USF Libraries

Resource Partitioning and Niche Differentiation of Nectarivorous Bats in a Monteverde Cloud Forest

MISSING IMAGE

Material Information

Title:
Resource Partitioning and Niche Differentiation of Nectarivorous Bats in a Monteverde Cloud Forest
Translated Title:
Repartición de recursos y el nicho de diferenciación de los murciélagos nectarívoros en un bosque nuboso de Monteverde ( )
Physical Description:
Book
Language:
text in English
Creator:
Cohen, D. Morris
Publication Date:

Subjects

Subjects / Keywords:
Plant and pollinator coevolution
Coevolución de las plantas y los polinizadores
Bats--Behavior
Murciélagos--Comportamiento
Niche partitioning
Compartimentación de nichos
Costa Rica--Puntarenas--Monteverde Zone--Monteverde
Costa Rica--Puntarenas--Zona de Monteverde--Monteverde
Tropical Ecology Spring 2003
Ecología Tropical Primavera 2003
Genre:

Notes

Abstract:
Plant-pollinator interactions are an important part of tropical ecosystem function and diversity. In this paper, I present the results of a month-long study comparing the three nectarivorous Phylostomid bat species, Anoura geoffroyi, Glossophaga commissarisi, and Hylonycteris underwoodi, and the their plant resource base within the cloud forest of Monteverde, Costa Rica. Bats were mistnetted, and their pollen loads were removed and later measured for richness, abundance, and diversity. The extent of Niche overlap between the three species was also assessed. I found G. commissarisi to carry the most diverse pollen loads (H'=0.648) and to also have the most consistent floral visitation patterns. Pollen load analysis revealed similar visitation patterns for the morphologically analogous H. underwoodi, and different patterns for the larger and more distinct A. geoffroyi. Niche overlap was high between H. underwoodi and G. commissarisi and low between both of these species and A. geoffroyi. Although capture rates were too low for significant conclusions regarding H. underwoodi, resource partitioning was found to occur between the other two species.
Abstract:
Las interacciones entre las plantas y sus polinizadores son una parte importante de la diversidad y el funcionamiento de los ecosistemas tropicales. En este ensayo, presento los resultados de un estudio que compara tres especies de los murciélagos nectarívoros (Anoura geoffroyi, Glossophaga commissarisi, y Hylonycteris underwoodi) y su base de recursos florales en el bosque nuboso de Monteverde, Costa Rica. Se agarraron los murciélagos y se quitaron sus cargas de polen para medir la riqueza, abundancia, y diversidad del polen. El estudio demuestra que G. commissarisi lleva las cargas de polen más diversas (H’ = 0.648) y también tiene patrones de visitación de flores más consistentes. El análisis de las cargas de polen también reveló que los patrones de visitación de flores son similares para la especies más parecida morfológicamente H. underwoodi, y patrones diferentes para la especies más grande y distinta A. geoffroyi. La superposición de nicho fue grande entre H. underwoodi y G. commissarisi, y pequeño entre ellas y A. geoffroyi. Aunque las tasas de captura de H. underwoodi eran demasiado pequeños para sacar conclusiones significativas, se encontró evidencia para la partición de recursos entre las otras dos especies.
General Note:
Student affiliation: Department of Wildlife Ecology, University of Wisconsin Madison

Record Information

Source Institution:
Added automatically
Holding Location:
Added automatically
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - M39-00433
usfldc handle - m39.433
System ID:
SFS0044347:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Resource Partitioning and Niche Differentiation of Nectarivorous Bats in a Monteverde Cloud Forest D. Morris Cohen Department of Wildlife Ecology, University of Wisconsin Madison _____________________________________________________________________________________ Abstract Plant pollinator interactions are an important part of tropical ecosystem function and diversity. In this paper, I present the results of a month long study comp aring the three Nectarivorous Phylostomid bat species, Anoura geoffroyi Glossophaga commissarisi and Hylonycteris underwoodi and their plant resource base within the cloud forest of Monteverde, Costa Rica. Bats were mistnetted and their pollen loads were removed and later measured for richness, abundance, and diversity. The extent of Niche overlap between the three species was also assessed. I found G. commissarisi to carry the most diverse pollen loads H€ = 0.648 and to als o have the most consistent floral visitation patterns. Pollen load analysis revealed similar visitation patterns for the morphologically analogous H. underwoodi and different patterns for the larger and more distinct A. geoffroyi Niche overlap was high b etween H. underwoodi and G. commissarisi and low between both of these species and A. geoffroyi Although capture rates were too low for significant conclusions regarding H. underwoodi resource partitioning was found to occur between the other two species Resumen Las interacciones de las plantas y sus polinizadores son una parte importante de la diversidad y el funcionando de las ecosistemas tropicales. En este ensayo, presento los resultados de un estudio comparando tres especies de los murcilagos nec tariviosos y su comida de plantas en el bosque nuboso de Monteverde, Costa Rica. Encontr que Anoura geoffroyi Glossophaga commissarisi and Hylonycteris underwoodi todos llevan el polen de especies plantas locales en grados diferentes y en el parte al mi smo tiempo. Encontr G. commissarisi llevar cargas de polen ms diversas H€ = .648 que los otros especies, y tambin tiene patrones de hbitos comiendo ms consistentes. Anlisis de las cargas de polen tambin revelo que los patrones de hbitos comiendo son similares para la especie simila r de morfologa, H. underwoodi y patrones diferentes para la especie ms grande y distinto, A. geoffroyi El parte al mismo tiempo de niche estuvo grande entre H. underwoodi y G. commissarisi y pequeo entre estas especies ambos y A. geoffroyi Aunque los razones de captura de H. underwoodi estaba ms pequeos para conclusiones significas, las divisiones de los recursos haba encontrados ser entre las dos otras especies. Introduction Classic resource based models of niche partitioning and competitive excl usion Hutchinson 1959 have been used to describe patterns and maintenance of species diversity in natural communities Terborgh 2002. These models assume that all organisms are uniquely adapted to use a specific subset of available resources. Similar sp ecies partition resources to allow coexistence within the community MacArthur 1958. Thus, each species evolves to occupy a specific niche, with corresponding morphological, behavioral, and physiological adaptations. Resource based models have been used to explore plant ‚ pollinator relationships in natural communities Feinsinger, 1983. Plants have evolved to produce rewards for pollinators who in turn carry pollen grains from flower to flower Bawa 1990; Lemke 1985. Having highly specialized flowers ca n lead to pollinator specialization, ensuring greater exchange of conspecific pollen Cunningham 1995; Feinsinger 1983. Resulting specialization increases foraging efficiency for pollinators whose resource is not being exploited by other species. Efficien cy is also increased because

