Citation
Differences in heterospecific pollen deposition across pollination syndromes in the Monteverde area

Material Information

Title:
Differences in heterospecific pollen deposition across pollination syndromes in the Monteverde area
Translated Title:
Diferencia en la deposición de polen heteroespecífico entre síndromes de polinización en Monteverde, Costa Rica
Creator:
Powell, Kara
Publication Date:
Language:
Text in English

Subjects

Subjects / Keywords:
Pollination ( lcsh )
Polinización ( lcsh )
Pollinators ( lcsh )
Polinizadores ( lcsh )
EAP Fall 2017
EAP Otoño 2017
Costa Rica--Puntarenas--Monteverde Zone
Costa Rica--Puntarenas--Zona de Monteverde
Genre:
Reports

Notes

Abstract:
Pollination is an important ecosystem service, increasing the ability of plants to adapt to changing environments through the process of sexual reproduction. The efficiency of pollination can be reduced due to the transfer of heterospecific pollen, which is pollen from another species. Flower characteristics vary in order to attract certain types of pollinators. This phenomenon is classified in groups of flower morphology characteristics known as pollination syndromes. Little research exists that summarizes heterospecific pollen transfer within these pollination syndromes. I investigated pollen deposition across pollination syndromes in Monteverde, Costa Rica in two different sites: San Gerardo and Curi Cancha. I mounted stigmas from each reserve onto microscope slides in order to count the amounts of both conspecific and heterospecific pollen. I then classified each species into a certain pollination syndrome using characteristics defined by the USDA. I divided the plant species in my data into two categories based on their strategy for attracting pollinators: generalist and specialist. I found that there was no significant difference in the amount of conspecific and heterospecific pollen deposition between generalist and specialist plant species. Pollen deposition is likely more affected by floral traits that are indistinguishable when classifying plants into generalist and specialist categories. Thus, heterospecific pollen transfer (HPT) is affected by more factors than the specialization of floral morphology. ( , )
Abstract:
La polinización es importante porque aumenta la diversidad genética de las plantas. La eficiencia de la misma puede ser reducida por polen heterospecífico, es decir de otras especies, que es depositado en el estigma. Las flores pueden presenter diferentes características según su polinizador. Este conjunto de características relacionados a la atracción del polinizador se conoce como síndrome de polinización. Existen pocos estudios que analizen el efecto de diferentes síndromes de polinización sobre la transferencia de polen heteroespecífico. En este proyecto investigué la deposición de polen heteroespecífico según el síndrome de polinización de diferentes flores en dos sitios de Monteverde, Costa Rica. Conté la cantidad de polen conspecífico y heteroespecífico en los estigmas de diferentes especies de flores. Clasifiqué cada especie según su síndrome de polinización basado en las características de las flores establecidas por el USDA. Categoricé las especies de plantas en dos categorías según su grado de especialización para atraer polinizadores: especialista y generalista. La cantidad de polen conspecífico y heterospecífico depositado en los estigmas no difirió entre las especies generalistas y las especialistas. Ese resultado sugiere que la deposición del polen heteroespecífico puede estar más determinada por características florales que no estén relacionadas con los síndromes de polinización, por ejemplo tamaño del estigma o largo del pistilo. Así mismo pueden existir otros factores no relacionados con características florales que tengan un efecto sobre la cantidad de polen heteroespecífico depositado, por ejemplo densidad de flores y biología de los polinizadores.
Biographical:
Student affiliation: Division of Biological Sciences, University of California, San Diego.

Record Information

Source Institution:
Monteverde Institute
Holding Location:
Monteverde Institute
Rights Management:
This item is licensed with the Creative Commons Attribution Non-Commercial No Derivative License. This license allows others to download this work and share them with others as long as they mention the author and link back to the author, but they can’t change them in any way or use them commercially.
Resource Identifier:
M39-00657 ( USFLDC DOI )
m39.657 ( USFLDC Handle )

USFLDC Membership

Aggregations:
Tropical Ecology Collection [Monteverde Institute]

