Records of the Western Australian Museum 17: 277-281 (1995). 
Evaporative water loss and colour change in the Australian desert tree frog 
Litoria rubella (Amphibia: Hylidae) 
P.C. Withers 
Department of Zoology, University of Western Australia, Nedlands, Western Australia 6907 
Abstract - The desert tree frog, Litoria rubella, is a small (2-4 g) frog found in 
northern Australia. These tree frogs typically rest in a water-conserving 
posture, and are moderately water-proof. Their evaporative water loss when 
in the water-conserving posture is reduced to 1.8 mg min' 1 (39 mg g' 1 h' 1 ) and 
resistance increased to 7.3 sec cm' 1 , compared with tree frogs not in the 
water-conserving posture (7.6 mg min' 1 ,173 mg g' 1 h' 1 ,1.1 sec cm 1 ). When in 
the water-conserving posture and exposed to dry air, the tree frogs 
dramatically change colour from the typical gray, brown or fawn, to a bright 
white. The toe-web melanophore index decreases from 3.8 for moist frogs, to 
2.3 for desiccated frogs. The high skin resistance to evaporation and white 
colour of tree frogs when exposed to desiccating conditions appear to be 
important adaptations to reduce evaporative water loss and prevent 
overheating when basking in direct sunlight. 
INTRODUCTION 
Many species of Australian tree frogs of the 
genus Litoria, are arboreal and frequently perch in 
exposed sites on vegetation. The desert tree frog, 
Litoria rubella (Gray 1842) is a small, robust species 
found throughout northern Australia (Cogger 1992; 
Tyler et al. 1994), It survives in arid areas, generally 
by sheltering in cracks between rocks, or other 
habitats that provide protection from wind and 
solar insolation, such as bore holes and water 
tanks. However, these tree frogs have been 
observed to perch in sites exposed to direct solar 
insolation (A.R. Main, pers. comm.). 
Perching in locations exposed to wind and solar 
insolation would potentially subject a small frog, 
such as Litoria rubella, to severe desiccation and 
high body temperatures. Some of the arboreal 
Australian Litoria species have been reported to 
have reduced rates of cutaneous evaporative water 
loss (Withers et al. 1984; Buttemer 1990) and this 
would minimise the potential for desiccation when 
basking. The objective of this study was initially to 
examine the rate of cutaneous evaporative water 
loss for the desert tree frog, but preliminary 
observations revealed that these tree frogs often 
blanched in colour to a bright white when exposed 
to dry air. Desert tree frogs are generally grey, 
brown or fawn in colour, with a broad dark 
vertebral band and a black stripe along the side of 
the head (Cogger 1992; Tyler et al. 1994). 
Consequently, the study was extended to 
investigate the potential roles of ambient humidity 
and background colour in controlling skin colour. 
MATERIALS AND METHODS 
Desert tree frogs were collected from a bore on 
Mallina Station (26° S, 114° E), in the arid Pilbara 
region of Western Australia. They were 
transported to the laboratory in Perth, for study. 
Body mass was measured to ± 1 mg; mass ranged 
from 1.6 to 4.3 g. 
The rate of evaporative water loss was 
determined for individual L. rubella at an ambient 
air temperature of 22-24 °C, using flow-through 
hygrometry. Compressed air was dried by passage 
through a column of silica gel, and flowed at a rate 
of 500 ml min' 1 through a glass tube containing a 
frog. The excurrent air was passed through a 
Western Digital DP1100 dewpoint hygrometer for 
determination of the water content of the air. The 
analog voltage output of the hygrometer was 
monitored at 30 second intervals by a PC, and the 
dewpoint converted to absolute humidity (mg H 2 0 
L 1 ) using the psychrometric equations of Parrish 
and Putnam (1977). The rate of evaporative water 
loss (EWL; mg H 2 0 min 1 ) of the frog was 
calculated from the absolute humidity of the 
incurrent and excurrent air, and the mass specific 
evaporative water loss (MSEWL; mg H 2 0 g' 1 h 1 ) 
was then calculated from the body weight 
(measured to ± 0.001 g). The exposed dorsal 
surface area of the frog (whether it was or was not 
in the water-conserving posture) was estimated to 
be 2 / 3 of the total body surface area (see Withers et 
al. 1982a; Buttemer 1990) calculated from body 
weight by the allometric equation of McClanahan 
and Baldwin (1969) i.e. cm 2 = 2 / 3 9.9 g 0567 . The 
