The Influence of Aggressive Ants on Fruit Removal in the Tropical Tree, Ficus capensis (Moraceae)

Abstract

This study examined the effect of aggressive arboreal weaver ants (Oecophylla longinoda) on the dispersal of fruits from the African fig, Ficus capensis. Compared with trees having no or few 0. longinoda, trees with large ant colonies had reduced levels of nocturnal fruit removal and an increased proportion of fruits that fell undispersed. Diurnal fruit removal did not differ significantly between trees with different densities of ants. This shift in fruit dispersal characteristics resulted from the clustering of aggressive 0. longinoda on the fruiting rama at night. Although ant-plant mutualisms are commonly thought to have evolved from the enhanced fitness that aggressive ants confer on occupied trees by reducing herbivory or competition, reduced seed dispersal in zoochoric species may be an important evolutionary block. ANYONE TRYING TO CLIMB OR CUT TREES in the tropics is well aware that the presence of aggressive biting or stinging ants can be a minor irritant or a major deterrent. Experimental and observational studies have shown that the activities of aggressive ants on trees can reduce herbivory and/or control plant competitors, and so confer a degree of protection on occupied individuals (e.g., Bequeart 1922; Wheeler 1942; Janzen 1966, 1967a, b, 1969a; Bentley 1977). In Africa, the aggressive behavior of Pachysima ants protects Barteriafistulosa (Passifloraceae), and in Central America Pseudomyrmex protect Acacia (Leguminosae) and Azteca protect Cecropia (Moraceae) in a similar fashion (Janzen 1966, 1969a, 1972). The aggressive response of ants to intruders or novel stimuli appears to be undirected and crosses size and taxonomic categories. Pachysima and Pseudomyrmex both respond aggressively by attacking insects and larger vertebrates (Janzen 1969a, b, 1972; McKey 1974). While these cases are examples of coevolved mutualistic ant-plant interactions, it is reasonable to suppose that antiherbivore protection is a widespread incidental result of occupation by aggressive ants, even in less conspicuously coevolved systems. The aggressive behavior of arboreal ants, resulting in reduced herbivory or competition, likely was the first step leading to the evolution of ant-plant mutualisms. Reduced levels of herbivory and the resulting increase in vegetative production are clearly advantageous to occupied plants; however, colonization by aggressive ants may also entail subtle costs. Many tropical plants produce edible fruits and rely on frugivores to effect seed dispersal. Any factors that act to reduce the attractiveness of a plant or its fruit crop to the point of reducing fruit removal, and hence seed dispersal, will have a direct negative impact on its fitness. The Pachysima-Barteria mutualism may serve as a hypothetical example (see Janzen 1972). The presence of Pachysima ants on B. fistulosa trees confers a degree of protection against herbivore damage. While foraging on cultivated fungi and honeydew from scale insects, the ants patrol occupied trees and respond to any disturbance by stinging. These ants are sufficiently aggressive to deter large vertebrates (monkeys) from entering occupied trees (McKey 1974). At the same time, however, B. fistulosa produces arrays of edible fruits that are dispersed by a variety of mammals and birds. Presumably these frugivores must enter the trees to access the fruits, and so must encounter the same ant aggression that deters herbivores. Although the evolution of a close mutualistic interaction between B. fistulosa and Pachysima suggests that the net benefit-cost balance (enhanced competitive ability vs potential reduction in seed dispersal efficiency) is positive, this may not always be the case. While studying the impact of fruit bats (Megachiroptera: Pteropodidae) on fruit crops and seed dispersal of the African fig, F. capensis (Moraceae), in Ivory Coast, it became apparent to me that the common weaver ant (0. longinoda) had a major influence on fruit removal. Although the relationship between F. capensis and 0. longinoda is not a tight, mutualistic one, it can serve as an initial model for interactions between any aggressive ant species and frugivores. In this paper, I present data on the interactions between F. capensis, diurnal and nocturnal frugivores, and 0. longinoda in West Africa.