The ecology of rivers in the Australian dry tropics

Dryland rivers are globally widespread and regionally important, and understanding their ecology is critical for sustaining ecosystem processes and biodiversity. The dry tropics are characterised by episodic summer rainfall, and rivers have a climate-driven hydrology. Most annual river flow occurs in a short period of time (e.g., 6–16 weeks), after which rivers typically contract into a series of waterholes of varying permanence and hydrological connectivity. The current paradigm of dryland river ecology is the 'boom-and-bust' model, whereby biological activity increases during high summer flows (the boom) and declines during the dry season (the bust). Dryland rivers drain a large proportion of the land area of the globe, yet have received disproportionately little research attention, until very recently. The rivers in Australia's northern dry tropics are among the most unregulated and ecologically intact in the world, thereby providing excellent opportunities to study the ecology of dry tropics rivers. This thesis aims to understand the ecological patterns and processes in dryland rivers that underpin monitoring and management activities.\ud \ud Fifteen study sites were located on three northern unimpounded tributaries of the Burdekin River (Keelbottom Creek, Basalt River, Cape-Campaspe River), and on the Burdekin River itself. Biophysical data and macroinvertebrate samples were collected from in-stream habitats during four seasons to reflect the complete hydrological cycle from 2008-2009. Biophysical variables (riparian condition, water quality, habitat characteristics) were spatially and temporally variable, and demonstrated the importance of seasonal hydrological events as well as geographical position in structuring in-stream habitats. Sites within each river differed substantially, even when connected by perennial baseflow, indicating the importance of local variables other than flow such as riparian condition, in-stream habitat, and land use.\ud \ud Seasonal environmental factors affected macroinvertebrate assemblage composition in dry-tropics rivers at the river, site, and in-stream habitat scales. I assessed biophysical characteristics, including water physicochemistry, riparian-zone condition, and macroinvertebrate assemblage composition at the same 15 sites, this time sampling across five seasons. Biophysical data were spatially and temporally variable, and all rivers and most sites were significantly different. Macroinvertebrate assemblages also differed among rivers, sites within rivers, habitats, and seasons. Assemblages from the same habitat but different sites were more similar than assemblages from different habitats within the same site. Riparian condition directly influenced the ecology of these rivers, e.g., shredder abundance was strongly correlated with the preponderance of canopy cover and leaf litter. In contrast to the boom-and-bust model, I found that macroinvertebrate diversity and abundance did not consistently decline during the dry season (December), with many sites actually experiencing an increase in diversity or abundance at this time. Additionally, peak-and post-flood samples at selected sites in December demonstrated that seasonal flooding can cause dramatic decreases in diversity and abundance. I found that the typical boom-and-bust cycle might be reversed (or partly reversed) in the upper Burdekin catchment, with the bust occurring during times of peak flood flow, and the boom occurring during the interflood stable period. Thus, the dynamic environment of dryland river systems drives naturally dynamic and variable macroinvertebrate assemblages across a range of scales.\ud \ud Seasonal changes in biophysical variables affected the temporal trajectories of in-stream habitats and macroinvertebrate assemblage composition. I described the dynamics of seven in-stream habitats with a particular focus on the three most enduring: sandy pools, edges and rocky runs. Trajectories differed among seasons, sites and habitats, even within the same habitat and/or river. Some assemblages cycled with temperature through the seasons, but some did not; and while temperature (and some other biophysical variables) clearly followed a seasonal cycle, the overall complex of environmental variables did not. Wet-season flooding did not appear to 'reset' assemblages, with post-wet season assemblages differing between years. I found no consistent pattern in species richness, turnover, or evenness with time, and sites within rivers showed no consistent convergence or divergence in macroinvertebrate assemblage composition. Biophysical variables that best explained changes in assemblages were associated with in-stream habitat variability and the rigours of the late dry season, highlighting the harsh and variable conditions of dryland rivers.\ud \ud I also investigated food web structure of 12 sites in the upper Burdekin catchment to determine if the variability in macroinvertebrate assemblage composition extended to the broader food web. Analyses of stable isotopes (δ¹³C and δ¹⁵N) and stomach contents were used to determine consumer trophic level, and a mixing model (SIAR) was used to identify basal sources at each site. Spatial variability afftected isotope values, basal source contribution, trophic position of individual taxa, and food web structure. The effect of spatial scale was also variable, with sites from the same river often as different as sites from different rivers. Results of SIAR for macroinvertebrates and fishes demonstrated that the majority of isotopic mixing models could not be resolved, despite consumer isotopic values being within the source geometry. Nevertheless, strong patterns emerged, especially for the macroinvertebrates: within-site source fidelity was high, with taxa feeding across multiple different basal sources, which was dependent on their site of origin. These results demonstrate that the variability in biophysical data and macroinvertebrate assemblages of these dryland rivers extends to the broader food web.\ud \ud Given that individual sites accommodate subsets of the available species pool, and that each site is expected to have unique combinations of biophysical characteristics, colonisation history, and biotic interactions, it is unsurprising that biotic assemblages differ between sites (and habitats). However, the level of variability, and even the inconsistency of the variability, was surprising, such that spatial and temporal variability were major defining features of these dry-tropics rivers. The differences among sites from the same river are of particular importance from a monitoring standpoint because they indicated that extrapolating whole-river condition from a few sampled sites would be difficult. The dynamic nature of dryland rivers presents major challenges to monitoring programs, and my results suggest a more complex scenario for monitoring and management than previously described