PAGE 2

specialized pollinators are adept at responding to the cues of specific flowers with minimum error Feinsinger 1983. At the community level, pollination specialization could ultimately lead to resource partitioning and increas ed local diversity. Futuyma 1997 documented examples of plant pollinator coevolution between Ficus and Agaonid wasp populations, showing that highly specialized mutualisms between specific Ficus and wasp species have led to increased diversity throughou t the Ficus genera and Agaonidae family. With 1001 known species, the bat order, Chiroptera is one of the most diverse within the class Mammalia LaVal and Rodriguez ‚ H 2002. Furthermore, bats are conspicuous and important pollinators of tropical forests. Nectarivorous bats are a particularly diverse group in the Neotropics and in the Monteverd e area there are seven species of four genera that share a reported 33 plant species Haber 2000. Cloud forest bats are known to partition their plant resources, but this has only been studied for a single forest with two congeners Muchhala and Jarrin V 2002. Perhaps the large amount of diversity within this group could be attributed to pollination related coevolution and specialization. The focus of this paper is to better understand bat foraging behavior and how it relates to coevolution. I also hope to examine how well resource based models of niche partitioning explain plant pollinator interactions for bats and their flowers. Methods Study Site The study was conducted between April 10, 2003 and May 5, 2003 at the Monteverde Hummingbird Gallery in M onteverde, Costa Rica. The Gallery is adjacent to the Monteverde Cloud Forest Reserve, which has an elevation of 1560m, and is classified as premontane to montane wet/rain forest sensu Holdridge. The gallery has ten hummingbird feeders that contain a suc rose and water solution on which three Nectarivorous bat species feed. Species Description Anoura geoffroyi is the largest of the three species forearm length 39 47mm. This species lacks a calcar tail membrane and the tail itself is very reduced. It i s known to visit the Monteverde Hummingbird Gallery in large numbers LaVal and Rodriguez H, 2002. Studies in Ecuadorian cloud forest have shown A. geoffroyi to carry pollen of 11 different plant species Muchhala and Jarrin ‚ V, 2002. Glossophaga commi ssarisi has been described by LaVal and Rodriguez H 2002 to have a forearm length of 32 36mm. This species has both tail and calcar present. Laval and Rodriguez ‚ H found it to carry pollen of tropical forest epiphytes Weberocereus tunilla Cactaceae and Marcgravia nepenthoides Marcgraviaceae as well as Musa acuminata Musaceae and Mucuna Papilionaceae. They also found this species to eat insects and fruit. Hylonycteris underwoodi is very similar in morphology to G. commissarisi but with a sli ghtly longer snout LaVal and Rodriguez ‚ H. Forearm lengths are reported as 31 36mm. Individuals have also been found carrying large amounts of pollen of unidentified species LaVal and Rodriguez 2002; Braun and Shelley 2002. Capture Methods Mistnetting occurred a total of eight nights between 6:00pm ‚ 9:00pm. Two nets one small 6m net and one larger 12m net were placed in open areas such as stairway corridors or along the patio floor of the gallery each night. Nets were continually monitore d and bats were taken out quickly after capture. Bats were placed in holding bags after retrieval from the net.

