USF Libraries

Evidence for floral mimicry in Epidendrum radicans (Orchidaceae) with Asclepias curassavica (Apocynaceae) and Lantana ca...

MISSING IMAGE

Material Information

Title:
Evidence for floral mimicry in Epidendrum radicans (Orchidaceae) with Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae)
Translated Title:
Evidencia para la mimica floral en Epidendrum radicans (Orchidaceae) con Asclepias curassavica (Apocynaceae) y Lantana cámara (Verbenaceae) ( )
Physical Description:
Book
Language:
English
Creator:
Dupre, Sarah
Publication Date:

Subjects

Subjects / Keywords:
Mimicry (Biology)   ( lcsh )
Pollination by animals   ( lcsh )
Costa Rica--Guanacaste--Cañitas   ( lcsh )
Mimetismo (Biología)
Polinizado por animales
Costa Rica--Guanacaste--Cañitas
Tropical Ecology Fall 2004
Ecología Tropical Otoño 2004
Genre:
Reports   ( lcsh )
Reports

Notes

Abstract:
Floral mimicry may occur between plant species with geographic overlap, similar flowering times and a shared set of pollinators. To demonstrate mimicry in such species, visitors must enhance fitness of individual flowering plants. This study focuses on potential floral mimicry for a Neotropical roadside weed, Epidendrum radicans (Orchidaceae), which produces no nectar but resembles nectar-producing, sympatric and phenologically similar plant species, Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae). Twenty pollinia of E. radicans were placed in four different groups. One containing the orchid alone, the second consisting of the orchid and A. curassavica, the third consisting of the orchid and L. camara and the fourth group containing all three species. Proximity to L. camara increased butterfly visitation in E. radicans nearly four-fold and more than doubled visitation when near A. curassavica (Friedman Test, Chi-squared = 22.23, df = 3, p = 0.0001; MCN Test, q = 5.962, df = ∞, p < 0.05). Likewise, pollinia removal increased as a result of proximity to both A. curassavica and L. camara (Friedman Test, Chi-squared = 24.825, df = 3, p = 0.0001; MCN Test, q = 6.261, df = ∞, p < 0.05). Simply doubling E. radicans flowers in monospecific groups to the same total flower number as mixed patches had no effect on visitation nor pollinia removal. Hence, the experiments presented here support the case for floral mimicry in E. radicans, though its impact on wild populations remains in doubt.
Abstract:
La mímica floral puede ocurrir entre las especies de plantas con traslape geográfico, épocas florecientes similares y un sistema compartido de polinizadores. Para demostrar la mímica en tales especies, los visitantes deben realzar la adaptabilidad de plantas florecientes individuales. Este estudio se centra en la mímica floral potencial para una mala hierba neotropical del borde de la carretera, Epidendrum radicans (Orchidaceae), que no produce ningún néctar pero que se asemeja a otras especies de plantas que producen néctar, son simpátricas y fenológicamente similares, Asclepias curassavica (Apocynaceae) y Lantana camara (Verbenaceae). Las polinias de veinte E. radicans se colocaron en cuatro grupos diferentes. Uno que contenía solamente la orquídea, el segundo que consistía en la orquídea y A. curassavica, el tercero que consistía en la orquídea y L. camara y el cuarto grupo que contenía las tres especies. La proximidad a L. camara aumentó las visitas de las mariposas a E. radicans casi cuatro veces y, cuando A. curassavica estaba cerca, las visitas a E. radicans se duplicaron. (Prueba de Friedman, Chi-cuadrada = 22.23, df = 3, p = 0.0001; prueba de MCN, q = 5.962, df =3, p = 0.0001; Prueba de MCN, q = 5.962, df = ∞, p < 0.05). De la misma manera, la remoción de la pollinia aumento debido a la proximidad de A. curassavica y L. camara (Prueba de Friedman, Chi-cuadrada = 24.825, df = 3, p = 0.0001; prueba de MCN, q = 6.261, df = ∞, p < 0.05). Cuando se duplico el número de flores de E. radicans en los grupos monoespecíficos al mismo número total de flores como los parches mixtos, no tuvo efecto en las visitas o en la remoción de polinias. Los experimentos reportados aquí dan evidencia en favor de la mímica floral en E. radicans, aunque su impacto en las poblaciones silvestres sigue en duda.
Language:
Text in English.
General Note:
Born Digital

