Citation
Adaptive radiation of oil-reward compounds among neotropical orchid species (Oncidiinae)

Material Information

Title:
Adaptive radiation of oil-reward compounds among neotropical orchid species (Oncidiinae)
Creator:
Silvera, Katia
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Bees ( jstor )
Biochemistry ( jstor )
Flowers ( jstor )
Ions ( jstor )
Lipids ( jstor )
Pollination ( jstor )
Pollinators ( jstor )
Pollinia ( jstor )
Principal components analysis ( jstor )
Species ( jstor )
Botany thesis, M.S ( lcsh )
Dissertations, Academic -- Botany -- UF ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent) ( marcgt )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
ABSTRACT: Oil-secreting flowers are found throughout the angiosperms and the production of floral lipid rewards is a polyphyletic trait. In this study, I explore variation in floral-oil composition among neotropical orchids of the subtribe Oncidiinae and a malpighiaceous species (Byrsonima crassifolia). Many Oncidium species are thought to be mimics of malpighiaceous species because members of both groups flower simultaneously and have structurally and functionally similar inflorescences, thus attracting and sharing pollinators. The chemistry of floral-oil secretions in these groups is compared and insight into the radiation of oil-related compounds and their relationship with oil-collecting bees is provided. Floral oils were analyzed by gas chromatography/mass spectrometry (gc/ms). Mass spectra of significant peaks were compared among species of the same genus and among different orchid genera. I provide a major survey of oil components with over a hundred different major compounds for members of the Oncidiinae, a subtribe with the most diverse type of pollination systems found in the Orchidaceae. The results show both quantitative and qualitative differences among orchid genera, but mainly quantitative differences within genera. There is biochemical convergence between oil-reward compounds among members of the Malpighiaceae and Orchidaceae. Müllerian mimicry between these groups seems to best describe this chemical convergence. The variability in oil chemistry found among different genera of orchids can be used to understand the exploitation of oil resources by oil-collecting bees and could potentially explain the high degree of pollinator specificity for many tropical orchid species.
Thesis:
Thesis (M.S.)--University of Florida, 2002.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
General Note:
Title from title page of source document.
General Note:
Includes vita.
Statement of Responsibility:
by Katia Silvera.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Silvera, Katia. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/1/2003
Resource Identifier:
029835446 ( ALEPH )
53465517 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

ADAPTIVE RADIATION OF OIL-REWARD COMPOUNDS AMONG NEOTROPICAL ORCHID SPECIES (ONCIDIINAE) By KATIA SILVERA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2002

PAGE 2

Copyright 2002 by KATIA SILVERA

PAGE 3

I dedicate this thesis to my husband Lou Santiago and to my parents Gaspar Silvera and Flor Silvera for their endless support and encouragement of my goals, and especially for caring about my personal growth.

PAGE 4

ACKNOWLEDGMENTS Many people were involved in different stages of this thesis and I am very grateful for their support and help. First, I thank my committee chair, Norris H. Williams, for his continuing support, encouragement and dedication to this project and most importantly for supporting Hispanic students and orchid research in Latin American. I also want to thank the committee members, W. Mark Whitten, Walter S. Judd and Stephen S. Mulkey, for their comments and leading discussions to the ideas presented in this thesis. Special thanks go to Bill Wcislo, who stimulated many new ideas through his challenging discussions and provided valuable lab and field equipment necessary for conducting the pollination and floral visitation section, and to David Roubik, who kindly identified all bee specimens used in my study. I thank Robert L. Dressler for providing several specimens of Paratetrapedia calcarata and arranging my first visit to the Herbarium at the University of Florida. I want to thank Lou Santiago, who substantially helped with statistical analysis. Thanks go to the Core Laboratory and Nancy Swabo, who conducted the gc/ms analysis, the American Orchid Society (AOS) 11 TH World Conference Scholarship Fund for funding this research, and to the Smithsonian Tropical Research Institute (STRI) for lab space and partial funding to conduct field work in Panama. Special thanks are due to Mark Whitten for his endless help in interpreting and analyzing mass spectra and providing interesting discussions about oil-reward chemistry. iv

PAGE 5

I thank Marc Frank, who took special care of my orchids while in the greenhouse, and the staff of the Herbarium at the University of Florida and the Botany Department for logistical support. Finally, I want to thank my lab mates for experiences and endless support. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT.........................................................................................................................x CHAPTER 1 INTRODUCTION...........................................................................................................1 Literature Review............................................................................................................1 Elaiophore as an Oil-Secreting Organ............................................................................2 Objectives of This Study.................................................................................................4 2 CHEMISTRY OF OIL-REWARD COMPOUNDS........................................................6 Introduction.....................................................................................................................6 Literature Review.....................................................................................................6 Floral Mimicry in Oil-Secreting Flowers.................................................................7 Chemistry of Floral Lipids in Oncidium..................................................................8 Material and Methods.....................................................................................................8 Study Species and Study Sites.................................................................................8 Chemical Analysis...................................................................................................9 Data Analysis.........................................................................................................11 Results...........................................................................................................................12 Chemistry of Oil-Flowers......................................................................................12 Principal Component Analysis (PCA)...................................................................13 Hierarchical Cluster Analysis (Cluster Tree).........................................................15 Discussion.....................................................................................................................17 Biochemical Convergence of Orchids and Malpighiaceous species......................17 Analysis of Oil-Related Compounds Using gc/ms and Multivariate Methods......19 3 FLORAL VISITATION AND POLLINATION OF ORCHIDS BY OIL-COLLECTING BEES...................................................................................................28 Introduction...................................................................................................................28 Literature Review...................................................................................................28 vi

PAGE 7

Floral Visitation and Bee-Behavior in Oil-Secreting Flowers...............................29 Material and Methods...................................................................................................30 Study Site and Oil-Flower Visitation.....................................................................30 Flower Temperature...............................................................................................31 Oncidium cheirophorum Flower Visitation...........................................................31 Fruit Production of Malpighiaceous Mimic Orchid Species..................................32 Results...........................................................................................................................32 Byrsonima crassifolia and Trichocentrum stipitatum Floral Visitation.................32 Flower Temperature of Oil-Flowers......................................................................35 Oncidium cheirophorum Floral Visitation.............................................................36 Female Reproductive Success in Oil-Secreting Orchids.......................................36 Discussion.....................................................................................................................37 Pollination of Byrsonima crassifolia and Trichocentrum stipitatum.....................37 Capsule Production and Reproductive Success in Oil-Secreting Flowers.............40 4 CONCLUSION..............................................................................................................45 APPENDIX A LIST OF COMPOUNDS AND ION ABUNDANCE PER MASS OF FRESH FLOWER.......................................................................................................................48 B LOCALITIES AND VOUCHER NUMBERS FOR PLANT TAXA...........................53 C LOCALITIES AND VOUCHER NUMBERS FOR BEE TAXA................................55 LIST OF REFENCES........................................................................................................56 BIOGRAPHICAL SKETCH.............................................................................................60 vii

PAGE 8

viii LIST OF TABLES Table page 2-1 List of species used in this st udy with references to their pollinator..........................10 2-2 Quantity of oil-related compounds given as ion abundance per unit of fresh flower mass in various oil-secreting species..........................................................................14 3-1 List of pollinators and visitors for species of oil-secreting flowers of Byrsonima crassifolia and Trichocentrum stipitatum ...................................................................34

PAGE 9

LIST OF FIGURES Figure page 2-1 Map of the Republic of Panama with collection localities.........................................22 2-2 Mass spectrum of nonacosane. Top panel represents the mass spectrum found in the sample.........................................................................................................................23 2-3 Scree plot of Eigenvalues and number of factors.......................................................24 2-4 Principal Component Analysis (PCA) performed on 130 different oil-related compounds and 16 variables......................................................................................25 2-5 Cluster tree of oil-related compounds........................................................................26 2-6 Cluster tree of oil-related compounds correlated with pollinator diversity................27 3-1 Flower temperature of two species of oil-secreting flowers measured during three consecutive days.........................................................................................................42 3-2 Paratetrapedia sp. gathering floral lipids from flowers of Oncidium cheirophorum in Fortuna, Panama.........................................................................................................43 3-3 Female reproductive success (capsule formation) of three orchid species that mimic members of the Malpighiaceae...................................................................................43 4-1 Single shortest tree for the Subtribe Oncidiinae from successively weighted analysis of combined data set (ITS/matK/trnL-F, Williams et al. 2001).................................47 ix

PAGE 10

x Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ADAPTIVE RADIATION OF OIL-REWARD COMPOUNDS AMONG NEOTROPICAL ORCHID SPECIES (ONCIDIINAE) By Katia Silvera December 2002 Chair: Norris H. Williams Major Department: Botany Oil-secreting flowers are found throughout the angiosperms and the production of floral lipid rewards is a polyphyletic trait. In this study, I explore variation in floral-oil composition among neotropical orchids of the subtribe Oncidiinae and a malpighiaceous species ( Byrsonima crassifolia ). Many Oncidium species are thought to be mimics of malpighiaceous species because members of both groups flower simultaneously and have structurally and functionally similar inflorescences, thus attracting and sharing pollinators. The chemistry of floral-oil secretions in these groups is compared and insight into the radiation of oil-related compounds and their relationship with oil-collecting bees is provided. Floral oils were analyzed by gas chromatography/mass spectrometry (gc/ms). Mass spectra of significant peaks were compared among species of the same genus and among different orchid genera. I provide a major survey of oil components with over a hundred different major compounds for members of the Oncidiinae, a subtribe with the most diverse type of pollination systems found in the Orchidaceae. The results show both

PAGE 11

quantitative and qualitative differences among orchid genera, but mainly quantitative differences within genera. There is biochemical convergence between oil-reward compounds among members of the Malpighiaceae and Orchidaceae. Mllerian mimicry between these groups seems to best describe this chemical convergence. The variability in oil chemistry found among different genera of orchids can be used to understand the exploitation of oil resources by oil-collecting bees and could potentially explain the high degree of pollinator specificity for many tropical orchid species. xi

PAGE 12

CHAPTER 1 INTRODUCTION Literature Review The discovery of pollination of orchids by oil-collecting bees opened a new area of research in pollination ecology (Vogel 1969, 1974; Buchmann 1987; Roubik 1989; Steiner 1989, 1991; Wcislo and Cane 1996). Oil-secretion by flowers among angiosperms has been proposed to be a highly successful evolutionary strategy to enhance pollination (Buchmann 1987), yet data on the evolution of oil flowers and its relationship with pollinators are still needed in order to understand radiation of the oil-pollination syndrome. The oil-pollination system was first reported by Stefan Vogel in 1969 and first presented at the International Botanical Congress in Seattle in the same year (“Flowers Offering Fatty Oils Instead of Nectar”). He studied this system in 1964 while conducting research in Argentina, observing species of Angelonia (Scrophulariaceae s.l.) in the field. In the oil-pollination system, oil-collecting bees gather floral lipids from oil-secreting flowers. The majority of oil-producing species rely exclusively on oil-collecting bees for pollination. Female bees collect these oils with specialized structures on their forelegs, absorb the oils by capillary action, and carry the oils in the scopae of their hind legs. It is important to note that the term floral oils is used to describe non-volatile oils rather than volatile essential oils (mainly terpenes) used as sexual or odor attractants in many insect-visited flowers. 1

PAGE 13

2 Vogel (1974) published the most detailed account on this syndrome, and included descriptions of the oil-secreting glands and a detailed table of all families and genera possessing this syndrome. Flower structures and pollination behavior of 24 plant species were explained in detailed for the first time. This publication became the framework for many other studies and the beginning of a new area of research in pollination ecology. Simpson, Neff and Seigler (1977), Simpson and Neff (1981), and Buchmann (1987) discussed this system in more detail. It is now known that the oil-pollination system is predominantly found in the neotropics, mainly within six families: Krameriaceae, Iridaceae, Scrophulariaceae s.s., Calceolariaceae, Malpighiaceae, and Orchidaceae. Flowers offering floral lipids are found in 8 orders, 11 families, 79 genera and 2402 species of flowering plants worldwide, but these numbers are increasing as we learn more about this system (Buchmann 1987). Gesneriaceae, Melastomataceae, Solanaceae, and Primulaceae contain oil-flower species but their pollination is unknown. Orchids that produce oils are well represented in the neotropics. Their flowers produce floral lipids in the form of oils that (in most cases) are secreted from structures called elaiophores and are visited exclusively by bees (Vogel 1969; Buchmann 1987). These floral lipids may serve as pollinator larval supplements since the energetic value of lipids is higher than that of nectar and pollen (Simpson et al. 1977; Roubik 1989; Wcislo and Cane 1996). Adult bees are not known to ingest lipids, but specialized bees can utilize these oils as nesting material (water-resistant cell linings) or accessory pollenkit (Buchmann 1987; Roubik 1989). Elaiophore as an Oil-Secreting Organ Vogel (1974) was the first to describe the different types of oil-secreting organs or “elaiophores” present in angiosperms. He recognized two different types of elaiophores:

