Effects of Elevation on Pollinaria Removal in Asclepias curassavica Asclepiadaceae Carolyn S. Weaver Department of Biology, Rhodes College, Memphis, Tennessee, 38110 U.S.A. RESUMEN En este proyecto se estudio la remocin de polen en tres grupos de plantas Asclepias curassavica Asclepiadaceae a diferentes elevaciones 815m, 1 175m, 1440m. Adems se tomo: n mero de floraciones, nmero de flores por floracin, n mero de polen removido por flor, nmero de frutos, presencia de herbvoros, y presencia de pulgones. Los datos se colectaron por nueve das durante Octubre y Noviembre del 2000. Los resultados indican una diferencia significativa entre todas las elevaciones con respecto a nmer o de polen removido por flor ANOVA P value < 0.0001; FisherÂ€s PLSD alta y intermedia, intermedia y baja, alta y baja, todos P values < 0.0001. La remocin de polen fue mas baja en la elevacin baja, intermedia en la elevacin intermedia y ms alta en la elevacin alta. Los resultados se pueden explicar por una presencia mayor de polinizadores en la elevacin alta, especficamente la Monarca Danaus plexippus. La elevacin alta tuvo el mayor nmero de frutas por planta y ms nmero de plantas con presenci a de herbvoros. La elevacin intermedia tuvo el mayor nmero de plantas con presencia de pulgones. El nmero de flores por planta fue mayor en la elevacin intermedia y menor en la elevacin alta. De cualquier modo, la elevacin alta tuvo el mayor nmero de las flores por floracin. La elevacin es un factor importante que afecta la remocin de polen en las flores de A. curassavica. ABSTRACT Pollinaria removal was studied in three populations of Asclepias curassavica located on an elevational gradient in and around Monte verde, Costa Rica 815m, 1175m, 1440m The following plant characteristics were also studied: number of inflorescences, number of flowers per inflorescence, number of pollinaria removed per flower, nu mber of pods. Pollinaria removal in plants with the presence of herbivory was compared to pollinaria removal in plants without herbivory at all three elevations. Pollinaria removal and the presence of aphids were compared the same way as herbivory. Data we re collected over a nine day period during the misty season October Â November. Results showed a significant difference in pollinaria removal between all three elevations ANOVA P value < 0.0001. Pollinaria removal was lowest at the low elevation site, intermediate at the middle elevation site and highest at the high elevation site. These results may be explained by an increase in pollinators, specifically the Monarch butterfly Danaus plexippus as elevation increases, or the density of the plants per site. The highest numbers of pods per plant were observed at the high elevation site, which may indicate those plants also have the highest fitness. Significant differences in number of pods per plant were seen between high and low, and low and middle ele vations ANOVA P value < 0.0001; FisherÂ€s PLSD both P values < 0.0001. Neither the presence of herbivory nor presence of aphids had a significant effect on pollinaria removal. The number of flowers per inflorescence was highest at the high elevation site and lowest at the low elevation significantly affects pollinaria removal in Asclepias curassavica
INTRODUCTION Asclepias curassavica commonly referred to as milkweed, is an herbaceous species found along roadsides and in clearings throughout the tropics. It flowers and fruits throughout the year Croat 1978. Along with other members of the Asclepiadaceae family, A. curassavica contai ns alkaloids and cardenolides, heart poisons that are toxic to most vertebrate herbivores Taiz & Zeiger 1991. Due to this characteristic, A. curassavica has developed a interesting relationship with a number herbivorous insects, a classic example of whic h involves the Monarch butterfly Danaus plexippus Monarchs lay a single egg under each of several leaves of A. curassavica The larvae then feed on the leaves and accumulate the cardenolides in their bodies. Both the caterpillars and the adults are toxi c to predators Opler & Krizek 1984. If a predator such as a bird should eat a Monarch, severe gastric distress and vomiting will result Raven et al. 1986. Asclepias curassavica benefits from this relationship because the adult Monarch butterflies serve as important pollinators. The success of flowering plants is largely dependent on their relationship with pollinators. Cross pollination provides plants with an evolutionary advantage by increasing variability within a species. Thus, plants can develop a daptations or modifications enabling them to maintain fertility and possibly increase their fitness Heywood 1993. Pollination success can be inferred by measuring the amount of pollen removed in flowers. Asclepias curassavica is