Journal of the Royal Society of Western Australia, 87(3), September 2004 
of emergent shrubs on quadrats (f 2 ]5 = 3.961, P = 0.042), 
with natural areas having more emergents than the areas 
around oil-pads (P = 0.0132). Also, mean vegetation 
height differed significantly between the three areas (f 2 g7 
= 8.724, P = 0.004), with vegetation in natural areas 
significantly taller than around oil-pads (P = 0.0002) and 
along roads (P = 0.0013). Thus, comparison of vegetation 
measurements between natural and disturbed areas 
varied with the specific habitat type. 
The census results conducted in the four vegetation 
zones were analyzed in two ways. First, across the four 
vegetation zones, there was no significant variation in 
either the number of species (P 3 g = 1.961, P = 0.198) or 
number of individuals (P 3 g = 1.620, P = 0.620) recorded 
according to area sampled (natural areas, oil-pads, 
roadsides). Second, we asked whether, regardless of 
vegetation zone, species richness and number of 
individuals varied between natural areas, oil-pads, and 
roadsides. Although there was no significant variation in 
number of species recorded (P 2 9 = 0.864, P = 0.454), there 
was significant variation in number of individuals 
recorded (P 2 9 = 4.922, P = 0.036). Roadsides supported 
significantly greater numbers of individuals than natural 
areas (P = 0.013) and oil-pads (although not significantly 
so, P = 0.077). The number of individuals did not differ 
significantly between natural areas and oil-pads (P = 
0.301). 
Discussion 
Avifauna 
Of the 110 bird species recorded on Barrow Island and 
reported by Sedgwick (1978), 48 were land birds. These 
48 species included 15 permanent residents, six regular 
migrants, and 27 irregular visitors. Of the 15 species that 
Sedgwick considered to be residents, the only species 
that we did not observe was the Southern Boobook 
(Nino:x novaeseelandiae). Our quadrat counts were 
conducted during daylight hours and thus it is not 
surprising that we not detect this species on the quadrats. 
Nevertheless, we did do considerable driving at night, 
and this species was neither heard nor observed. 
Additionally, W. H. Butler (pers. comm.) reports that the 
Southern Boobook is no longer resident on the island. Of 
the 18 species that we recorded, four species were 
considered by Sedgwick to be non-residents: the Zebra 
Finch and Black-eared Cuckoo were considered to be 
irregular visitors, and the Horsfield's Bronze Cuckoo and 
Sacred Kingfisher were considered to be regular 
migrants. Considering these species in turn, the Zebra 
Finch now appears to be a regular migrant to the island 
and breeds there every year (L. McClements, pers. 
comm.). The Black-eared Cuckoo is now known to be an 
irregular migrant, and Horsfield's Bronze Cuckoo a 
regular migrant (W. H. Butler and R. Johnstone, pers. 
comm.). Both species are widespread on the island 
during the breeding season, and both undoubtedly 
commonly breed there (see Pruett-Jones and Tarvin 
2001). We have no observations or other information to 
contradict the status of the Sacred Kingfisher as a regular 
migrant. 
Sedgwick recorded four other species as regular 
migrants that we did not observe: the Pallid Cuckoo 
(Cucuius pallidus), the White-throated Needletail 
(Hirundapus caudacutus) which is listed as the Spine-tailed 
Swift in Sedgwick (1978), the Fork-tailed Swift (Apus 
pacificus), and the White-winged Triller (Lalage tricolor). 
Records of these species have been scattered since 
ornithological observations on Barrow Island began and 
none of these species were recorded by Whitlock (1918, 
1919) or Serventy and Marshall (1964). It appears that the 
Pallid Cuckoo and White-winged Triller are both passage 
migrants to Barrow Island and the two swift species 
irregular visitors from Asia (R. Johnstone, pers. comm.). 
All other land birds listed by Sedgwick and not 
observed during this study or discussed above (25 
species total) are irregular visitors and do not breed on 
Barrow Island. This list includes some species that could 
occasionally breed on the island (e.g., Whistling Kite 
(Haliaster sphenurus) and Wedge-tailed Eagle (Aquila 
audax ; the record by Sedgwick is now thought to be a 
mis-identified juvenile White-breasted Sea-Eagle)) as well 
as species that are accidental occurrences on the island 
and do not breed there (e.g., Galah ( Cacatua roseicapilla) 
and Little Corella (C. pastinator)). The number of such 
accidental species visiting the island or blown over to the 
island during storms will likely increase as observations 
continue on the island. 
Sedgwick (1978) calculated total population estimates 
for all of the species he recorded on his quadrats, as did 
we in this study (Table 2). There are several points of 
comparison to make. First, we concur with Sedgwick that 
the estimates for the Osprey are too high. This is a 
maritime species that can be seen from inland habitats, 
but inland habitats such as is available on Barrow Island 
cannot be considered as regular habitat of this species. At 
the time of Sedgwick's survey, 23 Osprey nests were 
known on the island, and although we did not attempt to 
count Osprey nests we suspect that there are 
approximately this number of active nests still present. 
Similarly, we believe our estimate of the total population 
of White-breasted Sea-Eagle is too high, for the same 
reasons as outlined above. At the time of Sedwick's 
survey, six sea-eagle nests were known on the island, 
and during our work we had reported to us the locations 
of five nests. More accurate population estimates for the 
Osprey would be approximately 50 birds and for the 
White-breasted Sea-Eagle, 10-12 birds. 
For the three most common species (Spinifexbird, 
White-winged Fairy-wren, and Singing Honeyeater), our 
total population estimates and those of Sedgwick (1978) 
are comparable (Table 2). For many of the other species, 
there are striking differences. In some cases these 
differences are easily explainable. For example, the Zebra 
Finch is clearly now more common on Barrow Island 
than it was during the work of Sedgwick (W. H. Butler, 
pers. comm.), as reflected in our results and estimates. 
Also, if resident numbers of the migratory Black-eared 
Cuckoo and Horsfield's Bronze Cuckoo fluctuate from 
year to year, the differences between our estimates and 
those of Sedgwick (1978) may reflect annual variation in 
the size of breeding populations. 
For other species listed in Table 2, the differences 
between our estimates and those Sedgwick appear to 
reflect population declines. The most striking example is 
the Tree Martin, listed by Sedgwick as the fourth most 
abundant species on the island (but see Storr 1984), and 
106 