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - M39-00374
usfldc handle - m39.374
System ID:
SFS0001543:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00374
040
FHM
0 041
eng
049
FHmm
1 100
Dupre, Sarah
242
Evidencia para la mimica floral en Epidendrum radicans (Orchidaceae) con Asclepias curassavica (Apocynaceae) y Lantana cmara (Verbenaceae)
245
Evidence for floral mimicry in Epidendrum radicans (Orchidaceae) with Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae)
260
c 2004-11
500
Born Digital
3 520
Floral mimicry may occur between plant species with geographic overlap, similar flowering times and a shared set of pollinators. To demonstrate mimicry in such species, visitors must enhance fitness of
individual flowering plants. This study focuses on potential floral mimicry for a Neotropical roadside weed, Epidendrum radicans (Orchidaceae), which produces no nectar but resembles nectar-producing,
sympatric and phenologically similar plant species, Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae). Twenty pollinia of E. radicans were placed in four different groups. One containing the orchid alone, the second consisting of the orchid and A. curassavica, the third consisting of
the orchid and L. camara and the fourth group containing all three species. Proximity to L. camara increased butterfly visitation in E. radicans nearly four-fold and more than doubled visitation when near A. curassavica (Friedman Test, Chi-squared = 22.23, df = 3, p = 0.0001; MCN Test, q = 5.962, df = , p <
0.05). Likewise, pollinia removal increased as a result of proximity to both A. curassavica and L. camara (Friedman Test, Chi-squared = 24.825, df = 3, p = 0.0001; MCN Test, q = 6.261, df = , p < 0.05). Simply
doubling E. radicans flowers in monospecific groups to the same total flower number as mixed patches had no effect on visitation nor pollinia removal. Hence, the experiments presented here support the case for floral mimicry in E. radicans, though its impact on wild populations remains in doubt.
La mmica floral puede ocurrir entre las especies de plantas con traslape geogrfico, pocas florecientes similares y un sistema compartido de polinizadores. Para demostrar la mmica en tales especies, los visitantes deben realzar la adaptabilidad de plantas florecientes individuales. Este estudio se centra en la mmica floral potencial para una mala hierba neotropical del borde de la carretera, Epidendrum radicans (Orchidaceae), que no produce ningn nctar pero que se asemeja a otras especies de plantas que producen nctar, son simptricas y fenolgicamente similares, Asclepias curassavica (Apocynaceae) y Lantana camara (Verbenaceae). Las polinias de veinte E. radicans se colocaron en cuatro grupos diferentes. Uno que contena solamente la orqudea, el segundo que consista en la orqudea y A. curassavica, el tercero que consista en la orqudea y L. camara y el cuarto grupo que contena las tres especies. La proximidad a L. camara aument las visitas de las mariposas a E. radicans casi cuatro veces y, cuando A. curassavica estaba cerca, las visitas a E. radicans se duplicaron. (Prueba de Friedman, Chi-cuadrada = 22.23, df = 3, p = 0.0001; prueba de MCN, q = 5.962, df =3, p = 0.0001; Prueba de MCN, q = 5.962, df = , p < 0.05). De la misma manera, la remocin de la pollinia aumento debido a la proximidad de A. curassavica y L. camara (Prueba de Friedman, Chi-cuadrada = 24.825, df = 3, p = 0.0001; prueba de MCN, q = 6.261, df = , p < 0.05). Cuando se duplico el nmero de flores de E. radicans en los grupos monoespecficos al mismo nmero total de flores como los parches mixtos, no tuvo efecto en las visitas o en la remocin de polinias. Los experimentos reportados aqu dan evidencia en favor de la mmica floral en E. radicans, aunque su impacto en las poblaciones silvestres sigue en duda.
546
Text in English.
650
Mimicry (Biology)
Pollination by animals
Costa Rica--Guanacaste--Caitas
4
Mimetismo (Biologa)
Polinizado por animales
Costa Rica--Guanacaste--Caitas
653
Tropical Ecology Fall 2004
Ecologa Tropical Otoo 2004
655
Reports
720
CIEE
773
t Monteverde Institute : Tropical Ecology
856
u http://digital.lib.usf.edu/?m39.374



