PARENT SESSION
Contributed Oral Session 48: Landscape Ecology: Animal Dynamics
Tuesday, August 9, 8:00 AM - 11:30 AM, Meeting Room 520 B, Level 5, Palais des congrès de Montréal
Empirical evidence of landscape-level thresholds of habitat cover: woodland-dependent birds in rural landscapes in south-eastern Australia.
Radford, James*,1, Bennett, Andrew1, 1 Deakin University, Burwood, Victoria, Australia
ABSTRACT- Wildlife conservation in rural environments requires a sound understanding of how natural systems respond to land-use change at the landscape scale. One approach that has recently received considerable attention from theoreticians and modelers is the concept of ecological thresholds. However, empirical data pertaining to thresholds are rare, especially at a landscape scale. Here, we present empirical data collected using an innovative sampling strategy that tracks the response of woodland-dependent avifauna to habitat loss at the landscape level. Woodland birds were surveyed in 24 'landscapes' (each 100 km2) in temperate south-east Australia that sampled a gradient in tree cover from 60% down to 2% cover. Critically, the sampling unit was the 'landscape': all response and explanatory variables (e.g. tree cover, geo-climatic, land-use, configuration) were measured at the landscape scale. A variety of response types, including thresholds, were identified. Multivariate analyses showed that tree cover was the most important predictor of species richness of woodland-dependent birds, accounting for 55% of the variation among landscapes. Linear and non-linear models were fitted to examine the univariate relationship between tree cover and species richness. The best models showed a marked discontinuity (i.e. threshold) at 10% tree cover: above this value species richness appeared relatively stable; below 10% cover there was a disproportionate loss of species as the woodland bird community 'crashed'. The assemblage-level response was not necessarily mirrored by individual species, which showed a variety of responses to landscape change, including linear decline, curvilinear, step-threshold and quadratic. The level of tree cover at which population decline commenced, and at which local extinction occurred, differed between species. Significantly, the process of decline for many species commenced well above the 10% threshold identified for species richness. Management must therefore ensure the viability of individual species rather than focus on thresholds of species richness. In contrast to the 10% threshold, our data indicates that 30-35% native vegetation cover is necessary to preserve resilient populations of most woodland birds in rural landscapes of south-east Australia.
Key words: avifauna, landscape, threshold, Australia