Non-linear effects of non-host diversity on the removal of free-living infective stages of parasites.

Among the ecological functions and services of biodiversity is the potential buffering of diseases through dilution effects where increased biodiversity results in a reduction in disease risk for humans and wildlife hosts. Whether such effects are a universal phenomenon is still under intense debate and diversity effects are little studied in cases when non-host organisms remove free-living parasite stages during their transmission from one host to the next by consumption or physical obstruction. Here, we investigated non-host diversity effects on the removal of cercarial stages of trematodes, ubiquitous parasites in aquatic ecosystems. In laboratory experiments using response surface designs, varying both diversity and density at same time, we compared three combinations of two non-hosts at four density levels: predatory crabs that actively remove cercariae from the water column via their mouth parts and gills, filter feeding oysters that passively filter cercariae from the water column while not becoming infected themselves, and seaweed which physically obstructs cercariae. The addition of a second non-host did not generally result in increased parasite removal but neutralised, amplified or reduced the parasite removal exerted by the first non-host, depending on the density and non-host combination. These non-linear non-host diversity effects were probably driven by intra- and interspecific interactions and suggest the need to integrate non-host diversity effects in understanding the links between community diversity and infection risk.