PAGE 1

1 Evidence for floral mimicry in Epidendrum radicans (Orchidaceae) with Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae). Sarah Dupre Department of Secondary Education: Biology and General Science, University of Rhode Island ABSTRACT Floral mimicry may occur between plant species with geographic overlap, similar flowering times and a shared set of pollinators. To demonstrate mimicry in such species, visitors must enhance fitness of individual flowering plants. This study focuses on potential floral mimicry for a Neotropical roadside weed, Epidendrum radicans (Orchidaceae), which produces no nectar but resembles nectar producing, sympatric and phenologically similar plant species, Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae). Twenty pollinia of E. radicans were placed in four different groups. One containing the orchid alone, the second consisting of the orchid and A. curassavica the third consisting of the orchid and L. camara and the fourth group containing all three species. Proximity to L. camara increased butterfly visitation in E. radicans nearly four fold and more than doubled visitation when near A. curassavica (Friedman Test, Chi 0 .05). Likewise, pollinia removal increased as a result of proximity to both A. curassavica and L. camara (Friedman Test, Chi doubling E. radicans flowers in monospecific groups to the same total flower number as mixed patches had no effect on visitation nor pollinia removal. Hence, the experiments presented here support the case for floral mimicry in E. radicans though its impact on wild populations remains in doubt. R ESUMEN La mmica floral puede ocurrir entre las especies de plantas con traslapo geogrfico, pocas florecientes similares y un sistema compartido de polinizadores. Para demostrar la mmica en tales especies, los visitantes deben realzar la adaptabilidad de plantas florecientes individuales. Este estudio se centra en la mmica floral potencial para una mala hierba neotropical del borde de la carretera, Epidendrum radicans (Orchidaceae), que no produce ningn nctar pero que se asemeja a otras especies de plantas que producen nectar, son simptricas y fenolgicamente similares, Asclepias curassavica (Apocynaceae) y Lantana camara (Verbenaceae). Los polinios de veinte E. radicans se colocaron en cuatro grupos diferentes. Uno que contena solamente la orqu dea, el segundo que consista en la orqudea y A. curassavica el tercero que consista en la orqudea y L. camara y el cuarto grupo que contena las tres especies. La proximidad a L. camara aument las visitas de las mariposas a E. radicans casi cuatro veces y, cuando A. curassavica estaba cerca, las visitas a E. radicans se duplicaron. (Prueba de Friedman, Chi cuadrada = 22.23, df = 3, p = 0.0001; prueba de MCN, q = 5.962, df =3, p = 0.0001; Prueba de MCN, q = 5.962, df = la misma manera, la remocin del polinio aumento debido a la proximidad de A. curassavica y L. camara (Prueba de Friedman, Chi 0.05). Cuando se dobl el nmero de flores de E. radicans en los grupos monoespecficos al mismo nmero total de flores como los parches mixtos, no tuvo efecto en las visitas o en la remocin de polinios. Los experimentos reportados aqu dan evidencia en favor de la mmica floral en E. r adicans aunque su impacto en las poblaciones silvestres sigue en duda.

