The response of invertebrates to a gradient of flow reduction – an instream channel study in a New Zealand lowland river

Abstract

Summary1. The impacts of low flows on invertebrates were assessed in six experimental channels placed in a lowland stream. Each channel consisted of replicate stacked gravel‐filled baskets (buried 0–7, 8–14 and 15–21 cm deep) from which invertebrates were sampled. Samples were collected before, and 1 and 2 months after flows were reduced by varying amounts. We predicted that invertebrate communities would change most with the lowest flows that persisted longest. The effects of reduced flow on channel and intragravel‐velocities, temperature, dissolved oxygen and algal biomass were also monitored.2. Mean velocity (17 cm s−1) and depth (20 cm) were similar in all channels prior to flow reduction, but were reduced in five channels by 25% to 98%, a sixth channel serving as a control. Mean intragavel velocity prior to flow reduction (1.7 mm s−1) was also reduced from 17% to 51%. Flow reduction had no effect on temperature, which ranged from 11.2 °C (night) to 19.6 °C (day). Dissolved oxygen was generally high (mean = 94% saturation) but variable (range from 50% to 145%), and did not differ between channels after flow reduction. Periphyton biomass (chlorophyll a and AFDMalgae) increased over time and was positively related to velocity after 1 month of flow reduction. After 2 months, however, this relationship reversed, with higher chlorophyll and AFDMalgae in lower velocity channels. Organic matter biomass (AFDMtotal) was significantly higher in channels with lower velocities, and did not change with time. There was no relationship between AFDMtotal and velocity in baskets from different depths.3. Greater than 85% of animals were found in the upper baskets (0–7 cm deep), and flow reduction had no influence on the vertical distribution of invertebrates. After 1 month, invertebrate density had declined roughly in proportion to the magnitude of flow reduction in all channels where flows were reduced more than 25%. Densities recovered to pre‐reduction levels within 2 months. The amphipod, Paracalliope fluviatilis dominated all channels numerically prior to flow reduction, but its density declined markedly during the study. Densities of Ostracoda, Oxyethira albiceps and Cladocera increased dramatically after 2 months of flow reduction, especially in the lowest flow channels. Of the physical variables measured, chlorophyll a biomass and discharge best explained the temporal changes in the invertebrate community.4. The effects of increased duration and magnitude of flow reduction on invertebrate communities were restricted to changes in the relative abundances of just a few taxa. Our results suggest that invertebrates in this lowland stream were resistant to the experimental flow reduction, presumably because of their broad ecological tolerances. We also found that more prolonged low flows did not result in predictable changes. This finding may have implications in terms of using hydraulic‐habitat models to set minimum flows in lowland streams if invertebrates can persist at flows much lower than their supposed optima.