Ants on Cecropia Trees in Urban San Jose, Costa Rica

Abstract

The symbiosis between ants and Cecropia trees is among the best-studied antplant relationships (Belt 1874, Muller 1874, Janzen 1969, Longino 1989). ResidentAzteca ants, the commonest Cecropia symbionts, typically defend their host trees against herbivory and overgrowth by vines (Janzen 1969, Schupp 1986, Rocha and Bergallo 1992). Cecropia trees, in turn, provide resident ants with shelter within their trunks and with food in the form of nutrient-rich Mullerian bodies on the base of the petioles and pearl bodies on the undersides of the leaves (Rickson 1971). Resident ants also often gain additional nutrition through feeding on honeydew produced by homopterans which the ants tend within the Cecropia trunk (Belt 1874, Wheeler 1942). In the forests of Costa Rica, the majority of Cecropia trees, including Cecropia obtusifolia Bertol., are occupied by mutualistic Azteca ants (Longino 1989). In the present study, I surveyed ants on C. obtusifolia trees planted as ornamentals in the capital city of San Jose, Costa Rica, to determine whether these trees, isolated from native forest by several kilometers of urban areas, were occupied and protected byAzteca ants. In June 1996, 1 surveyed ants on all 27 C. obtusifolia trees planted in the plaza in front of the Costa Rican National Museum near the center of San Jos6, Costa Rica. Trees ranged from 0.5 to 7 m in height. For trees five meters or more in height (estimated to the nearest 1 m), I collected from the trunk all ants I could reach (up to -2.5 m). For trees three meters or less in height (estimated to the nearest 0.5 m), I examined every leaf for ants, and searched the entire trunk for any ant nest entrance holes. When surveying, I shook all trees to stimulate ant activity. Voucher ant specimens from this study are in the Museum of Comparative Zoology, Harvard University and the Smithsonian Institution. Ants were on twelve of the 27 C. obtusifolia trees. Ants were more common on larger trees: on nine of twelve large trees (> 2 m height), and three of fifteen small trees (< 2 m height) (XK = 8.3, P < 0.05). There were only three species of ants on the trees: Paratrechina longicornis (Latreille), Acromyrmex octospinosus (Reich), and Azteca sp. (Table 1). P longicornis was by far the most common and widespread ant, occurring on ten of the twelve trees with ants (Table 1). A. octospinosus, a leaf-cutting ant, was also common, with foragers carrying leaf fragments and pieces of other vegetable matter coming down six trees. There was leaf-cutting ant damage on the leaves of at least three additional trees without leaf-cutters present. One of these damaged trees had P longicornis and two were without any ants on them. Azteca ants were on only two Cecropia trees, both five meters in height. On these two trees, large numbers ofAzteca poured out of their nests within the trunk when I shook the tree. The Azteca workers could not be identified to species because queens are needed for such identification and I could not cut down and dissect these ornamental Cecropia trees to obtain the queens. Dense setae on the workers' scapes and tibiae indicated they did not belong to the alfari group (Longino 1991a, b.) Paratrechina longicornis occurred in both trees occupied byAzteca, but A. octospinosus occurred in neither. The P longicornis and A. octospinosus nested in the ground in the plaza, whereas the Azteca sp. nested within the trunks of the Cecropia trees. I noted tens to hundreds of unharvested Mullerian bodies on the leaf petioles of the small Cecropia trees.