Oral Presentation Australian Freshwater Sciences Society Conference 2024

What happens when the flow stops? Fragmentation on the Barwon-Darling River, Australia (112372)

Mark Southwell 1 , Paul Frazier 1 , Adrian Matheson 2 , David Ryan 2 , Niva Verma 1 , Tamara Kermode 1
  1. 2rog Consulting, Armidale, NSW, Australia
  2. NSW Department of Climate Change, the Environment, Energy and Water, Surface Water Science, Newcastle, NSW, Australia

The Barwon-Darling River is located in the Murray-Darling Basin in Southeast Australia. It drains a flat semi-arid landscape that is geomorphically complex, with a highly incised sinuous main channel, anabranch channels, deep pools, benches and bars representing the dominant in-channel habitats. Flows in the Barwon-Darling River are highly variable, and during droughts the river can stop flowing for extended periods (>500 days in extreme cases). During these periods, the river fragments into a series of isolated pools, with water quality deteriorating over time, placing stress on aquatic biota and town water supply. The current project mapped the presence of water along the channel over several drying events and attempted to link these drying patterns to ecological consequences of poor water quality and fish kills.

We used satellite imagery (Sentinel-2 and Landsat TM/ETM) and common water detection techniques to map water in the landscape during drying phases. This was problematic along some reaches due to pools being obscured by vegetation, shading or variations in water colour/clarity which restricted the further analysis of around 44% of images. Based on usable imagery, the distribution of residual pools varied down the length of the river, being smaller in size and evenly spaced in the upper reaches, larger and more contiguous in the mid reaches and smaller but isolated in the lower reaches. Drying patterns over time varied between reaches but were generally consistent between no-flow periods. In some reaches, distinct breakpoints could be identified for some metrics. These breakpoints ranged from 20 days since cease-to-flow in the upper reaches to 100 days in the mid reaches of the river. Unfortunately, available water quality and fish kill data were too sparse to allow a robust comparison with drying.