PAGE 3

Captured bats were weighed using a Pesola 100g scale, sexed, and identified to species. Reproductive condition was determined by the existenc e of enlarged testicles on males and visible teats on females. Areas of visible pollen on the bat were also noted. For each individual, one sterile, wet, cotton swab was used to remove pollen from the head, back, chest, and tail membrane or in the case of Anoura geoffroyi the hind legs areas. Swabbing efforts were aimed at removing the Full Pollen Load FPL from each individual. After pollen removal, sampled bats were released and cotton swabs were labeled and placed in wax paper envelopes to ensure pol len retention and limit contamination from other swabs. Initially, all captured bats were sampled, but early slide examination showed that bats that lacked visible pollen never had significant amounts of any pollen type. A significant amount of pollen was defined as the minimum amount of pollen present that indicates a flower visit. This study used three or more grains as this criterion, following methods used by Heithaus et al. 1975. Methods were thus modified and only bats with visible pollen were there after sampled. The same criterion was used for bats that did have visible pollen in order to limit the effect of residual pollen acquired from holding bags or feeders. I disregarded pollen types with less than three grains on these individuals. Pollen Cou nts Cotton swab samples were next transferred to a dry box and allowed to dry for at least two days. Once dry, pollen was exhaustively scraped from swabs onto glass slides using wooden toothpicks. Glass coverslips were applied with Permount TM Fischer Sci entific Company before being mounted onto pollen covered slides. Mounted slides were then placed in a dry box until Permount TM set. Slides were examined using a compound microscope Carl Zeiss ‚ Jenna model set at 100x. A pollen morphospecies reference figure Appendix A was created using digital photography. The figure was created using pollen sampled from known bat pollinated species in the area Mucuna urens Papilionaceae, Musa acuminata Musaceae, Nycandra physalodes Solanaceae, and Vriesea gladioliflora Bromeliaceae and unknown types found on sample slides from this study. Pollen morphospecies on each slide were matched to the reference figure when possible and the type and number of pollen grains per species were counted on each slide. R esults Three of seven species of n ectarivorous bats in the Monteverde cloud forest are utilizing artificial nectar feeders at the Monteverde Hummingbird gallery. I caught 175 individuals between these three species: A. geoffroyi n = 103, G. commissarisi n = 67, and H. underwoodi n = 5. Each species was found to have at least some individuals carrying pollen. Because of the low capture rate and absence of female samples, H. underwoodi was not used in some of the statistical analyses. Size and Reprodu ctive Condition Two of these Glossaphagine bats, H. underwoodi and G. commissarisi are morphologically similar in length, as shown by Laval and Rodriguez ‚ H 2002, and were not found to have significant differences in weight as measured in this study Fi scher€s PLSD, p > 0.05. The third species, A. geoffroyi is significantly larger ANOVA p < .0001 and lacks a calcar. Mean weights and standard deviation are shown in Figure 1. Times of reproductive activity are also apparently different, with A. geoffro yi males 66.6%, n = 42 and H. underwoodi males 40%, n = 3 being the only groups showing active reproductive status April May to a substantial degree Table 1. Reproductively active G. commissarisi were relatively low for both sexes 2.1% of males n = 48, 14.3% of females n = 7 and no reproductively active female A. geoffroyi were captured.

PAGE 4

Incidence of Pollen on Different Bat Species Table 2 compares the amount of individuals carrying pollen vs. those without pollen for male and female A. geoffroyi and G. commissarisi Chi squared analysis reveals significant differences in the number of individuals between both species and sex carrying pollen X 2 = 73.044, df = 3, n = 165. Individuals of G. comm issarisi 92.5% carry pollen, n = 67 are more apt to carry pollen than A. geoffroyi 24.5%, n = 98. Males 54.0% carry pollen, n = 126 are more likely to carry pollen than females 46.2%, n = 39. It should also be mentioned that 100% of individuals of H. underwoodi caught, were carrying pollen n = 5. Species Differences in Pollen Loads Abundance Total pollen counts for each species reveal that overall males carry a significantly greater number of pollen grains for both A. geoffroyi male abundance = 2157, female = 1099; X 2 = 627.038, df = 1, n = 18 and G. commissarisi male abundance = 2396, female = 899; X 2 = 190.078, df = 1, n = 18. Species do not differ in the total number of pollen grains found collectively X 2 = .174, df = 3, n = 36 Table 3. On an individual basis, bats of A. geoffroyi mean male pollen abundance = 269.6, SD = 673.8; mean female abundance = 109.9, SD = 130.6 and G. commissarisi mean male pollen abundance = 217.8, SD = 200.4; mean female abundance = 128.4 SD = 91.9 do not differ in mean pollen abundance between sexes One way ANOVA p ‚ value > .05. No significant difference in mean abundance exists between species A. geoffroyi mean = 180.89, G. commissarisi mean = 183.05; One way ANOVA p value > .05 or between species and sex Two way ANOVA p value > .05; Figure 3. An F test however, shows a significantly greater variance for an individual bat€s pollen load for A. geoffroyi compared to an individual of G. commissarisi F = 7.081, p = .0002. Therefore, A. geoffroyi individuals are more variable in number of pollen grains carried by an individual bat compared to G. commissarisi Richness The number of different pollen types found on each bat species does not differ between sex for both A. geoffroyi male richness = 3, female = 4; X 2 = .143, df = 1, n = 18 and G. commissarisi male richness = 9, female = 4; X 2 = 1.923, df = 1, n = 18 Table 4. I also found no significant difference in number of pollen types found between all three species A. geof froyi = 5, G. commissarisi = 9, H. underwoodi = 2; X 2 = 1.143, df = 2, n = 4 Table 5. Individual female bats mean pollen species richness = 1.47, SD = .624 do not differ from males mean richness = 1.68, SD = 1.11 in the number of pollen types that they carry ANOVA p > .05 ; however there are significant differences between species. Individuals of G. commissarisi mean richness = 2.00, SD = 1.08 tend to carry on average, one more type of pollen species than A. geoffroyi mean richness = 1.167, SD = .383 ANOVA p value = .0071; Figure 4. Diversity Figure 2 shows the differences in pollen load diversity found on each species using a Shannon weiner index of diversity. As a species, G. commissarisi H€ = .638 were found to have a more diverse pollen load than A. geoffroyi H€ = .307, which had a more diverse pollen load than H. underwoodi H€ = .217. The following Modified t test results between all three species reveal a significant difference i n pollen diversity in each case, A. geoffroyi vs. G. commissarisi v = 6444.405, t value = 42.869 G. commissarisi vs. H. underwoodi v = 704.115, t value = 32.907. A. geoffroyi vs. H. underwoodi v = 639.379, t ‚ value = 7.255.