PAGE 2

2 INTRODUCTION Floral mimicry is said to occur when two or more plant species converge on a common flower morphology. This might occur in two ways: Mllerian mimics converge to attrac t similar pollinators to their rewarding flowers, while Batesian mimics offer no reward and deceive the pollinator by resembling a rewarding plant species (Roy & Widmer 1999). For floral mimicry to occur, supposed mimics must overlap geographically and ph enologically, share individual pollinators and the resemblance must enhance fitness (Bierzychudek 1981; Roy & Widmer 1999). Though proposed for many sets of overlapping species, no case of floral mimicry has ever been verified (Roy & Widmer 1999). Condit ions for floral mimicry exist for three species of plants throughout Neotropical America: Epidendrum radicans (Orchidaceae), Asclepias curassavica (Apocynaceae) and Lantana camara (Verbenaceae). These species share common floral characteristics, have broa d overlapping geographical distributions and share a potential set of general butterfly pollinators (Bierzychudek 1981). They are all roadside weeds that grow singularly and together at an elevation between 1,000 and 2,000 meters. Both L. camara (Todzia 1983) and A. curassavica (Galil & Zeroni 1965; Wyatt 1980) are reported to produce large amounts of nectar, making them potential Mllerian mimics. E. radicans on the other hand, produces no nectar (Bierzychudek 1981; William Haber, personal communicatio n 2004), making it a candidate for Batesian mimicry. Several studies conducted in the past have failed to support the idea of floral mimicry in E. radicans A. curassavica and L. camara possibly because they dealt with too many variables that potentially impacted visitation and pollination success beyond floral resemblance. For example, Bierzychudek (1981), studied naturally occurring E. radicans show that density is an important determinant of vi sitation and pollinia removal (Deacon 2000; Wolfe 1987; Woo 2001). This study returns to test the impact of floral mimicry on visitation and pollinia removal in E. radicans taking care to control for patch size and floral density. MATERIALS AND METHO DS The study site was located in Caitas, Monteverde, Costa Rica, an elevation of 1,265 meters, which is Tropical Lower Montane Moist Forest (Hartshorn 1983). The different groupings of plants were located on an inclined embankment with a river running a long the base. The slope did not have any trees growing on it, however it did have forests starting at the bottom and continuing away from the slope. The only two flowering plants that grew on different parts of the embankment were A. curassavica and L. camara The study site received direct sunlight from 8:30 a.m. until approximately 3:30 p.m. Approximately one hundred E. radicans were collected along the roadsides in San Luis, Costa Rica. Plants were then taken to the study site in Caitas and four groups of these three species were made. Every group contained a patch of E. radicans that had a count of twenty flowers and hence, twenty pollinia, since each flower has one pollinium pollinator or was not a younger, yellow and orange flower, it was removed from the plant.

PAGE 3

3 E. radicans always contained twenty flowers while the other two species had approximately twenty to forty flowers, always looking similar in size and floral appearanc e. The first group included twenty flowers of only E. radicans The second group had a patch of orchids with a patch of A. curassavica The third group had a patch of orchids with a patch of L. camara The final group included all three of the species located in close proximity of each other. Each group was at least seven meters away from the next closest group. There were twelve one hour trial periods in which butterfly visitations were observed and recorded for each of the species in every group. After an hour of observation had passed, the pollinia that had been taken by its pollinators were counted, removed from the orchid and then replaced with the necessary new flowers to replace the removed ones. When the study was done on consecutive days, pollina removals were recorded from the time the site was left the previous day until the arrival to the site the following morning. Towards the end of the study, it was noted that E. radicans growing alone might have a disadvantage. Although E. radicans did have the same number of pollinia as the other orchid patches, it was still lacking an additional patch of coloration and the presence of a nectar producing neighbor ( A. curassavica and/or L. camara ) that the other groups had. For six days, an addition al patch containing only E. radicans with forty flowers and thus, forty pollinia were put on the embankment along with the other four groups. The additional group was observed and recorded the same way previous trials had been done. RESULTS There was a clear trend, displaying that butterflies were visiting E. radicans much more frequently when they were located with both A. curassavica and L. camara There was a slight increase of butterfly visitation when E. radicans was with L. camara rather than when it was with A. curassavica These observations are evident in Figure 1, where there was a significant difference in visitation rates (Friedman Test, Chi square = 22.23, df = 3, p = 0.0001). When the group of A. curassavica L. camara and E. radicans we re compared to E. radicans alone, there was also a significant difference in butterfly visitation rates, but there was not a sign ificant difference between any other groups in pollinia removal overall for E. radicans when it was with the other two species (Friedman Test, Chi squared = 24.825, d f = 3, p = 0.0001) and also when the three species were together were compared to when the orchid was alone (MCN Test, q = E. radicans is with the other nectar producing plants, its visitation rates and po llinia removal increases. For a period of five days, there were two different groups of E. radicans One group contained twenty flowers while the other group contained forty flowers. It is illustrated in Figure 2 that whether there were twice as many fl owers and thus, twice as many pollinia in a group of orchids, there was not a significant difference in visitation rates or in pollinia removal (Wilcoxon Signed Rank Test, (visitation) n = 5, z = 0.137, p = 0.891 and (pollinia removal) n = 5, z = 1.342 p = 0.18). Therefore, even a doubling of flower number did not impact visitation or pollinia removal.