Postcard Information

Format:
Book

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

Powell 1 Differences in Heterospecific Pollen Deposition a cross Pollination Syndromes in the Monteverde Area Kara Powell University of California, San Diego Division of Biological Sciences EAP Tropical Biology and Conservation, Fall 2017 15 December, 2017 ABSTRACT Pollination is an important ecosystem service, increasing the ability of plants to adapt to changing environments through the process of sexual reproduction. The efficiency of pollination can be reduced due to the transfer of heterospecific pollen, which is pollen from another species. Flower characteristics vary in order to attract certain types of pollinators. This phenomenon is classified in groups of flower morphology characteristics known as pollination syndromes. Little research exists that summarizes heterospecific pollen transfer within these pollination syndromes. I investigated pollen de position across pollination syndromes in Monteverde, Costa Rica in two different sites: San Gerardo and Curi Cancha. I mounted stigmas from each reserve onto microscope slides in order to count the amounts of both conspecific and heterospecific pollen. I t hen classified each species into a certain pollination syndrome using characteristics defined by the USDA. I divided the plant species in my data into two categories based on their strategy for attracting pollinators: generalist and specialist. I found tha t there was no significant difference in the amount of conspecific and heterospecific pollen deposition between generalist and s pecialist plant species. Pollen deposition is likely more affected by floral traits that are indistinguishable when classifying plants into generalist and specialist categories. Thus, h eterospecific pollen transfer (HPT) is affected by more factors than the specialization of floral morphology. Dife rencia en la Deposici—n de Polen H eteroespec’fico entre S’ndromes de P olinizaci—n e n Monteverde, Costa Rica RESUMEN La polinizaci—n es importante porque aumenta la diversidad genŽtica de las plantas. La eficiencia de la misma puede ser reducida por polen heterospec’fico, es decir de otras especies, que es depositado en el estigma. Las flores pueden presenter diferentes caracter’sticas segœn su polinizador. Este conjunto de caracter’sticas relacionados a la atracci—n del polinizador se conoce como s’ndrome de polinizaci—n. Existen pocos estudios que analizen el efecto de difer entes s’ndromes de polinizaci—n sobre l a trans ferencia de polen heteroes pec’fico En este proyecto investiguŽ la deposici—n de polen heteroespec’fico segœn el s’ndrome de polinizaci—n de diferentes flores en dos sitios de Monteverde, Costa Rica. ContŽ la cantidad de polen conspec’fico y hetero e spec’fi co en los estigmas de diferentes especies de flores. ClasifiquŽ cada especie segœn su s’ndrome de polinizaci—n basado en las caracter’sticas de las flores establecidas por el USDA. CategoricŽ las especies de plantas en dos categor’as segœn su grado de espe cializaci—n para atraer polinizadores: especialista y generalista. La cantidad de polen conspec’fico y heterospec’fico depositado en los estigmas no difiri— entre las especies generalistas y las especialistas. Ese resultado sugiere que la deposici—n del po len heteroespec’fico puede estar m‡s determinada por caracter’sticas florales que no estŽn relacionadas con los s’ndromes de polinizaci—n, por ejemplo tama–o del estigma o largo del pistilo. As’ mismo pueden existir otros factores no relacionados con cara cter’sticas florales que tengan un efecto sobre la cantidad de polen heteroespec’fico depositado, por ejemplo densidad de flores y biolog’a de los polinizadores.

