Iveson et ah Salmonella in an island marsupial 
identified from a total of 219 isolations, suggesting 
contact with other reservoir hosts, such as reptiles or 
seagulls (Lams novaehollandiae) (Iveson & Bradshaw 
1978). 
Movements of quokkas 
The record of animal movements confirmed that most 
quokkas were highly sedentary, moving only short 
distances to foraging areas close to diurnal rest sites 
(Kitchener 1972). A small number of individuals, 
however, showed evidence of high mobility, being 
recaptured at distances over I km on the small island. 
Five quokkas, for example, moved between Kingston 
Barracks and the camping ground. Forty nine animals 
(29 males and 20 females) were classified as 'mobile' and 
10 of these (20%) were positive and infected with S. 
adelaide, S. javiana and S. typhimiirium PT 202. 
Movements were mainly in the settlement areas and 
included quokkas moving from the beach to the camping 
area, from Bathurst Point to the Police Station, and from 
Garden Lake to Bickley Swamp - this latter displacement 
being approximately 1 km. There was no evidence, 
however, of quokkas moving short or long distances 
acting as carriers of Salmonella infections to new 
locations. One quokka sampled in the settlement 
classified as 'sick' and in poor condition showing 
evidence of wasting and scouring was positive for 
Campylobacter jejuni, an important zoonotic pathogen. 
Edwardsiella tarda, an opportunistic human pathogen, 
was also isolated from quokka droppings at Garden 
Lake. 
Population and density estimates 
The estimates of population sizes are of necessity only 
approximate, merging close sites and different collecting 
dates in order to overcome the problem of small sample 
sizes. They nonetheless give some idea of the 
extraordinarily high density of the quokka population in 
the settlement area. Estimates of the total number of 
quokkas on Rottnest Island are vague but the figure of 
12,000 is accepted by most commentators (Iveson & 
Bradshaw 1978; Miller & Bradshaw 1979; Pen & Green 
1983), giving an average density of 6.3 animals ha ' for 
the 1912 ha island. This needs to be contrasted, however, 
with the figure of 21 animals ha ' in the confined 
settlement area. The settlement area, because of its 
proximity to humans and supplementary sources of food, 
is obviously a preferred site for quokkas and this almost 
certainly adds to the likelihood of their acquiring and 
recycling infections. Animal densities in the wild of 
greater than 2 ha ' are unusual - for example in the 
Niassa Reserve in Mozambique this only occurs in small 
localised areas with densities of 0.5-2.0 individual ha ' 
being much more common (Gib.son 2000). Population 
densities at Kingston Barracks/Bickley Swamp and the 
Geordie Bay and Longreach settlements were lower, but 
still well above what would be needed to nullify the risk 
of recycling and re-infection between quokkas. 
Implications of the study 
Guidelines for the management of salmonellosis on 
Rottnest Island were established prior to the present 
study (Iveson & Hart 1983) and quokkas in subsequent 
year have been effectively excluded from the settlement 
area that is most frequented by members of the public. 
This has been done primarily by denying quokkas access 
to spaces beneath houses and buildings where they 
previously sought refuge during the daylight hours, 
emerging to feed at night. The one exception is the tourist 
bus stop at Lighthouse Swamp which the present study 
suggests does not pose a significant health risk to the 
public, as visitors to the site are supervised. These 
measures appear to have been very effective, with no 
documented recurrences of human cases of salmonellosis 
on the Island since 1991. 
The study provides insights into the potential hazards 
of tourist-type facilities that focus on close encounters 
with wildlife species. Rottnest Island is an iconic resort 
for West Australians, where escape from the pressures of 
modern living are sought specifically by a temporary 
retreat, and rejuvenating contact with nature is a primary 
aim of the visit (Seddon 1983). The prevalence of a large 
number of serotypes in a natural population of animals 
such as the quokka, with infection rates reaching 70% in 
some areas in the summer months, serves, however, to 
highlight the inherent dangers of too-close contact with 
wild species. Outbreaks of salmonellosis that can be 
traced to wildlife are quite common (e.g. an outbreak of 
Salmonella saintpaul in Queensland that was traced to 
frogs and mice (Taylor Sloan Morton & Hunter 2000)) 
and there is a need to educate the public of the 
importance of respecting wildlife and not treating it as a 
natural extension of household pets. There is also a need 
to better understand the source of Salmonella infections 
in mammalian species, where they appear to be at least 
potentially pathogenic, whereas in some vertebrates, 
such as reptiles, they appear to be natural members of 
the gastrointestinal flora. 
Acknowledgements: This paper is dedicated to the memory of the late 
Ray Hart who, as a PhD student of the other two authors, laid the 
groundwork for our current understanding of the fascinating dynamic 
nexus between the marsupial quokka and the many Salmonella serotypes 
established on Rottnest Island. Quokkas were collected under licence 
from the Department of Conservation and Land Management (CALM, 
now DLC) and the Rottnest Island Authority (RIA) facilitated the study in 
many ways. The joint investigations were funded by the Public Health 
Department of Western Australia and the Department of Zoology of the 
University of WA. The authors acknowledge the invaluable contribution 
to the study of Mr R. Curtis and Mr G. Mofflin. 
References 
Anon. Surveillance Report Wildlife. Weybridge, UK: Veterinary 
Laboratories Agency, 2006: 5. 
Aschfalk A, Hundertmark K ), Bendiksen H R, Arnemo J M & 
Elchen N 2003 Serosurvey for antibodies against Salmonella 
species in free-ranging moose (Alces alces) from Norway. 
Berl Miinch Tieraztl Wochenschr 116: 417-420 
Bailey N J 1952 Improvements in the interpretation of recapture 
data. Journal of Animal Ecology 21: 120-127 
Bakker H R & Main A R 1980 Condition, body composition and 
total body water estimation in the quokka, Setonix 
brachyurus (Macropodidae). Australian Journal of Zoology 
28: 395-406 
Bool P H & Kampelmacher E H 1958 Some data on the 
occurrence of Salmonella in animals in Surinam. Antonie van 
Leeuwenhoek Journal of Microbiology and Serology 24: 76- 
80 
Chambers D L & Hulse A C 2006 Salmonella serovars in the 
herpetofauna of Indiana County, Pennsylvania. Applied and 
Environmental Microbiology May 2006: 'i77\-'i77'i 
135 
