| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 2413738 | Agriculture, Ecosystems & Environment | 2015 | 9 Pages |
•Predation services enhanced where ant abundance and richness increase.•Predation did not correlate with agrochemical use.•Ant richness and abundance were strongest predictors of predation.•Local factors were stronger predictors of predation services than landscape factors.
Agricultural management and the landscape surrounding farms impact biological diversity and ecosystem services, such as predation, in agroecosystems. Diversified coffee agroecosystems harbor biodiversity, and maintain ecosystem services, especially when in complex landscapes, and when diversity of organisms providing services is maintained. But few have examined whether biological diversity, per se, or the local and landscape habitat features are stronger drivers of the services provided. We studied the relationships between local characteristics associated with agricultural management (vegetation complexity and agrochemical use), landscape surroundings of farms, abundance and richness of ants, and predation services provided by ants in a tropical coffee landscape. Specifically, we tested whether: (1) ants remove prey items and prey removal differs in the wet and dry season, (2) ant prey removal differs in farms that vary in vegetation complexity and agrochemical use, (3) ant prey removal differs with changes in ant abundance and species richness, and (4) ant prey removal differs with changes in local and landscape characteristics. We established sites across a range of coffee management systems, at varying distances to forest fragments, and in forest fragments and examined prey removal by ants on the ground, coffee branches, and shade tree trunks in the wet and dry season. Prey removal did not differ with season. Prey removal on the ground increased within increases in ant abundance and richness, and number of hollow coffee twigs. Prey removal on coffee plants decreased with vegetation complexity, but increased with ant abundance and richness, coffee density (a local factor), and rustic coffee within 200 m (a landscape factor). Prey removal on trees declined with vegetation complexity, but increased with abundance and richness of ants. Characteristics of the ant community were consistent predictors of prey removal, whereas local and landscape habitat features were less important, and only three habitat features correlated with prey removal. Our results demonstrate the positive effects of ant diversity for predatory function, and that promoting pest control services within coffee agroecosystems may be best accomplished by manipulating ant abundance and richness, or vegetation factors that correlate with ants.
