One of the oldest observations in ecology is that larger areas contain more species. But do larger areas contain different species? The trophic rank hypothesis argues that specialist predators will disproportionately occur in large areas, as large areas maximize the chance of predators co-colonizing with their prey species. This implies that predator diversity increases faster with area than will prey diversity.

We used the aquatic insects in bromeliads to examine how species-area relationships differ between predators and prey. Our dataset included complete censuses of nearly 300 bromeliads from three countries. Overall, predator diversity increased faster with bromeliad size than did prey diversity, consistent with the trophic rank hypothesis. However, bromeliad predators are not specialists, arguing against trophic specialization as a mechanism. Rather, trophic level effects in species-area relationships may reflect trait differences between predators and prey. Predators are often rarer and longer-lived than their prey. We show that any group of species with low abundance will necessarily have steeper species-area curves. The abundance mechanism explains the observed patterns in Puerto Rico and Dominica, but not Costa Rica. The patterns in Costa Rica are likely explained by another mechanism: predators have long larval lifespans, and thus are especially sensitive to the high disturbance rate that characterizes small bromeliads. In summary, trophic level effects on species-area relationships in this system reflect trait differences between predators and prey, not trophic differernces.