A critical challenge at the interface between ecology and the study of geomorphology lies in understanding how the dynamics of geomorphological and ecological processes interact to influence emergent phenomenon such as the characteristics of food webs and the maintenance of biodiversity.  Theory predicts that ecological stability (e.g., maintenance of a suite of species in a community over time) can arise from the dynamic geomorphological processes that create and maintain habitat heterogeneity.  In addition to classic niche-based explanations for this, such stability is thought to be supported by dimensions of foodweb complexity (e.g., variability in interaction strengths) as well as a broader set of “portfolio effects” that may be generated in the context of geomorphological patchiness. The study of streams and rivers and their longitudinal and lateral connections has been a natural context within which such phenomena have been investigated for decades, and these observations can be used to confront theoretical expectations.  Here I synthesize the findings from a suite of studies conducted with students and collaborators over the past 20 years that provide empirical evidence for relationships between geomorphological dynamics, meta food web characteristics and their dynamism, and the maintenance of biodiversity in associated metacommunities.  These studies have occurred in the context of both river network complexity and the shifting mosaics of river-floodplains.  For example, a combination of spatially extensive surveys and >35-years of macroinvertebrate community monitoring suggests that habitat heterogeneity occurring in domains within networks that are punctuated by multiple tributary confluences in close proximity sets the stage for both food web traits and metacommunity dynamics that appear to sustain beta and gamma diversity of these organisms.  Similarly, a sequence of studies of a river-floodplain whose mosaic of habitats shifts due to its natural flow regime showed that variability in food webs of local habitat patches combines to yield meta-food web traits (e.g., interactions strength distributions) that may be ecologically stabilizing for both populations of conservation concern (e.g., endangered salmon) and the diversity of broader communities.  I draw on such illustrations to provide some empirical critique of ecological theory, but also raise new hypotheses and suggest future approaches for study at the intersection between ecology and geomorphology.