Experimental venue and spatial scale in food-web manipulations

Trophic cascades are well known in ecology and are usually discovered only after manipulation of the existing trophic structure.  Food webs are often manipulated using enclosure/exclosure cages to restrict access to only one or two trophic levels.  Experiments using such manipulations are subject to criticism because results may not scale-up to the natural environment and cage artifacts can complicate interpretation of results.  Here we examine evidence for trophic cascades in freshwater marshes of the Florida Everglades and test outcomes of two different cage designs and two different spatial scales.  There was no evidence of trophic cascades involving large fish, intermediate consumers, and primary consumers but snails avoided areas with high densities of crayfish.  Cage designs did not manipulate the marsh environment equally.  One design appeared to attract crayfish for both predator refuges and controls, indicating the result was an artifact.  Similarly, results at the 1-m² scale did not scale-up to the 16-m² scale.  The implications for these results on future food web manipulations are discussed.