Journal of the Royal Society of Western Australia, 90(3), September 2007 
of the possibility of zoonotic infections (Kruse et al. 2004) 
with Taylor Latham & Woolhouse (2001) estimating that 
of 1,415 known human pathogens, 62% arc of zoonotic 
origin. An early paper by Bool & Kampelmacher (1958) 
reported that wild animals were carriers of Salmonella 
and Kourany Myers & Schneider (1970) identified 
amphibians and reptiles as common carriers. Everard et 
al. (1979) isolated 20 serovars of Salmonella from 44 out 
of 219 animals in Trinidad and Grenada with the toad, 
Bufo marinus, being the most frequently infected. 
Evidence of transmission of Salmonella typhimuriiim O; 
4-12 from wild passerine birds from to humans was 
reported in Norway in 1998 (Kapperud Stenwig & Lassen 
1998). A total of 33 Salmonella serovars was isolated from 
34 symptom-less reptilian species in Indiana County, 
Pennsylvania, adding support to the contention that 
Salmonella is a natural member of the intestinal flora of 
herpetofauna (Chambers & Hulse 2006; Minette 1984). 
Fifty nine different serotypes of Salmonella were isolated 
from 25 species of wildlife, including crocodiles, in 
Queensland over a 20-year period but the major serotype 
infecting humans in north Queensland was absent from 
the reptiles (Thomas Forbes-Faulkner Speare & Murray 
2001). Recent publications have documented the 
occurrence of Salmonella infections in moose (Alces alces) 
in Norway (Aschfalk Hundertmark Bendiksen Arnemo 
& Elchen 2003), the New Zealand endangered passerine 
bird, the Hihi {Notiomystls cincta) (Ewen Thorogood 
Nicol Armstrong & Alley 2007) and a wide range of 
wildlife species in the Nairobi National Park (Gitter & 
Brand 2005). 
Salmonella in the quokka 
The first isolation of a Salmonella serovar from the 
quokka was in 1972 (Iveson & Bradshaw 1973) and the 
fact that it was a rare south-east Asian serotype (S. 
javiana), unknown on the mainland of Australia, 
prompted the initial investigations of its ecology on 
Rottnest Island Hart Iveson & Bradshaw 1987; Iveson & 
Bradshaw 1978). The differences between infection rates 
found in the present study are consistent with the results 
of general area comparisons by Hart Bradshaw & Iveson 
(1985); and Iveson & Bradshaw (1973). This early work 
established that quokkas in the settlement area were in 
the best physical condition of animals on the island and 
exhibited the lowest rates of infection of approximately 
20% in all seasons compared with rates of 40-70% from 
other sites. Thus, paradoxically, infection rates are lowest 
in areas of greatest human contact, but this would appear 
to be due to the supplementary feeding that the quokkas 
receive from tourists in the settlement and the tourist site 
at Lighthouse Swamp, which helps to maintain their 
body condition during the adverse summer period, and 
resistance to infection. Hart et al. (1987) similarly 
established that quokkas collected at the Island rubbish 
disposal site had low infection rates year round and high 
physical condition, measured with a meristic condition 
index relating body mass to bone length (Bakker & Main 
1980). 
Variation in infection rates 
Over most of the island, away from settlement and 
recreation areas, there is a dramatic change in the 
infection rate, which cycles seasonally, approaching 70% 
in summer and falling to 0-30% in winter. Hart (1980) 
found that Bickley Swamp, which is near the Kingston 
Barracks, was intermediate. Lighthouse Swamp is also 
intermediate, reflecting probably the impact of daily 
feeding by tourists on the condition of the quokkas at 
this site. The lower results for the late summer of 1985 
included in the analyses are considered atypical because 
of unseasonable rains that fell on the island in the 
summer of 1984-85. No evidence of the marked rise in 
infections and epizootic of S. adelaide recorded earlier in 
the settlement area and at Lighthouse Swamp was found 
in subsequent studies and S. adelaide infections declined 
in incidence to previous levels. Salmonella javiana was 
isolated on 21 occasions in the combined settlement, 
Kingston Barracks and campsite area, but only once at 
Lighthouse Swamp. A similar pattern of isolations 
occurred with S. typhimurium. The vast majority of these 
strains were Phage Type 202 in quokkas inhabiting 
settlement areas, the camping area and Kingston 
Barracks. Analysis of infection rates in males, females 
and juvenile quokkas showed that only at one site. 
Lighthouse Swamp, was there any evidence of 
discrimination on the basis of sex and it was concluded 
that Public Health procedures should not focus on any 
particular age group or sex when collecting samples. 
Recapture rates 
Analysis of the fate of infections in recaptured animals 
revealed that infection rates conformed to the seasonal 
pattern and was estimated at 20% for the settlement, 28% 
for Lighthouse Swamp and 38% for the Kingston/Bickley 
Swamp population. These estimates are fairly consistent, 
despite differences in infection rates between the study 
areas. All of the Salmonella isolations and re-isolations 
recorded for comparative purposes were made in 
summer and the majority were within a two-week period 
following the primary isolation. The possible bias 
involving failure to detect small numbers of organisms 
and subordinate strains in mixed infections by the swab 
procedure has been investigated by Hart et al. (1982) and 
applying the minimum level of recovery it is estimated 
that only 50% of serotypes detected are still present after 
2 weeks. The numbers of organisms excreted are thus 
below the dosage likely to pose a significant hazard to 
humans, assuming contamination of foodstuffs, water 
supplies and coastal waters is avoided. 
There is some evidence that serotypes do persist in 
carrier animals over long periods of time. Quokkas No. 
198 with S. Chester and No. 389 with S. oranienberg at 
Lighthouse Swamp, for example, each retained a given 
serotype for 9 months. Evidence that these infections 
represent a carrier state in maintaining hosts is that the 
serotypes are relatively uncommon and are unlikely to 
be isolated repeatedly as transient infections. These 
results are consistent if the great bulk of infections 
recorded represent short-term transient episodes of 
infection. The few long-term infections may also be 
maintained by recy'cling in high infection areas such as 
contaminated seepage areas and other surface water used 
by quokkas and cohabiting wildlife converging for 
drinking. Quokkas have been shown to supplement their 
meagre supplies of water by drinking brackish water that 
seeps into the various salt lakes on the island (Jones 
Bradshaw Fergusson & Watts 1990). During the summer 
months these seepage areas are heavily contaminated 
with droppings and 24 different .serotypes have been 
134 
