|Much research has been devoted to describing the effects of streamflow regulation on food webs. However, only a few studies have thus far quantified how muting environmental variation may determine integrative ecological measures such as pathways of energy flow. Further, despite the great advances recently made by food web theory, there have still been very few attemps to test how theory may be able to predict food web stability in human-altered ecosystems such as regulated rivers. We liked this paper because it aimed at reconciliating food web theory and empirical research by studying a unique river discontinuum: the 386-km segment of the Colorado River that stretches between Glen Canyon Dam and Lake Mead (Arizona, USA). Moreover, the paper presents a small textbook worth of valuable data and synthesizes many of the classic ideas in stream ecology in one place. The authors described food webs at six locations across a spatial gradient of perturbation (i.e., distance from Glen Canyon Dam), both before and after an experimental flood. They report a comprehensive set of longitudinal patterns in food-web structure, energy flow, and species interaction strengths, with altered food webs being generally simpler and more inefficient. However, perhaps the more compelling result is that, as theory predicts, food webs closest to the dam (i.e., those more altered) were clearly less resistant to flood perturbation than food webs downstream of tributaries. We believe this is an excellent empirical example of how dams, as large-scale modifications of river ecosystems, influence the way that energy enters and flows through food webs.