Journal of the Royal Society of Western Australia, 90: 161-163, 2007 
Measurement of Southern 
Brown Bandicoot {Isoodon 
obesulus) body temperature 
using internal and external 
telemeters 
A Larcombe 
Zoology, School of Animal Biology M092, 
The University of Western Australia, Stirling Highway, 
Crawley, WA 6009, Australia 
E ebmocral@graduate.uwa.edu.au 
Manuscript received Jnmianj 2007; accepted ]une 2007 
Abstract. Two types of external temperature telemeter 
were designed and tested for their accuracy in measuring 
body temperature of southern brown bandicoots. Three 
attachment sites (groin, armpit and base of tail) and a 
number of methods of attachment were tested. Tire most 
effective was attachment to the base of the tail with 
surgical tape. Accuracy of external body temperatures 
was tested against data obtained from a surgically- 
implanted telemeter. External temperatures measured by 
telemetry did not accurately reflect core body 
temperature, but were instead closer to ambient 
temperature. As such, external telemetry is not 
recommended for use with animals of this size (~ 1000 g). 
Keywords: Body temperature telemetry, bandicoot, 
external telemeter, circadian rhythm. 
Introduction 
Measurement of body temperature (T^) is fundamental 
to studies of mammalian physiology and is of prime 
importance in assessing the thermal responses of animals 
which may exhibit heterothermy (Brown & Bernard 1991; 
Dausmann 2005). The most common method of T^, 
measurement is via the insertion of a rapidly responding 
thermocouple or thermistor into the cloaca or rectum. 
Although accurate if performed correctly, this method 
requires the study species be Captured and handled. Such 
procedures can be stressful for the study species and 
therefore may result in elevated T,^. Further, the 
technique does not allow for continual remote 
measurement of Tj^ in free-ranging species (Brown & 
Bernard 1991; Audet <& Thomas 1996). To combat this 
limitation, some researchers use surgically implanted 
telemeters to remotely monitor T^ (Muchlinski et al.; 1998; 
Refinetti 1999; Geiser & Drury 2003), however, this 
method also has it drawbacks. Tire most important is that 
implantation of telemeters requires the study animals to 
undergo at least one surgical procedure. This can be 
stressful and cause infection, both of which can alter 
thermoregulatory patterns. Surgical implantation also 
reduces the signal transmission distance of the telemeter 
© Royal Society of Western Australia 2007 
(as the signal must pass through the body wall), which is 
important in studies of highly-mobile species (Audet & 
Thomas 1996; Barclay etal. 1996). 
To both measure Tj_ remotely and avoid surgery 
related difficulties, external attachment of body 
temperature telemeters have been used on a number of 
species of endotherms (Audet & Thomas 1996; Barclay et 
al. 1996; Kortner & Geiser 2000; Kortner et al. 2001; 
Dausmann 2005). In these studies, small telemeters were 
either glued to the experimental animals, attached using 
an elastic harness, or as a collar. Individual studies have 
had mixed results, largely depending on the size of the 
subject species. External T,, telemetry worked well with 
small sized animals (- 20-100 g) where skin temperature 
(T^i^) was close to and linearly correlated with Tj^ (Audet 
& Thomas 1996; Dausmann 2005). However, in larger 
animals (380-550 g) the difference between T^^ and T^, 
increased by several degrees (Kortner et al. 2001). The 
only time external T,^ telemeters have been used on a 
mammal greater than 600 g was by Dawson & Bennett 
(1978), who measured pouch temperature at moderate 
ambient temperatures (TJ for a single female spectacled 
hare wallaby (Lagorchestes conspicillatus; average mass 
2660 g). In this experiment T was approximately 0.4- 
0.7°C lower than the simuftaneously measured rectal 
temperature. 
This study tested the validity of using externally 
attached temperature telemeters for measurement of T^^ 
for captive southern-brown bandicoots {Isoodon 
obesulus). Data were compared to measurements of T^^ 
obtained using a surgically implanted telemeter. 
Materials and Methods 
Two adult male I. obesulus were studied at the 
University of Western Australia. During study, 
bandicoots were maintained in sheltered outdoor 
enclosures or controlled temperature rooms (CTR) 
depending on the experiment and were provided with 
food and water ad libitum. 
Two types of single-stage EM external T^^ telemeters 
were custom made for this study (Titley Electronics, 
Ballina, Australia and Sirtrack Wildlife Tracking 
Equipment, Havelock North, New Zealand). Both were < 
20 mm x 15 mm x 10 mm in size and weighed < 3 g. Both 
had internal loop antennae and transmission longevity of 
~ 3 months. Prior to use, both telemeters w'ere calibrated 
by placing them in a water bath at various temperatures 
(measured by a reference mercury thermometer) and 
recording their pulse rate. The temperature of the water 
bath was increased in intervals of 2''C every thirty 
minutes between 10 and 45°C. Exponential equations 
relating telemeter frequency and temperature were 
calculated. Telemeters were initially used to test 
attachment sites (groin, armpit and base of the tail) and 
methods (including surgical tape and glue) on one 
bandicoot (mass 1035 g). These particular attachment 
sites were chosen as it was thought that T^,^ at these sites 
would be closer to core T^. Then, one telemeter was 
coated in purified beeswax and surgically implanted into 
another study animal (mass 1160 g). For implantation, 
the bandicoot was anaesthetised with 4% (induction) and 
1% (maintenance) halothane. The abdomen was incised 
161 
