Background/Question/Methods A rich body of theory has been developed regarding the manner by which intraguild predation (IGP, and trophic omnivory) can affect the structure and dynamics of ecological communities. The behavior of IGP systems is of interest not only because IGP is a common feature of natural food webs, but because IGP combines the processes of both predation and competition which can produce system outcomes different to those exhibited by systems consisting of only one of these processes. A key prediction of simple IGP models is that increases in basal-prey productivity lead to the decreased abundance and eventual extinction of the intermediate predator. This prediction is not qualitatively altered by more complex models that describe IGP interactions as embedded in larger food webs with potential alternate prey, or in models that incorporate the nonlinearity of trophic interactions. Here I test this prediction of current IGP theory by investigating the trophic interactions of a species-rich omnivorous food web along the large-scale productivity gradient of the New Zealand intertidal. Results/Conclusions Changes in community structure observed along this gradient are notably counter to the theory's predictions. I describe these differences in the context of the changing food web topologies and the empirically-measured species-specific strength and nonlinearity of trophic interactions at six sites across the productivity gradient. Results suggest that the disconnect between predictions and data are due to a failure of current IGP theory to account for productivity-dependent changes in (i) consumer trophic niche-widths, (ii) per capita interaction strengths, and (iii) competitive interactions between alternate basal prey species, but is unlikely to be due to the inherent nonlinearity of pair-wise trophic interactions.