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 M39-00314
Estmulo supernormal como una estrategia de imitacin en el caso de Epidendrum radicans (Orchidaceae)
Supernormal stimulus as a mimicry strategy the case for Epidendrum radicans (Orchidaceae)
Epidendrum radicans is a food deceptive Batesian mimic of its sympatric model species Asclepias curassavica and Lantana camara. Theoretically, food deceptive orchids should be rare with small inflorescences (Johnson et al. 1993; Weins 1978) yet in San Luis, Costa Rica E. radicans grows in large monotypic stands with individuals sporting up to 12 open flowers per inflorescence. E. radicans might attract pollinators using flower or inflorescence size as a visual supernormal stimulus also implicated in floral mimicry (Scheistl 2004). Pollinia removal of E. radicans was measured in plants with i)
inflorescence sizes of two and ten growing amongst model species in patches ii) inflorescence sizes of one through eight flowers on plants growing naturally in dense stands without models and iii) unmodified and
enlarged flowers. Pollinia removal was proportionally greater for flowers of smaller inflorescence sizes, and greater for unmodified flowers. E. radicans does not appear to use visual supernormal stimuli to attract pollinators. E. radicans may occur in large monotypic stands as a result of human disturbance. Continued pollinia removal in E. radicans may result from constant recruitment of nave pollinators to the area.
Epidendrum radicans es una planta que presenta mimetismo Batesiano de las especies modelos Asclepias curassavica y Lantana camara. Tericamente, el engao alimenticio en las orqudeas es poco comn y con las inflorescencias pequeas (Johnson et al. 1993; Weins 1978), an en San Luis, Costa Rica E. radicans crece en largos tallos monotpicos con individuos que contienen hasta 12 flores por inflorescencia. E. radicans puede atraer a los polinizadores usando el tamao de las flores o inflorescencias como un estmulo supernormal tambin implicado en el mimetismo floral (Scheistl 2004). La remocin de polinia de E. radicans fue medido en las plantas con i) tamao de inflorescencias de dos a diez flores entre especies modelo en parches ii) tamao de inflorescencias de uno a ocho flores en plantas creciendo naturalmente en parches sin modelos iii) y flores alargadas y no modificadas. La remocin de polinia es proporcionalmente mayor para las flores con un menor nmero de flores por inflorescencia y mayor para flores no modificadas. E.radicans puede ocurrir en parches largos monotpicos como resultado de los disturbios humanos. La remocin continua en E. radicans puede resultar por el constante reclutamiento de los polinizadores inexpertos en el rea.
Text in English.
Pollination by animals
Costa Rica--Puntarenas--Monteverde Zone--San Luis
Polinizado por animales
Costa Rica--Puntarenas--Zona de Monteverde--San Luis
Tropical Ecology Spring 2010
Ecologa Tropical Primavera 2010
t Monteverde Institute : Tropical Ecology
Supernormal Stimulus as a Mimicry Strategy: the case for Epidendrum radicans (Orchidaceae) Daniel Paul Department of Biology, Allegheny College, Meadville, Pennsylvania, U.S.A. ABSTRACT Epidendrum radicans is a food deceptive Batesian mimic of its sy mpatric model species Asclepias curassavica and Lantana camara. Theoretically, food deceptive orchids should be rare with small inflorescences (Johnson et al. 1993; Weins 1978) yet in San Luis, Costa Rica E. radicans grows in large monotypic stands with i ndividuals sporting up to 12 open flowers per inflorescence. E. radicans might attract pollinators using flower or inflorescence size as a visual supernormal stimulus also implicated in floral mimicry (Scheistl 2004). Pollinia removal of E. radicans was m easured in plants with i) inflorescence sizes of two and ten growing amongst model species in patches ii) inflorescence sizes of one through eight flowers on plants growing naturally in dense stands without models iii) and unmodified and enlarged flowers. Pollinia removal was proportionally greater for flowers of smaller inflorescence sizes, and greater for unmodified flowers. E. radicans does not appear to use visual supernormal stimuli to attract pollinators. E. radicans may occur in large monotypic stan ds as a result of human disturbance. Continued pollinia removal in E. radicans may result from constant recruitment of nave pollinators to the area. RESUMEN Epidendrum radicans es una planta que presenta mimetismo Batesiano de las especies modelos Asclepias curassavica y Lantana camara Tericamente, el engao alimenticio en orqudeas es poco comn y con inflorescencias pequeas (Johnson et al. 1993; Weins 1978), an en San Luis, Costa Rica E. radicans crece en largos