PAGE 5

A Mann Whitney U test reveals that individual bats of A. geoffroyi and G. commissarisi differ in diversity of pollen grains carried U = 84, U€ = 240, p = 0.006. G. commissarisi individuals tend to hold higher pollen diversity mean H€ = 0.317 than those of A. geoffroyi mean H€ = 0.007. Overlap A Morista Horn index of overlap Table 6 reveals that there is a high degree of niche overlap between the two morphologically similar species, G. commissarisi and H. underwoodi Index value = 0.592. There is no niche overlap between the larger and calcar lacking A. geoffroyi and H. underwoodi and very little overlap between A. geoffroyi and G. commissarisi Index value = 0.001. No species of plant is used by all three species and the two plants that are sha red by A. geoffroyi and G. commissarisi are not common food sources for either bat Table 7. Four plant species are visited only by G. commissarisi morphospecies a, c, e, and k and two species are unique to A. geoffroyi morphospecies f and g. All plan ts visited by H. underwoodi are visited by G. commissarisi but a larger sample size is needed to determine if H. underwoodi is visiting more flowers than those found in this study. Additional Observations I observed that most of the pollen present on G. commissarisi and H. underwoodi is on the calcar. A smaller proportion of individuals of these two species carry pollen on the neck or head region. Members of the larger and calcar lacking species A. geoffroyi tend to carry pollen on top of the head. From the pollen sampled from local bat pollinated flowers, only that of two plants occurred on bats captured for this study. Morphospecies c, Mucuna urens Papilionaceae, was being carried by G. commissarisi Morphospecies f, Vriesea gladioliflora Bromeliace ae, was being carried by A. geoffroyi Discussion This study illustrates an interesting situation where niche differentiation is apparently occurring between some species within the Monteverde bat and bat flower community, but overlap between others may be high. A. geoffroyi and G. commissarisi appear to partition resources in a way that limits interspecific competition, but from the results of this study, resource partitioning is not apparent between G. commissarisi and H. underwoodi Both similar morph ology between the two species and low sample size of H. underwoodi, could explain this discrepancy. Perhaps the low capture rate is evidence that they do not solely prefer nectar and they may in fact be feeding on insects or fruit. Though plausible, this e xplanation is doubtful because of high 100% pollen incidence rates of individuals caught and the high rates of capture success in the past Braun and Shelley 2002. A more likely explanation of low capture rates is that they are nectar specialists and ex hibit small scale migrations in response to flower availability LaVal pers. comm.. Low flower availability at the time of study would explain their absence and may also explain how they partition resources with G. commissarisi G. commissarisi known to feed on insects and fruit LaVal and Rodriguez H 2002; Nowak 1994 is probably more sedentary due to this generalist foraging strategy. This species does not need to move to areas with higher flower densities and it may actually benefit from low flower de nsity, because of less competition with morphologically similar H. underwoodi As H. underwoodi migrate out of these common grounds, G. commissarisi may actually begin focusing on the plants that are in flower. This might explain the high pollen incidence rate found on this species at the time of this study. When flower availability is high, G. commissarisi may switch its search image back to insects or fruit because of increased competition with H. underwoodi These patterns of resource switching and availability

PAGE 6

might explain why I found high overlap between both species, low numbers of H. underwoodi and high incidence of pollen on G. commissarisi The small amount of overlap between A. geoffroyi and both H. underwoodi and G. commissarisi is most likely due to differing morphology. A. geoffroyi a significantly larger species than the other two, is probably visiting larger flowers. Muchhala and Jarr in ‚ V 2002 showed that A. geoffroyi was visiting larger flowers than the smaller A. caudifera in Ecuadorian cloud forest. The two pollen types found on bats that were positively identified in this study evidence this same discrepancy in size. On A. geoff royi I found pollen from Vriesea gladioliflora Bromeliaceae, a much larger flower than that of Mucuna urens Papilionaceae, which I found on G. commissarisi. Another more interesting explanation of how resource partitioning is occurring between G. com missarisi and A. geoffroyi may relate to their differing foraging styles. Comparatively lower amounts of species richness of pollen on A. geoffroyi suggests that this species has more specific feeding habits than G. commissarisi When it is visiting flow ers, A. geoffroyi is commonly visiting two plants in particular type f Vriesea gladioflora and type g TABLE 7 although visitation rates overall are much lower than those of the other two bat species. In addition, pollen abundances vary widely, suggest ing that it is not always visiting multiple flowers in a night. As a pollinator, A. geoffroyi can thus be described as visiting relatively few species of flowers, and with little regularity. Lower incidence rates of pollen on A. geoffroyi might also be the result of feeder dependence, which would also explain high capture rates at the Gallery. The high proportion of reproductive males might suggest that they are spending less time foraging and are focusing on courtship, harem maintenance and reproduction. Their search image may have shifted to the feeders because of increased foraging efficiency, which would allow more time for reproductive activities. Although feeder dependence may be an explanation of why relatively few A. geoffroyi indi viduals are carrying pollen, it probably does not have a large effect because of the small amount of feeders and high levels of competition around them. Higher pollen species richness overall and more pollen species per individual, suggest that G. commiss arisi occupies a broader niche than A. geoffroyi In addition, this species often carries diverse pollen loads and carries pollen much more often than A. geoffroyi When it is acting as a pollinator, G. commissarisi can thus be described as less specialize d but visiting flowers with greater regularity and consistency. The differing foraging strategies of A. geoffroyi and G. commissarisi present an interesting dichotomy that may help to understand plant pollinator relationships in the Monteverde area. From a plant€s perspective, A. geoffroyi might be a good pollinator because it is less likely to be carrying foreign pollen. However, it is also fairly unreliable because of erratic foraging behavior and irregular visitation rates. On the other hand, a plant us ing G. commissarisi as a pollinator has more consistent visitation but may also have to deal with large amounts of foreign pollen coming in contact with the stigma. In responses to these differing selective pressures plants may have evolved to cope with th e shortcomings that come as a result of these differing pollinator feeding strategies. Further studies within this plant pollinator community could examine the possibility that coevolution is not only morphological but also physiological. Perhaps plants b eing visited by A. geoffroyi have not only evolved to match the relatively large body type of this species as shown by this study and that of Muchhala and Jarrin ‚ V 2002, but also to match its foraging strategy. Mechanisms to deal with erratic visitation rates, such as increased stigma longevity, may explain the relationship in greater detail. Similarly, G. commissarisi pollinated plants may have coevolved on a morphological level. Perhaps stigmas of these plants have evolved to deal with large amounts of heterotypic pollen. Further floral studies of the plants that A. geoffroyi and G. commissarisi are pollinating may reveal the extent of specialization within these relationships and the mechanisms by which this specialization occurs. Such studies might sh ow that pollinator behavioral traits can drive flower evolution and lead to

