Placing global stream flow variability in geographic and geomorphic contexts

The importance of hydrologic variability in sustaining natural riverine ecosystems is now well accepted. Over the last 15 years or so, many typologies and assessment tools have been developed to assist ecologists and managers in describing natural flow regimes in quantitative terms. In the course of this recent progress, however, some critical questions have arisen concerning the degree to which generalizations about flow regime characteristics are geographically dependent both within and among regions, and the degree to which flow variability alone captures critical environmental variability. In this paper we address these issues in a hierarchical framework that allows comparative statements about hydrologic variability to be made a multiple spatial scales, from local to global. First, we examined hydrologic variability among 463 readily available daily streamflow gauges from five continents/ countries around the world: Australia, New Zealand, South Africa, Europe, and the United States. Using ordination and clustering techniques, we identified similarities and differences among these gauges. We found that the US gauges exhibited the greatest overall flow variability among a suite of 66 hydrologic indicators, whereas Australian streams showed the greatest influence by interannual variability in flow. Similarities in overall flow regime were greatest between Australia and the US, whereas New Zealand streams were most regionally distinctive. These results support the idea of intercontinental distinction in streamflow variability at a global scale; however, they also point to important similarities in flow characteristics among continents/countries. Second, within the continental United States, we examined how hydrologic variability changes along river profiles as catchment area increases for five river basins arrayed across a gradient of hydroclimatic variation. Using historical streamflow records that precede river impoundment, we found that small 'headwater' streams exhibit the greatest similarity in flow characteristics across the basins, as compared to mid-sized and larger river reaches, which often diverged among the rivers. These results reveal the importance of more carefully defining the spatial domain of allowable hydrologic extrapolation from individual stream gauges and emphasize the need to stratify within basins when considering hydrologic variability at regional scales. Third, we used a modeling approach to illustrate how geomorphic setting provides a context for assessing the ecological consequences of flow variation at the local scale of stream reaches. For modeled channels having the same sediment size distribution but with either entrenched or floodplain morphology, we found that the effective regime of bed movement for three hydrologically distinct streams depended as much on geomorphic setting as on flow regime per se. These results emphasize the need to integrate hydrology with geomorphology to characterize 'disturbance regimes' at the channel reach scale to allow generation of spatially explicit mapping of flow-mediated habitat dynamics for entire drainage networks within specific regions. In sum, if riverine scientists wish to develop a general framework for comparing hydrologic variability across basins, regions, and continents, a hierarchical approach is advised. At very broad scales, intercontinental differences in flow regimes could allow a stratification of basins to identify similar hydroecological settings. Within continents or hydroclimatically similar regions, finer-scale spatial analysis of flow regime types would further assist in hydrologic stratification, based only on the regionally-relevant components of flow variability. Finally, within hydrologically homogeneous sub-regions, geomorphic stratification could be applied to identify stream reaches or segments having similar hydrogeomorphic properties. Copyright (C) 2006 John Wiley & Sons, Ltd.