PAGE 2

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 2 Pollination allows plants to achieve sexual reproduction, increasing their ability to adapt to changing environments. Without pollinators, plants would be forced to create clones of themselves in order to reproduce, which reduces the amount of genetic variability in a given population and increases the odds that a useful combination of gene s will be available in a time of need. Due to its importance in both natural and agricultural systems, pollination biology is a topic in need of further research. It is a complex explanation of pollination services since pollinators and their interactions with plants are extremely diverse (Waser 2006) Many flowers have morphological characteristics that are specialized to attract certain types of pollinators (Fenster 2004) This occurrence of specialization for certain types of pollinators led to the esta blishment of pollinator syndromes, which are groups of characteristics that are meant to cater to a certain type of pollinator. Examples of pollinator syndromes include bee pollinated flowers, hummingbird pollinated flowers, and moth pollinated flowers. So me of these pollination syndromes are more specific than others, restricting the amount of overlap between types of possible pollinators. Pollinators can be generalists, visiting many different species of flowering plants, or they can focus on one or a sma ll group of species as specialists Generalist flowers are those that are visited by generalist pollinators, and they are more likely to receive visits from a wider array of pollinators than specialist flowers (Waser 2006). Generalist flowers are those tha t have a reward, such as nectar, that is accessible and attractive to multiple types of pollinat ors. Bee pollinated flowers are considered to be more generalist than other types o f pollination syndromes (Menzel 1993). G eneralist flowers often resemble bee pollinated flowers in their morphological structure. Bees pollinate a larger number of plant families than other types of p ollinators. In contrast, specialist flowers harbor rewards that are only accessible and attractive to a handful of pollinators. H umm ingbird pollinated flowers are thought to be more specialist, since the visitors to the flowers are unable to reach the nectar rewards without specific morphologies such as l ong beaks or proboscis (Fenster 1991). One of the plant species in my study, Razis ea spicata is pollinated by long billed hummingbirds. In a study comparing the deposition of heterospecific pollen between patches of R. spicata of different densities, researchers found that heterospecific pollen deposition was sporadic in flowers pollinated by long billed hummingbirds (Feinsinger 1986). This study aims in part to build on this previous research and focus on other factors that were not measured, discussed below. Due to the diversity of flowers visited, generalist pollinators are mo re likely to deposit heterospecific pollen (pollen from a species other t han the flower being visited). H eterospecific pollen can reduce or inhibit reproduction success because high amounts of heterospecific pollen on pollinators can decrease the amount o f space for conspecific pollen. H eterospecific pollen on the stigma can form blockages that prevent pollen tubes from generating, altering female reprod uctive success (Wesselingh 2000). In a study that measured the effect of heterospecific pollen on the sti gmas of flowers found in the understory of the Costa Rican cloud forest, it was found that less pollen tubes were generated in the presence o f heterospecific pollen (Murcia 1996). The transfer of heterospecific pollen can also be detrimental to the male fi tness of a plant. Pollinators that visit multiple species will carry multiple types of pollen, reducing the amount of pollen for any given species that the pollinator has visited. This reduces the potential for male reproductive success for all species inv olved (McLernon 1996). In addition, a higher prevalence of heterospecific pollen has been shown to reduce seed output, thus redu cing plant fitness (Waites 2004). In a study conducted in an alpine meadow in Colorado in which two hummingbird pollinated speci es' flowering periods overlap for some time, seed set is reduced

PAGE 3

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 3 during th e period of overlap (Kohn 1985). This may be due to the factors described above, such as interspecific pollen competition and the presence of heterospecific pollen on the stigma. In an attempt to reduce the transfer of heterospecific pollen, plants within a community may have evolved characteristics meant to attract a certain type of pollinator and prevent others from exploiting the reward, such as nectar or perfu mes (Waser 1983). Th is phenomenon is termed the Sexual Architecture Hypothesis. This hypothesis can also refer to different sized floral structures involved in pollination, such as different leng th styles and filaments (Murcia 1995). R. spicata has similar length filaments (s tructures that support the pollen producing anthers) across individuals, but the lengths of the styles (structures that support the pollen receiving stigmas) differ between individuals. The lengths of the styles can be placed along a continuum, which exclu des the possibility of heterostyly. Previous studies have suggested that stigma position is an important floral trait in determining the amount and type of pollen received (Qiang 2013). The second objective of this study is to determine if the sporadic tra nsfer of heterospecific pollen in R. spicata is concentrated in certain lengths of styles. It is important to refrain from testing multiple species in order to reduce the amount of possible confounding variables in my study. Within natural systems, there m ight be natural selection pressure to evolve differing flower morphologies in order to reduce the amount of heterospecific pollen transfer. The third objective of this study is to compare the differing levels of heterospecific pollen present within flowers that belong to certain pollination syndromes. For example, hummingbird pollinated flowers are tubular and require the pollinator to be able to reach the nectar reward at the bottom of the tube. This entices pollinators with long beaks or probosces, thus e xcluding certain pollinator morphologies from being useful for pollination of flowers in this syndrome (an example of a specialist) The question that I am interested in answering with this study is: "Do generalist pollination syndromes tend to accumulate higher amounts of heterospecific pollen?" I can speculate that more specialized syndromes, such as hummingbird pollinated flowers, will contain less heterospecific pollen than syndromes that appeal to more generalist pollinators (as in bee pollinated flowe rs). Additionally, I am asking the question, "Does style length in R. spicata affect the amount of heterospecific pollen deposited?" My prediction was that style length will alter the presence and amount of heterospecific pollen deposition, with certain le ngths having higher amounts of heterospecific pollen than others. MATERIALS AND METHODS I conducted the study in Monteverde, Costa Rica, collecting only from reserves in order to prevent using species that were planted for use as ornamental foliage. I collected flower samples over a two week period from 13 November 2017 to 27 November 2017. I visited two sites, the San Gerardo Research Station (at the Children's Eternal Rainforest) and Curi C ancha Reserve (San Gerardo is Lower Montane Rain Forest at 1550 1850 m elevation, Curi Cancha is Lower Montane Wet Forest at 1450 1600 m elevation.) In each site I walked multiple trails and collected the flowers that were visible from the trail and not dangerous to acc ess (usually within two meters from the ground). I usually collected all flowers that were available, but if they were present in excess I would limit myself to approximately four flowers. I collected flowers from