PAGE 7

greater specialization. The result of these specializations is greater diversity and finer niche partitioning. Continuing research might also attempt to quantify niche differentia tion among all seven species of nectarivorous bats in the Monteverde area. These studies might increase understanding of how resource partitioning mechanisms elevate diversity within communities and ecosystems as well as shed more light on coevolution theo ries presented in this paper. Acknowledgements I would first like to thank Richard LaVal for his inspiration, vast bank of knowledge, and guidance. Thank you to Michael and Patricia Fogden for permitting me to use the Monteverde Hummingbird Gallery. Thank you to Bill Haber and Willow Zuchowski for plan t identification and help with methods. Thank you to the staff of Estacin Biolgica, the Monteverde community, and especially the Arguedas ‚ Villalobos family for their hospitality and support. Thank you to the University of Wisconsin Madison Department o f Wildlife Ecology. Thank you to Evan Adams, Sarah Bachman, and Katherine Ayers for their help in the field. Thank you to James Dugan for help in the field and critical review of proposal. Much gratitude and special thanks goes out to Alan Masters, Rick F. X. Smith X, and R. Andrew Rodstrom for whom without, this project would not have been possible. Thank you Alan for your invaluable advice, critical review, and really for keeping this project on the ground. Thank you Rick for your work servicing the nets, digital imagery, statistical help, entertainment, and the immense time and attentiveness you put into this project. Thank you Andrew for your experience, innovative ideas, handling of bats, innate composure in the field, and wide spanning general knowledge Finalmente, gracias a los Anoura geoffroyi para su cooperacin y su espritu tranquilo. Literature Cited Bawa, K.S. 1990. Plant Pollinator Interactions in Tropical Rain Forests. Ann. Rev. Ecol. Syst. 21: 399 422. Braun, E. and D. Shelley 2002. Influence of Musa acuminata Musaceae On Bat Pollination Systems in the Monteverde Area. CIEE Tropical Ecology and Conservation Fall. Cunningham, S.A. 1995. Ecological Constraints on Fruit Initiation by Calyptrogyne ghiesbreightiana Arecaceae: Floral H erbivory, Pollen Availability, and Visitation by Pollinating Bats. American Journal of Botany 82 12: 1527 1536. Feinsinger, P. 1983. Coevolution and Pollination. In D.J. Futuyma, M. Slatkin eds.. Coevolution. Sinauer Press: Mass. Futuyma, D.J. 1997. Th e Evolution and Importance of Species Interactions. In Meffe, G. and Ronald Corrol eds.. Principals of Conservation Biology, 2 nd Ed. Sinauer Press: Mass. Haber, W. 2000. Vascular Plants of Monteverde. In N.M. Nadkarni and N.T. Wheelwright eds.. Monteve rde: The Ecology and Conservation of a Tropical Cloud Forest. Oxford University Press, Inc. New York. Heithaus, R.E. 1982. Coevolution between bats and plants. In T.H. Kunz ed.. Ecology of Bats. Plenum Press, New York, New York. Hutchinson, G.E. 1959. Homage to Santa Rosalia or Why Are There So Many Kinds of Animals? The American Naturalist. XCIII: 145 159. LaVal, R.K. and B. Rodriguez H. 2002. Murcilagos de Costa Rica. Instituto Nacional de Biodiversidad, Costa Rica. Lemke, T.O. 1985. Pollen Carrying by the Nectar feeding Bat Glossphaga soricina in a Suburban Environment. Biotropica 172: 107 111. MacArthur, R.H. 1958. Population ecology of some warblers of northeastern coniferous forests. In L.A. Real and S.H. Brown eds.. Foundations of Ecology. Un iversity of Chicago Press, Chicago, IL. Muchhala N. and P. Jarrin ‚ V. 2002. Flower Visitation by Bats in Cloud Forests Western Ecuador. Biotropica 343: 387 395. Nowak, R.M. 1991. Walker€s Bats of the World. The John Hopkins University Press, London, Engl and. Terborgh, J. 1992. Diversity and the Rainforest. Scientific American Library, New York, New York.