PAGE 14

3 trichomal and epithelial elaiophores. Trichomal elaiophores are composed of glandular hairs or outgrowths from the epidermis, which consist of up to 50,000 glandular hairs that constantly secrete oils. There can be up to 3 mg of oil per individual flower, and oils can be secreted and accumulated directly from the trichomes, or it can be stored below the cuticle until removed by combs in the forelegs of female bees (Vogel 1974). Trichomal glandular hairs have been reported in the following families: Cucurbitaceae, Iridaceae, Primulaceae, Scrophulariaceae (s.s), Calceolariaceae, Solanaceae, and Orchidaceae (Vogel 1974; Buchmann 1987). The trichome is generally club-shaped and measures about 100 m long and 23 m wide [Calceolaria species, (Vogel 1974; Raman 1989)]. Vogel found that the trichomes were formed once the floral bud reached 2 mm in length. The floral oils start appearing one to two days after corolla opening and after eight to ten days the corolla then wilts (Vogel 1974; Molau 1988). The other type of oil-secreting gland, an epithelial elaiophore, is characterized by small compact tufts of glandular epithelial or epidermal cells, often well circumscribed, with one or two layered patches of columnar epithelia, capable of copious secretions. This type of elaiophore is the most widespread one and has been described for many species of the families Malpighiaceae and Krameriaceae and a few orchid species (Vogel 1974; Buchmann 1987). Even though nectar and pollen are still the primary sources of floral rewards in most plant communities, natural selection has promoted the elaboration of alternative floral rewards. Production of other sources of reward may be advantageous since it can promote strong selection towards utilizing specific suites of pollinators that can collect

PAGE 15

4 and use these oils in a given community. Natural selection may promote pollinator specificity. However, there are disadvantages as rapid extinction could occur if there are highly species-specific pollination relationships. In orchids, the pollen is packed into pollinia, and the size of the pollinator determines whether the pollinia may be easily removed and transported to another orchid flower. Selection may favor relationships with a limited number of specialized pollinators that can transport the pollinia correctly. In other words, by producing a highly energetic reward, such as floral lipids, orchids may be able to attract a limited number of oil-collecting bees that can benefit by this relationship and maintain orchid diversity by increasing pollinator specificity. Objectives of This Study In this study, I explore the oil-pollination syndrome radiation within neotropical orchids by relating variation in floral-oil reward compounds to phylogeny and to pollinator-flower specificity among neotropical orchid species. This study comprises three parts. In the first part (Chapter 2) I discuss the chemistry of oil-reward compounds and relationships with pollinators. The second part (Chapter 3) consists of a study on pollination observations of different oil-secreting species and a comparison of fruit production among species that vary in floral oil content. The last section and conclusion (Chapter 4) explores the evolution of the oil-pollination syndrome in orchids using a well supported phylogeny to speculate on the adaptive radiation of floral-oil reward compounds in neotropical orchids. A proposed hypothesis of evolutionary relationship for the subtribe Oncidiinae (Williams et al. 2001) is used as a framework for this study, as this tribe has the most diverse types of pollination systems found in the Orchidaceae (Williams 1982). The wide variety of pollination mechanisms in orchids makes them an

PAGE 16

5 appropriate group for studies of speciation and pollination evolution (Steiner 1994). Orchidaceae are the largest family of flowering plants and their members occupy a wide variety of habitats (Arditti 1992; Dressler 1993b). As a consequence, orchids show a large number of morphological, anatomical, ecological, and physiological characteristics that allow them to excel in a variety of environments. Orchids and pollinators have coadapted resulting in an array of highly specialized pollination systems that have contributed to the success of orchids in specialized habitats (Pijl and Dodson 1966; Dressler 1993a).

PAGE 17

CHAPTER 2 CHEMISTRY OF OIL-REWARD COMPOUNDS Introduction Literature Review In the neotropics, most flowers offer nectar or pollen to their pollinator organisms. Other flowers offer floral lipids, an unusual type of floral reward, instead of nectar or pollen. These floral oils are released from structures called elaiophores (sensu Vogel 1974) and are utilized by oil-collecting bees to waterproof the cell linings or mix with pollen as a larval provision (Cane et al. 1983). The relationship between flowers and their pollinators is a mutualistic one and many visual, chemical, and sensory cues are important in this relationship. From a chemical point of view, floral lipids are energetically more costly than nectar or pollen, and therefore, have a higher energy content when compared to carbohydrates. Oil-collecting bees may benefit from such a high quality food source for their larvae. Floral oils are a combination of saturated fatty acids, paraffins, glycerols, and esters. Other components related with elaiophore exudates are diglycerides, amino acids, glucose, carotenoids, phenolics, glycosides, nonvolatile isoprenoids, and saponins (Buchmann and Buchmann 1981; Neff and Simpson 1981; Buchmann 1987; Roubik 1989). The floral lipids are clear and usually colorless, sometimes yellow, without odor and light in viscosity (Vogel 1974; Buchmann 1987). The chemical composition of some floral oils have been identified in a few species, ranging from free fatty acids (Buchmann and Buchmann 1981; Buchmann 1987; Vinson et al. 1997) to acetyl-glycerol derivatives 6

PAGE 18

7 with -acetoxy-fatty acids (Reis et al. 2000) to -acetate-substituted free fatty acids or monodior tri-glycerides (Vogel 1976; Siegler et al. 1978; Simpson et al. 1979; Vinson et al. 1997). A study of the floral lipids from Calceolaria (Calceolariaceae) and Krameria (Krameriaceae) revealed a C 16 -C 20 -acetoxy substituted free fatty acids, and an unusual 3-hydroxy fatty acid (Siegler et al. 1978; Buchmann 1987). Floral Mimicry in Oil-Secreting Flowers In the neotropics, many Oncidium, Tolumnia (Orchidaceae: Oncidiinae) and relatives mimic the flowers of malpighiaceous species. Members of both groups flower simultaneously and have functionally similar inflorescences, consequently attracting and sharing similar pollinators (Ackerman 1986). However, it has not yet been established if the orchid oil compounds are a more specialized form of reward than the common malpighiaceous oils. In Panama, for example, Trichocentrum stipitatum (Lindl.) M. W. Chase & N. H. Williams mimic the flowers of the malpighiaceous species Byrsonima crassifolia (L.) Kunth, thus sharing a similar set of pollinators present in a given habitat. Some orchid species may produce floral lipids similar to those of malpighiaceous species, thus adding a chemical as well as a visual element to the mimicry between orchids and Malpighiaceae. Many malpighiaceous species, especially vines, are visited by Centris bees that gather oils from paired sepaline flower glands. Examples of orchid species mimicking malpighiaceous species, especially in the genus Tolumnia (formerly included within Oncidium) are T. guianensis (Aubl.) Braem, T. haitiensis (Leonard & Ames ex Ames) Braem, T. quadriloba (C. Schweinf), T. X osmentii (Moir) Braem, and T. compressicaulis (Withner) Braem. Female bees of the genus Centris are known to visit these species looking for floral lipids (Cingel 2001).

PAGE 19

8 Chemistry of Floral Lipids in Oncidium The first report on the presence of oils in a species of Oncidium included a brief description of the nature of the floral lipids found in Oncidium ornithorrhynchum Kunth. Thin layer chromatography was used to determine the presence of oils in the flowers, but no further chemical or spectroscopic characterization was provided (Vogel 1974). The first report on the pollination of an oil-secreting species of Oncidium using chemical evidence included the chemistry of floral oils in Oncidium pubes Lindl. (Reis et al. 2000). These oils were composed of diacylglycerols and triacylglycerols with one or two acetyl residues and one long chain fatty acid. The constitution of major compounds still requires further study. Floral lipids constitute a form of reward different from nectar, the most widespread form of reward. Oncidium pubes offers floral lipid rewards instead of nectar, and the pollination is by oil-collecting bees and not by an attraction–deception system found commonly in orchid species (Reis et al. 2000). Even though the pollination of many Oncidium species is known to be mediated by oil-collecting bees, biologists have failed to provide evidence of the chemical relationship between the pollinators and the oil-secreting flowers. Dodson (1967) observed Centris bees visiting Oncidium species, and Vogel (1974) described the oil-pollination system in detail. Since then, biologists have added little information about the chemistry of these flowers. Material and Methods Study Species and Study Sites The study was conducted at various sites throughout the Republic of Panama. Oil-orchid species were collected directly in the field at seven different sites (Fig. 2-1). The orchid species were transported to a greenhouse at the University of Florida and fresh

PAGE 20

9 flowers were collected directly from the plant for analysis. Table 2-1 is a list of the species used in this study, including a literature reference to the pollinator of each species where known. Additionally, three species from Ecuador were included in this study: Sigmatostalix portillae Kniger, Sigmatostalix oxyceras Kniger & J. G. Weinm., and Sigmatostalix marinii Kniger. I have also included a malpighiaceous tree species from Panama, Byrsonima crassifolia. The flowers of Byrsonima crassifolia were collected directly in the field in Panama and washed in hexane for 5 minutes. The samples were transported to the University of Florida for analysis. Vouchers of all plant species included in this study are deposited at the Herbarium of the University of Florida (FLAS, see Appendix B). In total, this study included sixteen species, comprising seven different genera (Table 2-1). Chemical Analysis In order to analyze oil composition among orchid species, 1 to 5 flowers were washed in hexane (1 ml. of hexane/flower) and samples were stored at C. The samples were analyzed using gas chromatography/mass spectrometry (gc/ms) at the University of Florida (Analytical Toxicology Core Laboratory). Samples were first methylated and then separated on a HP-5MS column. Gas chromatography parameters were: inlet temperature 280 C; 4 min hold at 40 C, ramped at 10 C/minute to 270 C (held for 5 minutes) and ramped at 25 C/minute to 300 C and held for 7 minutes. Mass spectra of significant peaks were compared to a library and identified whenever possible. Figure 2-2 shows the alkane nonacosane as an example of a mass spectrum from this study. The mass spectra and retention times were also compared manually against all species to determine if the same compounds were encountered across species. Isomers,

PAGE 21

Table 2-1. List of species used in this study with references to their pollinator. Species are separated by family. Family Orchidaceae Pollinator Reference Erycina cristagalli (Rchb.f.) N.H.Williams & M.W.Chase Unknown no reference Erycina pusilla (L.) N.H.Williams & M.W.Chase Unknown no reference Lockhartia micrantha Rchb.f. Unknown no reference Lockhartia oerstedii Rchb.f. Unknown no reference Oncidium cheirophorum Rchb.f. Paratetrapedia sp. Personal observation Oncidium klotzschianum Rchb.f. Unknown no reference Oncidium ornithorrhynchum Kunth. Unknown no reference Ornithocephalus bicornis Lindl. exBenth. Paratetrapedia calcarata Cingel (2001) Dressler, personal communication Ornithocephalus cochleariformis C.Schweinf. Paratetrapedia spp. Cingel (2001) Sigmatostalix picturatissima Krnzl Unknown no reference Sigmatostalix marinii Kniger Unknown noreference Sigmatostalix oxyceras Kniger & J.G.Weinm Unknown no reference Sigmatostalix portillae Kniger Unknown no reference Trichocentum ascendens (Lindl.) M.W.Chase & N.H.Williams Trigona nigra Parra-Tabla et al. (2000) Trichocentrum stipitatum (Lindl.) M.W.Chase & N.H.Williams Centris sp. Cingel (2001) Family Malpighiaceae Pollinator Reference Byrsonima crassifolia Rich. ex Kunth. Centris spp. Albuquerque et al. (1989) Barros (1992) 10

PAGE 22

11 for example, display similar fragmentation patterns and identification solely based on mass spectra may be misleading. Therefore, comparison of mass spectra and retention times is helpful for identification of compounds. The many unknowns were treated similarly to known compounds; by comparison of individual mass spectra and retention times it is possible to determine if a specific unknown is present across species. Total ion abundance (defined as the integrated ion abundance across a peak) of individual compounds was used as an indicator of composition amounts present in each species. The mass of 10 to 45 individual flowers per species was used to calculate an average of the mass of flowers. Total ion abundance of oil compounds per gram of fresh flower mass was used to calculate the average amount of floral lipids present per flower. Total ion abundance per unit of fresh flower mass of each species was used as an indicator of quantity of floral oil present per species. Data Analysis The ion abundance of individual compounds and total ion abundance per species were used to calculate percentages of individual compounds in each single species and across species. The results for Oncidium klotzschianum Rchb. f., Erycina cristagalli (Rchb. f.) N. H. Williams & M. W. Chase, and Erycina pusilla (L.) N. H. Williams & M. W. Chase were eliminated from further analysis since the total ion abundance present was less than 0.05 total ion abundance per gram of fresh flower in each case (Table 2-2). Replicates for Oncidium ornithorrynchum and Trichocentrum stipitatum were included in the analysis to test accuracy of the gc/ms analysis and variation within samples. The data were arcsine transformed to meet requirements of normality. The transformed data for all other species were used to perform a Principal Component Analysis (PCA) in order to reduce the large number of original variables to a smaller number of principal

