Environmental stress impedes predation and herbivory by limiting the ability of animals to search for and consume prey. We tested the contingency of this relationship on consumer traits, and specifically hypothesized that herbivore mobility relative to the return time of limiting environmental stress would predict consumer effects. We examined how wave-induced water motion affects marine communities via herbivory by highly mobile (fish) versus slow moving (pencil urchin) consumers at two wave-sheltered and two wave-exposed rocky subtidal locations in the Galapagos Islands. The exposed locations experienced 99th percentile flow speeds that were 2-5 times greater than sheltered locations, with mean flow speeds >33 cm/s vs <16 cm/s, 2-7 times higher standing macroalgal cover and 2-3 times lower cover of crustose coralline algae than the sheltered locations. As predicted by the Environmental Stress Hypothesis (ESH), there was a negative relationship between mean flow speed and urchin abundance and herbivory rates on Ulva spp. algal feeding assays. In contrast, the biomass of surgeonfishes (Acanthuridae) and parrotfishes (Labridae- Scarinae) was positively correlated with mean flow speed. Ulva assays were consumed at equal rates by fish at exposed and sheltered locations, indicating continued herbivory even when flow speeds surpassed maximum reported swimming speeds at a rate of 1-2 times per minute. Modeled variation in fish species richness revealed minimal effects of diversity on herbivory rates at flow speeds <40 cm/s, when all species were capable of foraging, and above 120 cm/s, when no species could forage, while increasing diversity maximized herbivory rates at flow speeds of 40 – 120 cm/s. Two-month herbivore exclusion experiments during warm and cool seasons revealed that macroalgal biomass was positively correlated with flow speed. Fish limited macroalgal development by 65-91% at one exposed location but not the second, and by 70% at the two sheltered locations. In contrast, pencil urchins did not affect algal communities at either exposed location, but reduced macroalgae by 87% relative to controls at both sheltered locations. We propose an extension of the ESH that is contingent upon mobility to explain species-specific changes in feeding rates and consumer effects on benthic communities across environmental gradients.