PAGE 8

Tables ________________________________________________________________________ TABLE 1. Reproductive condition of three phylostomid bats Males con sidered ƒactive„ with enlarged testicles, females with visible teats ________________________________________________________________________ Species and sex Reproductively active No sign of Activity Anoura geoffroyi males 42 21 Anoura geoffroyi females 0 30 Glossophaga commissarisi males 1 47 Glossophaga commissarisi females 1 6 Hylonycteris underwoodi males 2 3 ________________________________________________________________________ TABLE 2. Incidence of pollen on two species of Nectarivorous phylostomid bats. Number of individuals with pollen vs. number of individuals without pollen. X 2 = 73.044. ________________________________________________________________________ Species and sex With Pollen Without Pollen Anoura geoffroyi males 13 53 Anoura geoffroyi females 11 21 Glossophaga commissarisi males 55 5 Glossophaga commissarisi females 7 0 ______________________________________________________________________ TABLE 3. Pollen abundances of two species of nectarivorous phylostomid bats. Males have significantly higher pollen abundance levels than females in both A. geoffroyi X = 627.038 and G. commissarisi X = 190.078. Between species there is not a significant diff erence in pollen load abundance X = .174. ______________________________________________________________________ Species and sex N Number of pollen grains Anoura geoffroyi males 8 2157 Anoura geoffroyi females 10 1099 Glossophaga commissarisi males 11 2396 Glossophaga commissarisi females 7 899 _____________________________________________________________________________ TABLE 4. Pollen richness of sexes of two species of nectarivorous phylostomid bats. Number of pollen types carried are not significantly different between sex in both A. geoffroyi X = .143 and G. commissarisi X = 1.923. ______________________________________________________________________________ _____________________________________________________________________ Species and Sex N Number of pollen types Anoura geoffroyi males 8 3 Anoura geoffroyi females 10 4 Glossophaga commissarisi males 11 9 Glossophaga commissarisi females 7 4

PAGE 9

______________________________________________________________________ TABLE 5. Pollen richness of three species of nectarivorous phylostomid bats. Pollen load richness does not differ between species X = 1.143. ______________________________________________________________________ Species and sex N Number of pollen types Anoura geoffroyi 18 5 Glossophaga commissarisi 18 9 Hylonycteris underwoodi 5 2 ______________________________________________________________________ TABLE 6. Morista Horn Index of Overlap between three nectarivorous phylostomid bats. ______________________________________________________________________ Species Overlapping Morista Horn Index Of Overlap Value Anoura Geoffroyi vs. H. underwoodi 0 Anoura Geoffroyi vs. G. commissarisi 0.001382205 G. commissarisi vs. H. underwoodi 0.592444724

PAGE 11

__________________________________________________________________________________________ ________ FIGURE 3. Average pollen abundance load per individuals and standard deviation of two Nectarivorous phylostomid bats. ANOVA tests show P = .8873 between species, P = .2928 between sex, and P = .7646 between sex and species. __________________________________________________________________________________________________ ________________________________________ __________________________________________________________ FIGURE 4. Average pollen species richness per individual and standard deviation of two Nectarivorous bats. ANOVA tests showed P = .7804 between sex, P = .0071 between species, and P = .5883 between sex and species. __________________________________________________ ________________________________________________



PAGE 1

Resource Partitioning and Niche Differentiation of Nectarivorous Bats in a Monteverde Cloud Forest D. Morris Cohen Department of Wildlife Ecology, University of Wisconsin Madison ____________________________________________________________________________________ Abstract Plant pollinator interactions are an important part of tropical ecosystem function and diversity. In this paper, I present the results of a month long study com paring the three nectarivorous Phylostomid bat species, Anoura geoffroyi, Glossophaga commissarisi, and Hylonycteris underwoodi, and the their plant resource base within the cloud forest of Monteverde, Costa Rica. Bats were mistnetted, and their pollen loads were removed and later measured for richness, abundance, and diversity. The extent of Niche overlap between the three species was also assessed. I found G. commissarisi to carry the most diverse pollen loads H'=0.648 and to also have the most consistent floral visitation patterns. Pollen load analysis revealed similar visitation patterns for the morphologically analogous H. und erwoodi, and different patterns for the larger and more distinct A. geoffroyi. Niche overlap was high between H. underwoodi and G. commissarisi and low between both of these species and A. geoffroyi. Although capture rates were too low for significant conc lusions regarding H. underwoodi, resource partitioning was found to occur between the other two species. Resumen Las interacciones de las plantas y sus polinizadores son una parte importante de la diversidad y el funcionando de las ecosistemas tropicales. En este ensayo, presento los resultados de un estudio comparando tres especies de los murcilagos nectariviosos y su comida de plantas en el bosque nuboso de Monteverde, Costa Rica. Encontre que Anoura geoffroyi, Glossophaga commissarisi, and Hylonycteris underwoodi todos llevan el polen de e species plantas locales en grado s diferentes y en el parte al mismo tiempo. Encontr G. commissarisi llevar cargas de polen mas diversas H'=.648 que los otros especies, y tambin tiene patrones de hbitos comiendo mas consistentes. Anlisis de las cargas de polen tambin revelo que los patrones de hbitos comiendo son similares para la especie similar de morfologa, H. underwoodi y patrones diferentes para la especie

