Journal of the Royal Society of Western Australia, 86(4), December 2003 
Table 2 
Number of species and individuals recorded in each habitat 
group for Wanjarri Nature Reserve. Values are mean ± se; n is 
the number of sites in each habitat group; values with different 
superscript letters are significantly different. 
Habitat type 
n 
# of species 
# of individuals 
Groving or creekline mulga 
5 
8.60 ± 1.50* 
25.00 ±3.39 a 
Open mulga 
5 
3.60 + 1.50 b 
11.80 + 4.79 b 
Spinifex 
4 
2.25 ± 0.95 b 
5.25 ± 2.10 b 
Other communities 
7 
3.86 ± 0.74 b 
12.57 ±2.44 b 
Table 3 
Numbers of raptors and all birds sighted during this study at 
Wanjarri Nature Reserve, compared to Hall et al. (1994). All 
values are number of individuals sighted per person hour. 
This study Hall et al. (1994) 
September 
February 
May 
August 
1994 
1979 
1980 
1981 
Raptors 
0.06 
0.28 
0.48 
0.92 
All birds 
9.3 
17.3 
23.4 
37.4 
Discussion 
The data obtained in this survey at Wanjarri Nature 
Reserve support two notable aspects of faunal response 
to short-term dry conditions. The first aspect is that the 
number of bird species and individuals was low during 
the drought conditions of our study. Although we 
recorded most resident passerine species in the reserve 
(Moriarty 1972), the number of individuals that we 
recorded (Table 3) was low compared to the intensive 
study of the reserve conducted by Hall et al. (1994), even 
though we recorded all the resident passerine species 
they did. In addition, the number of raptors that we 
sighted (Table 3) was also low compared to Hall et al. 
(1994). Additional evidence of stress was the absence of 
breeding records during the survey, despite the survey 
having been conducted at a time when many species 
would be expected to breed. These observations suggest 
that that biological productivity was low during the 
drought conditions of our study. 
The second aspect was that the decline in avian species 
richness and abundance at Wanjarri differed 
considerably between groving/creekline mulga and 
other habitats. Areas of groving and creekline mulga 
supported a much higher number of birds, both species 
and individuals, than the other habitats. This suggests 
that these areas act as drought refugia for resident birds 
during drought conditions. The 14 bird species that were 
most common in groving and creekline mulga (Table 1) 
are considered to be resident (Blakers et al. 1984; Higgins 
1999; Higgins et al. 2001). They include nine species that 
can be reliably recorded in mulga communities across 
Australia (Galah, Red-capped Robin, Rufous Whistler, 
Grey Shrike-thrush, White-browed Babbler, Inland 
Thornbill, Chestnut-rumped Thornbill, Southern 
Whiteface and Singing Honeyeater) and four species that 
are considered peripheral in mulga communities 
(Common Bronzewing, White-browed Treecreeper, 
Yellow-throated Miner and Slaty-backed Thornbill; Cody 
1994). The high numbers of bird species and individuals 
recorded from groving and creekline mulga communities 
in this study were in sharp contrast to the spinifex- 
dominated habitats. 
So what are the habitat characteristics and parameters 
of groving and creekline mulga that allow birds to persist 
there in dry conditions? The most obvious characteristic 
is that they have a much greater vegetation density than 
other habitats. The five creekline and groving mulga 
communities had an average of 25 to 30% foliage cover, 
whereas open mulga communities had an average of 
only 15 to 20% (H Pringle, Department of Agriculture, 
unpublished data). Groving and creekline mulga areas 
probably have greater vegetation density than other 
habitats because of differences in soil structure and the 
greater availability of water and nutrients. In contrast to 
open mulga soils, the soils of groving and creekline 
mulga areas usually have a shallow A horizon and 
subsequent horizons in the same texture group as the A 
horizon leading to less resistance to root penetration (P 
Hennig, Department of Agriculture, unpublished data). 
In addition, soils of groving and creekline mulga have 
stable macropores that enable them to absorb up to 10 
times the amount of water of open mulga soils (Greene 
1992). This greater water availability is accentuated 
because groving and creekline mulga occupies low points 
in the landscape, where water and nutrients tend to 
collect (Tongway 1990; Tongway & Ludwig 1990; Ludwig 
et al. 1994). In contrast, open mulga communities tend to 
occur on hardpans and breakaway footslopes where 
there is less water and nutrients (Tongway & Ludwig 
1990; Greene 1992; Ludwig et al. 1994). So, plants that 
grow in areas of groving and creekline mulga have 
greater access to water and nutrients and can form a 
denser canopy. Their access to water and nutrients also 
means that they can maintain greater productivity during 
dry periods, which presumably supports more 
invertebrates and, hence, insectivorous birds. It has been 
shown that areas of groving and creekline mulga support 
higher densities of terrestrial invertebrates than other 
habitats (Whitford et al. 1992; Noble et al. 1996), which 
may be important as many mulga-inhabiting birds forage 
on the ground (Recher & Davis 1997). Unfortunately, no 
comparable information is available on the arboreal 
invertebrate fauna. 
Morton (1990) proposed that vertebrates in arid-zone 
ecosystems, especially mammals, were dependent on 
"uncommon pockets of relatively fertile, moist country 
that are scattered across the arid-zone" to act as refugia 
in times of drought. The significance of this to the 
conservation and management of arid-zone vertebrates is 
that the degradation of these drought refugia, by feral or 
introduced herbivores, changed fire regimes, or 
predation by introduced mammals, could cause species 
extinction. Indeed, Morton (1990) proposed that 
degradation of drought refugia could explain the 
extinction of large numbers of medium-sized mammals 
in the arid-zone. Although extinctions in the arid-zone 
have only occurred among mammal species, there are 
some indications that we may see loss of biodiversity 
among the arid-zone avifauna in the near future (Recher 
& Lim, 1990; Smith et al. 1994; Recher 1999). Thus, 
identification of drought refugia is imperative if we are 
to maintain avian biodiversity in Australia's arid-zone. 
To unequivocally identify drought refugia for birds in 
136 