well suited for this type of study because its flowers contain pollen grains packed into structures called pollinaria. The pollinaria are visible to the naked eye and thus easily observed and counted Wilson & Melampy 1983. Pollinaria consist of two pollinia containing the compac ted pollen, and connected via a wishbone shaped translator. The two arms of the translator, each referred to as a retinaculum, are joined by a structure called the corpusculum Proctor et al. 1996. The corpusculum is a visible dark spot, on either side of which are cavities containing the pollinia Wilson & Melampy 1983. Each flower of A. curassavica contains five pollinaria, located in five separate stigmatic chambers. Each chamber occurs between a cucullum hood of the corona and the adjacent anther Z omlefer 1994. The gynostegium, encompassed by the stigmatic chambers, consists of five stamens surrounding an enlarged stigmatic head. Nectar is produced by secretory cells located in the lining of the stigmatic chambers Somlefer 1994. The nectar accumu lates in pouches formed by the cuculli of the corona Proctor et al. 1996. The brightly colored flowers reddish Â orange corolla and yellow gynostegium attract a variety of pollinators. The main pollinators of A. curassavica are butterflies; however, b eetles, bees, wasps and flies also frequent the flowers Wilson & Melampy 1983. When an insect visits a flower, its legs slide down the slippery stigma apex to the space between the anthers. The translator is snagged by the spurs or hairs on the insectÂ€s legs, and the entire pollinaria is dislodged when the insect departs. When the insect visits another flower and its ladened leg slips into a stigmatic chamber, the pollinaria can be deposited. However, successful cross pollination will only occur if the mo re convex side of the pollinaria is properly inserted into the chamber Zomlefer 1994. A wind dispersed
pod develops following pollination, and multiple pods can, but not always, develop on a single plant Wilson & Melampy 1983. Many studies have been d one on pollinaria removal in A. curassavica flowers. A previous study in Monteverde found a higher rate of pollinaria removal in plants from lower elevations as compared to those at higher elevations Wilson & Melampy 1983. In addition to elevation, the p resence of herbivory and aphids may also affect pollinaria removal. The purpose of this study is to determine the effect of elevation on pollinaria removal in A. curassavica In addition, the effects of elevation on other factors relating to pollinaria rem oval are also discussed. METHODS This study was conducted over a nine day period, from October 25 to November 2, 2000. The three study sites in and around Monteverde, Costa Rica, were selected on an elevational gradient. The low elevation site 815m was located in San Luis Abajo, the middle elevation site 1175m was located in San Luis Arriba, and the high elevation site 1440m was lo cated in the community of Monteverde. At each site 40 Asclepias curassavica plants were tagged and the following characteristics recorded: number of inflorescences, number of flowers per inflorescence, number of pollinaria removed per flower, and number of pods. Also, the presence or absence of herbivory and aphids were recorded for each plant. Over the length of the study, initial budding and death of flowers were also recorded. Because the flowers on each inflorescence do not bloom simultaneously, data we re collected for every flower independently. In order to monitor every flower over the length of the study, each inflorescence was drawn and each flower assigned a letter. The plant tags were oriented so that inflorescences were viewed from the same perspe ctive each time the site was visited, to reduce the probability of confusing flowers. The study sites were visited on a rotational basis over the nine day period, so that each site was visited four times. The average length of time between visitations for a given site was two days. Over the nine day study, inflorescences often died, in which case a second would almost always bloom from the same plant. When this happened, data for the second inflorescence were also recorded. All data were analyzed using ANO VA and FisherÂ€s PLSD post hoc tests. The following factors were each compared to elevation: pollinaria removal per flower, number of pods, number of flowers per inflorescence, and number of flowers per plant. For each elevation, the number of pollinaria re moved from plants on which aphids were present was compared to the number of pollinaria removed from plants without aphids. Finally, the number of pollinaria removed from plants on which herbivory was present was compared to the number of pollinaria remove d from plants without herbivory for each elevation.