tallos monotpicos con indivi duos que contienen hasta 12 flores por inflorescencia. E. radicans puede atraer polinizadores usando el tamao de flores o inflorescencias como un estmulo supernormal tambin implicado en mimetismo floral (Scheistl 2004). La remocin de polinia de E. ra dicans fue medido en plantas con i) tamao de inflorescencias de dos a diez flores entre especies modelo en parches ii) tamao de inflorescencias de uno a ocho flores en plantas creciendo naturalmente en parches sin modelos iii) y flores alargadas y no mod ificadas. La remocin de polinia es proporcionalmente mayor para flores con un menor nmero de flores por inflorescencia y mayor para flores no modificadas. E.radicans puede ocurrir en parches largos monotpicos como resultado de disturbios humanos. La remocin continua en E. radicans puede resultar por el constante reclutamiento de polinizadores inexpertos en el rea. INTRODUCTION Animal mediated pollination relies on floral attractants coupled with a reward ( Johnson, S.D. et al. 2003 ). However, the production of these rewards, which can include nectar, pollen, and oil, is energetically costly (Dafni 1984). Some plants avoid making expensive rewards and, instead, are pollinated by deception (Dafni 1984). In this case, pollination occurs when pollina tors are duped by floral attractants mimicking rewarding species (food deception) or flowers mimicking potential mates (sexual deception) (Dafni 1984; Scheistl 2005).
One third of all orchid species are food deceptive (Sheistl 2005). Orchid flowers c an fool pollinators by having similar scents, floral shapes or colors of sympatric rewarding plants (Johnson 1994; Weins 1978). Such deception saves energy and increases outcrossing ( Jersakova et al. 2006) however food deceptive orchids face the challeng e of continually attracting pollinators that may lose interest in or learn to avoid deceptive flowers (Gumbert & Kunze 2001; Smithson and MacNair 1997). Experienced Apis mellifera bees foraging on the food deceptive orchid Orchis boryi visited the orchid less often than nave workers (Gumbert & Kunze 2001). It is important, therefore, to ensure the continued pollination of food deceptive orchids by maximizing pollinator attraction while minimizing avoidance learning (Dafni 1984). In short, orchid deceit pollination relies upon the orchid being relatively rare but in close proximity to its model Johnston 1998; Sheistl 2005; Weins 1978). Epidendrum radicans is a food deceptiv e Batesian mimic of the two sympatric, rewarding model species Asclepias currasavica (Asclepiadaceae) and Lantana camara (Verbenaceae) (Roy & Widmer, 1999; Weins, 1978) All three have overlapping geographical and ecological ranges, are pollinated by but terflies, and have similarly colored orange and yellow flowers on inflorescences (Figure 1: Wiens 1978). However, E. radicans is not rare, often growing in large monotypic stands of thousands away from models. Despite this, E. radicans in these population s are continually pollinated as indicated by pollinia removal and capsule formation (Bierzychudek, 1981 ). The puzzling growth habits of E. radicans can possibly be explained by pollinator responses to a supernormal stimulus (Scheistl 2004). A supernorm al stimulus is an exaggerated character that is met with a correspondingly exaggerated response (Christy 1995). Supernormal stimuli in floral mimics exaggerate attractants used by the model (Scheistl 2004). For example, Chilogottis trapeziformis is a sex ually deceptive orchid that produces up to 100 times more pheromone and has a flower larger than the females of its model wasp species Neozeleboria cryptoides (Scheistl 2004). While supernormal stimuli are known in sexually deceptive orchids, it has not b een studied in food deceptive orchids. E. radicans does not produce scents and its floral coloration is similar to its models. Nonetheless, its floral display could be a visual supernormal stimulus as its individual flowers and inflorescences are both la rger than the model species. Large inflorescences correlate with higher pollinia removal and seed pod production in the orchid Brassavola nodosa (Schemske 1980) and could act as super attractants that compensate for or even override avoidance learning. Here, I determine if pollination success of E. radicans is impacted by a visual supernormal stimulus, and whether inflorescence size or petal size is the primary contributing factor to the stimulus. To do this I measure pollinia removal of E. radicans wit h i) artificially enlarged flowers ii) inflorescence sizes of ten flowers versus two flowers for plants with models nearby and iii) inflorescence sizes ranging from one flower to eight flowers for plants in large monotypic stands.
