Field ecology has generally focussed on the quantification of patterns in the distribution of biota and attempted to explain these by comparison with variations in habitat conditions. This approach generates hypotheses about causal mechanisms which can be tested experimentally. Such approaches have been used in stream ecology and these have highlighted the role that natural physical disturbance may have in determining patterns in the distribution and abundance of taxa among streams of different regions. In recent years, there has been a great increase in effort to understand factors controlling the distribution and abundance of stream periphyton communities. Awareness has grown of the importance of these communities as food for invertebrates (and thus as a fundamental energy base for streams), as excellent indicators of environmental change, and on a less positive note as causes of degradation in water quality and aesthetic values. The importance of flood disturbance as a determinant of periphyton development has been clearly shown in many early studies (see Chapter 1). In-fact from first principles alone it can easily be argued that streams with a high frequency of flood disturbances are unlikely to suffer from enrichment and nor are their communities going to be subject to heavy invertebrate grazing control. Conversely, nutrients and grazers are likely to become increasingly important with time since the last disturbance event and thus require long interdisturbance stability to be a significant determinant of community development. It is highly possible that the disturbance regime is the most important determinant of community development in streams of the temperate region. However, much of the research undertaken between 1989 and 1995 has focussed on grazing and nutrient enrichment processes in hydrologically stable, continental, streams. Only a limited number of studies have extended our understandings of the importance of disturbance as a major factor controlling periphyton community structure, biomass and production in streams. The goal of the following set of studies was therefore to more clearly define the role of hydraulic disturbance in determining the development of periphyton in streams, and to provide a physical basis upon which disturbance intensity and frequency could be defined. Using this approach I hoped to obtain a much clearer understanding, and ultimately to produce a predictive model, to explain why periphyton communities of streams in one region can be so different from those in another. Following a general review of patterns in periphyton development in natural streams, and an elucidation of the potential importance of hydraulic disturbance in determining these patterns (Chapter 1), I describe a broadscale study of 16 New Zealand streams which investigates the importance of differing disturbance frequencies in controlling average periphyton biomass over a whole year. The contribution, and interactive effects, of nutrient resource supply in determining this pattern are also investigated (Chapter 2). In the third chapter, I describe an experiment designed to determine what shear stresses are required during flood events to dislodge different types of periphyton communities, and thus also determine removal kinetics and resistance properties. In Chapters 4 and 5, I investigate community redevelopment as functions of spatial variation in hydraulic conditions in a stream. Hypotheses are tested regarding differential colonisation, growth and sloughing dynamics in relation to time since a disturbance, and the interaction of these dynamics with hydraulic conditions on the falling stage of the hydrograph. In Chapter 6, I attempt to establish a conceptual basis for understanding the effects of disturbance on periphyton community development and to explain how this interacts with limiting resource supply and invertebrate grazers as major components of the habitat matrix of stream periphyton. This conceptual model is tested using both field (Chapter 7) and experimental (Chapter 8) studies. The field study involved monitoring periphyton development at multiple locations which varied in disturbance and nutrient supply regimes within a single catchment. Average development over a two year period was then compared with predictions made under the disturbance - resource supply - grazer conceptual model. In the experiment, periphyton communities were grown under varying degrees of light and nutrient resource stress, and the effects of a single, simulated disturbance measured in terms of community resistance, resilience and mean biomass. These results are compared with the predictions of the conceptual model. The chapters in this thesis are presented in the form of manuscripts that have been, or will be, submitted to various scientific journals. Inevitably, this has resulted in some repetition of material (particularly review material in the Introductions), and some differences in format including the citation of references. Nevertheless, I have attempted to standardise the layout of chapters as much as possible, and have provided some continuity to the thesis by including linking sections in the form of short prefaces before each chapter.