RESULTS Data were recorded for 40 plants from each site and a total of 1087 flowers low N= 395, 73 inflorescences; middle N = 356, 78 inflorescences; high N = 336, 57 inflorescences. There was a signi ficant difference between all elevations regarding pollinaria removal per flower ANOVA P value < 0.0001; Fig, 1. Results showed a significant difference between high and middle, high and low, and middle and low elevations FisherÂ€s PLSD all P values < 0. 0001. There was a significant difference between the number of pods per plant and elevation ANOVA P value < 0.0001; Fig. 2. The low elevation was significantly different from both the middle and high elevations FisherÂ€s PLSD both P values < 0.0001, a nd there was not a significant difference between the high and middle elevations FisherÂ€s PLSD P value = 0.90. Twenty seven of the 40 plants in the high elevation showed the presence of some herbivory. Herbivory was present on 11 plants in the middle e levation and only one plant in the low elevation. There was not a significant difference regarding pollinaria removal between plants with herbivory and plants without herbivory ANOVA P value = 0.2232; FisherÂ€s PLSD P value = 0.6951; Fig. 3. Aphids were present on 12 plants at the middle elevation, two plants at the high elevation, and four plants at the low elevation. There was not a significant difference in pollinaria removal between plants with aphids and plants without aphids ANOVA P value = 0.1966; FisherÂ€s PLSD P = 0.7405; Fig. 4. Elevation did have a significant affect on the number of flowers per plant ANOVA P value = 0.0041; Fig. 5. The number of flowers was highest in the middle elevation, followed by the low elevation, and then the high elevation. There was not a significant difference between numbers of flowers per plant in the high and low elevations FisherÂ€s PLSD P value = 0.79. However, there were significant differences between both high and middle elevations, and middle and low el evations FisherÂ€s PLSD P value = 0.0027, P value = .0060. Elevation also had a significant effect on the number of flowers per inflorescence ANOVA P value = 0.0084, Fig. 6. Again, there was no significant difference between high and middle elevations FisherÂ€s PLSD P value = 0.74. Significant differences in number of flowers per inflorescence were seen between high and low, and middle and low elevations FisherÂ€s PLSD P value = 0.0066, P value = 0.0093. DISCUSSION The results of this study show that elevation has an effect on pollinaria removal in Asclepias curassavica Fig. 1. A possible explanation for these results may be the locations of the plants in each site. Plants in the low elevation were located in a pasture. Plants in this site also occur red at the lowest density, meaning the most area had to be covered to obtain 40 plants. Plants in the middle elevation were located by a gravel road and occurred at an intermediate density compared with the other sites. Plants in the high
elevation were lo cated in a small dense patch next to a lake. The high density of plants in an area, also known as clumping, has been suggested to affect pollinaria removal Hayon 1996. The flowers of plants found close together could be expected to receive a high number of visitations by pollinators because many flowers would be available in a small area, and the pollinators would not have far to go to reach the next flower. The flowers of plants that are less clumped could be expected to have a lower rate of visitation by pollinators because flowers would be more spread out. Clumping of plants in this study did increase from low to middle to high elevation, and although not test was done, increasing density of plants may be related to increasing polli naria removal. A previous study Hayon 1996 found there to be no significant difference in pollinaria removal in A. curassavica between single plants and plants occurring in clumped patches; however, these results may not accurately reflect trends on a la rger scale due to the small sample size used in the study N=45. Clumping affects on pollinaria removal is something to be researched in further studies. The biology of the Monarch, a major pollinator of A. curassavica, may also help explain the results of this study. Monarchs are common in all open areas on both slopes of Costa Rica, and are especially abundant at and above 1500m. Also, unlike their migratory North American relatives, Central American Monarchs are thought to remain sedentary De Vries 19 87. The high elevation site, at 1440m, was the closest site to the preferred elevation of Monarchs. Therefore, there may have been a greater abundance of these pollinators at that site. Monarchs also live in groups, and the high density of the plants at t he high elevation would provide the pollinators with many easily accessible plants. Since Monarch populations are also sedentary, flowers of A. curassavica would have a continual supply of pollinators. A previous study in Monteverde, Costa Rica also found that butterfly abundance increased with elevation Araujo 1999. Therefore, the results of the present study, which report an increase in pollinaria removal with elevation, may be due to an increase in butterfly abundance. Although the presence of aphids varied with elevation, it did not significantly affect pollinaria removal in A. curassavica Only in the middle elevation did plants without aphids have a higher number of pollinaria removals. In both the high and low elevations, plants with aphids had mor e pollinaria removed than those without aphids Fig. 4. Therefore, it cannot be concluded that presence of aphids reduces pollinaria removal in the flowers. The location of the sites may explain the differences in presence of herbivory between sites. Her bivory was present on the most plants at the high elevation. The high density at which these flowers occurred may account for these results. Because these plants were located in a dense patch, herbivores, just like pollinators, would have access to many pl ants in a small area at one time. The lowest occurrence of plant herbivory occurred in the middle elevation. These plants were located along a gravel road and occurred among many other weedy species so that A. curassavica was not the most common or most de nsely occurring plant. Because herbivory was dispersed over many different, but closely occurring species, the frequency of herbivory per plant would be reduced.