PAGE 4

4 When the study was conducted on consecutive days, pollinia were counted and recorded from the time the site was left on the previous day until the nex t morning of arrival. It is apparent in Figure 3 that there was not a significant difference in the overall pollinia removal (Freidman Test, Chi squared = 5.475, df = 3, p = 0.132) (MCN Test of pollinia between the hours of leaving the site the previous day (~1:00 p.m.) and the arrival time of the following day (~8:30 a.m.). Visitation rates were compared between A. curassavica and L. camara In Figure 4, there was a significant difference in the amount of butterfly visits for the two species (Wilcoxon Signed Rank Test, n = 12, z = 3.065, p = 0.002). L. camara received many more visits than A. c urassavica in every single trial that was performed. In Figure 5, there was not a significant difference for the average number of butterfly visitations to L. camara when it was with E. radicans and then when it was with E. radicans and A. curassavica (Wil coxon Signed Rank Test, n = 12, z = 0.551, p = 0.582). Additionally, there was not a significant difference for the average number of butterfly visitations to A. curassavica when it was with E. radicans and then with E. radicans and L. camara (Wilcoxon S igned Rank Test, n = 12, z = 0.707, p = 0.48). Additional Observations The most common butterflies observed on the study site were Anartia fatima Papilio polyxines Heliconius clysonymus Siproeta epaphus epaphus and Lycorea cleobaeae atergatis. Duri ng the observation periods, it was noted that these butterflies would stay on A. curassavica and L. camara for longer periods of time rather than when they were on E. radicans Also, throughout the duration of this study, a unique association was observed between a wasp and E. radicans The wasp would travel up and down the the top of the plant and nudge the butterfly off of the flower with its head. It was noted that th e wasp seldom visited the other two plant species and that it would fly around the orchid in a territorial way, flying after any butterflies that would try to land on the flowers. This would only occur occasionally, since the wasps were frequently swatted away by the observer during the study periods. DISCUSSION Unlike other studies that fail to support the concept of floral mimicry for E. radicans with A. curassavica and L. camara data here support that E. radicans increases its fitness as a result of proximity to the other two species. Because E. radicans produces no nectar, this is the first field test of Batesian mimicry in a plant. Gigord et al. (2002) show similar patterns for a non nectar producing orchid in Europe, but only under restricted gre enhouse conditions where flower selection by nave bees were heavily frequency dependent. When E. radicans is growing in close proximity with A. curassavica and L. camara it is more frequently visited by butterflies and therefore having higher pollinia re moval than when it is growing by itself. The presence of the other nectar producing plants is thought to be a reason for higher rates of visitation and pollinia removal. The

