Science to inform habitat restoration for woodland bird communities of the Tasmanian Midlands

Agricultural intensification over the last 50 years has been a major cause of global biodiversity decline and continues to result in habitat loss, fragmentation and degradation. The agriculture sector must increase production to feed a rapidly growing human population but, to be sustainable, must also maintain the ecosystem services provided by the biodiversity on which it relies. Just how to achieve this is the current focus of much scientific debate. What is clear, is that the world's current estate of protected areas is no longer sufficient to conserve healthy wildlife populations. Ecological restoration of degraded farmland and patches of remnant habitat will be necessary if we are to halt, and ultimately reverse, trends of biodiversity loss. Populations of many birds are declining in agricultural landscapes around the world. Agricultural intensification has been suggested as the primary cause of these declines through processes such as increased pesticide use, mechanisation of farming practices and the removal of critical habitat features including hedgerows, large old trees, wetlands and coarse woody debris. The impacts of land use intensification on avifauna have been best studied in the farmland birds‚ÄövÑvp of Europe and North America, but in Australia it is woodland birds‚ÄövÑvp that are of most conservation concern. Woodland birds are increasingly threatened because of their restriction to ever smaller patches of remnant habitat surrounded by agriculture, where they are vulnerable to edge effects, exotic predators and higher levels of interspecific competition. A primary goal of the thesis is to describe current patterns of patch occupancy and abundance of birds in the agricultural Midlands region of Tasmania, Australia, and explore the processes that underlie those patterns. Despite having high biodiversity value, very little is known about the current state of terrestrial avifauna in Tasmania, especially in the Midlands. The data collected in this thesis will be used to guide local landscape restoration efforts and inform decisions on which habitat features are most important to restore for Tasmanian birds, what species could be selected as targets for restoration, and how birds might respond to future environmental change, whether that be continued habitat loss or revegetation. In the first data chapter, I surveyed birds at 72 sites across the Midlands including in woodlands that had previously been surveyed 20 years ago. I used new statistical techniques to determine which elements of habitat are most important in influencing the composition of woodland bird communities and compared contemporary survey data with historical records to understand how birds have responded to land use change over the last two decades. The amount of woodland cover at survey sites, structural complexity of vegetation and the presence of an aggressive honeyeater species, the noisy miner (Manorina melanocephala), had the strongest effects on birds. Small to medium-sized arboreal foragers appear to have declined while large-bodied granivorous birds have increased. I make practical recommendations for the restoration of habitat for local avifauna, paying particular attention to the threat that noisy miners pose to ongoing restoration efforts. In the second data chapter, I seek a deeper understanding of how noisy miners might exclude other bird species from suitable woodland habitat. Noisy miners are well recognised in Australia as a leading cause of population decline in small woodland birds. Indeed, in Chapter Three, changes in the abundance of noisy miners was identified as a key factor in explaining changes in species richness at survey sites and their presence had a agricultural habitat loss and fragmentation and now dominate remnant woodlands throughout the eastern states. I tested the hypothesis that interference competition with miners could result in chronic stress among cohabiting bird species, with the potential to force individuals to abandon miner-dominated habitat or otherwise reduce fitness such that they can no longer persist. I captured 86 individuals of a model passerine species, the superb fairy-wren (Malurus cyaneus) from six woodland sites, three with miners and three without, and used heterophil-to-lymphocyte (H:L) ratios to infer relative levels of chronic stress. H:L ratios were 1.8 times higher in fairy-wrens from remnant woodlands occupied by miners, suggesting higher levels of physiological stress. H:L ratios were also negatively associated with the residual mass of fairy-wrens and positively associated with the presence of a blood parasite, Haemoproteus spp. Noisy miner presence might also be correlated with other potential stressors, such as food scarcity, in small patches of degraded habitat. I suggest further experiments to clarify whether conflict with miners is the proximate cause of chronic stress in fairy-wrens living in remnant woodlands. Finally, I applied new techniques to test a long-hypothesised mechanism for bird declines in agricultural landscapes: a high frequency of nest predation due to changes in habitat structure or an increased abundance of nest predators. I used motion-sensor cameras to monitor 84 nests of brown thornbills (Acanthiza pusilla) and superb fairy-wrens, and terrestrial LiDAR to quantify the three-dimensional structure of vegetation at nesting sites. A diverse range of predators were recorded preying on nests, but overall rates of predation were no higher than previously recorded for these species living in non-agricultural landscapes. Daily survival rates (DSR) of nests were influenced by the amount of nearby edge habitat, nest height, surrounding woodland cover and the density of vegetation, although the direction of these effects was not always as predicted. DSR declined with increasing vegetation clutter at nest sites and was higher in areas with more edge habitat. I suggest that nest predation in fairy-wrens and thornbills is higher in larger and more intact woodlands because this habitat supports a greater abundance of native predators. My results indicate that restoration practitioners working in the Midlands should focus on restoring structural complexity of vegetation in remnant habitat, particularly in the midstorey, to provide small and medium-sized birds with protection from noisy miners, safe nesting sites and suitable foraging habitat. Connecting remnant woodlands through planting wildlife corridors will benefit some birds simply through increasing levels of wooded cover but this should be done with caution to mitigate the risk of noisy miners further increasing their distribution. I strongly encourage a continued long-term monitoring effort in restored areas as they mature to ensure that restoration sites do not constitute an ecological trap, whereby birds prefer these habitats but their fitness is reduced. While any form of bird data collected in the Tasmanian Midlands will prove useful, it is important that future monitoring moves beyond measuring the occurrence of species and extends to the survival, growth and reproductive success of individual birds.