PAGE 2

mas grande y distinto, A. geoffroyi. El parte al mismo tiempo de niche estuvo grande entre H. underwoodi y G. com missarisi y pequeo entre estas especias ambos y A. geoffroyi. Aunque los razones de captura de H. underwoodi estaba mas pequeos para conclusiones significas, las divisiones de los recursos haba encontrados ser entre las dos otras especies. Introduction Classic resource based models of niche partitioning and competitive exclusion Hutchinson 1959 have been used to describe patterns and maintenance of species diversity in natural communities Terborgh 2002. These models assume that all organisms are uniquely adapted to use a specific subset of available resources. Similar species partition resources to allow coexistence within the community MacArthur 1958. T hus, each species evolves to occupy a specific niche, with corresponding morphological, behavioral, and physiological adaptations. Resource based models have been used to explore plant pollinator relationships in natural communities Feinsinger, 1983. Pla nts have evolved to produce rewards for pollinators who in turn carry pollen grains from flower to flower Bawa 1990; Lemke 1985. Having highly specialized flowers can lead to pollinator specialization, ensuring greater exchange of conspecific pollen Cun ningham 1995; Feinsinger 1983. Resulting specialization increases foraging efficiency for pollinators whose resource is not being exploited by other species. Efficiency is also increased because specialized pollinators are adept at responding to the cues of specific flowers with minimum error Feinsinger 1983. At the community level, pollination specialization could ultimately lead to resource partitioning and increased local diversity. Futuyma 1997 documented examples of plant pollinator coe volution between Ficus and Agaonid wasp populations, showing that highly specialized mutualisms between specific Ficus and wasp species have led to increased diversity throughout the Ficus genera and Agaonidae family. With 1001 known species, the bat order Chiroptera is one of the most diverse within the class Mammalia LaVal and Rodriguez H 2002. Furthermore, bats are conspicuous and important pollinators of tropical forests. Nectarivorous bats are a particularly diverse group in the neotropics and in th e Monteverde area there are seven species of four genera that share a reported 33 plant species Haber 2000. Cloud forest bats are known to partition their plant resources, but this has only been studied for a single forest with two congenerars Muchhala and Jarrin V 2002. Perhaps the large amount of diversity within this group could be attri buted to pollination related coe volution and specialization. The focus of this paper is to better understand bat foraging be havior and how it relates to coe volution. I also hope to examine how well resource based models of niche partitioning explain plant pollinator interactions for bats and their flowers. Methods Study Site The study was conducted between April 10, 2003 and May 5, 2003 at the Monteverde Hummingbird Gallery in Monteverde, Costa Rica. The Gallery is adjacent to the Monteverde Cloud Forest Reserve, which has an elevation of 1560 m, and is classified as premontane to montane wet/rain forest sensu Holdridge. The gallery has ten hummingbird feeders that contain a sucrose

PAGE 3

and water solution on which three nectarivorous bat species feed. Species Description Anoura geoffroyi is the largest of the three species forearm length 39 47mm. This species l acks a calcar tail membrane and the tail itself is very reduced. It is known to visit the Monteverde Hummingbird Gallery in large numbers LaVal and Rodriguez H, 2 002. Studies in Ecuadorian cloud forest have shown A. geoffroyi to carr y pollen of 11 different plant species Muchhala and Jarr in V, 2002. Glossophaga commissarisi has been described by LaVal and Rodriguez H 2002 to have a forearm lengths of 32 36mm. This species has both tail and calcar present. Laval and Rodriguez H found it to carry pollen of tropical forest epiphytes Weberocereus tunilla Cactaceae and Marcgravia nepenthoides Marcgraviaceae as well as Musa acuminata Musaceae and Mucuna Papilionaceae. They also found this species to eat insects and fruit. Hylonycteris underwoodi is very similar in morphology to G. commissarisi but with a slightly longer snout LaVal and Rodriguez H. Forearm lengths are reported as 3 l 36mm. Individuals have also been found carrying large amounts of pollen of unidentified s pecies LaVal and Rodriguez 2002; Braun and Shelley 2002. Capture Methods Mistnetting occurred a total of eight nights between 6:00pm 9:00pm. Two nets one small 6m net and one larger 12m net were placed in open areas such as stairway corridors or along the patio floor of the gallery each night Nets were continually monitored and bats were taken out quickly after capture. Bats were placed in holding bags after retrieval from the net. Captured bats were weighed using a Pesola 100 g scale, sexed, an d identified to species. Reproductive condition was determined by the existence of enlarged testicles on males and visible teats on females. Areas of visible pollen on the bat were also noted. For each individual, one sterile, wet, cotton swab was used to remove pollen from the head, back, chest, and tail membrane or in the case of Anoura geoffroyi, the hind legs areas. Swabbing efforts were aimed at removing the Full Pollen Load FPL from each individual. After pollen removal, sampled bats were released and cotton swabs were labeled and placed in wax paper envelopes to ensure pollen retention and limit contamination from other swabs. Initially, all captured bats were sampled, but early slide examination showed that bats that lacked visible pollen never had significant amounts of any pollen type. A significant amount of pollen was defined as the minimum amount of pollen present that indicates a flower visit. This study used three or more grains as this criterion, following methods used by Heithaus et al. 1975. Methods were thus modified and only bats with visible pollen were thereafter sampled. The same criterion was used for bats that did have vis ible pollen in order to limit th e effect of re sidual pollen acquired from holding bags or feeders. I disregarded pollen types with less than three grains on these individuals. Pollen Counts Cotton swab samples were next transferred to a dry box and allowed to dry for at least two days. Once dry, pol len was exhaustively scraped from swabs onto glass slides using wooden