PAGE 4

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 4 the ground as long as the parent plant cou ld be identified. I covered 10.7 kilometers of trail in San Gerardo and 6.2 kilometers of trail in Curi Cancha. After collecting the flowers and taking photos and samples of the plant specimens, I mounted the anthers and stigmas onto microscope slides usi ng clear nail polish. I created a preserved anther library in order to be able to identify the types of pollen present on each stigma as either conspecific or heterospecific. I then pressed the stigmas from each flower onto the slide using a drop of nail p olish and a coverslip. I recorded the type of habitat that the plant was located in (forest, pasture, forest edge) For the species R. spicata I also recorded the length of the style. When collecting stigmas, I collected as many stigmas as were available from the individual plant and mounted them on the slide together. This will be referred to as a sample. In order to normalize the data, I divided the raw counts by the number of stigmas collected in each sample, and then by the total number of individuals collected in each site. Upon returning from the field, I counted the number of conspecific grains and heterospecific grains present in each sample using a light microscope. To normalize the amounts of pollen present in each sample, I divided the total co unts for both con and heterospecific pollen by the number of stigmas in each sample. I identified the species of flower to the best of my ability using A Field Guide to Plants of Costa Rica ( Garguillo 2008 ) and A Guide to Tropical Plants of Costa Rica (Z uchowski 2006) I enlisted the help of a local conservationist, Eladio Cruz to assist in identifying the flowers that I was unable to. I divided the plant species that I collected into four different pollination syndromes (bee, beetle, bird, and butterfl y). I classified them using the information available in the literature for previously studied species that already had a known pollination syndrome and then I classified the remaining unstudied species using the criteria listed in the appendix (Table 4). I did not witness many pollinators on the flowers, so I was unable to use that additional information. Once the plant species were divided in this way, I categorized the pollination syndromes into two groups: specialist and generalist. Specialist flowers i nclude bird pollinated flowers and butterfly pollinated flowers, while generalist flowers include bee pollinated flowers and beetle pollinated flowers. This is because pollinators require specific morphologies in order to access the nectar rewards of speci alist flowers, which birds and butterflies have (long proboscis). Specialist flowers are defined by the restriction of access to the reward, such as long floral tubes that only allow access to nectar for species with the appropriate morphological character istics. Generalist flowers are those that do not fit into the above description or have been recorded with a large diversity of floral visitors open flowers, accessible rewards, etc RESULTS General Results In total, I collected 581 stigmas, 251 from San Gerardo and 330 from Curi Cancha. I collected 20 unique species from San Gerardo and nine species from Curi Cancha. I collected from 167 individual plants. Most of the specialist flowers in my dataset are bri ght red and tubular, thus, pollinated by hummingbirds. R. spicata was the most common flower available to

PAGE 5

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 5 collect at the time of my study, which is reflected in the amount of stigmas and individuals from this species in my dataset. There were a few species that only were present in one of the sites. Cococicelum pubescens and Fabaceae were only present in Curi Cancha, while Blachea Calathea Commelinaceae, Melastomataceae, Miconia Monochaetum vulcanicum Solanum and Phytolacca were only present in San Ger ardo. The distribution of stigmas and individuals collected in both sites can be found in Figure 1: Figure 1: Number and sample summary of individuals collected across both study sites. Number of stigmas collected per taxonomic group across both study sites. The amount and type of pollen grains found varied greatly between species. The highest amounts of conspecific pollen deposition occurs in Centropogon Columnea consanguinea other Gesneriaceae, Poaceae, and Malvaviscus palmanus Heterospecific pollen deposition in San Gerardo was concentrated in Centropogon Columnea consanguinea and other species of Columnea which are specialist flowers. Centropogon received approximately 11 times more conspecific pollen than heterospecific pollen Conspecifi c pollen transfer was approximately 11 more common than heterospecific pollen transfer. The majority of species received little to no pollen at all. Figure 2 shows these amounts.

