FLORAL MIMICRY BY EPIDENDRUM IBAGUENSE (ORCHIDACEAE) IN PANAMA.

Abstract

Mimicry in animals is a well-known field of research with an extensive literature. In contrast, the subject of mimicry in plants is an area which has received little attention by investigators. Wickler (1968) and Barlow and Wiens (1977) give examples of plant mimicry in their recent reviews, but experimental data demonstrating deception and its advantage to the plant are, in most instances, scant or lacking. An interesting case of floral similarity occurs in Central America which involves three plant species: Lantana camara L. (Verbenaceae), Asciepias curassavica L. (Asclepiadaceae) and Epidendrum ibaguense HBK (Orchidaceae). These species are frequently sympatric, have flowers that are similar in color (orange and yellow) and are visited primarily by butterflies and hummingbirds and to a lesser extent, bees (Dodson, 1962; van der Pijl and Dodson, 1966). The flowers of L. camara and A. curassavica have nectar while those of E. ibaguense do not. These facts have led researchers (C. H. Dodson, H. G. Baker and others, pers. comm.) to hypothesize that E. ibaguense may be a mimic of the other two species, although no published accounts are available. By deceiving flower visitors with mimicry, plants of E. ibaguense are, in theory, spared the cost of producing nectar as a reward for pollinators. While examining populations of these species in Panama in 1976, I had an opportunity to test this hypothesis. Specifically, I wanted to discover what rewards were offered by these plants, whether flower visitors were deceived by the mimicry, if such it was, and if there was a means by which pollen could be transferred to the correct species. I sampled flowers of these three species for nectar near Boquete, Panama. Sugar content of the nectar was measured in the field with a Bausch and Lomb hand refractometer. Only new flowers were sampled. The nectar of L. camara averaged 19% sugar by weight (N = 25), while that of A. curassavica averaged 35% by weight (N = 32). I obtained no nectar from the flowers of E. ibaguense (N 30). I witnessed only two visits to E. ibaguense flowers by insects in the Boquete area. Both of these visitors were monarch butterflies (Danaus plexippus L.). To determine whether or not individuals were visiting flowers of all three species, I captured ten monarchs and examined them with a dissecting scope. Each of these butterflies had pollen of L. camara on its head and thorax; one or more pollinia of A. curassavica attached to its legs; and one or more pollinia of E. ibaguense attached to the upper one-third of its proboscis. In all individuals the pollinia of A. curassavica and E. ibaguense were confined to these specific areas. Pollen of L. camara was generally on the head and ventral part of the thorax of D. plexippus but could sometimes be found on other parts of the body. These observations, although limited, suggest that E. ibaguense may be a bona fide mimic of L. camara and A. curassavica and that monarch butterflies are deceived by the mimicry. The species-specific placement of the pollinia (and pollen) of these three species on different areas of the body of D. plexippus imply proper intraspecific pollen transfer, data which