More for less: a study of environmental flows during drought in two Australian rivers

1. In rivers affected by drought, flow regulation can further reduce flow and intensify its effects. We measured ecological responses to environmental flows, during a prolonged drought in a regulated river (Cotter River), compared with a drought affected, unregulated river (Goodradigbee River) in south‐eastern Australia. 2. Environmental flows in the regulated Cotter River were reduced from a monthly average base flow of 15 MLd^?1 to only 5 MLd^?1, which was implemented as two test flow regimes. Initially, flows were delivered in cycles of 14 days at 3 MLd^?1 followed by 3 days at 14 MLd^?1 and then another 14 days at 3 MLd^?1 to make up the monthly average of 5 MLd^?1. This flow regime continued for 6 months, after which a preliminary ecological assessment indicated deterioration in river condition. Consequently, the flow regime was altered to a cycle of 2 MLd^?1 for 28 days followed by 20 MLd^?1 for either 3 or 4 days. This new flow regime continued for another 5 months. 3. The ecological outcomes of the test flow regimes were assessed in terms of (i) the provision of available habitat (wetted channel) for aquatic biota; (ii) the accumulation of periphyton; and (iii) the structure and richness of macroinvertebrate assemblages. 4. Flow of 20 MLd^?1 covered most of the streambed in the Cotter River, thus providing more wetted area and connectivity between habitats than flows of 2, 3 or 14 MLd^?1. Depth and velocity were always less in the Cotter River than in the unregulated Goodradigbee River. Periphyton decreased in the Cotter River during the 2/20 MLd^?1 flow regime, which combined the lowest and greatest test flow volumes, while periphyton did not change significantly in the unregulated river. 5. The reduced flow in the Cotter River resulted in fewer macroinvertebrates than expected (13) compared with unregulated Goodradigbee sites (19), although the magnitude of the differences did not depend on the test flow releases. Macroinvertebrates in the Cotter River became numerically dominated by Diptera and Oligochaeta, while Ephemeroptera, Plecoptera and Trichoptera decreased in abundance. 6. In the Cotter River, the monthly average flow of 5 MLd^?1 (exceeded 97% of the time pre‐regulation) was insufficient to maintain the macroinvertebrate assemblages in reference condition, regardless of release patterns. However, short‐term ecological objectives were achieved, such as reduced periphyton accumulation and increased habitat availability, and the environmental flows maintained the river’s ability to recover (resilience) when higher flows returned.