PAGE 23

12 components (SYSTAT, Wilkinson et al. 1992). The PCA was performed in SYSTAT version 9, using no rotation, same number of variables as factors, and a matrix of extraction using correlations among the variables. A Scree Plot was performed to evaluate the relationship between the Eigenvalues and number of factors that explain the variance. The total variance explained by the three most important factors, Factor 1, Factor 2, and Factor 3 were used to construct a factor-loading plot. A hierarchical clustering analysis (cluster tree) was also performed using SYSTAT version 9. I used the same data as the PCA analysis, an average linkage method (unweighted pair group method with arithmetic mean: UPGMA), and Euclidean distances as a distance metric. Finally, the cluster tree was used to evaluate individual relationship between clusters within the tree with pollinator diversity. The pollinator of each species in the cluster tree was determined by a combination of literature review, personal communications and personal observations. Results Chemistry of Oil-Flowers One hundred thirty different compounds were found by gc/ms analyses. The majority of these compounds were unknown (72%). I found 93 unknowns and 37 known compounds. Even though unknowns were the most common, ion abundance showed an opposite pattern: unknowns constituted only 44% of the total abundance, whereas known compounds constituted 56% of the total abundance. Of the known compounds, the most abundant were C 14 -C 44 alkanes (36%). I also found esters (15%) mostly from methylations of fatty acids, aldehydes (2%), phenols (2%), alcohols (0.8%), and alkenes (0.2%). Nonacosane alone constituted 11 % of the total abundance and was the most

PAGE 24

13 common compound found in this study (Fig. 2-2). Interestingly, 31 out of 97 unknown compounds (32%) display a fragment m/z 157 characteristic of diacyl glycerols. Quantity of oil varied among members of the same genus (Table 2-2). Sigmatostalix picturatissima Krnzl, Sigmatostalix oxyceras, and Ornithocephalus bicornis Lindl. ex Benth have the largest amounts of oil-related compounds per flower (Table 2-2). Although Ornithocephalus and Sigmatostalix species have legitimate oil-reward flowers, they are not floral mimics of malpighiaceous species, in contrast with the other orchid species that possess large yellow flowers, oil-reward compounds in the flowers, and similar inflorescences to malpighiaceous species. There is a gradient of amounts of oil-related compounds among these different species. Oncidium klotzschianum, Erycina cristagalli, Erycina pusilla produce almost no oils (ion abundance/g fresh flower <0.005); Trichocentrum species produce intermediate amounts of oils (ion abundance/g fresh flower from 0.17 to 0.68); and Sigmatostalix picturatissima and Ornithocephalus bicornis produce the highest amounts of oils (ion abundance/g fresh flower of 4.69 and 2.18, respectively, Table 2-2). Principal Component Analysis (PCA) The entire data set for the PCA analysis consisted of 130 different compounds and 16 plant taxa (see Appendix A). A Scree Plot showed an inverse relationship between Eigenvalues and number of factors (Fig. 2-3). Using the scree factor graph, Factor 1, Factor 2, and Factor 3 were selected as the most influential factors, accounting for the majority of the variance. Factor 1, Factor 2, and Factor 3 accounted for 64.6% of the totalvariance (Factor 1=34.47%, Factor 2=16.89, and Factor 3=13.2%, Fig. 2-3). The PCA showed four main groups (Fig. 2-4).

PAGE 25

14 Table 2-2. Quantity of oil-related compounds given as ion abundance per unit of fresh flower mass in various oil-secreting species. Species Ion abundance/g of fresh flower mass (x10 11 ) Byrsonima crassifolia 0.400 Erycina cristagalli 0.031 Erycina pusilla 0.016 Lockhartia micrantha 1.995 Lockhartia oerstedii 0.210 Oncidium cheirophorum 0.845 Oncidium klotzschianum 0.041 Oncidium ornithorrhynchum 0.453 Ornithocephalus bicornis 2.183 Ornithocephalus cochleariformis 0.578 Sigmatostalix picturatissima 4.687 Sigmatostalix marinii 0.450 Sigmatostalix oxyceras 2.158 Sigmatostalix portillae 0.788 Trichocentrum stipitatum 0.165 Trichocentrum ascendens 0.679 The first group, represented by open circles, is composed of the Ecuadorian species: Sigmatostalix marinii, Sigmatostalix oxyceras, and Sigmatostalix portillae (Fig. 2-4). The second group, represented by open squares, is the largest group and is comprised mostly of Panamanian orchid species and the central american species, Oncidium ornithorrhynchum. Other species in this group are: Lockhartia micrantha Rchb. f., Lockhartia oerstedii Rchb. f., Ornithocephalus bicornis, Ornithocephalus cochleariformis C. Schweinf., Trichocentrum stipitatum, Trichocentrum ascendens (Lindl.) M. W. Chase and N. H. Williams (formerly called Oncidium teres), Sigmatostalix picturatissima, and Oncidium cheirophorum Rchb. f. (Fig. 2-4). The third group, represented by an open triangle, is composed of the oils from a bee’s leg (Paratetrapedia sp.), and the fourth group, represented by an open diamond is composed

PAGE 26

15 of the oils from a malpighiaceous species, Byrsonima crassifolia (Fig. 2-4). Paratetrapedia sp. is the suggested pollinator for Oncidium cheirophorum, an orchid species from the second group, and it was collected gathering lipids from fresh opened flowers of Onc. cheirophorum. Hierarchical Cluster Analysis (Cluster Tree) The hierarchical clustering analysis of oil-related compounds revealed slightly different results from the PCA analysis. First, the groupings found by the cluster analysis do not represent phylogenetic relationships, in fact, oils from some orchid species are more similar to those of the malpighiaceous species, Byrsonima crassifolia. In other words, orchid species that are close relatives (see Fig. 4-1) did not necessarily appear in the same clusters. For example, oils from Sigmatostalix picturatissima were more similar to the oils of Ornithocephalus spp. than to other Sigmatostalix species (Fig. 2-5). In general, I found two main phenetic clusters. The first was composed of the oils of Ecuadorian Sigmatostalix species, a group which was also found in the PCA analysis. The second group is larger and is composed of the oils from the malpighiaceous species Byrsonima crassifolia which is similar to all the rest of the orchid species (Fig. 2-5). In this second group, we also found that Oncidium ornithorrhynchum is more similar to a group composed of all Panamanian orchids than it is to Byrsonima crassifolia. Interestingly, there seems to be a geographical correlation in the pattern of oil compositions in these orchid species, since oils from the Ecuadorian species form a distinctive separate cluster. The oils from the Panamanian species form a distinct cluster as well, and the oils from the Central American species Oncidium ornithorrhynchum are most similar to those produced by the Panamanian species. Finally, oils from Oncidium cheirophorum form a group with oils from the bee’s leg (Paratetrapedia sp.) and these

PAGE 27

16 are embedded in the group of all Panamanian species (Fig. 2-5). This result was expected since this bee was collected gathering lipids from Onc. cheirophorum; however, the PCA analysis places the oils from the bee’s leg individually and outside all orchid groups, suggesting the presence of oils from taxa other than orchids (Fig. 2-4). The geographical separation found in the cluster analysis appears to be related to pollinator diversity, and pollinator diversity may partially explain the groups found in the Cluster Analysis. Figure 2-6 shows the relationship between pollinator diversity and cluster groups. I divide the pollinator diversity into four categories: The first category is composed of pollinators for Byrsonima crassifolia (Malpighiaceae). It is known that many species of Centris bees pollinate flowers of Byrsonima (Albuquerque and Rego 1989; Vinson et al. 1997) mention at least 14 species of Centris bees responsible for collecting oils in flowers of Byrsonima crassifolia in Costa Rica. This first category, therefore, is represented by many species of Centris bees (Fig. 2-6). The second category is represented by territorial male Centris bees, with a smaller number of bee species than the first category. Species of rat tail Oncidiums (now treated as Trichocentrum spp.) are known to be pollinated by territorial male Centris bees (Nierenberg 1972; Cingel 2001). Pollinators for Lockhartia oerstedii and Lockhartia micrantha are unknown. However, it is likely that these species are all pollinated by species of Centris bees because of their similarity in oil-related compounds (Fig. 2-6). Oncidium ornithorrhynchum is still a mystery. It is not clear in which category it should be placed, and for lack of placement, it will be included in both Centris categories, but cautiously since it is phenetically distinctive and could be pollinated by a different species of oil-collecting bee (Fig. 2-6). The third category is represented by the Paratetrapedia species.

PAGE 28

17 These small bees are known to pollinate small oil-producing flowers such as those of Ornithocephalus bicornis and Ornithocephalus cochleariformis. The group formed by Oncidium cheirophorum and the oils from the bee’s leg provide evidence that the unknown Paratetrapedia sp. may pollinate this species. I failed to find pollinia attached to the body of the bee. However, based on my observations, and the size of the bee, I propose that this bee is a pollinator of the orchid Oncidium cheirophorum. Paratetrapedia sp. collects the floral reward, has the appropriate size to detach the pollinia from the flower and spends enough time collecting the oils from the flower, increasing the possibilities of pollinia removal. The last category is represented by pollinators of the Ecuadorian Sigmatostalix species (Fig. 2-6). There is no published information on the pollination of these Sigmatostalix. All Sigmatostalix species produce oils from elaiophores, but there are no records of pollination observations (Cingel 2001). This group needs further study. Discussion Biochemical Convergence of Orchids and Malpighiaceous species The biochemical convergence among members of the Malpighiaceae and certain orchid species is evidence for Mllerian mimicry in plants. Mllerian mimicry is a system in which both plant species mutually benefit because both species have similar flowers and inflorescences that can share similar pollinators (Dafni 1984). In this study, not only do the majority of the orchid flowers mimic members of Malpighiaceae, but they also produce similar chemical compounds. There seems to be convergence, first to produce lipids in the orchid flowers and second, to produce similar oil-reward compounds and flowers that are yellow and resemble those of Malpighiaceae. This is the first report of Mllerian mimicry involving orchid species.

PAGE 29

18 Mimicry among malpighiaceous species and orchid species has been poorly documented in general. Most of the orchid mimicry studies to date involved Batesian mimicry, in which one of the species (the mimic) is rewardless and imitates the flowers of a reward species (the model) (Boyden 1980; Dafni and Ivri 1981a, b; Nilsson 1983; Kjellsson et al. 1985; Beardsell et al. 1986; Johnson 1994, 2000). The majority of these studies involve terrestrial mimic orchid species that do not produce rewards. Batesian mimicry in Oncidium cosymbephorum C. Morren (mimic) and a malpighiaceous species Malpighia glabra L. (model) has been studied by looking at the reproductive success of Oncidium cosymbephorum when Malpighia glabra is blooming in the same area (Carmona-Diaz 2001). This is the only study that provides substantial evidence for this mimicry and discusses the effects of population density, floral density, and reproductive success between Malpighiaceae and Oncidium species. The pollination mechanism of equitant species of Oncidium (now placed in Tolumnia) in the Caribbean that mimic malpighiaceous flowers but do not offer any rewards (deceit flowers) has also been described (Nierenberg 1972), but evidence for the mimicry was provided solely on the basis of both species blooming at the same time, occupying the same areas, and having similar floral UV patterns, which can attract the same pollinators. In this study, there is only chemical evidence for Mllerian mimicry. Field investigation would be necessary to confirm the system. Floral Mllerian mimicry, in general, is less commonly accepted than Batesian mimicry (Roy and Widmer 1998) and four conditions should be established: (1) both species should have overlapping phenologies and distributions, (2) the pollinator should be necessary for fruit set, (3) both species should share same pollinator species and the pollinators should be able to move

PAGE 30

19 freely between individuals of both species, and (4) the similarities between species should enhance plant fitness. The last condition would determine whether the similarity is an adaptive condition (Roy and Widmer 1998). Analysis of Oil-Related Compounds Using gc/ms and Multivariate Methods Nonacosane is the most abundant compound found by gc/ms. Alkanes, in general, constituted the bulk of general compounds found across oil-producing species. These alkanes form fatty acids that join with glycerols to form glycerides. Additionally, the majority of the unknown compounds in this study display a fragment m/z of 157 (31 compounds and 18.7% of total ion abundace). This fragment is characteristic of diacyl glycerols. -acetoxy free fatty acids and diacylglycerols derived from them have been reported as common in oil-secreting flowers (Simpson and Neff 1983). Esters, the second most common compounds found in this study, were present as methyl esters that resulted from methylations of fatty acids with long carbon chains. For example, oleic acid, one of the three most common fatty acids present in triacylglycerols, was observed through fatty acid methylations (9-octadecenoic acid, methyl ester). Although alkanes were the most abundant compound found, the exact oil-compositions of the species used in this study were not determined because samples were methylated and no further analysis was performed. By methylating samples one can enhance volatility, predict the mode of cleavage of larger compounds and obtain characteristic peaks (Silverstein et al. 1974), but on the other hand, one compromises the integrity of the whole compound. Many fragments obtained from methylated samples can rearrange and obtaining the molecular formula of the original compound becomes difficult. The results here are preliminary findings that can be used for a more elaborate process that requires Nuclear Magnetic Resonance (NMR) procedures and repeating gc/ms with non-methylated material using a