PAGE 6

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 6 Figure 2. Types and amounts of pollen deposition across the two sites in the study. Each graph displays only the species that were present in that particular site. Error bars display standard deviation. Pollen Deposition in Specialist and Generalist Flowers Figure 3 shows that f lowers classified as generalist received relatively the same amount of conspe cific and heterospecific pollen

PAGE 7

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 7 Figure 3. A comparison of type and amount of pollen deposition between generalist and specialist plant species. Error bars display standard deviation. Pollen Deposition in Razisea s picata In total, I collected 56 stigmas from R. spicata I divided the stigmas into categories based on the length in centimeters that the styles protruded past the anthers. Figure 4a shows the number of stigmas collected from each length category. There were four categories total: equal length with anthers, 0.5 cm past anthers, 1 cm past anthers, and 2 cm past anthers. The anthers were in relatively the same position across individuals plants. Individual plants with equal length styles and filaments and i ndividual plants with styles extending one centimeter past the anthers were the most common, while p lants with a two centimeter difference between these two flower structures were the least common. Figure 4b shows the number of pollen grains deposited in each length category. The stigmas in the "equal length" category received the most conspecific pollen. There is a decreasing trend as the distance between the stigma and the anthers increases. The stigmas in the two centimeter category received very littl e conspecific pollen. The only category that presented heterospecific pollen is the "equal length" category, although there was not much present in the sample.

PAGE 8

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 8 a) b) Figure 4. An outline of the results of my sub question regarding heterospecific pollen transfer in Razisea spicata Figure 2a summarizes the number of stigmas collected in each category of style length. Figure 2b summarizes the amount and type of pollen de position within the two categories. DISCUSSION Differences in Heterospecific Pollen Transfer Between Species R. spicata boasts the largest number of individuals and stigmas collected in my dataset because it was the most abundant plant species in bo th sites with flowers that were feasible to collect. Other species may have been more abundant during this time, but unable to be collected, such as epiphytic species. Species composition differs between the two sites. Heterospecific pollen deposition in San Gerardo was concentrated in hummingbird pollinated species with similar floral morphology: Columnea consanguinea Columnea sp ., and Centropogon Columnea is a genus in the family Gesneriaceae giving this family the highest amount of pollen transfer in my dataset. This could mean that Gesneriaceae is more efficient at attracting pollinators. Columnea consanguinea has been witnessed to have hum mingbird pollinators (RodrÂ’guez 2009). It is safe to assume that related plants are likely visited by hummingbir ds as well due to similar flower morphologies. Hummingbirds likely visited each of these types of flowers within a single foraging period in order to transfer this amount of heterospecific pollen. In Curi Cancha, the flowers that received the most heterospecific pollen were previously classified as generalists. Poaceae and Ruvos both have small, white flowers that are attractive to bees. Heterospecific Pollen Transfer in Generalist and Specialist Pollination Syndromes

PAGE 9

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 9 There are many factors to co nsider when thinking about why heterospecific pollen deposition is relatively low across most species. It could be argued that the most important factor is phenology. The tropics provide an environment in which plants have the resources to produce flowers virtually year round, which allows plants to flower at different times of the year in order to reduce competition for space on pollinators. Another factor is different floral morphologies, which is discussed in this paper. The diversity of pollinators and their foraging habits also plays a role in reducing the amount of heterospecific pollen transfer. Monteverde is a relatively diverse area; even within the hummingbird pollinators there are two separate guilds of pollinators. Long billed hummingbirds and sh ort billed hummingbirds have different foraging behaviors and thus carry diff erent species of pollen (Murray 1987). This study also notes that many flowers in the area are unable to reach their maximum seed set with the amount of pollen deposited on their stigmas. This low amount of conspecific pollen deposition is reflected in my data, for which very few species have large amounts of any type of pollen deposition. Specifically, the majority of my samples of R. spicata did not have pollen at all. Hummingbir d pollinated flowers are known to have extr emely low visitation rates (Kay 2003), which may have contributed to this result. Figure 3 summarizes the res ults pertaining to the comparison between specialist and generalist syndromes in the amount of heterospe cific pollen received The amount of heterospecific pollen deposited does not display a statistically significant difference. However, there is a possible explanation. Many of the specialist flowers in my dataset had large stigmas that extended past the op ening of the flower ( Centropogon and Malvaviscus especially). Large, protruding stigmas are more likely to come into contact with the bodies of pollinators and thus are more exposed to pollen deposition. In Figure 2, the species that receive the majority o f pollen in the sample are Centropogon and Malvaviscus palmanus Finding more pollen on larger stigmas has occurred in a previous study on a diverse alpine m eadow in southwest China (Qiang 2013). A second possible explanation for why specialist flowers re ceived more heterospecific pollen than I expected is their distribution within the forest. Many of the specialized flowers in this study were dispersed intermittently throughout the forest, with individuals being separated from each other for quite some di stance (with the exception of R. spicata which was found in intermittent clusters). Since specialist flowers are more spread out in their distribution, a specialist pollinator likely visits many flowers of a different species as they travel through the fo rest. Specialization has been thought to evolve in the circumstance of low focal plant density in order to reduce heterosp ecific pollen transfer (Sargent 2006). However, my study found that the amount of heterospecific pollen was not significantly differen t between specialist and generalist flowers. Perhaps without specialization, flowers with intermittent distributions would have a much higher rate of heterospecific pollen transfer. Specialization may reduce HPT to relatively the same level experienced by generalist flowers. Flowers that were classified as generalist may not have been visited by as many generalist pollinators as I originally expected. Even if these flowers were visited by generalist pollinators, there may have been unique opportunities for the pollinators to forage on the same species for extended periods of time. Most of the generalist flowers in this study were found in large clusters where a pollinator could visit multiple individuals of the same species, eliminating the possibility of h eterospecific pollen transfer. Open areas of forest tended to harbor large populations of generalist flowers, usually clustered by species. Due to the way flowers were