Materials and Methods STUDY SITE All observations for the experiment were made in the San Luis Valley, Costa Rica (Premontane Moist Forest) from April 4 th to April 29 th E. radicans were found in stands along the La Trocha hillside, while A. curassavica and L. camara plants were in disturbed areas throughout the San Luis Valley (Figure 1). STUDY ORGANISMS A B C Figure 2. (A) E. radicans is a food deceptive Batesian mimic of its two rewarding model species that have similar floral coloration (B) A. curassavica and (C) L. camara Figure 1. The experimental patches of E. radicans and its model species were located ar ound six sites within the San Luis Valley. The dense stand of E. radicans plants was located along the road leading into the valley called La Trocha.
E. radicans is a common orchid found in disturbed areas around the Monteverde region. It flowers year round and produces infloresce nces of one to twelve flowers (Bierzychudek 1981). E. radicans has similar habitats as its model species A. currassavica and L. camara but only rarely do they occur in close proximity to one another in San Luis. All three plants prefer areas of human dis turbance such as roadsides and pastures (Bierzychudek, 1981). The three species also share common pollinators ( Bierzychudek, 1981; Boyden 1980). Butterflies including Anartia fatima and Danaus plexippus are the most common pollinators. While bees and fl ies visit the flowers also, pollinia are only known to be transported by butterflies (Bierzychudek, 1981). EFFECTS OF INFLORESC ENCE SIZE ON POLLINI A REMOVAL OF E. RADICANS WITH MODELS Twenty patches of either L. camara or A. curass avica (defined as all model plants within 3 m 2 of each other) were used to measure the effects of inflorescence size on pollinia removal of E. radicans (Figure 1) Twenty seven E. radicans plants were placed amongst the patches in densities of one to four plants, and pollinia removal measured daily for sixteen days. E. radicans were grown with models to maximize pollinia removal according to the magnet species effect described by Johnson et al. 2003. The inflorescence size of all E. radicans was either t wo flowers per inflorescence or ten flowers after the fifth, tenth, and thirteenth days of observation to account for effects of patch placement. Though plants with ten flowers per inflorescence occur naturally, they are rare. Therefore, plants for the i nflorescence size of ten group were created by tying together inflorescences of multiple plants. After a plant was pollinated, it was replaced with another plant of the same inflorescence size with intact pollinia. EFFECTS OF INFLORESC ENCE SIZE ON POLLIN IA REMOVAL OF E. RADICANS WITHOUT MODELS Eighty E. radicans growing in large roadside monotypic stands were used for twelve days to measure the effects of inflorescence size on pollinia removal without models. Here, a range of inflo rescence sizes ten plants for each inflorescence size of one through eight were observed and followed with replacement. Pollinia removal was measured daily. EFFECTS OF A SUPERNO RMAL FLOWER SIZE ON POLLINIA REMOVAL OF E. RADICANS Ten patches of Asclepias and E. radicans plants of an inflorescence size of two were created by placing one E. radican plant within each Asclepias patch. Pollinia removal of each E. radicans was measured daily for ten days. Plants with pollinia removed were replaced th e same day. The flowers of E. radicans in five patches were modified by enlarging the flowers with paper. A yellow paper square 1 cm x 1 cm was taped to the center of an orange paper square 2.5 cm x 2.5 cm. This square was fitted behind each flower and t aped to the top sepal of each flower of the E. radicans plant. The E. radicans flowers of the other five patches were not modified. Each patch was switched between
