Chemical Leaf Repellency to an Attine Ant: Seasonal Distribution Among Potential Host Plant Species

Abstract

The chemical repellency of leaves to the leaf—cutting, fungus—growing ant Atta cephalotes was surveyed for 42 plant species randomly selected from the dry forest of Santa Rosa national Park, Guanacaste, Costa Rica. The sample represents about one—sixth of the potential host plant species available to the ants in the secondary, semideciduous forest at Santa Rosa. Repellencies of leaf extracts were measured by bioassay. A laboratory colony of A. cephalotes collected from Santa Rosa was used. Three—quarters of the species exhibited significantly repellent nonpolar (lipid—soluble) extractables (e.g., terpenoids, steroids, and waxes), and half of the species exhibited repellent polar extractables (e.g., phenols, flavonoids, and glycosides). The occurrence of significant extractable repellents, particularly in the lipid—soluble fraction, correlated closely with species that were avoided in leaf—preference tests and with seasonal patterns of host—plant selection by ant colonies in the Santa Rosa forest. There were no clear correlations of repellency with plant growth form or with evergreen vs. Deciduous habit. Further chemical isolation of ant—repellent substances from a number of avoided species revealed that most of the repellent compounds are terpenoids, some of which have already been shown to be highly toxic to the ant's fungus and to many other fungi as well. A dramatic decline in the amount of extractable repellent occurred in almost all plants species in the latter half of the wet season, 1—2 mo before the dry season began. Because this decline was prevalent in evergreen as well as deciduous species, we tentatively concluded that the ant—repellent substances in the leaves are mainly the result of selection in plants for fungal resistance, which incidentally confers resistance to attack by leaf—cutting ants. We suggest that the decline in the amount of repellent substance is due to a reduction in the synthesis of antifungal secondary compounds in the dry season, when the risk of fungal attack is low.