Ecologists have long acknowledged the complexity of real communities (even the difficulty of defining them), but historically sought to build understanding through work on simplified or experimentally tractable sub-systems. A more recent trend across sub-disciplines in ecology has been to embrace, and often focus on, the very complexity that might in the past have been considered noise around some theoretical ideal. Not surprisingly, this has resulted in fervent interest in such “messy factors” as the movements of materials and organisms. Some of the best empirical examples come from past and recent investigations of land-water linkages (e.g., leaf and wood inputs, reciprocal fluxes of invertebrates) and the importance of consumer movement (e.g., fish migrations) among aquatic habitats. I review how such movement links what have been considered discrete, local communities into a broader spatial fabric, now termed a “metacommunity.” Yet, temporal complexity in community membership and species interactions created by movement has not been addressed. Moreover, new concepts should not simply point out complexity, but offer new general predictions and testable hypotheses. From riverine examples, I describe a simple, “community-as-kaleidoscope” analogy that may mesh with the spatial metacommunity concept to provide the basis for a set of theoretical generalizations.