PAGE 5

5 orchid has been noted to possibly mimic the other two plant species in order to attract its pollinators since it does not provide a reward itself. If it imitates the other nectar reward. These ideas support the belief that E. radicans is indeed a floral mimicry complex, unlike the study done by Paulette Bierzychudek (1981) that gave insufficient evidence of such an idea. Because of this floral mimicry, the orchids are able to be pollinated more often than if they did not look so similar to A. cur assavica and L. camara When there was forty flowers in the orchid patch verses the patch that contained twenty flowers, there was not a significant difference of butterfly visitation and pollinia removal. This may have resulted due to the absence of o ther nectar producing plants. It is suggested that even though there was twice as many flowers in the patch that had forty flowers, there still was not an increase of visitation and pollinia removal because the pollinators may have noticed the absence of A. curassavica and L. camara and chose not to visit and thus, remove pollinia from the orchids. Pollinators visiting flowers lacking rewards, such as nectar, are less likely to visit an adjacent flower and more likely to leave the patch (Ackerman 1994). This proposes the idea that nectar producing plants are visited first and then the pollinators move onto another flower until it reaches an orchid at which time it will pollinate the flower. On the other hand, in a patch consisting of solely E. radicans the pollinator is less likely to visit flower after flower perhaps because the pollinator realizes the lack of nectar. On days when the study was consecutive, a pollinia count was recorded the following day. There was not a significant difference in poll inia removal, which is logical since butterflies are most active between 10:00 a.m. and 2:00 p.m. (Drummond 1976) and thus, pollination decreases. The study was conducted at 9:00 a.m. and usually finished by 1:00 p.m. This only left the hours between 1:0 0 p.m. to 8:30 a.m. the following day to allow pollinia removal. Since butterflies tend to be less active as it gets later in the day and also in the early hours of the morning, it is reasonable that the pollinia removed during those hours would be incons equential. Additionally, there was a significant difference in the visitation rates between A. curassavica and L. camara L. camara was visited more frequently every single trial period. On the study site, A. curassavica plants were taller and had large r flowers than the L. camara plants. The flowers were more numerous, causing more of an attractive appearance for pollinators. If the butterflies were more attracted to flowers, they would have visited and pollinated them more fre quently than A. curassavica One other explanation for such high rates of visitation and pollinia removal could be that L. camara has a greater nectar percentage than A. curassavica (Haber, personal communication 2004). Most importantly, there was not a significant difference of butterfly visitations to L. camara when it was with E. radicans or then when it was with E. radicans and A. curassavica Also, no significant difference was found in butterfly visitations to A. curassavica when it was with E. rad icans or when it was with both. This is crucial because it demonstrates A. curassavica and L. camara are Mllerian mimics. If there were a significant difference between A. curassavica and L. camara butterfly visitations, this would mean that the two spe cies do not mimic each other and thus, do not enhance their fitness via Mllerian mimicry. Simultaneously, the data support the idea that E.

