Journal of the Royal Society of Western Australia, 90(3), September 2007 
and the telemeter inserted into the peritoneum. The 
incision was stitched and the animal left in a CTR at 28°C 
for 2 weeks to recover. Another telemeter was attached 
to the base of the tail of the same animal using surgical 
tape for comparison between external and internal 
telemetry data. This bandicoot was placed in a CTR on a 
12 hour lightrdark cycle. The room was set to 10°C (night) 
/ 20°C (day) for one week (Cycle 1) then 20°C (night) / 
30°C (day) for one week (Cycle 2) to test the influence of 
T^. After this time, the bandicoot was returned to its 
sheltered, outdoor enclosure and T^, was measured again 
for approximately 2 weeks. These data are shown in 
Figure 1. 
Telemetry data were recorded using an AR8000 radio 
receiver, CU8232 interface, antenna and personal 
computer running AR8000 Temperature Telemeter 
Logging Companion (© 1997 Stig OTracey Spiney 
Norman Systems). Raw data were exported to MS Excel 
for conversion and analysis. 
Results 
Attachment of external telemeters proved to be 
difficult at all sites, with the base of the tail proving the 
best in terms of attachment duration, lowest discomfort 
for the experimental animals, and greatest signal 
strength. Attachment in the groin or armpit using 
surgical tape inhibited the normal movement of the 
bandicoots and individuals were easily able to reach 
these sites to remove the telemeter (reducing attachment 
time to less than 24 hours). Telemeters stayed attached to 
the tail for 5 to 14 days. 
A distinct daily pattern in Tj, was observed for /. 
obesulus based on the data from the surgically implanted 
telemeter (Fig. 1). Under Cycle 1, the average core T^^ 
measured was 36.8±0.1°C (max 39.7°C, min 32.6°C, n = 
4547). Both the highest and lowest Ti^s generally occurred 
in the early hours of the morning (Fig. 1). During the 
inactive phase (0600 to 1800 hrs), T^, was more stable than 
during the night (1800 to 0600hrs - active phase); 
however, there was no difference in average T^ between 
day (37.8 ± 0.1 °C) and night (37.3 ± 0.1°C) throughout the 
controlled temperature study. Under Cycle 2, average T,^ 
Figure 1. Body temperatures of a single male southern brown 
bandicoot. Black line = core T^, measured by a surgically 
implanted telemeter for a bandicoot under ambient conditions; 
dark grey line = external telemetry data under a 10-20°C 
temperature cycle; light grey line = external telemetry data 
under a 20-30'’C cycle. 
was 36.0 + 0.2°C (n = 2502). Base of the tail external 
telemetry data appeared to simply mirror T^ throughout 
both trials. Under Cycle 1, average T^,. was 17.8 ± 0.1 °C 
(max 25.8°C, min 13.6°C, n = 2041). The inactive phase 
T^i. (19.3 ± 0.1 °C) was marginally higher than the active 
phase T^ (16.5 ± 0.1°C). Inactive phase T^^ (19.3°C) was 
virtually identical to T^ during that time. Under Cycle 2, 
average T^^ was 26.8 + 0.1 °C. Once again, the inactive 
phase T^i^ (26.5 ± 0.2°C) was only marginally higher than 
the active phase T^,^ (24.3 ± 0.2°C). 
Discussion 
Most previously trialled methods of attachment of 
external telemeters were deemed inappropriate in this 
study as (i) bandicoots are large enough for the thermal 
gradient between T^^ and T^,^ to become an issue if the 
telemeter was attached at any site except the most 
insulated, (ii) I wanted to be able to remove and reattach 
the telemeters easily and (iii) male (pouchless) animals 
were used. The resources to custom manufacture 
specialised telemeter attachment devices were not 
available, so surgical tape was used as the method of 
attachment. This was because it is cheap, simple to use, is 
easily removable and will fall off on its own after a while 
(important in the field). The 'armpit', groin and base of 
the tail were used, with the base of the tail proving the 
best in terms of ease of attachment / removal, strength of 
signal, reduced bandicoot discomfort and duration of 
attachment. The fact that base of the tail external 
telemetry data appeared to simply mirror T^ throughout 
both trials suggests that the gradient between T^,^ and T^, 
for a mammal of this size (ie over ~ 1000 g) is too great 
and that external body temperature telemetry is not 
viable for species of this size. 
Average T|j measured using internal telemetry was 
marginally higher than previously reported T|^s of resting 
bandicoots at 30°C (33.7 ± 0.2 to 36.1± 0.1 °C; Hulbert & 
Dawson 1974; Withers 1992, Larcombe & Withers 2006) 
and the resting T^, of I. obesulus at 30°© (35.0 + 0.1°C; 
Larcombe 2002). The slightly higher Tj^ measured here 
was expected, as the bandicoots were not resting when 
T|j was measured, but instead continued their normal 
activity. Increased activity results in an increase in T^^ 
(Brown & Dawson 1977). The mean Tj^ of the closely 
related northern brown bandicoot (Isoodon macroiinis) 
under T^s of 12-22°C was 36.2°C (range 34.2 to 38.6°C), 
which is almost exactly the same as the 36.8°C measured 
in this study under Cycle 1 (Gemmell et al. 1997). 
Similarly, the maximum (38.6°C) and minimum (34.2°C) 
T^ measured for I. macrounis arc close to those measured 
in this study (39.7°C and 32.6°C, respectively). This 
shows that I. obesulus, like /. macrourus has a relatively 
labile T^ with the T^, of both varying by ~ 5°C daily. 
The differences in T.s measured in the active and 
b 
inactive phases of the bandicoots natural circadian cycle 
can be explained by two factors. Firstly, a more stable 
inactive phase T|_ may be because the animals were 
intermittently moving during the night/active phase. 
During the inactive phase, Tj_ would be expected to be 
fairly stable as the animals activity levels were relatively 
constant (ie while they slept) however, this constancy 
would be lost during the active phase as the animals 
would have varying levels activity depending on what 
162 