the unmodified and modified flower arrangement after five days to account for local effects on pollinia removal of each patch. RESULTS The number of pollinia removed/plant was greater for the larger inflorescence size of ten flowers compared to plants with two flowers per inflorescence (Figure 3: Wilcoxon Signed Rank Test, Z = 3.425, P = 0.0006, n = 20). However, as inflorescence size increased, pollinia removal/flower decreased (Figure 3: Wilcoxon Signed Rank Test, Z = 3.347, P = 0.0008, n = 20). A B Figure 4. Pollinia rem oval over twelve days of E. radicans plants with inflorescence sizes ranging from one to eight flowers per inflorescence growing in dense roadside stands in San Luis (n = 10 and 12 for the inflorescence size of one) (A) Number of pollinia removed is a rati o of the total number of pollinia removed to the total plants for each treatment. (B) Pollinia Removed/Flower is the ratio of total pollinia removed to the flowers offered by all plants of each treatment. A B Figure 3. Mean (SD, n = 20) of the occurrences of pollinia removal over sixteen days in E. radicans plants growing amongst its model species in twenty patches. X axis bar labels represent inflorescence (A) Number of pollinia removed/plant observed in each patch is the average across twenty patches of the ratio between total occurren ces of pollinia removed and the number of E. radicans plants in each patch (B) Number of pollinia removed/flower is the average across twenty patches of the ratio between total occurrences of pollinia removal and total flowers in each patch.
For E. radicans growing in dense stands, the number of pollinia removed per plant increases as inflorescence size increases (Figure 4: Speaman Rank Test, Spearman Rho = 0.7904, P = 0.0195, n = 10, 12 for the inflorescence size of one). However, as inflorescence size increases in E. radicans, pollinia removal per flower decreases (Figure 4: Speaman Rank Test, Spearman Rho = 0.7381, P = 0.0366, n = 10, 12 for the inflorescence size of one). For E. radicans growing in patches amongst its model species, unmodified control flowers experi enced 24% greater total pollinia removal compared to the artificially enlarged treatment flowers, although the difference was not statistically significant ( 2 = 1.190, P = 0.275). Pollinia removal often occurred in bursts from one pollinator at a time fo r E. radicans in the monotypic stands. Flowers of a small area (3 m 2 ) remained unpollinated for days until all flowers were recorded as pollinated in a single morning (pers. obs.). DISCUSSION Food deceptive flowers grow in low densities in close proxim ity to their model species and have small inflorescence sizes to maximize pollinator attraction while minimizing avoidance learning (Dafni 1984; Scheistl 2005). Weins 1978 studied the pollination success of the food deceptive orchid Cephalanthera, which u sually grows in low densities compared to its model the bellflower. Pollination success is higher when Cephalanthera grows along side its model compared to growing alone (Weins 1978). Figure 5. Total pollinia removal over ten days in E. radicans plants growing alongside A. curassavica in ten patches (n=5: 5 patches of unmodified flowers and 5 patches of enlarged flowers). Both unmodified and enlarged flower sizes had an inflorescence size of 2. Numbers above the bars represent the occurrences of pollinia removal as a percentage of the total pollinia removed. 62 38
However, E. radicans does not exhibit mimic growth patterns possibl y because its flower or inflorescence size is a supernormal stimulus (Scheistl 2004). Artificially large flowers are visited by pollinators less frequently compared to natural flowers. This indicates that larger flower size is not a supernormal stimulus i n E. radicans A small flower size possibly influences pollinator attraction because it minimizes avoidance learning. Deceptive plants have increased pollinator visitation when they have flowers closely resembling model species (Gumbert & Kunze 2001; Johns ton et al. 2003). Pollinators may not recognize the artificially enlarged flower and preferentially visit the flowers of the model species or the unmodified flowers in nearby patches. The primary pollinator of E. radicans Danaus plexippus has evolved very closely with A. currassavica (DeVries 1987; Goode 1999) and may only