PAGE 6

6 radicans displays Batesian mimicry because E. radicans has higher visitations in the presence of either model. The data taken throughout this study could have been more accurate if the individual flowers of A. curassavica and L. camara were counted and made equivalent to the number of E. radicans flowers. If this was done, butterfly visitation and pollinia removal co unts would not have been affected by the size or coloration differences between the different species. Another factor that may have influenced the data collected was that another study was being conducted at the same time and place as this study. Butterf lies were being collected and not released back into their habitat, which reduced the amount of butterflies that foraged on the field site. For future investigations relating to this study, it is suggested that there be more trials conducted. In addition more individual groups should be used, including additional groups of A. curassavica and L. camara by themselves. It would also be interesting to measure how much nectar is taken from each of these species and compare the success of pollination of all f our groups. ACKNOWLEDGMENTS This study would not have been possible without the permission of Licho Perez to use his land. I would Masters. I would like to like to thank Matt Gasner for all his patience and support regarding my statistics. And I cannot forget to thank Katrina Welch for her unconditional support and help throughout my entire study. Thank you to my family back home for all of your encouragement and being an ear for me to cry to during rough times. Finally, I am most grateful to my advisor, Alan Masters. Thank you so much for your patience and guidance; you truly ar e a gift. LITERATURE CITED Ackerman, J.D., J.A. Rodriguez, and E.J. Melndez.1994. A meager nectar offering by an epiphytic orchid is better than nothing. Biotropica 26 (1): 44 49. Bierzychudek, P. 1981. Asclepias Lantana and Epidendrum : A floral mimic ry complex? Biotropica 13: 54 58. Deacon, N. 2000. Pollinia removal and visitation in Epidendrum radicans (Orchidaceae) and Asclepias curassavica (Asclepiadaceae). CIEE Program. Fall 2000. Drummond, B.A. III. 1976. Comparative ecology and mimetic relatio nships of Ithomiine butterflies in Eastern Ecuador. PhD Dissertation. University of Florida. Galil, J. and M. Zeroni. 1965. Nectar system of A. curassavica Bot. Gaz. 130: 1 4. Gigord, L. et al. 2002. The potential for floral mimicry in rewardless orchid s: an experimental study. Proc. R. Soc. Lond. B 269, 1389 1395. Haber, W. 1984. Pollination by deceit in a mass flowering tropical tree Plumeria rubra (Apocynaceae). Biotropica 16 (4): 269 275 Haber, W. 2004. Personal communication. Hartshorn, G.S. 1983 Holdridge Life Zone. In D.H. Janzen. Costa Rican Natural History. University of Chicago Press.

PAGE 7

7 Roy, B.A. & Widmer, A. 1999. Floral mimicry: a fascinating yet poorly understood phenomenon. Trends Plant Sci. 4, 325 330. Todzia, C.A. 1983. Epidendrum radi cans (Bandera Espanola, Gallito). In D.H. Janzen. Costa Rican Natural History. University of Chicago Press. Wolfe, L.M. 1987. Inflorescence size and pollinaria removal in Asclepias curassavica and E. radicans Biotropica :19(1): 86 89. Woo, J. The effects of spatial distribution on the proportion of pollinia removed in Epidendrum radicans (Orchidaceae). CIEE Program. Spring 2001. Wyatt, R. 1980a. The impact of nectar robbing ants on the pollination system of Asclepias curassavica Bull. Torrey Bot. Cl. 10 7: 24 28.

PAGE 8

8 FIGURE 1. (A) Average butterfly visitation rates for E. radicans in the four different plant groups (Friedman Test, Chi square = 22.23, df = 3, p = 0.0001) (MCN Test, q = 5.9 2 E. radicans (Friedman Test, Chi A B

PAGE 9

9 FIGURE 2. (A) Average visitatio n rates for both E. radicans patches (Wilcoxon Signed Rank Test, n = 5, z = 0.137, p = 0.891). (B) Average pollinia removal rates for both patches (Wilcoxon Signed Rank Test, n = 5, z = 1.342, p = 0.18). A B

PAGE 10

10 FIGURE 3. Morning pollinia removal counts for E radicans (Friedman Test, Chi squared = 5.475, df = 3, p = 0.132) (MCN Test, q = 3.2 FIGURE 4. Average butterfly visitation rates for the two different plant species (Wilcoxon Signed Rank Test, n = 12, z = 3.065, p = 0.002). 0 5 10 15 20 25 30 35 Lantana camara Asclepias curassavica Plant Species Number of Visitations

PAGE 11

11 F IGURE 5. (A) Average number of visits to L. camara when it was with E. radicans and with A. curassavica and E. radicans (Wilcoxon Signed Rank Test, n = 12, z = 0.551, p = 0.582). (B) Average number of visits to A. curassavica when it was with E. radicans and with L. camara and E. radicans (Wilcoxon Signed Rank Test, n = 12, z = 0.707, p = 0.48). A B