Impacts of agrochemical intensification on the assembly and reassembly of a mainland-island model metacommunity

Abstract

Many lentic aquatic environments are found embedded in agricultural fields, forming complex metacommunity structures. These habitats are vulnerable to contamination by agrochemicals, which can differentially affect local communities depending on the intensity and variability of species dispersal rates. We conducted a field experiment to assess how agrochemical intensification simulating the conversion of savannas into managed pastures and sugarcane fields affects freshwater community structure at different levels of spatial isolation. We constructed forty-five 1,200-L artificial ponds in a savanna landscape at three distances from a source wetland (30 m, 120 m, and 480 m). Ponds were spontaneously colonized by aquatic insects and amphibians and treated with no agrochemicals ( savanna treatment), fertilizers ( pasture treatment), or fertilizers and a single pulse of the insecticide fipronil and the herbicide 2,4-D ( sugar cane treatment) following realistic dosages and application schedules. The experiment encompassed the entire rainy season. Pasture communities were only slightly different from controls largely because two predatory insect taxa were more abundant in pasture ponds. Sugarcane communities strongly diverged from other treatments after the insecticide application, when a decrease in insect abundance indirectly benefitted amphibian populations. However, this effect had nearly disappeared by the end of the rainy season. The herbicide pulse had no effect on community structure. Spatial isolation changed community structure by increasing the abundance of non-predatory insects. However, it did not affect all predatory insects nor, surprisingly, amphibians. Therefore, spatial isolation did not change the effects of agrochemicals on community structure. Because agrochemical application frequently overlaps with the rainy season in many monocultures, it can strongly affect temporary pond communities. Ponds embedded in pastures might suffer mild consequences of fertilization by favoring the abundance of few predators through bottom-up effects. Ponds in sugarcane fields, however, might experience a decline in the insect population, followed by an increase in the abundance of amphibians tolerant to environmental degradation. Furthermore, we found no evidence that isolation by distance can change the general effects of chemical intensification, but future experiments should consider using real crop fields as the terrestrial matrix since they can represent different dispersal barriers.