respond to deceptive flowers of a size similar to Asclepias Also, when artificially large flowers are visited, pollinators may learn to avoid those flowers faster because of their distinctiveness. The increase in pollinia removal as inflorescence size increases for E. radicans in monotypic stands and near its models indicates that more flowers increase the likelihood that any single pollinium is removed from a plant simply becaus e more pollinia are available to be removed. However, as inflorescence size increases, pollinator visits decrease per flower. Therefore, plants with larger inflorescences receiving more visits per plant are not receiving disproportionately high visitation s, suggesting that inflorescence size is not a supernormal stimulus utilized by E. radicans Increasing the number of flowers produced increases the likelihood of pollination, however, each flower has a smaller chance of being pollinated. Also, negative c onditioning of pollinators visiting plants with large inflorescences may decrease pollinator visitation for plants with large inflorescences. Rather, visitation per flower increases disproportionately as inflorescence size decreases. Therefore, having sm all inflorescence sizes maximizes pollination and is consistent with plant mimicry growth strategies (Weins et al. 1978). Despite this, sexually reproducing E. radicans grow in large stands, often with large inflorescence sizes. Because pollinators learn to avoid mimics the pollination success of food deceptive flowers relies on nave pollinators (Boyden 1982; Johnson 2000). Flower production of the food deceptive orchid Calypso bulbosa var. americana is timed closely with the emergence of its queen bumble bee pollinator in late spring (Boyden 1982). Pollination relies upon the initial deception of the pollinator because after time, the pollinator learns avoidance of the flower (Boyden 1982). While the pollinator avoids the mimic eventually, plant populati ons are supported by the limited occurrences of pollinations during the late Spring season. E. radicans flowers and pollinators are present year round, peaking in the dry season, suggesting that pollinators of E. radicans are not nave, however, recruitme nt of new nave pollinators to the area is most likely high. The two main pollinators of E. radicans are Anartia fatima the most common butterfly in Costa Rica, and Danaus plexippus (Bierzychudek 1981; DeVries 1987). The respective life spans of the rep roductively active adult A. fatima butterfly is two weeks and two to six weeks for D. plexippus (DeVries 1987) meaning there is a high recruitment of young nave pollinators year round to the area. Also, 80% of Pacific lowland butterfly species in Costa R ica are migratory (Stevenson and Haber 2000). Butterfly populations consisting of first generation offspring migrate to the Pacific slope
2000) contributing more naiv e pollinators to the area. Therefore, there is constantly recruitment of nave pollinators to the study site that most likely contribute to the continual pollination of E. radicans in stands. All data suggest that E. radicans should utilize typical mimic growth forms that maximize pollinator visitation while minimizing avoidance learning sparse densities, small flowers, and small inflorescence sizes (Weins 1978). Historically, E. radicans followed this pattern and did not grow in dense stands. It is l ikely that the state of E. radicans scale human disturbance, E. radicans most likely quickly colonized areas to form dense stands. Human disturbance also li kely increased the prevalence of Anartia fatima and Danaus plexippus, both are adapted to disturbed habitats. Further as E. radicans the models A. currassavica and L. camara However, E. radicans likel y increased disproportionately compared to its models because of its ability to regenerate from fragments. Continuous chopping of the plants increases E. radicans density dramatically and decreases the density of its models, which the dense monotypic stan ds. Despite this, E. radicans populations still prevail because of a human