The extent of human-induced change and damage to riverine habitats has meant that interest and investment in river restoration projects are growing throughout the world, with a concomitant increase in the publication of material on the theories, principles and practices of restoring degraded landscapes (Bernhardt et al. 2005, Lake 2005). Despite the growing consensus among restoration ecologists on the need for an improved scientific basis for restoration projects, there is little afreement among the restoration community on what characterises success and how it is measured (Ryder and Miller 2005). At present, river restoration projects often aim to restore systems to a predetermined state that resembles a reference condition or, less often, towards a more realistic state that represents the most desirable outcome, given the available resources (Palmer et al. 2005). In either process, goals must be stated up front and standards established, against which the success of restoration can be assessed. To do this, a suite of indicators is typically used to assess the 'health' of the river (Boulton 1999) and to focus restoration efforts towards achieving the best possible ecological outcome. Indicators of ecosystem structure are commonly used to monitor restoration success. Their selection is usually based on an understanding of the biology (e.g. genetic structure, abundance, richness and diversity) and habitat requirements - such as geomorphic setting or hydrologic requirements - of an organism or population.