The low number of pods in the low elevation may be a reflection of the low pollinaria removal in that site. Low pollinaria removal would result in low rates of pollination, thus producing few pods. However, the high and middle elevations showed almost the same number of pods, so factors other than elevation, perhaps random ones, are affecting pod development. A low number of pods may result in reduced fitness for the plant because it will have reduced dispersal ability. The more seeds a plant can disperse the greater chance it will have for successful establishment in other, perhaps more desirable areas. A plant having several pods can produce more offspring than a plant with one or no pods. The number of flowers per plant and numbe r of flowers per inflorescence Fig. 5 & Fig. 6 did vary between elevations, but did not show the same trend as polli naria removal per flower. These results can be explained by random variations. However, previous studies have suggested that inflorescence size number of flowers per inflorescence affects the number of pollinaria removed per flower Wolfe 1987. The res ults of the present study are not consistent with those of a previous study, that found pollinaria removal per flower decreased with increasing altitude Wilson & Melampy 1983. The previous study, conducted in Monteverde during the start of the dry season January Â February, reports an average of 1.72 N=352 pollinaria removed per flower in San Luis 900m, as compared with results from the present study, which found an average of 0.54 N=395 pollinaria removed per flower. A difference in the time of y ear in which the studies were conducted may help explain the contrasting results. Wilson & Melampy suggest that climate differences between study sites may have been a major factor affecting pollinaria removal. During the transition and dry seasons, the Sa n Luis Valley experiences little wind driven precipitation and minimal cloud immersion Clark et al. 2000. In the previous study, the middle and high elevations were subjected to more mist and lower temperatures than the low elevation, which was considerably warmer and drier Wilson & Melampy 1983. The drier climate of the low elevation may have provided more favorable conditions for pollinators and less favorable conditions for potential herbivores. The present study was conducted during the misty season October Â November, and dramatic climate differences between sites were not noted. Further studies are needed to help determine whether differences between this study and the previous one are merely a result of seasonal differences or if other factors are involved. ACKNOWLEDGEMENTS I thank Mauricio Garcia for his patienc e and direction in the development and completion of this project. I would also like to thank my host family, Eduardo Leitn, Xinia Villalobos, Ral, Cindy, and Eduardito Leitn for providing both emotional and physical nourishment during this study. Thank s to my fellow peers for their moral support, especially Karen Sonnenblick. Thanks to Tim Kuhnan and Andrew Rodstrom for reading and commenting on an earlier draft of this paper, and all those at the Estacin Biolgica de Monteverde.
LITERATURE CITED Arau jo, E. 1999. Elevation and vertical feeding preferences of cloud forest butterflies. CIEE Fall: 151 162. Clark, K.L., R.O. Lawton, and P.R. Butler. 2000. The physical environment. In N.M. Nadkarni and N.T. Wheelwright Eds.. Monteverde: ecology and conser vation of a tropical cloud forest, pp. 17 20. Oxford Univ. Press, New York, NY. Croat, T.B. 1978. Flora of Barro Colorado Island, p. 716. Stanford Univ. Press, Stanford, CA. DeVries, P.J. 1987. The butterflies of Costa Rica and their natural history, pp. 2 09 212. Princeton Univ. Press, Princeton, NJ. Hayon, R. 1996. Large patch size of Asclepias curassavica as a function of favorable living conditions vs. communal living benefits. CIEE Fall: 41 50. Heywood, V.H. 1993. Flowering plants. In V.H. Heywood Ed. Flowering plants of the world, pp. 7 13. Mayflower Books, Inc., New York, NY. Opler, P.A. and G. O Krizek. 1984. Butterflies east of the great plains, pp. 193 195. The John Hopkins Univ. Press, Baltimore, MD. Proctor, M.P. Yeo, and A. Lack. 1996. The na tural history of pollination, pp. 185 186. Timber Press, Portland, OR. Raven, P.H., R. F. Evert, and S. E. Eichhorn. 1986. Biology of plants, pp. 616 618. Worth Publishers Inc., New York, NY. Taiz, L. and E. Zeiger. 1991. Plant physiology, pp. 327 328. The Benjamin/Cummings Publishing Inc., Redwood City, CA. Wilson, M. F. and M.N. Melampy. 1983. Asclepias curassavica Bailarina, Mata Caballo, Mal Casada, Milkweed. In D. H. Janzen Ed.. Costa Rican natural History pp. 191 193. Univ. of Chicago Press, Chicago, ILL. Wolfe, L. M. 1987. Inflorescence size and pollinaria removal in Asclepias curassavica and Epidendrum radicans Biotropica 191: 86 89. Zomlefer, W. H. 1994. Guide to flowering plant families, pp. 239 243. Univ. of North Carolina Press, Chape l Hill, NC.