Predators have large effects on ecosystem properties by changing plant diversity, not plant biomass.

Cascading effects of predators on total plant trophic-level biomass tend to be weaker in terrestrial than in aquatic systems. Accordingly, it is hypothesized that top predator effects on terrestrial plant diversity and on ecosystem function should likewise be weak or unimportant. This report presents a test of this hypothesis using data from a long-term field experiment. The five-year experiment manipulated the trophic structure of an old field ecosystem by excluding either predators or predators and herbivores relative to an unmanipulated, natural control. Long-term manipulations led to systematic treatment effects on community properties (plant trophic-level biomass, plant species biomass, plant species evenness) and on ecosystem properties (supply rate of solar radiation, N mineralization rate). The strengths of top predator effects on community properties were modest compared with nonterrestrial systems. But, predator-caused changes in plant community structure via alteration of plant dominance, and hence plant species evenness, strengthened effects on ecosystem properties. Counter to the hypothesis, weak trophic cascades do not necessarily lead to weak indirect effects of predators on ecosystem properties.