BIOLOGICAL INVENTORY FOR CONSERVATION EVALUATION .3. RELATIONSHIPS BETWEEN BIRDS, VEGETATION AND ENVIRONMENTAL ATTRIBUTES IN SOUTHERN AUSTRALIA

Relationships between diurnal, terrestrial birds, vegetation and envir onmental attributes were investigated in the open Eucalyptus forests o f south east Australia. Of the 83 bird species recorded, the probabili ty of occurrence of the Yellow-faced Honeyeater, Superb Fairy-wren, Go lden Whistler, Fan-tailed Cuckoo and Shining Bronze-Cuckoo was modelle d. The Yellow-faced Honeyeater was the only taxon modelled whose distr ibution could be best explained using climatic variables. The other fo ur species were related to habitat features at the site level. The mod els developed for these species were specific to each data set used an d are unlikely to be appropriate for other data collected at other tim es of the year. Statistical modelling was used to investigate the resp onse of bird assemblages to environmental and biological regimes. The models reflect differences in the bird species composition of sites in relation to elevation and soil moisture/nutrient gradients across the region, Birds favouring higher elevations were separated from birds f avouring mid and lower elevations. Birds favouring wetter environments were separated from those using drier habitats although these species were also present in the wetter environments. Differences in the dist ribution of birds related to elevation were explained by the environme ntal attributes minimum temperature of the coldest month and mean prec ipitation of the driest quarter which also affect vegetation formation s and associations across the study region. Vegetation floristics and topographic position, which can reflect the moisture and nutrient stat us of sites, were important in characterising the specific habitat req uirements of a number of birds. Physical and biological regimes were s ignificantly related to bird species richness, Species richness was fo und to be greatest in moist forests found at mid and low elevations th roughout the year. A stratified field survey was successful in recordi ng the majority of forest birds that can occur in the eucalypt forests of south east Australia but sufficient data for reliably modelling th e distribution and habitat quality of birds were obtained for only abo ut 10% of the birds recorded. Even when data from two census periods w ere aggregated, data were sufficient for only 27% of the total number of birds recorded to develop a habitat model, Suggested modifications to the survey design based on the data and analyses used in this study , are discussed in terms of the intended purpose of the field survey.