An analysis of artificial riffles and enhancement of benthic community diversity by physical habitat simulation (PHABSIM) and direct observation

Critical to river rehabilitation decisions is the prediction of the benefits of certain procedures. In low-order systems, planning should focus on flow requirements of lotic organisms, especially benthic species. We examined the value of placement of artificial riffles in Holly Fork, a low-order tributary of the West Sandy River (west Tennessee). The objective was to determine if the instream flow incremental methodology (IFIM) and its component computer model, PHABSIM, could adequately predict the habitat value of the riffles for benthic macroinvertebrates and if, after colonization, this ‘value’ was reflected by increased diversity of benthic macroinvertebrates. Holly Fork is a severely head-cut channel with 2 m high vertical embankments and a substrate dominated by sand and fines, with occasional gravel riffles. We chose to conduct the analysis of each stream reach using macroinvertebrate diversity as the ‘target’. Habitat suitability criteria, developed using data from adjacent watersheds, predicted a range of suitable physical conditions that supported highest benthic community diversity. A traditional IFIM analysis of a reach of the Holly Fork indicated that, below 0.2 m3/s (a flow exceeded 10% of the time), less than 5% of the wetted area contained adequate habitat for benthic macroinvertebrates. At optimum flows (0.4 m3/s and higher), only 15% of the wetted area, primarily across small gravel bars, was adequate to support high community diversity. Two artificial riffles, composed of large cobble and boulder keystones, with leading and trailing aprons of medium cobble and gravel, were placed at 35 m intervals in the test section. After hydrological stabilization and time for colonization by macroinvertebrates, the artificial riffles were re-analysed. The simulation predicted that this reach contained significantly higher amounts of available benthic habitat at low flows (more than tripled), and over 40% of the total wetted area should support high benthic community diversity at optimal flows. The presence of artificial riffles contributed most of this habitat enhancement. A plot of cell-by-cell composite habitat suitability and sample diversity from these cells revealed a significant correlation between PHABSIM predictions and actual community diversity. This technique can be an aid in demonstrating the value of certain restoration structures during the rehabilitation planning process. Our data suggest that benthic community diversity is an appropriate target for evaluation of instream flow values that sustain ecosystem integrity.