PAGE 31

20 higher temperature program that can detect glycerols more accurately. Here I used Cluster Analysis and PCA analysis of the oil-related compound matrix as a way of reducing the data to clusters (or groups) utilizing the variability generated by the data. This is the first report on the chemistry of these orchid species, except for Oncidium ornithorrhynchum whose oil chemistry has been reported using a different method of oil analysis than the one described here (Vogel 1974). PCA and Cluster analysis results were somewhat different. Both phenetic techniques use different algorithms and use different assumptions to calculate groupings. The hierarchical cluster analysis seems to be more effective in joining groups according to pollinator diversity. The PCA analysis showed 4 main groups (Fig. 2-4) and the Cluster analysis showed 2 main groups (Ecuadorian Sigmatostalix spp. and all the rest, Fig. 2-5, Fig. 2-6). These groups are different, except for the Ecuadorian Sigmatostalix group, which remains the same across these two phenetic techniques. Factor 1, Factor 2, and Factor 3 from the PCA analysis accounted for over 50% percentage of the variance, therefore looking at the PCA is an effective way of selecting clusters (Manly 2000). PCA makes no assumptions about group memberships but gives the fewest possible dimensions to examine the data while maximizing its variation (Manly 2000). Both the PCA analysis and the Cluster Analysis found patterns that seem to correlate with geographical regions and pollinator diversity and do not correlate with phylogenetic relationships. Clusters that can be explained by pollinator diversity suggest that there is biochemical convergence, and that oil-collecting bees seem to be the mediators of this convergence. Further evidence supports that Byrsonima crassifolia, a malpighiaceous species, has oil-related compounds that are very similar to those of

PAGE 32

21 mimicking orchids, and that both malpighiaceous species and orchids can be pollinated by the same oil-collecting bees. In the cluster tree (Fig. 2-6), two of the categories discerned involved Centris bees, and the orchid species in these groups have large yellow flowers with oil-related compounds; presumably they are mimics of malpighiaceous species.

PAGE 33

22 Figure 2-1. Map of the Republic of Panama with collection localities.

PAGE 34

23 Figure 2-2. Mass spectrum of nonacosane. Top panel represents the mass spectrum found in the sample. Bottom panel represents library match.

PAGE 35

24 0 5 10 15 20Number of Factors 0 1 2 3 4 5 6Eigenvalue Figure 2-3. Scree plot of Eigenvalues and number of factors. Factor 1, Factor 2, and Factor 3 account for the majority of the variance.

PAGE 36

25 -1012-101-101Factor 2 (16.9%)Factor 3 (13.2%)Factor 1 (34.5%) Figure 2-4. Principal Component Analysis (PCA) performed on 130 different oil-related compounds and 16 variables. Open circles represent group formed by Sigmatostalix spp. (Ecuadorians). Open squares represent group formed by Onc. ornithorrhynchum, Onc. cheirophorum, Trichocentrum spp., Sigmatostalix picturatissima, Ornithocephalus spp. and Lockhartia spp. Open triangle represents oils from bee’s leg (Paratetrapedia sp.). Open diamond represents oils from Byrsonima crassifolia (Malpighiaceae).

PAGE 37

26 0.00 0.05 0.10 0.15Distances Trich. stipitatum B . crassifolia (Malpighiaceae)Onc. cheirophorum P aratetrapedia sp. (Bee)Sigm. portillaeSigm. mariniiSigm. oxyceras Trich. ascendens L ock. micranthaOrnith. cochleariformisOnc. ornithorrhynchumSigm. picturatissimaOrnith. bicornis L ock. oerstedii Figure 2-5. Cluster tree of oil-related compounds. Red color represents the Ecuadorian cluster. Black represents Panamanian cluster.

PAGE 38

27 0.00 0.05 0.10 0.15Distances Trich. stipitatum B . crassifolia (Malpighiaceae)Onc. cheirophorum P aratetrapedia sp. (Bee)Sigm. portillaeSigm. mariniiSigm. oxyceras Trich. ascendens L ock. micranthaOrnith. cochleariformisOnc. ornithorrhynchumSigm. picturatissimaOrnith. bicornis L ock. oerstediimaleCentris spp. Centris spp. ? P aratetrapediaspp. ? ? Figure 2-6. Cluster tree of oil-related compounds correlated with pollinator diversity. Colors represent different categories according to pollinator groups. Question marks represent unknowns for pollinators. References for pollinator diversity are given in Table 1-1.

PAGE 39

CHAPTER 3 FLORAL VISITATION AND POLLINATION OF ORCHIDS BY OIL-COLLECTING BEES Introduction Literature Review Bees that can collect and utilize floral lipids occur in the New and Old World, but occur most commonly in the neotropics, especially in subtropical regions of North and South America (Buchmann 1987). The ability to collect and utilize floral oils has been found primarily in five tribes and four families of bees (Roubik 1989; Steiner 1998). The majority of these bees are female, solitary bees that provision the nests with mixtures of pollen and nectar. Most of these bees belong to the families Melittidae, Ctenoplectridae, Anthophoridae, and Apidae (Buchmann 1987; Roubik 1989). The most important oil-collecting bee family in the neotropics appears to be Anthophoridae with three tribes, seven genera, and over 400 species adapted to collect floral lipids. Anthophorid tribes with oil-collecting bees are: Centridini (Centris, Ptilotopus, and Epicharis), Exomalopsini (Paratetrapedia, Tapinotaspis, and Chalepogenus) and Tetrapediini (Tetrapedia) (Roubik 1989). Many species (e.g. Centris spp.) are known to collect floral lipids from a diverse number of neotropical oil-secreting plant species. Oil-collecting bees have specialized structures on the fore-legs (combs) used to harvest oils from floral elaiophores. These combs have been found exclusively in some members of the families of oil-collecting bees mentioned above, and are mainly modifications of basitarsal setae. These setae have been modified into rows of combs and 28

PAGE 40

29 setal pads of different size, located on the forelegs and midtarsi of the bees. They also have a unique set of combs located on the frontlegs to collect the oils directly from the flowers. These modified combs are usually flattened, curved at the tip, with additional two to five large setae that are called the “hood organ” (sensu Vogel 1974). The hood organ is located on the ventral surface of the anterior and middle basitarsi and is used to rupture the elaiophores so that oils may flow out of flowers. Floral Visitation and Bee-Behavior in Oil-Secreting Flowers Territorial behavior from male Centris bees in Trichocentrum stipitatum (formerly Oncidium stipitatum) has previously been observed by Robert L. Dressler in Panama (Cingel 2001). Male Centris bees use flower clusters as the focal point where they meet females, and guard them against other male Centris bees. Trichocentrum stipitatum has long pendent inflorescences that move gracefully with wind, and the flowers appear to imitate flying male bees in the focal point, thus arousing aggression from the territorial male Centris bees. When the flower moves with the wind, the bee hits the flower in an attempt to defend his territory. The bee does not land but receives pollinia on the front of its face and transports it to the next flower in the following aggression episode (Dodson and Frymire 1961; Cingel 2001). This mode of pollination has been described as “pseudotrespassing”. Female bees are known to collect the oils from Byrsonima crassifolia in the same region where Trichocentrum stipitatum grows (personal observations). I therefore argue that female bees may also be effective pollinators of this species since Trichocentrum stipitatum produces floral lipids. Both species grow in the same area, produce similar inflorescences, and flowers have similar temperature regimes that may be related to bee’s oil-collecting activity. In this study, I discuss the pollination of Oncidium cheirophorum,

PAGE 41

30 based on personal observations of floral visitations in a cloud forest of Panama. Finally, I include the results of a brief study on fruit production of three orchid species that have large yellow flowers and are probable mimics of malpighiaceous species: Oncidium klotzschianum, Trichocentrum stipitatum (formerly Oncidium stipitatum) and Oncidium cheirophorum using a floral-oil abundance gradient. Material and Methods Study Site and Oil-Flower Visitation This study was conducted in the Republic of Panama, at various sites, including lowland and cloud forests. Observations on the floral visitation of both Byrsonima crassifolia and Trichocentrum stipitatum were performed in Chilibre, at 50 m above sea level, in lowland Panama (see Fig. 2-1 for locality). I focused my observation on one emergent tree of Byrsonima crassifolia that had approximately 878,000 flowers when in full bloom (flowers per inflorescence = 69 18; inflorescences per braches = 138.4 30.8; branches per tree = 92 14). Two blooming plants of T. stipitatum were placed on separate branches of Byrsonima crassifolia, so that observations could be done simultaneously. T. stipitatum grows naturally on trees of Byrsonima crassifolia in the wild (personal observations). B. crassifolia starts blooming in February and stays in bloom from March to June with flowers that last about 6 weeks. Floral visitations in fresh flowers were observed during nine non-consecutive days: February 17, 21, 23, 24, March 21 to 23 and April 2, 3, and 9 of 2001 for a total of 34 hours. The observations where conducted from early morning hours (7:00 a.m.) to midday. However, when pollinator activity was high, observations lasted until 4:00 p.m. When B. crassifolia was in full bloom (March), the peak of pollinator activity was between 9:00 a.m. and 11:00 a.m.

PAGE 42

31 Vouchers for Byrsonima crassifolia and Trichocentrum stipitatum are deposited in the herbarium of the University of Florida (FLAS, see Appendix B). Insects gathering floral lipids from flowers of both species were collected using an insect net. Most bees that gather floral lipids from B. crassifolia are extremely fast flyers, so whenever possible, bees were collected, pinned and stored in insect boxes for identification. Vouchers of all insects collected will be deposited in the Smithsonian Tropical Research Institute (STRI) insect collection (see Appendix C). Floral visitation in Trichocentrum stipitatum was extremely rare, therefore observations were very limited. Flower Temperature Flower temperatures were measured in two flowers of two different inflorescences of both Trichocentrum stipitatum and Byrsonima crassifolia using copper-constantin thermocouples attached to a dataloger (Model LI-COR 1000, Li-Cor Inc., Linconln, NE, USA). Measurements were recorded every five minutes for three days (February 21-23, 2001). Variation in flower temperature was compared with bee’s oil-collecting activities and visitation activity was compared among plant species. Oncidium cheirophorum Flower Visitation Observations of Oncidium cheirophorum flower visitations were performed in a healthy population of about 50 plants at Fortuna, 1,061 m above sea level, in a cloud forest in Panama (see Fig. 2-1 for locality). Plants had an average of seven bulbs including old bulbs from previous years. Plants start blooming in late November and early December, with peak of blooming at the end of December. Flowers produce oils from epithelial and trichomal elaiophores throughout petals, sepals and the callus. Flowers are greenish yellow with an average 12 7 flowers per inflorescence. Observations were performed from 7:00 a.m. to 1:00 p.m. on December 20-21,

PAGE 43

32 2001 for a total of 12 hours. Bees collecting oils from flowers were collected with an insect net, pinned, and identified. Fruit Production of Malpighiaceous Mimic Orchid Species Female reproductive success (capsule formation) was measured in three orchid species that are mimics of malpighiaceous species: Oncidium klotzschianum growing in Fortuna at 1,061 m above sea level, Trichocentrum stipitatum growing in Chilibre at 50 m above sea level, and Oncidium cheirophorum growing in Fortuna at 1,060 m above sea level. An oil-abundance gradient among these 3 species was determined using the total ion abundance of compounds found in each species by analysis of gas chromatography and mass spectrometry. Number of flowers per plant and number of capsules produced per plant were measured in plants growing in natural populations for each species in order to calculate the mean capsule production per species. Eighteen plants for Oncidium klotzschianum with up to 200 flowers per plant, 25 plants for Trichocentrum stipitatum with up to 400 flowers per plant, and 20 plants for Oncidium cheirophorum with up to 85 flowers per plant, were used, for a total of 63 plants. Results were analyzed using the non-parametric Kruskal-Wallis test statistics. Results Byrsonima crassifolia and Trichocentrum stipitatum Floral Visitation Flowers of Byrsonima crassifolia were mostly visited by members of the tribe Centridini (Centris and Epicharis), Exomalopsini (Paratetrapedia) and to a minor extent Tetrapediini (Tetrapedia) and Trigonini (Trigona). Table 3-1 shows a list of the species found on different days of observations. All results are based on observations of only collected individuals. There were many observations in which I was not able to collect the individual observed, and these observations were eliminated from the study since

PAGE 44

33 identification of bees was uncertain. Paratetrapedia sp. 1 was the most commonly collected oil-gathering bee for both B. crassifolia and T. stipitatum because this species tends to stay for long periods of time on the flowers collecting oils and is therefore easy to observe and collect. Large quantities of Centris species visited B. crassifolia at the end of March when the tree was in full bloom, strongly suggesting that the species of Centris are the primary pollinators of Byrsonima crassifolia. Observation of visitation of Trichocentrum stipitatum was extremely rare, even when the plants were in full bloom. Trigona corvina Cockerell, Trigona muzoensis Schwarz, and Paratetrapedia sp. 1, were observed approaching flowers and spending from forty seconds to a minute if not disturbed, probing the flower and gathering oils with their forelegs. On only one occasion, Paratetrapedia sp. 1 approached the flower in hovering movements, landed on the callus of the flower, gathered oils with the forelegs, flew a few centimeters away from the flower, passed the oils from frontlegs to the hindlegs in mid air, and landed on the callus for a second round of oil gatherings. It performed this sequence of movements two more times before leaving the flower. Paratetrapedia spends more time in the flowers than other oil-collecting bees, but it is a very small bee, and on no occasion did I observe pollinia removal. This suggests that it may not be the pollinator of Trichocentrum stipitatum but merely a visitor because it is too small to remove the pollinia efficiently. On the other hand, Centris inermis Friese was observed approaching the flower, but it would only spend a second before moving to Byrsonima flowers. Centris bees, in general, spend just one to three seconds at or on each flower before moving to another one (personal observations), and when Byrsonima.