PAGE 10

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 10 distr ibuted in the sites that I visited, i t is possible that floral distribution may hav e an more of an effect on heterospecific pollen deposition than I originally thought. Pollinator foraging behavior is another possible factor that may have influenced my results. In a study comparing the foraging behaviors between a generalist pollinator, the bumblebee, and a specialist pollinator, the butterfly, each type of pollinator had different foraging strategies. Bumblebee flights remained relatively short, with individuals preferring to visit the next closest flower to the one previously visited. In contrast, butterflies were more likely to skip flowers and fly longer distances in between floral visits (Schmitt 1980). Due to this behavior, the generalist pollinator may be reducing the amount of heterospecific pollen transfer if the flowers are clus tered by species. Behavior differences between generalist and specialist pollinators were factors that I did not consider when making my origina l prediction. Additionally, the possibility that my classifications of specialist and generalist flowers may be too broad to be able to accurately make predictions on heterospecific pollen transfer. Taking these explanations into consideration, we must accept that there are more factors influencing heterospecific pollen transfer than just the level of specialization of floral morphology. There is debate in the field of pollination biology as to whether or not the specialization of floral syndromes is effective in limiting the types of pollinators th at visit the flowers (RodrÂ’guez 2009). Other factors, other than the morphological characteristics that separate flowers into different pollination syndromes, may have an effect on pollinator attraction. According to Stein Joar Hegland (2005), flower size is a characteristic independent of pollination syndrome that alters the attractiveness of a flower to a pollinator. Pollinators tend to prefer larger flowers rather tha n smaller ones (Conner and Rush 1996). This controversy must be taken into consideration when trying to explain why the levels of heterospecific pollen diff er between pollination syndromes. Pollen Deposition in Razisea spicata While conducting my study, I became interested in the heterospecific pollen within R. spicata due to the continuum along w hich style length differs. I found that t he amount of polle n deposition decreases as the length of the style increases. This could be due to deposition of pollen from the anthers of the same flower. Self pollination reduces the amount of genetic diversity in the population by preventing the possibility of genetic exchange. As a breeze passes through the forest or an animal brushes against the plant, pollen can be shaken off the anthers, and it is more likely to be deposited on the stigma if the stigma is closer. Previous studies have shown that R. spicata receives large amou nts of self pollen (Lindhart 1987). Since R. spicata is pollinated by long billed hummingbirds that forage over long distances, pollen dispersal is widespread. However, it must be taken into account that not all flowers in this sample experienced deposition solely with pollen originating from the same plant. Since the "equal length" category received some heterospecific pollen, there is proof that pollinators are transferring pollen between individuals. There is a possibility that the conspecific pollen in the "equal length" category contains some grains that were transferred through a pollinator. The likelihood of pollen transfer increases as the style length inc reases, since a stigma that dangles two centimeters or greater from the anthers is unlikely to come into contact with pollen from its own flower. Future Research Possibilities

