Scientists have studied the structure and dynamics of food webs and other ecological networks using both empiricism and theory since the 1970s. On the empirical side, much effort has been invested in compiling and analyzing ecological networks for particular habitats, resulting in a set of stand-alone networks. By comparing the structural properties of those networks across ecosystems, researchers have identified generalizable network properties associated with ecosystem stability and functioning. Yet, potential changes in network properties within ecosystems have mainly been predicted from simulations, rather than experiments that test causation. Recently, another type of ecological network data and analysis has begun to emerge. The researchers in this session are compiling databases consisting of replicate networks for particular systems across various kinds of gradients. These “gradient-based ecological networks” are at the heart of a powerful, flexible, and enriched research agenda that promises to provide compelling new insights into the causes and consequences of interaction complexity in relation to natural and anthropogenic change.