Biological inventory for conservation evaluation .2. Composition, functional relationships and spatial prediction of bird assemblages in southern Australia

Data collected from a systematic, stratified regional survey of the open Eucalyptus forests of south east Australia were used to determine if predictable patterns exist in the bird assemblages identified in the region and the extent to which these associations could be used to predict the potential complement of bird species in unsampled areas of the region. Eighty three bird species were recorded in the open Eucalyptus forests considered in this study. Approximately 30 to 40% of the bird species were recorded in a range of environments and appeared to be widely distributed across the region. Less than 20% of the species recorded occurred in more than half of the sites censused. The birds recorded did not appear to form consistent, discrete and predictable assemblages. Few birds were consistently associated with each other throughout the year. However, two assemblages of species consistently occurred together: a group of 11 of the most frequently recorded and widespread species and two relatively sedentary species restricted to wetter environments. The use of the ubiquitous bird species as indicators of the likely species complement at a particular site or area is limited. The Yellow-tufted Honeyeater and Bell Miner were consistently associated with each other throughout the year and were found in wet forests in gullies and along creeks and rivers. These environments also had the highest species richness of birds. These birds may be useful as indicators of particular types of environments and areas of relatively high species richness. Broad trends in the distribution of birds were associated with perceived elevation and soil moisture/nutrient gradients across the study region. BIOCLIM provided a means of characterising these broad trends in the distribution of birds by defining climatic envelopes for clusters of sites. Of the variables used to define these climatic envelopes, temperature variables were the most useful in discriminating between groups of sites. However, only a limited number of discrete assemblages of birds associated with groups of sites could be used for predicting the occurrence of birds in unsampled areas in the landscape.