PAGE 45

34 Table 3-1.List of pollinators and visitors for species of oil-secreting flowers of Byrsonima crassifolia and Trichocentrum stipitatum. Observations of Cassia fistula and Plumeria sp were included for relevant discussion. Date (2001) Observation time Plant species Insect Activity Feb 17 60 min Byrsonima crassifolia Paratetrapedia sp.1* Paratetrapedia moesta Cresson co-g co-g Feb 21 395 min Byrsonima crassifolia Plumeria sp. Paratetrapedia sp.1* Trigona corvina Cockerell* Paratetrapedia moesta Cresson Scaptotrigona luteipennis De la Torre co-g cr cr cr Feb 23 345 min Byrsonima crassifolia Plumeria sp. Trichocentrum stipitatum Paratetrapedia sp. 1* Trigona angustula La Treille Paratetrapedia moesta Cresson Eulaema nigrita Lepeletier Trigonia corvina Cockerell* Trigona muzoensis Schwarz co-g cp co-g a cr co-c Feb 24 182 min Byrsonima crassifolia Trichocentrum stipitatum Paratetrapedia sp. 1* Paratetrapedia moesta Cresson Trigona corvina Cockerell co-g co-g co-c March 21 242 min Byrsonima crassifolia Paratetrapedia sp. 1* Paratetrapedia moesta Cresson co-g co-g March 22 60 min Byrsonima crassifolia Cassia fistula Trichocentrum stipitatum Paratetrapedia sp. 1* Paratetrapedia moesta Cresson Tetrapedia cf. maura Cresson Centris bicornuta Mocsary* Frieseomelitta nigra Lepeletier Epicharis maculata Smith Centris trigonoides Lepeletier* Cephalotrigona capitata Smith Trigona fuscipennis Friese male Centris trigonoides Lepeletier Paratetrapedia sp. 1* co-g co-g co-g co-g cpo co-g co-g f f t co-g March 23 153 min Byrsonima crassifolia Paratetrapedia sp. 1* Epicharis maculata Smith* co-g co-g April 2 240 min Byrsonima crassifolia Trichocentrum sitpitatum Paratetrapedia sp.1* Paratetrapedia moesta Cresson Trigona muzoensis Schwarz Trigona fulviventris GurinMneville Centris inermis Friese* Centris longimana Fabricius* Trigona mozoensis Schwarz* Centris inermis Friese* co-g co-g co-g co-g co-g co-g co-g co-g April 5 120 min Byrsonima crassifolia Epicharis rustica Olivier* Paratetrapedia sp. 1* co-g co-g April 9 60 min Byrsonima crassifolia Paratetrapedia sp. 1* Epicharis rustica Olivier* co-g co-g * species commonly found visiting flowers at specific observation days; co-g: collecting oils from sepal glands; cr: collecting resins in stems; cp: collecting pollen; a: approaching flowers; co-c: collecting oils in callus; cpo: collecting pollen and oils; f: foraging; t: territorial behavior, striking pendent inflorescences

PAGE 46

35 flowers are present, they do not seem interested in spending time at or on flowers of Trichocentrum stipitatum. Territorial behavior from male Centris bees was not observed in relationship to orchids in this study. On one occasion, I observed territorial behavior of male Centris trigonoides Lepeletier hitting and chasing flowers of Cassia fistula L., a tree that grows in the same area where all other observations were conducted. Cassia fistula has pendent bright yellow inflorescences that can imitate other male Centris bees when moved by the wind. Females of Centris trigonoides are known to collect oils from malpighiaceous species and it was a common species found gathering oils from Byrsonima crassifolia. Flower Temperature of Oil-Flowers Flower temperature regimes among Byrsonima crassifolia and Trichocentrum stipitatum appear to be similar (Fig. 3-1B; 3-1C). There is an increase in air temperature as the day progresses and reaches a peak at midday with temperatures of 33 C, then decreases until 0600 h of the following day when it starts increasing again (Fig. 3-1A). Interestingly, each of the peaks found above the zero line in Fig. 3-1B and 3-1C relates with pollinator activity. These peaks represent points at which the oil-flowers were slightly warmer than air temperature. For example, the peak of insect visitation for Byrsonima crassifolia is between 8:00 a.m. and 11:30 a.m., just before noon, the hottest time of the day. Between these hours flower visitation is at peak and flowers are slightly warmer than air temperature. Additionally, the last peaks found on Fig. 3-1C for Trichocentrum stipitatum relates with the observations found in February 23, 2001 for this species. I observed Trigona muzoensis visiting flowers of T. stipitatum exactly at the time of the last peak in Panel 3-1C (mean flower temperature= 30 C, 15:26 p.m.). There

PAGE 47

36 seems to be insect attraction at the surface of fresh oil-secreting flowers when the flower temperature is slightly warmer than air temperature. Oncidium cheirophorum Floral Visitation Oncidium cheirophorum produces flowers that are visited by an undescribed species of Paratetrapedia. Floral visitation was observed on fresh opened flowers of Oncidium cheirophorum at Fortuna. Activity started at 10:30 in the morning and bees were actively gathering oils for about two hours. The bee approached the flower and gathered oils from the sepals and petals for up to one minute, then left the flower and transferred the oils from the forelegs to the hindlegs in the air. The bee returned to the same flower once more and repeated the same movements. Bees also probed and visited other flowers instead of returning to the same one. One bee can visit many flowers even on different inflorescences if they are close to each other. This method of gathering lipids is similar to the one observed in pollinators of Byrsonima crassifolia and Trichocentrum stipitatum. Figure 3-2 shows a sequence of photos related to the visitation of Oncidium cheirophorum by Paratetrapedia sp. I was able to observe more floral visitation events in Oncidium cheirophorum than in any other species of oil-secreting orchids, which allowed time to photograph some of the visitation events. Female Reproductive Success in Oil-Secreting Orchids Orchids that produce oils may benefit by an increase in visitation and as a consequence, an increase in fruit production. Here I found that as oil abundance increases, there is an increase in fruit production in orchids that mimic members of the Malpighiaceae (Fig. 3-3). These differences are significant (Kruskall-Wallis non parametric test statistic = 49.105, p<0.001).

PAGE 48

37 Discussion Pollination of Byrsonima crassifolia and Trichocentrum stipitatum Flowers from B. crassifolia were visited by many species of the family Anthophoridae, especially bees from the Tribe Centridini. It has been found that the genus Centris is the primary pollinator of species of Byrsonima (Albuquerque and Rego 1989). Additionally, other genera also visit the flowers and harvest the oils as well. It is interesting that bees that visited flowers of Trichocentrum stipitatum are known to pollinate B. crassifolia. This suggests that these species share similar visitors and pollinators. However, when B. crassifolia is in full bloom, visitation in Trichocentrum stipitatum is limited. In natural populations, Trichocentrum stipitatum blooms a few weeks before the beginning of the blooming season for B. crassifolia. I have observed T. stipitatum blooming in branches of B. crassifolia that are just developing flowering buds. This non-overlap in blooming time may be an advantage in attracting pollinators and reducing competition, which otherwise would occur if both species bloomed at the same time. Even though I did not find pollinia attached to the body of the bees collected, I suspect that a pollinator for Trichocentrum stipitatum is Centris inermis, because it has the appropriate size to detach the pollinia when gathering oils from the callus. Trigona corvina and Trigona muzoensis were also found gathering oils from T. stipitatum, but I suspect these species may be only visitors because their size may be too small to detach pollinia efficiently. Also, Paratetrapedia sp 1. was found gathering oils from the callus of the T. stipitatum. This may also be a pollinator, but after spending many seconds hovering and harvesting oils, it did not approach the pollinia. There is a possibility that these bees may remove the pollinia by chance alone, if frequent visits are done to

PAGE 49

38 different flowers. More observations are needed to confirm whether or not these species of bees are pollinators for T. stipitatum. Female bees of Centris versicolor Fabricius are known to visit flowers of Tolumnia lucayana (Nash ex Britton & Millsp.) Braem in Grand Bahama (Nierenberg 1972). The bees hover in front of the flower and land in the center of the flower looking for food or reward and leave a few seconds after finding out that the flower offers no reward. If the bee is positioned correctly, it will detach the pollinia from the orchid and carry it on the body. The similarity between orchid flowers and those of malpighiaceous species in the area could allow the bee to make another mistake confusing an orchid flower with a malpighiaceous species and transport the pollinia into the stigma of the orchid flower. The mimicry of malpighiaceous species by orchid species would apply to female bees, since they are looking for oils and can be deceived. Also, the similarity between these flowers, as shown by ultraviolet absorption patterns, appears to be key in the attraction-deception of female Centris bees (Nierenberg 1972). Other species of Oncidium are pollinated by Anthophorid bees (mostly Tetrapediini) of the genera Paratetrapedia and Tetrapedia. For example, in Southern Brazil, Paran, Oncidium paranaense Kraenzl. has been shown to be pollinated by Tetrapedia diversipes Klug, gathering oil from two elaiophores on the lip (Cingel 2001). Buchmann (1987) mentioned that Tetrapedia species are perhaps the probable pollinator of most oil-secreting orchids in the neotropics. Tetrapedia was not observed visiting flowers of T. stipitatum, however, it did visit flowers of Byrsonima crassifolia. Trichocentrum ascendens in Yucatn, Mexico is pollinated by Trigona nigra Cress, a native bee that searches for resin (Parra-Tabla et al. 2000). In Panama, two species of

PAGE 50

39 Trigona (T. corvina and T. muzoensis) were observed gathering lipids from both B. crassifolia and T. stipitatum. Trigona corvina was commonly found collecting resins from stems of Plumeria sp. in the same area as B. crassifolia, suggesting that this species specializes in oil and resin collecting. The fact that Trichocentrum stipitatum produces pendent inflorescences, blooms adjacent to malpighiaceous species, and shares similar visitors and pollinators suggest that this group may constitute a mimetic system, in which Byrsonima crassifolia is the model and Trichocentrum stipitatum the mimic. There are two additional conditions that were not tested but are necessary to conclude that there is Mllerian mimicry between these species: (1) the pollinator should be necessary for fruit set, (2) the similarities between species should enhance plant fitness. Similarities would determine whether there is an adaptive condition. Pollinators and visitors seem to move freely between species and phenology overlaps enough to exploit local pollinators while minimizing competition. Physiologically, both species are capable of attracting oil-gathering bees. It is interesting that as flowers become slightly warmer than air temperature, bees are attracted to the flowers. This may be due to the oils being volatilized and available for insects at this specific point. There seems to be a relationship between bee visitation and the point in which flowers are slightly warmer than air temperature. It is possible that bees detect changes in flower temperature and this difference plays an important role in oil-secreting flower pollination. Evidence suggests that the pollination of Trichocentrum stipitatum is truly based on a reward system mediated by oil-collecting bees, and does not constitute a deceit system as is found in the majority of neotropical orchid species. Trichocentrum stipitatum

PAGE 51

40 flowers produce rewards for oil-gathering bees and the mechanisms of attraction are similar to those of the model species Byrsonima crassifolia. Capsule Production and Reproductive Success in Oil-Secreting Flowers Orchids that mimic members of the Malpighiaceae can also produce an oil-reward. I found that oil production is positively correlated with fruit production. As mimic orchids produce more oil-reward compounds, there is an increase in capsule production, which may be related to an increase in pollinator visitation. This gradient goes from mimic species that produce little to no oil (e.g., Oncidium klotzschianum) to species that produce large amounts of oil-reward compounds (e.g., Oncidium cheirophorum). Interestingly, there seems to also be a gradient of mimicry systems. Orchid species can go from Batesian mimicry, producing no reward and exploiting the pollinators of malpighiaceous species (e.g., proposed mechanism for Onc. klotzschianum), to Mllerian mimicry, producing reward compounds that are utilized by oil-collecting bees that also pollinate malpighiaceous species. This mimicry gradient suggests that pollinator availability may be acting as a selective force in determining which mimicry system may be favored. There seems to be one critical question related to this mimicry gradient. Why would orchid species develop deceit systems (Batesian mimicry) if it negatively influences plant reproduction by decreasing chances of pollinator visitation and pollination events? Deception mechanisms can be seen as a two way process: (1) long distance attraction has to occur and (2) pollinator behavior needs to be stimulated. Long distance attraction can be gained by flower coloration alone (Johnson 1994). For example, mimic orchids may produce flowers that resemble in color and shape the flowers of malpighiaceous species. Eliciting collecting behavior in bees is the second process, and color alone will not provoke oil-gathering behavior (Johnson 1994). One

PAGE 52

41 explanation is that it is energetically advantageous to not produce a costly reward at the expense of reduced pollination success (Johnson 1994). Orchid flowers may produce areas in the callus of the flower that resembles food, some non-oil producing species may also look shiny, deceiving the pollinator into thinking there is oil. Orchids have the advantage of having pollen packed into pollinia that can be attached to the body of the bees and by fooling the pollinator a few times, orchids can get the pollinia transferred among different flowers. The real task is to determine if reproductive success is being favored by deceit mechanisms in the long term. The majority of orchids seem to have evolved deceit mechanisms. The loss of nectar or rewards in general, combined with an effective mimicry system, could result in increase in plant fitness.

