remaining rabbits. Unless rabbits are 
eradicated from fenced bush remnants, there will be an 
ongoing need to conduct regular control programs to 
prevent/reduce the detrimental effects of rabbit grazing. 
Benefits and costs of rabbit-proof fencing 
Fencing that excludes rabbits from 'prime habitat' can 
decrease the effects of rabbit grazing by limiting their 
capacity for population growth. If conducted on a 
sufficient scale, then this in turn reduces the number of 
rabbits that can potentially inflict damage to the 
surrounding crops and pastures. The benefits of rabbit- 
proof fencing can outweigh the expenditure and become 
cost-neutral to agricultural producers in the medium 
term (^2 years). The protection gained is also long-term. 
With a high value crop such as canola, costs can be 
recovered within two seasons, depending on the amount 
of fencing required, and the crop yields and returns 
obtained. The benefit-to-cost ratio of rabbit-proof fencing 
to eliminate rabbit damage is likely to be even more 
favourable for high return horticultural crops/market 
garden enterprises. This is particularly so as a poisoning 
program does not provide the same absolute protection 
compared to that achieved with rabbit-proof fencing. 
Small crop losses may well continue to occur following a 
baiting program unless some technique is used to remove 
any remaining rabbits (e.g. shooting). 
Implications for rabbit management and bush 
remnants 
Provided that some means are used to reduce residual 
rabbits, then rabbit-proof fencing of remnant vegetation 
of high conservation value would be well worthwhile. 
Obviously the benefits and costs depend on the size of 
the area that needs to be protected. Nevertheless, we do 
have some caveats. In the short term (one year), the effect 
of confining rabbits within remnant vegetation, thereby 
reducing available rabbit habitat, may be both positive 
and negative. Such effects were not always easy to 
define. However, based on the results of our longer term 
experiment, the effect of confining rabbits within bush 
remnants over a much longer term (say, 15 years) will 
almost certainly be negative unless steps are taken to 
remove the residual rabbits completely (i.e. eradication is 
achieved). This would be quite achievable using a 
combination of poison baiting and shooting, particularly 
if control efforts were undertaken when other food is in 
relative short supply (e.g. during summer). If the long¬ 
term viability of the remnant bush is to be maintained, 
then every effort needs to be made to ensure that residual 
rabbits are completely eradicated. If this is not achieved, 
then the bush remnants themselves may need to be 
subjected to regular poisoning campaigns, in addition to 
104 
