STREAM COMMUNITY STRUCTURE IN RELATION TO SPATIAL AND TEMPORAL VARIATION - A HABITAT TEMPLET STUDY OF 2 CONTRASTING NEW-ZEALAND STREAMS

1. The physical characteristics of two contrasting streams, and habita t types within these streams, are described in terms of a two-dimensio nal physical habitat templet in which disturbance frequency and the av ailability of spatial refugia are the temporal and spatial axes. 2. It is predicted that habitats experiencing a high disturbance frequency and low refuge availability will be characterized by a low invertebrat e species diversity, a low biomass of epilithic algae and particulate organic matter and a community made up of mobile, weedy species. Habit ats having a low disturbance frequency and high refuge availability wi ll be characterized by a diverse community containing sedentary and sp ecialist species, with high algal and particulate organic matter level s. 3. A lower median substrate particle size and higher shear stress r egime in Timber Creek were indicative of a higher disturbance frequenc y than in the Kyeburn. Substrate diversity was lower in Timber Creek t han in the Kyeburn and indicated that the availability of refugia was lower in Timber Creek. In both streams, pools were found to have a hig her disturbance frequency and lower availability of refugia than riffl es. 4. Invertebrate species diversity, the biomass of epilithic algae and particulate organic matter and the representation of sedentary spe cies, filter feeders and shredders were higher in the more temporally stable and spatially heterogeneous Kyeburn. The community of Timber Cr eek, frequently disturbed and having low refuge availability, had a hi gh proportion of mobile and weedy species, with the highly mobile, gen eralist-feeding Deleatidium spp. (Ephemeroptera; Leptophlebiidae) bein g the most dominant organisms. 5. The predictions made about stream co mmunity structure and species characteristics in relation to disturban ce frequency and the availability of spatial refugia are generally sup ported. Now a larger scale investigation is required to test the gener ality of the predictions. We conclude that the habitat templet approac h offers a sound framework within which to pose questions in stream ec ology.