From Mesocosms to the Field: The Role and Value of Cage Experiments in Understanding Top-Down Effects in Ecosystems

Abstract

Ecologists routinely use field experiments to gain predictive insights about ecosystem structure and function. However, experiments conducted at a whole ecosystem scale are often too crude to provide the detailed causal understanding needed for prediction. Ecologists instead try to gain such causal understanding by conducting experiments in small-scale enclosures or cages. Enclosure cages offer the fine-scale resolution and control needed to isolate certain combinations of species and derive a detailed understanding of interaction mechanisms among the species. However, the applicability of such insights to predicting whole ecosystem function depends critically on satisfying some key design criteria. These criteria include ensuring that enclosure experiments are conducted in natural field environments, as opposed to artificial laboratory settings; that behaviour of mobile species is not seriously hampered; and that experiments are conducted over time scales that represent species' natural life cycles. I detail here how enclosure cage experiments can be an effective tool in an endeavour to predict effects of perturbations on whole ecosystem function. I first provide a detailed explanation of the design of enclosure cages used in studies of herbaceous plants and arthropod herbivores and predators. I also provide guidelines for conducting experiments in ways that do not seriously distort the experimental conditions from those of natural environments. I then illustrate how cage experiments lead to predictive insights using an example from my own research on trophic interactions among spider predators, leaf-chewing and sap-feeding insect herbivores and grass and herb plants in a New England meadow ecosystem. I show that fulfilment of critical design criteria can allow one to isolate the dominant predator and herbivore species in this ecosystem and determine the nature and strength of top-down control of plant species composition and biomass production. I then show how predictions about top-down control of plant species diversity and biomass production are tested and confirmed using large-scale, unenclosed field plots that experimentally manipulate both predator and herbivore trophic levels of the meadow ecosystem.