Understanding scale-dependent processes linking environmental variability to the spatial structure of communities is a major issue in ecology. One goal of such understanding is to better predict and understand regional biotic pattern formation from local interactions between organisms and their environment. I adopted a multidisciplinary approach to elucidate processes linking scales of topographic heterogeneity to the spatial structure of an intertidal benthic community : (1) Controlled observations around natural boulders allowed to highlight and quantify the scale-dependent cascade linking topographic heterogeneity, hydrodynamics and benthic communities. My results showed that local infuence of hydrodynamics on the abundance of dominant species (mussels and Fucus macroalgae) depends on topographic scale (boulder size from 0.5 to 2.5 m diameter) generating hydrodynamic patterns. I thus highlighted and quantifed a scale-dependent cascade of events linking environmental scales to the structure of benthic communities (i.e. topographic heterogeneity -> hydrodynamics -> benthic communities). (2) Experiments using artifcial cylinder reefs of different sizes were carried out to test processes (growth, aggregation) responsible for observed patterns, and to determine the relative importance of hydrodynamic and shading patterns as mediating physical factors determining alternative scaling rules linking topographic complexity to benthic communities. (3) In order to generalize our experimental results to a natural landscape, I designed a low-cost blimp-based remote-sensing device allowing spatiotemporal data acquisition of physical (topography, hydrodynamics) and biological (density and biomass) intertidal variables from the mm to the km scale. I present a photogrammetric analysis of overlapping images allowing topographic restitution at a resolution of 10 cm. Calibration of red-infrared photographs allowed mapping of macroalgal biomass over topography. Multiscale analysis of topographic and algal biomass data further reveals the value of this method by allowing to generalize previous local experimental results to the landscape level. (4) A spatially-explicit, individual based model was developed and coupled to a finite-element hydrodynamic model. Using this model, I explore the importance of spatial scales of topographic complexity and their infuence on local biological processes for the regional dynamics of intertidal benthic communities. I also explore the influence of large scale external forcing (climate, hydrodynamic) on local and regional patterning of the community.