Determining the factors that generate and maintain patterns of biodiversity is a challenge central to ecology. To effectively manage heterogeneous landscapes for biodiversity conservation there is a need to understand the mechanisms driving the relationship between spatial heterogeneity and species richness at multiple scales. Disturbances such as fire affect landscape heterogeneity and are important influences on the distribution and abundance of biota across a range of scales. A key challenge for biodiversity conservation in fire-prone ecosystems is thus to understand the influence of fire on biotic communities.
Contemporary ecological fire management resounds with the term ‘pyrodiversity’. Efforts to achieve pyrodiversity are usually motivated by the rationale that increasing gamma diversity relies upon maintaining a mosaic of patches of different fire histories. From a global review of literature, this thesis revealed that despite widespread acceptance of this paradigm, the evidence supporting it is extremely sparse. This dearth of knowledge of how heterogeneity affects biota limits effective application of the mosaic burning approach. Studies that consider the spatial and temporal attributes of the fire regime mosaic, and investigate the relationship between these and multiple taxa at different spatial scales, are urgently required.
To respond to this knowledge gap, this thesis investigated the effects of fire history and habitat heterogeneity on structure and composition of floral and avifaunal communities. This thesis tests hypotheses about the importance of variable fire regimes, spatial heterogeneity, extent of habitat, and local- and landscape-responses on birds and their habitat in the fire-prone sub-tropical woodlands of central Queensland. Two methods were used to collect the data: i) digital mapping and characterisation of the fire regime mosaic; and ii) surveys of woody vegetation and avifaunal communities.
The post-fire response of vegetation reflects not only a single fire event but is the result of cumulative effects of previous fires in the landscape. Fire history categories based on unique combinations of time since last burn, fire frequency and season of last burn were developed and used to map the fire regime mosaic of the study area. This thesis establishes that fire history category strongly influences richness and abundance of mid-storey trees and of individual plants currently comprising the mid-level strata. The particular elements of fire history, time since fire and fire frequency, had the strongest influence on vegetation structure. A longer time since fire (> 4 years since last burn) combined with infrequent fires (< 2 fires in 12 year period) appeared to promote a dense mid-storey with the opposite conditions promoting more-open woodlands. This is of particular relevance for conservation management of woodland birds that have been shown to respond strongly to habitat structural variables in the study area.
Landscape heterogeneity and extent of habitat are each expected to contribute to species richness at intermediate- or landscape-scales (~1-100s km2) through elevating beta and mean alpha diversity, respectively. Understanding of their relative importance, however, is limited. I investigated the relative importance of diversity of habitat types (both natural and fire-mediated vegetation types), extent of habitat and indices of landscape configuration on landscape-scale (1 km2) bird richness. At this scale, species richness was most strongly associated (positively) with measures of landscape heterogeneity, in particular topographic complexity and diversity of vegetation types. Despite being relevant to management, fire-mediated diversity was of limited importance at this scale. Total estimated richness (Chao2) and richness of small passerines, a group of conservation concern, were also negatively associated with the mean abundance of noisy miners (Manorina melanocephala), a hyper-aggressive native species. Despite the relationship between heterogeneity and species richness, mean alpha diversity made a greater contribution to landscape-scale bird species richness than did beta diversity. These findings support environmental heterogeneity as a primary driver of species richness at the landscape-scale, and imply that it acts through increasing both turnover and mean alpha diversity.
Taxa respond to their environment at a variety of scales. An increasing number of studies have considered the impact of heterogeneity measured across spatial scales. Few studies, however have considered how the response of fauna to landscape elements varies when the response variable was measured at different scales. I investigated the relationship between properties of the fire regime mosaic, and the structure and composition of avian assemblages measured at both the site- (1 ha) and the landscape-scale (1 km2). I found topographic complexity was consistently important for bird species richness and composition; however, its explanatory power varied with spatial scale and the component of diversity under investigation. For foraging guilds different correlates of richness became evident depending on the scale at which the response was measured. The extent of unburnt habitat was the most important variable for the landscape-level richness of frugivores, insectivores and canopy feeders, whereas environmental heterogeneity was most important for site-level richness of these foraging guilds. These findings suggest that depending on the scale at which a biodiversity conservation goal is conceptualised – maximising richness at a site, or across a landscape – different landscape management approaches would be considered optimal.
The thesis outcomes have important implications for our understanding of the drivers of patterns of biodiversity and for conservation management in fire-prone landscapes. The main findings are: (i) a variable fire regime will maintain a heterogeneous vegetation mosaic through its influence on species in the mid-level strata in sub-tropical woodlands; (ii) environmental heterogeneity is an important driver of landscape-scale bird richness and acts through increasing both turnover and mean alpha diversity; (iii) the extent of unburnt habitat is also important, particularly influencing species composition and assemblage structure at the landscape scale; (iv) higher fire-mediated diversity of vegetation does not increase species richness at spatial scales of 1 km2; (v) landscapes with a higher abundance of noisy miners had fewer bird species, particularly small passerines; and (vi) landscape elements driving bird assemblages vary depending on the scale at which the response is measured and the foraging guild considered.