PAGE 53

42 Flower temperature air temperature(C) -1.0-0.50.00.51.0 Time (hour) 600120018002400600120018002400600120018002400 -1.0-0.50.00.51.0 Trichocentrum stipitatum Byrsonima crassifolia Air temperature (C) 1820222426283032 ABC * Figure 3-1. Flower temperature of two species of oil-secreting flowers measured during three consecutive days. Panel A represents air temperature measurements. Panel B and C represent the difference between air temperature and flower temperature for B. crassifolia and T. stipitatum respectively. * Represents temperature peak that matches pollination observation.

PAGE 54

43 Figure 3-2. Paratetrapedia sp. gathering floral lipids from flowers of Oncidium cheirophorum in Fortuna, Panama.

PAGE 55

Figure 3-3. Female reproductive success (capsule formation) of three orchid species that mimic members of the Malpighiaceae; IA=Ion abundance. Mean capsule production (%) 0 10 20 30 40 50 Kruskal-Wallis test statistics= 49.105 P<0.001 Onc. klotzschianum IA=0.0407 Trich. stipitatum IA=0.1914 Onc. cheirophorum IA=0.8450

PAGE 56

CHAPTER 4 CONCLUSION In conclusion, I present a brief summary of the most important results relating to the chemistry, pollination, and visitation of oil-secreting species. I conclude by giving insight into the evolution of oil-secretions in flowering plants. The biochemical convergence found among various groups of orchids and malpighiaceous species suggests an evolutionary pressure towards utilizing a specific suite of pollinators. In other words, species containing similar oil-compounds share similar pollinators in a geographical region. Geographical regions and pollinator diversity are, therefore, the best predictors of biochemical convergence in this study. Phylogeny cannot be used to explain biochemical convergence of oil-reward compounds in these orchid species. Chemically, both malpighiaceous species and oil-secreting orchids may produce similar oil-related compounds that can attract similar pollinators in a given habitat. This type of mutualistic relationship can best be described as Mllerian mimicry. There is convergence, first in producing floral lipids in different clades, and second in producing flowers and oil-reward compounds similar to those of malpighiaceous species. Biochemical, physiological and behavioral evidence suggests that the pollination of orchids with oil-secreting flowers is truly based on a floral-lipid reward system rather than an attraction-deception system, as is commonly found in orchids. Mllerian mimicry between orchids and malpighiaceous species seems to best describe the biochemical relationship between those groups. More field studies and further chemical analysis of 45

PAGE 57

46 flowers of Malpighiaceae and orchid species that grow in the same habitat may confirm the patterns found in this study. The production of oils in orchid species is a polyphyletic trait, which has evolved at least five times in the subtribe Oncidiinae (Fig. 4-1). This independent production of floral lipids in different clades of orchids can be explained as an evolutionary strategy to enhance pollination by oil-collecting bees. This same pattern has been observed among angiosperms. Oil-secreting species have evolved independently in non-related angiosperm families, and in all cases seem to be an adaptation to attract oil-collecting bees (Buchmann 1987). It is possible that nectar, as a form of reward, evolved first and gave rise to floral lipids as a derived form of reward. It would be interesting to test this hypothesis using different phylogenies of families with oil-secreting species. Future studies should address a combination of chemical analysis with pollination observations in the field to understand the radiation of oil-secreting flowers in angiosperms and especially in orchids, a group with a high number of pollination syndromes.

PAGE 58

47 * Fernandeziasp. Ornithocephalus inflexus Sphyrastylissp. Dipteranthus grandiflorusHofmeisterella eumicroscopicaStellilabium pogonostalixTelipogon parvulusTrichoceros sp.Miltonia flavescensMiltonia candidaAda aurantiacaAda sp.Mesospinidium sp.Ada alleniiBrassia arcuigeraBrassia caudataBrassia gireoudianaBrachtia andinaAspasia epidendroidesCischweinfia dasyandraSysteloglossum acuminatumRhynchostele bictonensisAmparoa costaricenseMesoglossum londesboroughianumErycina diaphanaErycina echinataPsygmorchis pusillaPsygmorchis pumilioOncidium crista-galliTolumnia calochilaTolumnia variegataTolumnia tuerckheimiiTolumnia henekeniiComparettia macroplectronIonopsis utricularioidesIonopsis satyrioidesKonantzia minutifloraRodriguezia lanceolataRodriguezia delcastilloiRodriguezia satipoanaNotylia barkeriiMacroclinium bicolorOncidium onustumCapanemia superfluaOncidium dasystyleOncidium flexuosumRodrigueziella gomeziodesGomesa planifoliaMiltoniopsis warscewicziiOdontoglossum edwardiiOncidium annulareNeodryas herzogiiOncidium cimiciferumOtoglossum chiriquenseOncidium reichenheimiiOncidium ghiesbreghtianumOncidium sphacelatumOncidium leucochilumOncidium ornithorrhynchum Oncidium cheirophorum Symphyglossum sanguineumOdontoglossum harryanumCochlioda noezlianaOdontoglossum multistellareSigmatostalix picta Lockhartia oerstediiLockhartia amoena Oncidium cebolletaOncidium jonesianum Trichocentrum pfavii Oncidium lanceanum Oncidium splendidumLeucohyle subulataTrichopilia maculataHelcia sanguinolentaNeoescobaria brevisPsychopsis papilioPescatorea lehmaniiDichaea neglecta5019121031732250715635893953109689473711120416218101042851613829136131632423294820295420241213152237234026615889608864794310263812234717129322050251240537993523652366776497208294619362171133146316233538483357124169181719288610131113181312127158222666174247241237613818249132721213166729299/970/5100/35100/52100/4285/5100/13100/38100/19100/7100/1373/199/763/6100/965/290/190/2100/11100/6100/898/13100/8100/21100/63100/23100/684/2100/27100/695/391/689/092/693/2100/33100/995/2100/1881/063/099/3100/15100/1385/2100/682/085/2100/12100/894/370/197/797/399/589/2100/11100/1199/998/3100/11100/1290/099/7100/1388/097/3100/35100/18100/2495/299/5*/098/481/091/0ABCDEFGHI OutgroupsI * * * * Figure 4-1. Single shortest tree for the Subtribe Oncidiinae from successively weighted analysis of combined data set (ITS/matK/trnL-F, Williams et al. 2001). * Representatives of oil-secreting orchid species. Major clades are indicated by letters. Fitch branch lengths (ACTRAN optimization) are indicated above the branches and bootstrap support percentages are indicated below the branches. More details are explained in Williams et al. (2001) (Reprinted with permission of the author).

PAGE 59

APPENDIX A LIST OF COMPOUNDS AND ION ABUNDANCE PER MASS OF FRESH FLOWER

PAGE 60

49 Name of compound TS1 TS2 BC OCH BEE SPO SMA SOX OKL TAS LMI OCO ECR OO1 OO2 SPI OBI LOE EPU tetradecanoic acid, methyl ester 0 0 0 0 0.16 0 0 0 0 0 0 0.86 5.20 0 0 0 0.35 0.86 0 unknown 1 0 0 0 0.26 0 0 0 0 0 0 1.48 0 0 0 0 0.60 0 0 0 unknown 2 0 0 0 1.16 1.85 0 0 0 0 0 0 0 0 0 0 0 0 0 0 hexadecanal 0 0 0 2.20 1.42 0 0 0 0 0 0 0.46 0 11.81 3.82 0.45 0.34 0 0 2,4 Diphenyl-4-methyl-2(E)pentene 0 0 0 0.69 0 0 0 0 0 0 1.68 0.69 0 0 0 0.29 0 0 0 phenol, 4-(1-methyl-1-phenylethyl)0 0 0 0 0 0 0 0 0 0 1.11 0 0 0 0 0 0 0 0 1,2-Benzenedicarboxylic acid, dibutyl ester 0 0 0 0 0 0 0 0 0 0 0 0 0 0.75 3.06 1.02 0.36 0 0 unknown 3 0 0 0 3.28 3.93 6.61 3.28 0 0 0 0 3.55 0 0 0 5.82 4.47 0 0 octadecanal 0 0 0 1.47 1.13 0 0 2.09 0 0 0 0.49 0 0.90 1.69 5.28 1.98 0 0 2-t-Butyl-4-(dimethylbenzyl)phenol 0 0 0 1.66 0 0 0 0 0 0 4.31 2.07 0 1.57 1.53 0.74 0.45 0 0 unknown 4 0 0 0 0.55 0 0 0 0 0 0 0 1.11 0 0.63 0 0 0 0 0 unknown 5 0 0 0 0 1.61 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 6 0 0 0 1.15 0 0 0 0 0 0 1.38 0 0 0 0 0 0 0 0 unknown 7 0 0 0 4.11 2.83 0 0 0 0 0.44 4.45 3.09 0 1.88 1.69 1.10 0.33 0 0 unknown 8 0 1.20 1.44 0 0.86 0 0 0 0 0.4 0 0 0 0.85 0 0 0.56 0 56.34 heneicosane 0 0 0 0 0 0 0 0 0 0 0 2.12 0 0 0 0 0 0 0 unknown 9 0 0 0 0 0 0 0 0 0 0 3.76 0 0 0 2.57 0.92 0 0 0 9-Octadecenoic acid (Z)methyl ester 0 5.28 0 3.91 1.69 0 0 0 0 1.89 0 0 0 0 0 0 0 3.56 43.65 unknown 10 0 0 0 0 0 0 1.78 0 0 0 0 0 0 0 0 0 0 0 0 unknown 11 0 0 0 0 0 0 2.14 0 0 0 0 0 0 0 0 0 0 0 0 unknown 12 0 0 0 0 0 0 0 0 0 0 1.75 0 0 0 0 0.47 0 0 0 unknown 13 0 0 0 3.60 0 0 0 0 0 0 0 0 0 7.04 3.22 0 0 0 0 unknown 14 0 0 0 0 2.34 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 15 (isomeres) 0 0 7.49 0 0 0 0 0 0 0 0 1.98 0 0 0 1.47 4.34 0 0 docosane 0 0 0 1.56 0 0 0 0 0 0.52 2.83 2.15 0 0.69 1.85 1.11 0.90 0 0 unknown 16 0 0 0 0.39 0.56 0 0 0 0 0 0 0 0 0 0 1.16 0.81 0 0 unknown 17 0 0 0 0 0 9.47 3.93 0 0 0 0 0 0 0 0 2.29 0.45 0 0 unknown 18 0 0 0 0 0 6.86 3.29 5.99 0 0 0 0 0 0 0 0 0 0 0 unknown 19 0 0 0 0 2.53 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 20 0 0 0 0 0 0 0 0 0 0.42 0 0 0 0 0 0 0 0 0 octadecanoic acid, 3-hydroxy-methyl ester 0 0 0 5.32 3.01 0 0 0 0 0 0 0 0 15.73 0 1.04 0 0 0