PAGE 11

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 11 There are several things that could be done to extend this study This study was limited in the amount of time and resources that were available. Given more of each, the question could be investigated further and several sources of error may be avoided. For example, I was only able to collect flowers that I could reach. This metho d of collection excludes a large portion of the flowers in the forest, especially those of trees and epiphytes. Further research should be completed with a more diverse array of flower collection methods. This study could also benefit from being repeated i n different seasons in order to capture information about flowers that bloom in different times of the year. Conclusion Overall the amount of heterospecific pollen that I found was low across most pollination syndromes. I expected to find more heteros pecific pollen because of my own previous exper ience and other research (Qiang 2013). However, it is important to note that these other studies were conducted in temperate zones where flowering is usually a seasonal occurrence. In this situation, heterospe cific pollen transfer may be a more influential factor in plant reproductive success than in the tropics. Since there is less seasonal variation in weather in tropical forests, plants are able to produce flowers throughout the entire year. This reduces the amount of overlap in flowering periods where heterospecific pollen transfer is possible. There was little heterospecific pollen deposition in R. spicata I originally thought that the style lengths in this species likely differed in order to reduce the amount of heterospecific pollen transfer. After learning that heterospecific pollen transfer is a rare occurrence, I now wonder if it is an attempt to reduce self pollination. The stigmas that are separated from the anthers by more than one centimeter had a relatively low chance of self pollen deposition compared to stigmas that were placed directly next to the source of pollen. Further research is needed, per haps investigating seed set in the different morphotypes, in order to confirm that this is why the different style lengths exist. ACKNOWLEDGMENTS I would like to give acknowledgement to the EAP instructors (Frank Joyce, Federico Chinchilla, Sof’a Fl ores and FŽlix Salazar), especially my advisors, AndrŽs Camacho and Emilia Triana, and Eladio Cruz for assistance in flower identification. Special thanks to the staff at the Estaci—n Biol—gica and at the reserves, and to the Rodr’guez Horner family for ho using me during my data collection. I am grateful to Noelle Pruett for peer reviewing my paper. Thanks to Miguel Ochoa for accompanying me to San Gerardo a second time. I was also lucky to have a positive working environment and support network thanks to m y fellow students, they inspired and motivated me throughout the duration of the program.

PAGE 12

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 12 LITERATURE CITED Bawa, K. S. 1990. "Plant pollinator interactions in tropical rain forests." Annual Review of Ecology and Systematics 21: 399 422 Conner J.K., Rush S. 1996. "Effects of flower size and number on pollinator visitation to wild radish, Raphanus raphanistrum ." Oecologia 105:509 516 Feinsinger, P., Murray, K. G., Kinsman, S., & Busby, W. H. 1986. "Floral neighborhood and pollination suc cess in four hummingbird pollinated cloud forest plant species." Ecology 67(2), 449 464. Fenster, C. 1991. "Selection on floral morphology by hummingbirds." Biotropica 23:98 101. Fenster. 2004. "Pollination syndromes and floral specialization." Annual Review of Ecology, Evolution, and Systematics. 35:375 403 Garguillo, M.B., Kimball, L., and B.L Magnuson. 2008. A Field Guide to the Plants of Costa Rica. Oxford University Press, New York, NY. Hegland, Totland. 2005. "Relationships between speci es' floral traits and pollinator visitation in a temperate grassland." Oecologia 145:586 594 Holdridge, L. R. 1967. Life Zone Ecology. Tropical Science Center, San Jose, Costa Rica. Kay, Shemske. 2003. "Pollinator assemblages and visitation rates for 1 1 species of neotropical Costus ." Biotropica 35:198 207. Kohn, J., Waser, N. 1985. The effect of Delphinium nelsonii pollen on seed set in Ipomopsis aggregata, a Competitor for Hummingbird Pollination American Journal of Botany 72:1144 1148. Linhart. 1987. "Forager behavior, pollen dispersal, and inbreeding in two species of hummingbird pollinated plants". Evolution 41:679 682. McLernon, Murphy, Aarssen. 1996. "Heterospecific pollen transfer between sympatric species in a mid successional old field community." American Journal of Botany 83: 1168 1174. Menzel R., Greggers U., Hammer M. 1993. "Functional organization of appetitive learning and memory in a generalist pollinator, the honey bee." In: Papaj D.R., Lewis A.C. (eds) Insect Learning. Springe r, Boston, MA Murcia, C., and P. Feinsinger. 1996. Interspecific pollen loss by hummingbirds visiting flower mixtures: effects of floral architecture. Ecology 77:550 560.

