In December 2005, a wildfire burnt a large area of semi-arid eucalypt woodland along ~10 km of the Norseman-
Coolgardie Road north of Norseman in the Great Western Woodlands (GWW), Western Australia. Few birds used the
burnt area in the first year after the fire and these were mainly ground and shrub foraging insectivores. There was no
influx of seed-eaters or open-country species as reported for post-fire habitats elsewhere in southern Australia. The
greatest number of individuals and species of birds occurred in the second year post-fire when ground and shrub
vegetation was floristically most diverse. Canopy foragers were attracted to the burnt area in the second year by an
outbreak of psyllid insects on seedling eucalypts. At the same time, bark dwelling arthropods associated with the standing
stems of fire-killed eucalypts attracted bark-foragers. From the third year, small insectivorous ground, shrub, and canopy
foragers dominated the avifauna on the burnt area. These foraged on fire-killed shrubs, as well as living vegetation,
including the lignotuberous regrowth of eucalypts. Bark foragers were uncommon after the second year.
Throughout the study, the burnt area had fewer species and individuals than adjacent unburnt habitats. Compared
with unburnt woodlands there were few differences in how species foraged on the burnt plots, but most species foraged
lower reflecting the stature of the vegetation in the burnt woodland. Nectar-feeders, fruit-eaters, large insectivores, raptors,
and parrots, although common in the unburnt woodland, were absent or rare in the burnt area. This reflected the limited
regrowth of vegetation on the burnt area, which lacked the structural and floristic complexity of nearby unburnt woodlands.
Ground foragers probably commenced nesting on the burnt area in the first year, with shrub and canopy foragers
nesting from the second year. However, after five years, there was no evidence of large insectivores, nectar-feeders,
raptors, seed-eaters, or foliage-eaters (i.e., parrots) nesting despite their abundance in adjacent unburnt woodland.
Some of the unburnt woodlands monitored in this study were even-aged regeneration estimated to be 30–50 years
post-fire or logging. Regardless of origin, these even-aged plots lacked the diverse avifauna associated with mature
woodlands and suggest that post-fire recovery of birds and vegetation in these woodlands is likely to take decades
and probably more than 100 years. If so, human activities that increase fire frequency in the GWW, including climate
change and fuel-reduction burns, will have long-term adverse impacts on regional biodiversity exceeding those associated
with wildfires in less arid forests and woodlands where rates of recovery are more rapid.