PAGE 61

50 Name of compound TS1 TS2 BC OCH BEE SPO SMA SOX OKL TAS LMI OCO ECR OO1 OO2 SPI OBI LOE EPU unknown 21 0 0 0 5.68 5.30 0 0 0 0 1.49 0 4.20 0 0 0 6.95 4.37 0 0 tricosane 1.49 2.74 0 4.73 2.01 0 0 0 0 3.08 4.95 6.54 0 4.92 4.05 3.9 4.06 2.04 0 eicosanoic acid, methyl ester 0 0 1.16 0 0 0 0 0 0 0.91 0.60 0 0 0 0 0 0 1.09 0 unknown 22 0 0 0 0.43 0.51 0 0 0 0 0 0 0 0 0 0 0.65 0.44 0 0 unknown 23 0 0 0 3.82 2.30 0 0 0 0 0 0 0 0 9.01 2.81 0 0.41 0 0 unknown 24 0 0 0 0 0 0 0 0 0 0 0 1.13 0 0 0 0.46 0 0 0 unknown 25 0 0 0 2.43 3.57 0 0 0 0 0 0 0 0 0 0 1.1 0.81 0 0 unknown 26 0 0 1.05 0 0 0 0 0 0 0 0 0 0 0 0 0 0.45 0 0 unknown 27 0 1.89 0 0 0.27 0 0 0 0 0 0 0 0 0 0 0 0 0 0 tetracosane 0 0 0.65 3.58 0 0 0 0 0 2.45 6.11 3.86 0 2.27 7.20 4.27 3.77 2.29 0 unknown 28 0 0 0 0 0 0 0 0 0 0 0.68 0 0 0 0 0 0 0 0 unknown 29 0 0 0 0 0 0 0 0 0 3.40 0 0 0 0 0 0 0 0 0 unknown 30 0 0 0 2.24 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 31 0 0 0 0 3.42 0 0 0 0 0 0 0 0 0 0 0 0 0 0 hydroxy fame 0 0 0 0 0 6.67 6.97 8.93 0 0 0 0 0 0 0 0 0 0 0 unknown 32 2.06 1.3 0 0 0 0 0 0 0 0.35 0 0 0 0 0 0 0 0 0 unknown 33 0 0 0 2.54 6.18 0 0 0 0 0 0 0 0 0 0 4.32 0 0 0 unknown 34 0 0 8.44 0 0 0 0 0 0 5.07 0 1.03 0 0 0 0 3.94 0 0 docosanoic acid, methyl ester (1) 0 0 0 0 0 0 1.12 0 0 0 0 0 0 0 0 0 0 0 0 pentacosane 4.21 4.32 0 0 0 0 0 0 0 6.57 6.56 5.88 7.62 5.61 12.24 7.73 7.43 4.02 0 docosanoic acid, methyl ester (2) 0 0 0 0 0 0 0 0 0 0 0 0.9 0 0 0 1.30 2.12 0 0 unknown 35 0 0 0 1.71 4.71 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 36 0 0 0 0 0 0 0 3.06 0 0 0 0 0 0 0 0 0 0 0 unknown 37 0 0 0 0 0 4.76 2.65 0 0 0 0 0 0 0 0 0 0 0 0 unknown 38 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4.56 0 0.96 0 0 unknown 39 0 0 0 0 1.44 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 40 0 0 0 3.47 2.56 0 0 0 0 0 0 0.80 0 0 0 1.06 1.73 0 0 docosanoic acid, methyl ester (3) 1.25 1.52 2.97 0 0 0 0 0 9.59 0.98 0 0 0 0 0 0 0 0 0 unknown 41 0 0 0 3.10 6.25 0 0 0 0 0 0 0 0 0 0 4.02 3.08 0 0 unknown 42 0 0 0 0 0 0 1.14 1.45 0 0 0 0 0 0 0 0 0 0 0 hexacosane 1.71 2.43 1.02 4.38 0 0 0 0 0 5.22 6.267 6.83 0 3.46 11.48 9.75 6.21 2.63 0 unknown 43 0 0 0 3.99 4.89 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 44 0 0 0 0 0 0 0 0 0 1.03 0 0 0 0 0 0 0 0 0 unknown 45 0 0 0 4.71 5.42 0 0 0 0 0 0 3.29 0 0.86 0 5.45 3.97 0 0 unknown 46 0 0 0 0 0 0 0 0 0 2.94 0 0 0 0 0 0 0 0 0 unknown 47 0 0 0 0 0 28.92 33.8 26.37 0 0 0 0 0 0 0 0 0 0 0 1,13-Tetradecadiene 0 0 0 0 0 0 0 0 0 1.75 0 0 0 0 0 0 0 0 0 unknown 48 0 0 0 0 0 0 0 0 0 0 1.52 8.03 0 0.59 0 0 4.28 1.63 0 unknown 49 3.20 1.62 0 0 0.80 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1-Eicosanol 0 0 0 0 0 0 0 0 0 1.42 0 0 0 0 0 0 0 0 0 unknown 50 0 0 0 0.42 0 0 0 0 5.7 0 0 0 0 2.09 0 0 0 0 0 unknown 51 3.04 1.45 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1.31 0 unknown 52 0 0 0 0 0 0 0 0 0 0.55 0 0 0 0 0 0 0 0 0 unknown 53 0 1.69 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

PAGE 62

51 Name of compound TS1 TS2 BC OCH BEE SPO SMA SOX OKL TAS LMI OCO ECR OO1 OO2 SPI OBI LOE EPU unknown 54 0 0 0.88 0 2.02 0 0 0 0 3.71 0 0 0 0 0 0 0 0 0 tetracosanoic acid, methyl ester 0 0 0 0 0 0 1.55 0 0 0 0 0 0 0 0 0 0 0 0 unknown 55 0 0 0 0 0 15.29 15.3 17.29 0 0 0 0 0 0 0 0 0 0 0 unknown 56 0 0 0 0 1.69 0 0 0 0 0 0 0 0 0 0 0 0 0 0 heptacosane 14.8 8.58 1.44 3.63 0 0 0 0 12.87 9.80 5.61 4.25 4.46 5.60 12.23 7.19 6.30 3.66 0 11,13-Eicosadienoic acid, methyl ester 0 0 1.51 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 57 0 0 0 0.37 0 0 0 0 0 0 2.00 0 0 0 0 0.706 0 0 0 unknown 58 0 0 0 0 0 13.96 12.9 12.18 0 0 0 0 0 0 0 0 0 0 0 tetracosanoic acid, methyl ester 3.74 3.44 6.29 2.77 5.67 0 0 0 9.25 1.78 0 2.86 31.82 0 0 4.148 7.779 0 0 unknown 59 0 0 0.88 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 60 0 0 0 0.86 7.43 7.56 17.67 0 0 0 0 0 0 0 0 0 0 0 unknown 61 0 0 0 0.97 1.57 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 62 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1.152 1.48 0 0 unknown 63 0 0 1.08 0 0 0 0 0 0 0 0 0 0 0 0 0 3.27 0 0 unknown 64 0 0 6.11 0 0 0 0 0 0 0 0 0 0 0 0 0 0.82 0 0 unknown 65 0 0 0 0 1.52 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 66 0 0 0 0 0 0 0 0 0 1.20 0 0 0 0 0 0 0 0 0 octacosane 3.34 3.77 6.15 0 0 0 0 0 0 5.18 3.37 0 0 3.31 6.31 7.25 0 2.84 0 unknown 67 0 0 0 3.74 4.90 0 0 0 0 0 0 5.17 0 0 0 0 0 0 0 unknown 68 0 0 0 0 0 0 0 0 0 2.01 0 0 0 0 0 0 0 0 0 unknown 69 0 0 7.42 0.54 1.09 0 0 0 0 0 0 1.03 0 0 0 0.53 3.24 0 0 unknown 70 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2.02 0 0 0 0 unknown 71 0 0 0 0 1.35 0 0 0 0 3.65 0 0 0 0 0 0 0 0 0 unknown 72 0 0 0 0 0 0 0 0 0 0 1.82 9.82 0 0.94 0 0 4.86 2.58 0 nonacosanol 3.97 0 0 0 0.58 0 0 0 0 1.4 0 0 0 0 0 0 0 0 0 unknown 73 5.11 2.42 0 0.55 0 0 0 0 12.86 0.92 0 0.70 2.05 2.64 0 0 0 1.90 0 nonacosane 21.8 19.3 0 1.66 0 0 2.42 0 49.72 7.6 18.89 4.28 0 4.92 8.52 2.46 3.79 64 0 unknown 74 0 0 2.18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 75 0 0 0 0 0 0 0 0 0 0 2.37 0 0 0 0 0 0 0 0 hexacosanoic acid, methyl ester 2.13 0 0 0 0.15 0 0 0 0 0.99 0 0 18.02 0 2.35 0 0.85 0 0 unknown 76 0 0 9.13 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 77 0 0 3.27 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 78 0 0 0 0 0 0 0 0 0 1.52 0 0 0 0 0 0 0 0 0 triacontane 2.54 3.70 0 1.47 0 0 0 0 0 0 1.78 2.74 0 2.77 2.85 1.33 1.15 2.22 0 nonadecane 0 0 0 0 0 0 0 0 0 2.92 0 0 0 0 0 0 0 0 0 unknown 79 0 0 4.69 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 80 0 0 1.19 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2,4,6-Tri (dimethylbenzyl)phenol 0 0 0 0.48 0 0 0 0 0 0 1.04 0.84 0 0 0 0 0 0 0 unknown 81 0 0 0 0 0.39 0 0 0 0 2.30 0 0 0 0 0 0 0 0 0 unknown 82 0 0 0 0 0 0 0 0 0 2.28 0 0 0 0 0 0 0 0 0 unknown 83 0 0 0 0.88 0 0 0 0 0 0 1.54 1.13 0 0 0 0 0 0 0 1-Dotriacontanol 2.74 0 0 0 0 0 0 0 0 1.73 0 1.76 0 0 0 0 1.45 0 0 unknown 84 0 0 0 0.98 0 0 0 0 0 0 1.04 0 0 0 0 0 0 0 0 unknown 85 0 0 0 0 0 0 0 0 0 0 0 0 0 3.79 0 0 0 0 0 unknown 86 0 0 0 0.93 0 0 0 0 0 0 1.64 0.97 0 0 0 0 0 0 0 0

PAGE 63

52 Name of compound TS1 TS2 BC OCH BEE SPO SMA SOX OKL TAS LMI OCO ECR OO1 OO2 SPI OBI LOE EPU hexatriacontane 19.2 24.6 1.84 0.54 0 0 0 0 0 5.94 0.90 1.48 0 2.80 2.95 0 0.64 2.59 0 tetratriacontane 1.05 2.78 0 0.64 0 0 0 0 0 1.01 0 0.90 0 1.36 0 0 0 0 0 unknown 87 0 0 1.3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 88 0 0 1.50 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 89 0 0 1.94 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 90 0 0 0 0 0 0 0 4.93 0 0 0 0 0 0 0 0 0 0 0 tetratetracontane 2.48 5.76 0.53 0 0 0 0 0 0 1.31 0 0 0 1.07 0 0 0 0 0 unknown 91 0 0 0 0 0 0 0 0 0 0 1.05 0 0 0 0 0 0 0 0 unknown 92 0 0 3.86 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 unknown 93 0 0 0 0 0 0 0 0 0 0 0 0 30.81 0 0 0 0 0 0 Each value represents the per cent composition of a given compound for that species. Species codes: OS1=Trichocentrum stipitatum replicate 1; TS2=Trichocentrum stipitatum replicate 2; BC=Byrsonima crassifolia; OCH=Oncidium cheirophorum; BEE=Oils from a bee’s leg (Paratetrapedia sp.); SPO=Sigmatostalix portillae; SMA=Sigmatostalix marinii; SOX=Sigmatostalix oxyceras; OKL=Oncidium klotzschianum; TAS=Trichocentrum ascendens; LMI=Lockhartia micrantha; OCO=Ornithocephalus cochleariformis; ECR=Erycina cristagalli; OO1=Oncidium ornithorrhynchum replicate 1; OO2=Oncidium ornithorrhynchum replicate 2; SPI=Sigmatostalix picturatissima; OBI=Ornithocephalus bicornis; LOE=Lockhartia oerstedii; EPU=Erycina pusilla.

PAGE 64

APPENDIX B LOCALITIES AND VOUCHER NUMBERS FOR PLANT TAXA

PAGE 65

54 Taxon Localities Voucher* Byrsonima crassifolia Rich. ex Kunth. Chilibre Silvera-30 Erycina cristagalli (Rchb.f.) N.H.Williams & M.W.Chase La mesa del Valle, Panama Silvera-87 Erycina pusilla (L.) N.H.Williams & M.W.Chase Gamboa, Panama Silvera-64 Lockhartia micrantha Rchb.f. Chilibre, Panama Silvera-72 Lockhartia oerstedii Rchb.f. Whitten-99235 Oncidium cheirophorum Rchb.f. La Fortuna, Boquete, Panama Silvera-15 Oncidium klotzschianum Rchb.f. La Fortuna, Panama Silvera-20 Oncidium ornithorrhynchum Kunth. Whitten s.n. Ornithocephalus bicornis Lindl . exBenth. Campana Silvera-55 Ornithocephalus cochleariformis C.Schweinf. El Valle Silvera-43 Sigmatostalix marinii Kniger Ecuador Whitten-99264 Sigmatostalix oxyceras Kniger & J.G.Weinm Ecuador Whitten-99302 Sigmatostalix picturatissima Krnzl La Fortuna, El Valle Silvera-66 Sigmatostalix portillae Kniger Ecuador Silvera-s.n. Trichocentum ascendens (Lindl.) M.W.Chase & N.H.Williams Sona, Panama Silvera-89 Trichocentrum stipitatum (Lindl.) M.W.Chase & N.H.Williams Chilibre, Panama Silvera-88 *All vouchers are deposited at the University of Florida Herbarium (FLAS).

