Ants are ubiquitous insects in terrestrial ecosystems that act as generalist predators of other arthropods or as mutualists of plant-sucking insects. There are many reports of  negative effects of ants on the abundance of carnivore arthropods and positive ones on plant-sucking insects, but much less is known about their effect on the diversity of the arthropod community and on plant performance. We conducted an ant-exclusion experiment on citrus canopies in NE Spain to investigate the effects on the abundance and diversity of canopy arthropods and on fruit yield. We expected to find a general increase of arthropod abundance and diversity in ant-excluded trees but a reduction of ant-mutualists groups like aphids. If these effects cascaded down to plant level we could also anticipate a greater fruit yield in ant-excluded trees. Arthropods were sampled monthly using beating trays and all individuals were classified to order or to species (Coleoptera, Heteroptera, Araneae, Aphididae, Formicidae). As some results were rather unexpected and some others were variable among years, the experiment was repeated seven years (2002 to 2008) using a different group of control and experimental trees each year.

Results/Conclusions

As expected, the exclusion of ants increased the abundance of most arthropod groups. The more detailed analysis within some groups showed that the effect was not widespread among all families or species within the group, but limited to a few taxa. Ant-exclusion also increased the diversity of some arthropod groups, like Araneae and Coleoptera. However, contrary to expectations, the exclusion of ants from the canopies did not reduce aphid attack, but clearly increased it. This result was surprising as all aphid species were ant-tended and the grove had several aphid-tending ants. The most likely explanation for this phenomenon is the concurrent exclusion of the European earwig Forficula auricularia with ants. Earwigs are known to predate on aphids and in this particular site the negative effect of earwigs on aphid populations seems to be stronger that the positive effect of aphid-tending ants. Finally, as expected, ant (and earwig) exclusion slightly increased fruit yield over the seven years of study.