Influence of drainage divides versus arid corridors on genetic structure and demography of a widespread freshwater turtle,Emydura macquarii krefftii,from Australia

The influence of Pleistocene climatic cycles on Southern Hemisphere biotas is not yet well understood. Australia's eastern coastal margin provides an ideal setting for examining the relative influence of landscape development, sea level fluctuation, and cyclic climatic aridity on the evolution of freshwater biodiversity. We examined the impact of climatic oscillations and physical biogeographic barriers on the evolutionary history of the wide‐ranging Krefft's river turtle (Emydura macquarii krefftii), using range‐wide sampling (649 individuals representing 18 locations across 11 drainages) and analysis of mitochondrial sequences (~1.3‐kb control region and ND4) and nuclear microsatellites (12 polymorphic loci). A range of phylogeographic (haplotype networks, molecular dating), demographic (neutrality tests, mismatch distributions), and population genetic analyses (pairwise F_ST, analysis of molecular variance, Bayesian clustering analysis) were implemented to differentiate between competing demographic (local persistence vs. range expansion) and biogeographic (arid corridor vs. drainage divide) scenarios. Genetic data reveal population genetic structure in Krefft's river turtles primarily reflects isolation across drainage divides. Striking north‐south regional divergence (2.2% ND4 p‐distance; c. 4.73 Ma, 95% higher posterior density (HPD) 2.08–8.16 Ma) was consistent with long‐term isolation across a major drainage divide, not an adjacent arid corridor. Ancient divergence among regional lineages implies persistence of northern Krefft's populations despite the recurrent phases of severe local aridity, but with very low contemporary genetic diversity. Stable demography and high levels of genetic diversity are inferred for southern populations, where aridity was less extreme. Range‐wide genetic structure in Krefft's river turtles reflects contemporary and historical drainage architecture, although regional differences in the extent of Plio–Pleistocene climatic aridity may be reflected in current levels of genetic diversity.     

Dynamics of habitats and macroinvertebrate assemblages in rivers of the Australian dry tropics

1. The dry tropics are characterised by episodic summer rainfall such that the majority of annual river flow occurs in a short period of time. This dryland hydrological cycle leads to variably connected channels and waterholes along the length of a river bed. 2. We investigated the seasonal changes in biophysical characteristics and macroinvertebrate assemblage composition in dry‐tropics rivers at 15 sites on four rivers, each sampled five times (representing one annual hydrological cycle), in the Burdekin catchment, north Queensland, Australia. 3. Assemblages and their temporal trajectories differed among seasons, sites and habitats, even within the same habitat and/or river. Wet season flooding did not appear to ‘reset’ assemblages, with post‐wet season assemblages differing between years. 4. We found no consistent pattern in taxonomic richness over time, and sites within rivers showed no consistent convergence or divergence (i.e. turnover) in macroinvertebrate assemblage composition. However, biophysical variables associated with the rigours of the late dry season had significant effects on macroinvertebrate assemblages, highlighting the variable and often harsh conditions of dry‐tropics rivers. Underlying these patterns were different resistance and resilience traits of invertebrates (such as colonisation and establishment abilities), as well as the local‐scale effects of biophysical variables. 5. The dynamic nature of dryland rivers presents major challenges to monitoring programmes, and our results suggest a more complex scenario for monitoring and management than previously described.     

Widespread omnivory and low temporal and spatial variation in the diet of fishes in a hydrologically variable northern Australian river

This article examines the trophic ecology of freshwater fishes (22 species in 15 families) in a wet and dry tropical Australian river of high intra‐annual and interannual hydrological variability. Seven major trophic groups were identified by cluster analysis; however, four food items (filamentous algae, chironomid larvae, Trichoptera larvae and Ephemeroptera nymphs) comprised almost half of the average diet of all species. The influence of species, fish size, spatial effects and temporal effects on food use was investigated using redundancy analysis. Size, time and space accounted for little of the perceived variation. Ontogenetic changes in diet were minor and limited to a few large species. Spatial variation in trophic composition of the fish assemblages reflected the effects of the Burdekin Falls and dam, a major geographic barrier, on species distributions. Little spatial variation in diet was detected after accounting for this biogeographical effect. Temporal variations in flow, although marked, had little effect on variations in fish diet composition due to the low temporal diversity of food resources in physically monotonous sand and gravel channels. Species identity accounted for<50% of the observed variation in food choice; omnivory and generalism were pronounced. The aquatic food web of the Burdekin River appears simple, supported largely by autochthonous production (filamentous and benthic microalgae, and to some extent, aquatic macrophytes). Allochthonous food resources appear to be unimportant. The generalist feeding strategies, widespread omnivory and absence of pronounced trophic segregation reported here for Burdekin River fishes may be common to variable and intermittent rivers of subtropical and tropical northern Australia with similar fish communities and may be a general feature of rivers of low habitat diversity and characterized by flow regimes that vary greatly both within and between years.     

Development of a stable isotope index to assess decadal-scale vegetation change and application to woodlands of the Burdekin catchment, Australia

Forty-four study sites were established in remnant woodland in the Burdekin River catchment in tropical north-east Queensland, Australia, to assess recent (decadal) vegetation change. The aim of this study was further to evaluate whether wide-scale vegetation ‘thickening’ (proliferation of woody plants in formerly more open woodlands) had occurred during the last century, coinciding with significant changes in land management. Soil samples from several depth intervals were size separated into different soil organic carbon (SOC) fractions, which differed from one another by chemical composition and turnover times. Tropical (C₄) grasses dominate in the Burdekin catchment, and thus δ¹³C analyses of SOC fractions with different turnover times can be used to assess whether the relative proportion of trees (C₃) and grasses (C₄) had changed over time. However, a method was required to permit standardized assessment of the δ¹³C data for the individual sites within the 13 Mha catchment, which varied in soil and vegetation characteristics. Thus, an index was developed using data from three detailed study sites and global literature to standardize individual isotopic data from different soil depths and SOC fractions to reflect only the changed proportion of trees (C₃) to grasses (C₄) over decadal timescales. When applied to the 44 individual sites distributed throughout the Burdekin catchment, 64% of the sites were shown to have experienced decadal vegetation thickening, while 29% had remained stable and the remaining 7% had thinned. Thus, the development of this index enabled regional scale assessment and comparison of decadal vegetation patterns without having to rely on prior knowledge of vegetation changes or aerial photography.