PAGE 66

APPENDIX C LOCALITIES AND VOUCHER NUMBERS FOR BEE TAXA Taxa Locality Vouchers * Centris bicornuta Mocsary Chilibre, Panama Silvera-9,12 Centris inermis Friese Chilibre, Panama Silvera-25,32,34,24,42 Centris longimana Fabricius Chilibre, Panama Silvera-28 Centris trigonoides Lepeletier Chilibre, Panama Silvera-15,63 Cephalotrigona capitata Smith Chilibre, Panama Silvera-13 Epicharis maculata Smith Chilibre, Panama Silvera-11,20 Epicharis rustica Olivier Chilibre, Panama Silvera-39,38 Eulaema nigrita Lepeletier Chilibre, Panama Silvera-44 Frieseomelitta nigra Lepeletier Chilibre, Panama Silvera-3,10,11 Paratetrapedia moesta Cresson Chilibre, Panama Silvera-36,23,3,1,2,14,6, Silvera-22,36,35,24,34,42 Paratetrapedia sp.1 Chilibre, Panama Silvera-7,23,36,55,57,21, Silvera-19,8,60,40,31,29, Silvera-4,21,56,30,58 Paratetrapedia sp. (undescribed) La Fortuna, Panama Silvera-F1, F2 Scaptotrigona luteipennis De la Torre Chilibre, Panama Silvera-49 Tetrapedia cf. maura Cresson Chilibre, Panama Silvera-17,8 Trigona angustula La Treille Chilibre, Panama Silvera-53,47,41,52 Trigona corvina Cockerell Chilibre, Panama Silvera-54,2,28,962,15,61, Silvera-5,31,33,45,18,46 Trigona fulviventris Gurin-Mneville Chilibre, Panama Silvera-5,12,30 Trigona fuscipennis Friese Chilibre, Panama Silvera-14,16,25 Trigona muzoensis Schwarz Chilibre, Panama Silvera-12,21,41,51,22 *Vouchers are kept in personal collection to be deposited at the Smithsonian Tropical Research Institute (STRI) Panama. 55

PAGE 67

LIST OF REFENCES Ackerman, J. D. 1986. Mechanisms and evolution of food-deceptive pollination systems in orchids. Lindleyana 1(2):108-113. Albuquerque, P. M. C., and M. M. C. Rego. 1989. Phenology of bees visiting murici trees (Byrsonima crassifolia, Malpighiaceae). Boletim do Museu Paraense Emilio Goeldi Serie Zoologia 5(2):163-178. Arditti, J. 1992. Fundamentals of orchid biology. J. Wiley and Sons, New York. Barros, M.A.G. 1992. Fenologia da florao, estrategias reproductivas e polinizao de espcies simptricas do gnero Byrsonima Rich (Malpighiaceae). Revista Brasileira de Biologia 52:343-353 Beardsell, D. V., M. A. Clements, J. F. Hutchinson, and E. G. Williams. 1986. Pollination of Diuris maculata R. Br. (Orchidaceae) by floral mimicry of the native legumes Daviesia spp. and Pultenea scraba R. Br. Australian Journal of Botany 34:165-173. Boyden, T. 1980. Floral mimicry by Epidendrum ibaguense (Orchidaceae) in Panama. Evolution 34:135-136. Buchmann, S. L. 1987. The ecology of oil flowers and their bees. Annual Reviews of Ecology and Systematics 18:343-369. Buchmann, S. L., and M. D. Buchmann. 1981. Anthecology of Mouriri myrtilloides (Melastomataceae: Memecyleae), an oil flower in Panama. Biotropica (Reproductive Botany Supplement) 13(2):7-24. Cane, J. H., G. C. Eickwort, F. R. Wesley, and J. Spielholz. 1983. Foraging, grooming and mate-seeking behaviors of Macropis nuda (Hymenoptera, Melitiidae) and use of Lysimachia ciliata (Primulaceae) oils in larval provisions and cell linings. The American Midland Naturalist 110(2):257-264. Carmona-Diaz, G. 2001. Mimetismo floral entre Oncidium cosymbephorum Morren (Orchidaceae) y Malpighia glabra L. (Malpighiaceae). Maestro en Neuroetologia. Universidad Veracruzana, Xalapa, Veracruz. Cingel, N. A. van der. 2001. An atlas of orchid pollination: America, Africa, Asia and Australia. A.A. Balkema, Rotterdam. 56

PAGE 68

57 Dafni, A. 1984. Mimicry and deception in pollination. Annual Reviews of Ecology and Systematics 15:259-278. Dafni, A., and Y. Ivri. 1981a. Floral mimicry between Orchis israelitica (Baumann and Dafni) (Orchidaceae) and Bellevalia flexuosa Boiss. (Liliaceae). Oecologia 49:229-232. Dafni, A., and Y. Ivri. 1981b. The flower biology of Cephalanthera longifolia (Orchidaceae)-pollen imitation and facultative floral mimicry. Plant Systematics and Evolution 137:229-240. Dodson, C. H. 1967. Relationships between pollinators and orchid flowers. Atas do Simpsio sbre a Biota Amaznica 5 (Zoologia):1-72. Dodson, C. H., and G. P. Frymire. 1961. Natural pollination of orchids. Missouri Botanical Garden Bulletin 49(9):133-152. Dressler, R. L. 1993a. Field guide to the orchids of Costa Rica and Panama. Cornell University Press, Ithaca, New York. Dressler, R. L. 1993b. Phylogeny and classification of the orchid family. Dioscorides Press, Portland, Oregon. Johnson, S. D. 1994. Evidence for Batesian mimicry in a butterfly-pollinated orchid. Biological Journal of the Linnean Society 53:91-104. Johnson, S. D. 2000. Batesian mimicry in the non-rewarding orchid Disa pulchra, and its consequences for pollinator behaviour. Biological Journal of the Linnean Society 71(1):119-132. Kjellsson, G., F.N. Rasmussen, and D. Dupuy. 1985. Pollination of Dendrobium infundibulum. Cymbidium insigne (Orchidaceae) and Rhododendron lyi (Ericaceae) by Bombus eximius (Apidae) in Thailand: a possible case of floral mimicry. Journal of Tropical Ecology 1:289-302. Manly, B. F. J. 2000. Multivariate statistical methods. A primer. Second edition. Chapman and Hall/CRC, New Zealand. Molau, U. 1988. Flora Neotropica. Scrophulariaceae. Part I. Calceolariaeae. New York Botanical Gardens, New York. Neff, J. L., and B. B. Simpson. 1981. Oil-collecting structures in the Anthophoridae (Hymenoptera): morphology, function, and use in systematics. Journal of the Kansas Entomological Society 54(1):95-123.

PAGE 69

58 Nierenberg, L. 1972. The mechanism for the maintenance of species integrity in sympatrically occurring equitant Oncidiums in the Caribbean. American Orchid Society Bulletin 41:873-882. Nilsson, L. A. 1983. Mimesis of bellfowers (Campanula) by the red helleborine orchid Cephalanthera rubra. Nature 305:799-800. Parra-Tabla, V., C. F. Vargas, S. Magaa-Rueda, and J. Navarro. 2000. Female and male pollination success of Oncidium ascendens Lindey (Orchidaceae) in two contrasting habitat patches: forest vs agricultural field. Biological Conservation 94:335-340. Pijl, L. van der., and C. H. Dodson. 1966. Orchid flowers: their pollination and evolution. University of Miami Press, Miami. Raman, S. 1989. The trichomes on the corolla of the ScrophulariaceaeII: Tribes Hemimeridae and Calceolarieae. Beitrage zur Biologie der Pflanzen 64:141-155. Reis, M. G., A. D. Faria, V. Bittrich, M. E. Amaral, and A. J. Marsaioli. 2000. The chemistry of flower rewards -Oncidium (Orchidaceae). Journal of the Brazilian Chemical Society 11(6):600-608. Roubik, D. W. 1989. Ecology and natural history of the tropical bees. Cambridge University Press, New York. Roy, B. A., and A. Widmer. 1998. Floral mimicry: a fascinating yet poorly understood phenomenon. Trends in Plant Science 4(8):325-330. Siegler, D. S., B. B. Simpson, C. Martin, and J. L. Neff. 1978. Free 3-acetoxy fatty acids in floral glands of Krameria species. Phytochemistry 17:995-996. Silverstein, R. M., G. C. Bassler, and T. C. Morrill. 1974. Spectrometric identification of organic compounds. Third edition. John Wiley and Sons, New York. Simpson, B. B., and J. L. Neff. 1981. Floral rewards: alternatives to pollen and nectar. Annals of the Missouri Botanical Garden 68:301-322. Simpson, B. B., and J. L. Neff. 1983. Evolution and diversity of floral rewards. Pages 142-159 in C. E. Jones and R. J. Little, editors. Handbook of experimental pollination biology. Scientific and Academic Editions, Van Nostrand Reinhold, New York. Simpson, B. B., J. L. Neff, and D. Siegler. 1977. Krameria, free fatty acids and oil-collecting bees. Nature 267:150-151.

PAGE 70

59 Simpson, B. B., D. S. Siegler, and J. L. Neff. 1979. Lipids from the floral glands of Krameria. Biochemistry, Systematics and Ecology 7:193-194. Steiner, K. E. 1989. The pollination of Disperis (Orchidaceae) by oil-collecting bees in southern Africa. Lindleyana 4:164-183. Steiner, K. E. 1991. Oil flowers and oil bees: further evidence for pollinator adaptation. Evolution 45(6):1493-1501. Steiner, K. E., V. B. Whitehead, and S. D. Johnson. 1994. Floral and pollinator divergence in two sexually deceptive South African orchids. American Journal of Botany 81(2):185-194. Steiner, K. E. 1998. The evolution of beetle pollination in a South African orchid. American Journal of Botany 85(8):1180-1193. Vinson, S. B., H. J. Williams, G. W. Frankie, and G. Shrum. 1997. Floral lipid chemistry of Byrsonima crassifolia (Malpighiaceae) and a use of floral lipids by Centris bees (Hymenoptera: Apidae). Biotropica 29(1):76-83. Vogel, S. 1969. Flowers offering fatty oil instead of nectar (Abstract). XI Proceedings of the International Botanical Congress:229. Vogel, S. 1974. lblumen und lsammelnde Bienen. Akademie der Wissenschaften und der Literatur Mathematisch-Naturwissenschaftliche Klasse, Mainz, Franz Steiner Verlag, Wiesbaden. Vogel, S. 1976. Lysimachia: lblumen der Holarktis. Naturwissenschaften 63:44-45. Wcislo, W. T., and J. H. Cane. 1996. Floral resource utilization by solitary bees (Hymenoptera: Apoidea) and exploitation of their stored foods by natural enemies. Annual Reviews of Entomology 41:195-224. Wilkinson, L., M. Hill, and E. Vang. 1992. SYSTAT: Statistic, Version 5.2 edition, Evanston, Illinois. Williams, N. H. 1982. The biology of orchids and euglossine bees. Pages 119-171 in J. Arditti, editor. Orchid biology: reviews and perspectives II. Cornell University Press, Ithaca, New York. Williams, N. H., M. W. Chase, T. Fulcher, and W. M. Whitten. 2001. Molecular systematics of the Oncidiinae based on evidence from four DNA sequence regions: expanded circumscriptions of the Cyrtochilum, Erycina, Otoglossum and Trichocentrum and a new genus (Orchidaceae). Lindleyana 16(2):113-139.

PAGE 71

60 BIOGRAPHICAL SKETCH Katia Itzel Silvera was born in Panama City, Republic of Panama, on January 31, 1974. As a university student, she began working with the Smithsonian Tropical Research Institute (STRI), and shortly af terward, in 1998, she received a Bachelor of Biology with specialization in plant biology from the University of Panama. Following graduation, she worked as a research assistant with Dr. Jim Dalling in Barro Colorado Island (BCI) who served as a co-advisor for her undergraduate thesis. Katia has been involved in several projects in STRI including plant physiology at the canopy crane, seedling ecology and water relations among others. She has also been involved in the visitors program working as a natural history interpreter for the public at BCI for several years. Her positive undergraduate experience, encouragement from her advisors, and interest in orchids brought her to the Florida Museum of Natural History to volunteer with Dr. Robert Dressler working in orchid systematics. She remained in the University of Florida Department of Botany to pursue a master’s degree with Dr. Norris H. Williams, Dr. Robert Dressler and Dr. Mark Whitten, who are all world experts on neotropical orchid evolution and systematics. Katia plans to continue studying the evolution of orchids and their pollinators in the tropics, especially in Panama.