Journal of the Royal Society of Western Australia, 87(2), June 2004 
surplus was shaken off to leave a film of powder of 
similar thickness to that on the trunks. These, and five 
clean filter papers, were individually placed in petri 
dishes. A drop of water was added to maintain moisture. 
Five worker ants (Iridomyrmex cJmsei) were then placed in 
each petri dish and the lids were replaced. Dishes were 
maintained in darkness at 20°C, with a drop of water 
added daily, and observed at 12 h intervals to record the 
numbers of live and dead animals. This experiment was 
conducted at Curtin University using locally-collected 
ants. However, I. chasei also occurs at Dryandra and is 
known to ascend the trunks of various tree species, 
where it tends sap-sucking insects (Hemiptera). 
A further experiment was performed in the vicinity of 
Curtin University in which 20 cm long, 13 mm diameter 
wooden dowels were lightly dusted with powder from 
the powderbark wandoo trees and the upper end was 
coated with dilute honey bait. Ten powdered and 10 
powder-free dowels were then 'planted' vertically in the 
ground and continuously observed over a 24 h period. 
Ants that climbed the dowels to reach the honey were 
scored and collected for subsequent identification. This 
experiment was performed on 5-6 April 2003 and again 
on 1-2 May 2003; the second experiment differed from 
the first in that the dowels were placed at an angle of 45° 
from the vertical in order to make them easier to climb. 
Results 
Powderbark wandoo had a much thinner exfoliated 
bark than wandoo (mean ± SD: 0.28 ± 0.20 vs 1.16 ± 0.45 
mm). Examination of the powderbark wandoo bark 
surface under a stereo-microscope revealed a coating of 
peridermal cells that were sloughing off, either in clumps 
or singly. Examination of the powder that had been 
brushed from the trunk indicated that this comprised of 
sloughed-off peridermal cells, with powder grains 
measuring on average 35 pm in length. The cells 
appeared to be empty, although some contained 
structures that may be fungal hyphae. Plating of the sub¬ 
sample of powder revealed the presence of a range of 
common saprophytic, epiphytic fungi (Scytalidium sp., 
Aspergillus sp., Ulocladium sp., Aureobasidium sp.), whose 
presence would be expected on decaying organic matter. 
The chemical analysis of the powder confirmed the 
presence of triterpenoids. Wandoo bark was similar in 
Figure 1. Mean mortality (number of ants per dish ± SD) of 
Iridomyrmex chasei ants (n = 5 replicates of 5 ants per dish) 
exposed to discs of filter paper that have, or have not, been 
coated with powder from the bark of powderbark wandoo trees. 
appearance and also exhibited sloughing peridermal 
cells, which did not accumulate to form a powder. 
Ant mortality was much greater on the powdered 
paper discs than on the control discs (Fig. 1), with all 
ants on the powdered discs having died after 72 h, 
compared with only 1 death per disc in the controls. 
Inspection of dead ants indicated that the powder had 
adhered to their body surfaces, with the greatest 
concentration around the mouthparts and on the tarsi. 
All ant species that ascended the dowels were 
common species that also occurred at Dryandra. In the 
first baiting experiment, single individuals of I. chasei, 
Pheidole ampla perthenesis and Paratrechina longicornis 
ascended the clean dowels to reach the bait; although 
several ants attempted to climb the powdered dowels, 
only one ant (J. chasei) ascended a powdered dowel, and 
this ant fell off after climbing half the distance. In the 
second experiment, in which dowels were set at a 45° 
angle, ants reached the honey bait on seven of the 10 
clean dowels (mean ± SD: 2.5 ± 2.2 ants), but no ants 
successfully ascended those that had been powdered; all 
ants were P. longicornis. 
Discussion 
Our findings suggest that the powder of powderbark 
wandoo trunks may be responsible for the lower usage 
of trunks by arthropods, when compared with the similar 
appearing, and co-occurring wandoo tree species. 
The powder of powderbark wandoo trunks may 
influence arthropods in a number of ways. The 
triterpenoids or other unknown chemicals in the powder 
may have a repellent or toxicity effect against arthropods. 
Repellency is a possibility that we did not test directly, 
although it is noteworthy that the powder was odourless 
to humans. Our bioassay with 1. chasei shows that the 
powder is capable of causing direct insect mortality, 
which could have been caused by the toxic triterpenoids 
or other chemicals in the powder. Evidence against this 
is the fact that ants attempted to climb the dowels, even 
when powder was present. Toxic or repellent effects of 
chemicals in the powder could be tested in the future by 
exhaustive solvent extraction of the powder before 
conducting boassays. 
Secondly, the powder may have blocked the spiracles 
of certain arthropods, hence interfering with their normal 
respiration. Close inspection revealed that this did not 
happen, as the powder grains are at least double the 
diameter of the spiracles of /. chasei, which are of a size 
that is typical of many bark-associated arthropods. 
Thirdly, the powder may render the bark difficult to 
walk on. This is supported by the concentrations of 
powder that had accumulated on the tarsi of ants which 
were exposed to this material, and also the fact that few 
or no ants were able to climb the powdered dowels. A 
similar mechanism has been reported in the pitcher plant, 
Sarracenia purpurea, in which wax inside the pitcher 
sloughs off when insects attempt to climb out, causing 
them to fall into the fluid within (Juniper 1986). Wax 
blooms on the stems of various species of Macaranga have 
also been found to prevent ants from walking across 
them (Federle et al. 1997), and wax layers on the surface 
of juvenile Eucalyptus nitens and E. globulus prevent 
82 
