Variation in the composition of buloke woodland bird assemblages : theoretical and management implications

Over the past few decades, declines in bird population densities have been recorded in agricultural regions worldwide. These declines are generally attributed to the related threatening processes of habitat loss, fragmentation and degradation, which ultimately are in turn driven largely by the intensification and spread of agriculture. In order to take action to effectively manage threatened avifauna and halt population declines in these regions we require the ability to predict the consequences of habitat alteration on species with reasonable accuracy. The Wimmera Plains area of western Victoria, Australia is a highly fragmented agricultural landscape, from which approximately 95% of the original native vegetation has been removed. The central question of this study is: how can we predict the consequences of possible future management actions on the avifauna of buloke Allocasuarina leuhmannii woodland remnants in the Wimmera in order to best manage and restore habitat for the region's native birds? I investigated inter-annual variability in the composition of the bird assemblages of a series of buloke woodland remnants and its influence on the potential of nested-subsets theory and the identification of bird-habitat relationships to contribute to conservation management of native birds in the Wimmera region. I also investigated the potential for the identification of more proximate mechanisms (in this case, through observations of foraging behaviour) determining patterns of habitat use by birds to characterise bird-habitat relationships. I compared the results of two year-long surveys of the bird assemblages of a series of buloke woodland remnants in the Wimmera, the first conducted in 1994-1995 and the second in 2001-2002. The bird assemblages of remnants varied substantially between these two survey periods. The bird species richness of each remnant was unrelated between the two periods, with an average of 63% minimum species turnover recorded. On average, the bird assemblage's composition in each remnant bore greater resemblance to those of other remnants in the same year than to that of the same site in the other survey period. The change in distribution of most species did not differ significantly from that expected if the species had redistributed at random among the sites. It therefore appears that although the remnant vegetation of the area is highly fragmented with minimal inter-patch connectivity, numerous bird movements among remnants occurred between the two survey periods. The results of this comparison suggest that the avifauna of this region is probably highly variable from year to year, and predictions made on the basis of species distributions as recorded during a single year's surveying could be quite misleading. The biotas of a suite of neighbouring patches of remnant vegetation often form a series of nested subsets, in which the species present in species-poor patches are proper subsets of those present in richer patches. I used the Temperature Calculator of Atmar and Patterson (1995) to predict likely avian colonization and extinction events in buloke woodland remnants on the basis of the pattern of nestedness of the assemblages measured in 1994-95. Although the degree of nestedness increased between the 1994-95 and 2001-02 survey periods, there was no linear relationship between the generated probability of extinctions or colonizations and the accuracy of the predictions, and the predictions derived from the Calculator were no more accurate than a second set generated by a simple non-nested model. For the nestedness matrix to generate accurate predictions, an increase in nestedness must be due to a minimization of unexpected species' presences and absences rather than an extensive redistribution of species among remnants, such as I found. A subset of the bird species present in the remnant buloke woodlands is known to be declining in other fragmented landscapes. I investigated spatial and temporal variation in the relationships between landscape and habitat characteristics of the remnants and the presence of this suite of species. Habitat relationships detected for these species were compared with a) those detected for the same species in the same remnants based on surveys undertaken seven years previously, and b) those reported from other fragmented habitats in southern Australia. Most of the statistical relationships with remnant size, isolation and habitat degradation that have been displayed by the study species in other landscapes were not apparent in the buloke system. The observed changes in distribution led to temporal differences in the detection of landscape and habitat relationships for several species. Consequently, results from the two survey periods presented very different pictures of the habitat relationships exhibited by this suite of declining species. For the buloke habitat system studied, it appears unwise to rely on correlative relationships between bird species' presence and habitat parameters, without having an understanding of the mechanisms through which the habitat factors exert their influence. In such dynamic habitat systems, we must be cautious when extrapolating from the findings of short-term studies to longer temporal scales, especially in relation to making conservation management decisions. An improved understanding of the processes causing distributional patterns will facilitate better predictions of species' responses to habitat change. Investigation of the influence of habitat fragmentation and degradation on factors such as food availability is likely to contribute to a greater understanding of bird-habitat relationships in the Wimmera. Ground-foraging insectivores are disproportionately represented among the 'woodland decliner' suite of bird species, and reduced food availability in remnants where the ground layer has been significantly disturbed might be partly responsible for their decline. I studied the foraging behaviour and microhabitat use of seven species of groundforaging insectivorous birds in native, grazed and ungrazed but weed-infested buloke woodland remnants, in order to identify the mechanisms through which habitat degradation might be affecting their habitat use. Evidence was found of a negative impact of one or both of these types of degradation (i.e. grazing or weed infestation) on five species. Three species that employed a range of foraging manoeuvres to attack prey used more potentially energetically-expensive aerial manoeuvres significantly more frequently in weedy than in native or grazed remnants. Two species selectively foraged near trees in grazed sites, and three avoided foraging in microhabitats with a high percentage cover of exotic grasses in weedy sites. Brown treecreepers Climacteris picumnus were also less likely to be present in weedy sites that had been protected from grazing than in either grazed or native sites. These results suggest that although grazing appears to have a detrimental impact on foraging habitat for ground-foraging birds in remnant buloke woodland, its exclusion from previously disturbed remnants is not sufficient on its own to restore habitat values for these birds. This study has demonstrated that even small and degraded buloke woodland remnants appear to have substantial conservation value for woodland birds in western Victoria. It has also brought into question our ability to accurately extrapolate from short-term patterns of bird distribution, either spatially or temporally, particularly for assemblages whose composition varies markedly over time.