278 
P.C. Withers 
surface-area specific evaporative water loss was 
calculated and converted to resistance by assuming 
that the gradient in water vapour density was the 
difference between the vapour density for air 
saturated at the ambient temperature and the 
vapour density of excurrent air. The resistance for 
frogs not in the water-conserving posture was not 
recalculated although their exposed surface area 
would have been higher than that of frogs in the 
water-conserving posture; this means that their 
resistance was potentially underestimated by as 
much as 33% (if all of the skin were exposed). The 
calculated resistance includes some respiratory 
water loss, hence the actual cutaneous resistance is 
slightly lower than the calculated value. No 
correction was made for the air boundary layer to 
resistance. 
For behavioural observations, eight frogs were 
placed individually in small plastic containers, 
either white or black in colour, with a transparent 
lid. The frogs were transferred to the glass tube, 
with either humid air or dry air flow, and kept in 
darkness for about 1 hour. The frogs were then 
removed from the container, and their colour was 
noted, and eight chromatophores on the 
transparent toe webbing of a hind foot were 
quantified by examination under a dissecting 
binocular microscope for their chromatophore 
index (Cl; Hogben and Slome 1931). 
A dark-background and a light-background 
adapted frog were sacrificed, and their dorsal skin 
removed and placed in a diffuse reflectance 
accessory of a Varian dual-beam 
spectrophotometer (DMS-80), and the reflectance 
of the skin determined for wavelengths from 200 to 
900 nm. 
(ft* 
*4 ' 
Figure 1 Light-adapted (right; chromatophore index = 
2) and dark-adapted (left; chromatophore 
index = 4) desert tree frogs, Litoria rubella. 
RESULTS 
The rate of evaporative water loss of desert tree 
frogs was relatively high, about 6 mg min 1 or 120 
mg g 1 h l , if they did not adopt a water-conserving 
posture in the hygrometer (Table 1), or were placed 
on a mesh platform to prevent them from 
concealing their ventral surface from the dry air 
stream. The skin resistance to evaporative water 
loss was about 1.1 sec cm 1 for these frogs, which is 
similar to the resistance of a free water surface 
(about 1 sec cm 1 ). In contrast, tree frogs which 
adopted a water-conserving posture and concealed 
their ventral surface from the dry air stream, had a 
considerably lower evaporative water loss (1.2 mg 
min 1 , 26 mg g 1 h 1 ) and a higher resistance of 7.3 
sec cm 1 . 
The desert tree frogs were observed to change 
colour dramatically, varying from bright white to a 
dark brown/black (Figure 1). The chromatophore 
index varies from 1 (fully aggregated) to 5 (fully 
dispersed; see Figure 2), and the index of toe-web 
melanophores was generally found to reflect the 
dorsal colour of the frog. The index changed 
according to the general body colour, from 2 
(almost completely aggregated for pale frogs) to 5 
(fully dispersed for dark frogs). For frogs kept on a 
white background, the Cl was 2.5 ± se 0.3 (n=9), 
which was significantly different (t 15 = 4.7, 
P<0.0002) from the Cl of frogs on a dark 
background of 3.9 ± se 0.2 (n=9). For frogs from 
black or white backgrounds and maintained in the 
dark, in moist or dry air, the chromatophore index 
(Table 2) was significantly related (2-factor 
ANOVA) to whether the air was moist or dry 
(P<0.0005) but not to the original background 
colour (P>0.80). Frogs in dry air had lower 
melanophore indices than frogs in moist air i.e. 
Table 1 Body mass and rates of evaporative water loss 
and resistance for the desert tree frog Litoria 
rubella, when not in, and in, the water- 
conserving posture. Values are mean ± 
standard error, with the number of 
observations (n). 
Not in 
In 
Water Conserving Water Conserving 
Posture 
Posture 
(n=8) 
(n=ll) 
Body mass (g) 
3.12 ± 0.23 
2.64 ± 0.26 NS 
EWL (mg min' 1 ) 
7.63 ± 1.91 
1.84 ±0.14* 
MSEWL (mg g 1 h' 1 ) 
173.2 ± 31.6 
38.8 ± 5.9* 
Resistance (sec cm 1 ) 
1.09 ±0.13 
7.29 ± 0.58* 
* Significant difference between posture groups 
(P<0.05), by Student t-test. 
NS No significant difference between posture groups 
(P>0.05), by Student t-test. 
