Flow-regime management: a paradigm for urban stream protection and restoration

Abstract

Healthy streams are important. They support a range of biota and ecosystem services. Yet, many urban streams are degraded; primarily by stormwater runoff, delivered through conventional stormwater drainage systems. Standard approaches to stormwater management for environmental protection do not address the ecologically important changes to hydrology caused by stormwater drainage. My thesis proposed and explored an alternative approach to stormwater management—flow-regime management—which emphasizes the protection or restoration of hydrologic (and water quality) processes at small scales with the aim of returning catchment-scale flow regimes towards their natural condition. Building on this concept, the thesis assessed the degree to which SIGNAL (an aquatic macroinvertebrate assemblage composition index) is predicted by hydrologic indicators and a landscape-scale indicator of stormwater runoff (attenuated imperviousness [AI]). SIGNAL was best predicted by AI, while a model that included hydrologic indicators characterising high- and low-flow hydrology was only marginally less plausible. It was postulated that AI integrates all aspects of hydrologic alteration and other stressors, such as reduced in-stream water quality. The analyses confirmed the importance of stormwater runoff conveyed by stormwater drainage systems as a primary source of degradation to receiving waters. The thesis developed and evaluated a framework for setting stormwater management objectives, covering three components: (i) the proportion of rain falling on impervious areas that should be lost (evapotranspired and/or harvested), (ii) the proportion that should be infiltrated, and the (iii) equivalent initial loss, which characterizes the probability of surface runoff from a rain-event, with the aim of restoring pre-development levels of runoff retention within the catchment. Numeric targets for each objective were calculated for Melbourne, Australia as a case study. At land-parcel (site) and streetscape scales, empirical data and modelling were used to investigate the hydrologic performance of stormwater management strategies (rainwater tanks and rain-gardens) against the proposed objectives. Twelve tanks at the land-parcel (site) scale were monitored. Of these, only three achieved an equivalent initial loss that approached the proposed target—a consequence of limited demand and small tank capacity. The work demonstrated the important synergy between the water supply substitution benefits of rainwater tanks and their stormwater retention performance. The performance of various design configurations of rainwater tanks and rain-gardens was then modeled. Configurations where tanks – connected to a large range of internal uses – overflowed to rain-gardens were often able to achieve all three hydrologic restoration targets. The hydrologic performance of a rain-garden at the streetscape scale was also monitored and found not to be ideal, producing frequently untreated overflows. The rain-garden performed poorly because its area was only ~1% of its upstream impervious catchment. Modelling a larger rain-garden and domestic rainwater tanks in the streetscape still resulted in a water balance dissimilar to pre-development conditions. The results confirmed the importance of finding a means of losing much of the excess volume generated by impervious areas. Restoring flow regimes at small scales is thus primarily limited by demand for stormwater. However, at the household scale the amount of harvestable impervious runoff is similar to the demands which could be provided by stormwater, meaning that with the right combination of uses, adequate volume reduction should be possible. Finding demands at the streetscape scale is more challenging, particularly given the lower water quality and limited space for storage. Modelling was also undertaken to examine if stormwater management strategies applied at small scales could return more natural catchment scale flow regimes. The scenarios which treated only allotment scale impervious roof areas generally performed poorly, with a strategy addressing all catchment imperviousness being shown to be required. My thesis has demonstrated both the potential and challenges of incorporating flow-regime management into the control of urban stormwater. Such an approach could readily be applied in new urban areas and would likely protect receiving waters. However, research to confirm this hypothesis is urgently needed. Flow regime management could also be applied to degraded urban streams, albeit with comparatively more challenges. These challenges are however, not insurmountable, and it is clearly important to persevere given the values provided by urban streams.