The representation of ecological communities as networks of interacting species is currently a particularly widespread approach in community ecology. The prevalence of networks has led to a broad literature that sheds light on the links between ecological network structure and stability, the explanatory role of traits and evolutionary history with regard to interaction patterns, and the persistence and dynamics of networks across time and space. In addition to the long-established study of trophic and competitive interactions, the importance of networks of positive interactions to ecological communities is also becoming increasing apparent. Despite these many advances, it is often sidestepped that each interaction network---be it a food-web, pollination network, or host-parasite network---is but one component in a greater, connected whole. That is, each and every ecological community is actually composed of multiple interaction networks that do not operate independent from one another. Indeed, predators and pollinators can undeniably influence the population dynamics of both seed-plant and seed-disperser species just as parasites may reduce the energetic budget of those same species. Perhaps most importantly, the stability of ecological communities and the ability of perturbations to cascade through those communities is likely determined by their integrated structure rather than one of their constituent layers. While the integration of differing interaction types is not without difficulties, movement towards a `whole-community' approach to ecological networks is beginning to emerge. Building off recent developments in social, technological, and neural networks and their analysis using tools from graph theory and network science, the study of these whole-community or multiplex ecological networks may well represent the next major advance in community ecology.