Journal of the Royal Society of Western Australia, 86(1), March 2003 
Immigrations and extinctions 
A few species had become extinct before the first flora 
list was published (Storr 1962). Two species were recorded 
by Preiss in 1839 but not since, Amyema melaleucae (see 
Frewer et al. 1985) and Acacia truncata (WA Herbarium 
specimen). Heliophila pusilla, Reseda alba and Waitzia nitida 
were only collected in about 1928, and Sorghum bicolor only 
in 1946-47 (WA Herbarium specimens). 
There was a comparatively rapid turnover of species 
on the island between the late 1950s and late 1990s. Forty- 
four of the total 201 species (22%) known to have been 
present in the late 1950s were not recorded at the end of 
the century, while 38 additional species were found. The 
82 species that either became extinct or immigrated are 
listed in Table 2. 
The overall rate of turnover (immigrations plus 
extinctions) was almost 2 species per annum over the 42 
years between the surveys, or 1% pa. Exotics had a 
turnover rate twice that of native species (0.7% species pa 
compared with 0.3% pa). The number of extinctions and 
immigrations was significantly higher for exotics than 
for native species (x, 2 = 4.06, P <0.05). 
Overall, there was a tendency for native species to 
give way to exotics and for annuals to give way to 
perennials. The proportion of native species dropped 
from 62% to 58%, and the proportion of annuals dropped 
from 41% to 37%. The majority of native species recorded 
were perennial (73%, or 83 in a total 113 native species in 
the recent survey) but only a minority of the introduced 
species (36% or 30/83) were perennial. 
The difference between the number of annuals that 
immigrated/became extinct compared with perennials 
was not statistically significant (x, 2 = 2.86, P > 0.05). 
Annuals had an extinction rate higher than the 
immigration rate (0.3% pa compared with 0.2% pa), while 
perennials had a higher immigration than extinction rate 
(again 0.3% pa compared with 0.2% pa). 
Argyranthemum frutescens (marguerite) was recorded 
for the first time in 1986. This garden escape was 
eradicated at that time (Keighery 1986) and is a good 
example of successful eradication following rapid 
response to a naturalisation event. This species was the 
only recognised example of cryp’to-tumover in this study, 
being present between the two censuses but not recorded 
at either. 
Deleted species names are provided in Appendix 3. 
Appendix 4 lists 35 exotic species cultivated outside 
gardens, which have not become naturalised. 
Discussion 
The revision necessary to achieve our Rottnest flora 
list involved the deletion of 122 species names, which are 
given in Appendix 3 together with explanatory notes. 
Half of the discarded names reflected simple taxonomic 
changes, with the older lists particularly requiring 
updating. Some of the revisions were complicated by two 
name changes. Others involved the splitting of one taxon 
into two; in cases where no material from Rottnest Island 
had been preserved by earlier collectors it was not 
possible to determine which of the two new categories 
had been recorded. A number of species listed by Storr 
(1962) and others who drew on his work appeared to 
have been misidentified. These were species that were 
not supported by herbarium specimens; where a species 
was not re-collected in the recent survey but a similar 
species was found, often in the location recorded 
previously, we assumed that the two species were the 
same. Nomenclatural problems that could not be 
resolved are addressed in Appendix 1. 
The total number of extant species recorded changed 
little between the late 1950s and the end of the century, 
with 201 and 196 species respectively. The stability in the 
total number of species found on Rottnest Island over the 
past half century supports MacArthur & Wilson's (19 67) 
theory of island biogeography, which states that the 
number of terrestrial species on islands tends to remain 
in equilibrium, with immigrations balancing extinctions. 
The total species number, according to their theory, U 
determined by the size of the island, the time since 
isolation and the distance from the nearest landmass. 
Comparable flora studies (Table 3) carried out on sever\ 
neighbouring, albeit much smaller islands, in the late 
1950s showed that total species number did reflect the 
area of each island, as predicted by MacArthur & Wilson 
(1967). However, by the late 1990s most of the smaller 
islands had experienced drastic declines in biodiversity. 
This loss of species was associated with increasing 
numbers of nesting seabirds and invasion by Malva 
dendromorpha and other coprophilid alien plant species 
(Rippey et al. 2002). These influences were very restricted 
on Rottnest Island itself, but important on its satellite 
Table 3 
Number of plant species on islands off Perth in the late 1950s and the late 1990s (Garden Island is excluded from this comparison 
ecause no recent figures are available). Isolation is distance to mainland or nearest island. 
Area (ha) 
Isolation 
(kilometres) 
Species in 
late 1950s 
Species in 
late 1990s 
% change 
Rottnest Island 
1900 
17 
201 
196 
-2 
Carnac Island 
16 
8 
82 a 
62 b 
-24 
Penguin Island 
12 
0.7 
74 a 
76 c 
+3 
Seal Island 
1.2 
1 
41a 
17 c 
-59 
Bird Island 
0.9 
0.3 
31a 
14 c 
-55 
Dyer Island 
0.6 
1 
18 a 
17 
-6 
Shag Island 
0.4 
1.1 
24 a 
16 c 
-33 
Green Island 
0.1 
0.1 
18 a 
14 
-20 
* fOF 1959 (Abbott 1977); b for 1995 (Abbott et al. 2000); c for 1998 (Rippey et al. 1998) 
10 