PAGE 13

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 13 Murray, Feinsinger, Busby, Linhart Beach, Kinsman. 1987. "Evaluation of character displacement among plants in two tropical pollination guilds." Ecology 68:1283 1293. Qiang Fang, Shuang Quan Huang. 2013. "A directed network analysis of heterospecific pollen transfer in a biodiverse com munity." Ecology 94: 1176 1185. RodrÂ’guez, S. 2009. "Evaluation of pollination syndromes in Antillean Gesneriaceae: evidence for bat, hummingbird, and generalized flowers." Journal of Ecology 97:348 359. Schmitt, J. 1980. "Pollinator foraging behavior a nd gene dispersal in Senecio (Compositae)". Evolution. 34:934 943 USDA. 2014. "Animal Pollination." USDA Forest Service. United States Department of Agriculture.
PAGE 14

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 14 APPENDIX Appendix 1: San Gerardo Collected Sample Most Specific Name Stigmas Collected Hetero/Stigma# Con/Stigma# Columnea consanguinea 8 148.5 350 Malvaviscus palmanus 20 0 423.66 Gesneriaceae 18 2.444 348 Tradescantia zanonia 7 2 7.5 Solanum 1 0 0 Aphelandra 19 0 43.083 Asteraceae 38 0 96.217 Monochaetum vulcanicum 1 20 1 Centropogon 8 92 966 Drymonia conchocalyx 2 0 0 Poaceae 2 0 0 Melastomataceae 3 0 0 Razisea spicata 74 0.591 6.05 Calathea 5 0 0 Miconia 19 0 2.916 Blachea 5 0 0.4 Tibouchina urvilleana 2 0 0 Columnea 7 25 75 Phytolacca 12 0 22 Idalgoa cuaternata 5 0 0 Total 251 290.535 2375.159 Overall Totals 581 377.332 4159.094 Appendix 2: Curi Cancha Reserve Collected Sample Most Specific Name Stigmas Collected Hetero/Stigma# Con/Stigma# Malvaviscus palmanus 16 0 816 Razisea spicata 170 0 50.95 Columnea consanguinea 1 0 200 Columnea 5 0 181.667 Poaceae 16 69.036 432.643 Asteraceae 56 1.109 54.642 Ruvos 44 16.652 0 Fabaceae 17 0 20.033 Cococicelum puvesens 5 0 28 Total 330 86.797 1783.935

PAGE 15

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 15 Overall Totals 581 377.332 4159.094 Appendix 3: Life Zones of the Reserves (Holdridge, 1967) Life Zone Elevation (m) Mean Annual Rainfall (cm) Mean Annual Temperature ( ) Reserve Lower Montane Rain Forest 1550 1850 3600 8000 12 17 San Gerardo Lower Montane Wet Forest 1450 1600 1850 4000 12 17 Curi Cancha Appendix 4: Pollinator Syndrome Identification Chart -USDA Trait Ant Bat Bee Beetle Bird Butterfly Fly Moth Wind Color Inconspi cuous White, green, purple Bright white, yellow, blue, UV White, green Scarlet, orange, red, white Bright red or purple Pale, dark brown, or purple Pale red, purple, pink, white Pale green, brown, or colorless Nectar Guides none present none none present none none none Odor none/fai nt Strong, musty, emitted at night Fresh, mild, pleasant None to strongly fruity or foul none Faint, fresh Putrid Strong, sweet, emitted and night none Nectar Abunda nt, some what hidden Usually present Someti mes present Ample, hidden Ample, hidden Usually absent Ample, hidden none Pollen Ample Limited, often sticky, scented Ample Limited Limited Limited Limited Abundant, small, smooth Flower Shape Small, low growing close to stem Bowl shaped, closed during the day Shallow with landing platform tubular Large and bowl shaped, cluster of small flowers or large and solitary Large, funnel like, strong perch support Narrow tube with spur, wide landing pad Shallow, funnel like or complex with trap Regular tubular without lip Regular and small

PAGE 16

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 16 Appendix 5: Plant and Pollen Identification Photos Most Specific Name Plant ID Photo Pollen ID Photo Columnea consanguinea Malvaviscus palmanus Tradescantia zanonia Solanum

PAGE 17

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 17 Aphelandra Asteraceae Monochaetum vulcanicum Centropogon Drymonia conchocalyx No image No photo

PAGE 18

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 18 Poaceae Melastomataceae Razisea spicata Calathea No photo

PAGE 19

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 19 Miconia Blachea No photo Phytolacca Idalgoa cuaternata

PAGE 20

A Comparison of Pollination Deposition in Different Pollination Syndromes Powell 20 Ruvos Fabaceae Cococicelum puvesens No photo