Simple ecological models that predict trophic responses to bottom-up forcing are valuable tools for ecosystem managers. Traditionally, theoretical ecologists have used resource-dependent functional responses to explain the modification of food chains exposed to bottom-up perturbations. These models predict alternating positive, negative and zero responses at each trophic level. More recently, ratio-dependent functional response models that predict proportional increases at each level have challenged this paradigm. The present study tested the predictions of the 2 hypotheses empirically by comparing the relative biomasses of 4 trophic levels of an estuarine seagrass food chain in relatively undisturbed, low-nutrient catchments and 'developed' catchments subjected to a prolonged period of nutrient enrichment. We found that nutrient-enriched sites had significantly greater biomass of both epiphytic algae and grazing invertebrates; however, the bottom- up forcing of nutrients was attenuated at higher trophic levels (occupied by juvenile and piscivorous fish), with no significant effect of catchment development. This disconnect in the upward cascade of energy may be due to a number of possible reasons including high levels of diversity and omnivory, trophic subsidy within the system or the strength or nature of perturbations. Although the predictions of both hypotheses failed to hold across all trophic groups, ratio dependence was prevalent at the lower levels of the food chain, which has implications for catchment management.