Effects of an herbicide and an insecticide on pond community structure and processes

Virtually all species live within complex food webs, and many of these organisms are exposed to contaminants. However, we know little about how community processes, such as competition and predation, influence susceptibility to contaminants or how different types of contaminants shape communities. The objective of our study was to determine how realistic concentrations of the herbicide atrazine and the insecticide endosulfan influence the structure of a pond community when the presence of common community members was manipulated. We employed a factorial design in mesocosms to evaluate the effects of pesticide treatments (25 mu g/L of atrazine, 10 mu g/L of endosulfan, solvent control; two pulses separated by two weeks) and the presence or absence of wood frog tadpoles (Rana sylvatica), adult snails (Planorbella trivolvis), and caged dragonfly larvae (Anax junius) on a freshwater community. Tadpoles, snails, and chironomid larvae, Polypedilum sp. (Dipterans), all competed for periphyton. As a result, tadpoles reduced the survival, mass, and reproduction of snails; snails reduced the growth, development, inactivity, and dragonfly avoidance of tadpoles; snails and tadpoles reduced the abundance of chironomid larvae; and chironomid larvae reduced snail mass. The adverse effect of snails on tadpole growth and behavior was greater in the presence of the caged tadpole predator, A. junius. Neither pesticide affected dragonfly survival, but endosulfan directly reduced zooplankton (Daphnia), and atrazine indirectly reduced chironomid abundance. Atrazine also directly decreased periphyton, and endosulfan decimated chironomid larvae, resulting in indirect increases and decreases in competition for both snails and tadpoles, respectively. Consequently, relative to endosulfan, atrazine tended to decrease snail mass and reproduction and. reduce tadpole mass, development, inactivity, refuge use, and dragonfly avoidance. However, the indirect effects of pesticides depended upon the presence of heterospecifics. The indirect benefit of endosulfan on snail mass was greater in the presence of caged dragonfly larvae, and endosulfan's indirect benefit on tadpole mass was greater in the absence of snails. The effect of pesticides on tadpole activity depended on both caged dragonflies and snails. Thus, environmentally realistic concentrations of pesticides directly and indirectly shaped species responses and community composition, but the initial composition of the community influenced these pesticide effects. These results emphasize the importance of quantifying the effects of contaminants within complex natural communities.