induced increase in optimal growth habitat and nave pollinators. Visual supernormal stimuli are not utilized by E. radicans as a strategy to attract pollinators. Rather, their p atterns of pollination support typical mimic growth strategies. Because selective pressures acting on the dense E. radicans stands most likely differ from those for E. radicans growing more sporadically in less disturbed habitats, studies should compare ho w E. radicans fitness has changed. For instance, a primary benefit to being a food deceptive mimic is increased outcrossing pollinators quickly detect a plant is not rewarding and move to a new area ( Jersakova et al. 2006). However the observed pattern of pollinia removal in the stands decreased out crossing because all flowers within a small area of a stand were pollinated by a single pollinator, causing a high degree of selfing amongst plants with inflorescence sizes above one. While abun dance has changed relative to its models and pollinators because of human disturbance it still maintains dense monotypic populations that sexually reproduce because of a high continuous recruitment of nave pollinators. ACKNOWLEDGEMENTS I thank CIEE T ropical Ecology and Conservation and Alan for their invitation into their study abroad program. I also thank CIEE for providing materials for the project, Anjali Kumar, Pablo Allen, Jos Caldern, and Yimen Araya, for project input and help with statistic al tests. I would also like to thank Alan Masters especially for being my project advisor. Thanks also to the many students of the program for providing help with writing the manuscript and with tips on formatting for Word. LITERATURE CITED Bierzych udek, P. 1981. Asclepias, Lantana, and Epidendrum: A Floral Mimicry Complex? Biotropica 13: 54 58. Boyden, T. 1980. Floral Mimicry by Epidendrum ibaguense (Orchidaceae) in Panama. Evolution 34: 135 136.
-------------1982. The Pollination Biology of Calypso bulbosa var. Americana (Orchidaceae): Initial Deception of Bumblebee Visitors. Oecologia 55: 178 184. Christy, J. 1995. Mimicry, Mate Choice, and the Sensory Trap Hypothesis. The American Naturalist 146: 171 181. Dafni, A. 1984. Mimicry and deception in pollination. Annu Rev Ecol Syst 15: 259 278. DeVries, P. 1987. The Butterflies of Costa Rica, pp. 178 179, 212. Princeton University Press, Prinston. Goode, M. 1999. An Introduction to Costa Rican Butterflies. Pp. 25 26. Mark Richa rd Good, San Jos. Gumbert, A., Kunze, J. 2001. Color similarity to rewarding model plants affects pollination in a food deceptive orchid, Orchis boryi. Biological Journal of the Linnean Society 72: 419 433. Jersakova, J., Johnson, S.D., Kindlmann, P. 20 06. Mechanisms and evolution of deceptive pollination in orchids. Biological Reviews 81: 219 235. Johnson, S.D., 2000. Batesian mimicry in the non rewarding orchid Disa pulchra, and its consequences for pollinator behaviour. Biological Journal of the L innean Society 71: 119 132. Johnson, S.D., Craig, I. Peter, L. Nilsson, A., Argen, J. 2003. Pollination Success in a Deceptive Orchid Is Enhanced by Co Occurring Rewarding Magnet Plants. Ecology, 84: 2919 2927. Roy, B.,Widmer, A. 1999. Floral mimicr y: a fascinating yet poorly understood phenomenon. Trends in Plant Science Perspectives 4: 1360 1385. Scheistl F.P. 2004. Floral evolution and pollinator mate choice in a sexually deceptive orchid. J. Evol. Biol. 17: 67 75. ------------. 2005. On t he success of a swindle: pollination by deception in orchids. Naturwissenschaften 92: 255 264. Schemske, S.W., 1980. Evolution of Floral Display in the Orchid Brassavola nodosa Evolution 34: 489 493. Smithson, A., MacNair, M.R. 1997. Negative Freq uency Dependent Selection by Pollinators on Artificial Flowers Without Rewards. Evolution 51: 715 723. Stevenson, R, Haber, W. 2000. Migration of Butterflies through Monteverde In Nadkarni, N. M., Wheelwright N. T. (Eds.). Monteverde: The Ecology and C onservation of a Tropical Cloud Forest, pp 118 119. Oxford University Press, New Yourk, New York Wiens, E. 1978. Mimicry in plants. Evol Biol 11: 365 403.