Journal of the Royal Society of Western Australia, 86(3), September 2003 
x 
( 1 ) 
-o 
C 
2.5 
2.0 « 
O - S & R mean - Open unfenced area (n = 10) 
□ S & R mean - Open fenced area (n = 10) 
■A " S & R mean - Exclosures unfenced area (n = 5) 
• S & R mean - Exclosures fenced area (n = 5) 
^ S' 
Figure 3. Changes in the abundance index of the seedlings and reshoots category as a result of rabbit grazing within fenced and 
unfenced areas of remnant vegetation at Parsons' site. An abundance index scale of 1 to 7 was used to score the 5 quadrats within each 
plot (see methods). Values are mean (± se) of the abundance index scores for the plots in each area. 
the tall (over-storey) shrub category during the trial 
period. Rabbits would not be generally expected to have 
a severe short-term impact on small and tall shrubs 
unless rabbit densities were very high. 
Long-term effects of rabbits on remnant vegetation 
At the start of this experiment in 1999, there were no 
significant differences in the vegetation parameters 
measured between the exclosure (ungrazed) and the 
open (grazed) plots. The resulting P-values for the 
comparison between ungrazed and grazed plots were; 
shrub height, 0.144; shrub density, 0.206; cover - small 
shrubs (SS), 0.977; cover - sedges and grasses (SG), 0.917; 
other monocotyledonous plants (OM), 0.353; seedlings 
and reshoots - quadrats, 0.154; seedlings and reshoots - 
score (square root transformation), 0.204. This indicates 
that the variety and abundance of the plants were similar 
in these plots before the rabbit-proof fence was erected to 
exclude rabbits from the exclosure plots. 
There were no significant differences in shrub height 
or density between the ungrazed and grazed plots over 
the two years of measurement after the erection of the 
rabbit-proof fences to exclude rabbits from the ungrazed 
plots. However, there were some interesting differences 
for many of the other vegetation measurements taken. 
The ANOVA used was a split plot design with a blocking 
factor (the two sites, df = 1), and with the corresponding 
1999 pre-treatment levels used as a covariate. This 
approach did not test for differences between sites as the 
main interest was in the response of the remnant 
Figure 4. Changes in the percentage cover of the sedges and grasses category as a result of rabbit grazing in the fenced and unfenced 
areas of remnant vegetation at Parsons' site. The results from rabbit exclosure plots are included for comparison. Values are mean (± se) 
percentages of these plants in the plots at each area. 
100 