PAGE 4

toothpicks. Glass c overslips were applied with Perm ount€ Fischer Scientific Company before being mounted onto pollen covered slides. Mounted slides were then placed in a dry box until Perm ount€ set Slides were examined using a compound microscope Carl Zeiss Jenna model set at 100 x. A Pollen morphospecies reference figure Appendix A was created using digital photography. The figure was created using pollen sampled from known bat pollinated species in the area Mucuna u rens Papilionaceae, Musa acuminata Musaceae, Nycandra physalodes Solanaceae, and Vriesea gladioliflora Bromeliaceae and unknown types found on sample slides from this study. Pollen morphospecies on each slide were matched to the reference figure when possible and the type and number of pollen grains per species were counted on each slide. Results Three of seven species of nectarivorous bats in the Monteverde cloud forest are utilizing artificial nectar feeders at the Monteverde Hummingbird gallery. I caught 175 individuals between these three species: A. geoffroyi n=103, G. commissarisi n=67, and H. underwoodi n=5. Each species was found to have at least some individuals carrying pollen. Because of the low capture rate and absence of female samples, H. underwoodi was not used in some of the statistical analyses. Size and Reproductive Condition Two of these Glossaphagine bats, H. underwoodi and G. commissarisi, are morphologically similar in length, as shown by Laval and Rodriguez H 2002, and were not found to have significant differences in weight as measured in this study Fischer's PLSD, p>0.05. The third species, A. geoffroyi is significantly larger ANOVA p<. 0001 and lacks a calcar. Mean weights and standard deviation are shown in Figure 1. Times of reproductive activity are also apparently different, with A. geoffroyi males 66.6%, n=42 and H. underwoodi males 40%, n=3 being the only groups showing active reproductive status April May to a substantial degree Table 1. Reproductively active G. commissarisi were relatively low for both sexes 2.1% of males n=48, 14 3% of females n=7 and no reproductively active female A. geoffroyi were captured. Incidence of Pollen on Different Bat Species Table 2 compares the amount of individuals carrying pollen vs. those without pollen for male and female A. geoffroyi and G. commissarisi. Chi squared analysis reveals significant differences in the number of individuals between both species and sex carrying pollen X 2 =73.044, df=3, n=165. Individuals of G. commissarisi 92.5% carry pollen, n=67 are more apt to carry pollen than A. geoffroyi 24.5 %, n=98. Males 54.0% carry pollen, n=126 are more likely to carry pollen than females 46.2%, n=39. It should also be mentioned that 100% of individuals of H. underwoodi caught, were carrying pollen n=5. Species Differences in Pollen Loads Abundance Total pollen counts for each species reveal that overall males carry a significantly greater

PAGE 5

number of pollen grains for both A. geoffroyi male abundance=2157, female=1099; X 2 =627.038, df=l, n=18 and G. commissarisi male abundance= 2396, female=899; X 2 =190.078, df=l, n=18. Species do not differ in the total number of pollen grains found collectively X = .174, df=3, n=36 Table 3. On an individual basis, bats of A. geoffroyi mean male pollen abundance=269.6, SD=673.8; mean female abundance= 109.9, SD=130.6 and G. commissarisi mean male pollen a bundance =217.8, SD=200.4; mean female abundance=128.4, SD=91.9 do not differ in mean pollen abundance between sexes One way ANOVA p value>.05. No significant difference in mean abundance exists between species A. geoffroyi mean=180.89, G. commissarisi mean= 183.05; One way ANOVA p value>.05 or between species and sex Two way ANOVA p value>.05; Figure 3. An F test however, shows a significantly greater variance for an individual bat's pollen load for A. geoffroyi compared to an individual of G. commissarisi F=7.081, p=.0002. Therefore, A. geoffroyi individuals are more variable in the number of pollen grains carried by an individual bat compared to G. commissarisi. Richness The number of different pollen types found on each bat species does not differ between sex for both A. geoffroyi male richness=3, female=4; X 2 =.143, df = l, n=18 and G. commissarisi male richness=9, female=4; X 2 =1.923, df=l, n=18 Table 4. I also found no significant difference in number of pollen types found between all three species A. geoffroyi=5, G. commissarisi = 9, H. underwoodi=2; X = 1 .143, df=2, n=41 Table 5. Individual female bats mean pollen species richness=1.47, SD=.624 do not differ from males mean richness=1.68, SD=1.11 in the number of pollen types that they carry ANOVA p > .05, however there are significant differences between species. Individuals of G. commissarisi mean richness = 2.00, SD = 1.08 tend to carry on average, one more type of pollen species than A. geoffroyi mean richness= 1.167, SD=.383 ANOVA p value=.0071; Figure 4. Diversity Figure 2 shows the differences in pollen load diversity found on each species using a Shannon wiener index of diversity. As a species, G. commissarisi H'=.638 were found to have a more diverse pollen load than A. geoffroyi H'=.307, which had a more div erse pollen load than H. underwoodi H'=.217 The following Modified t test results between all three species reveal a significant difference in pollen diversity in each case, A, geoffroyi vs. G. commissarisi v = 6444.405, t value = 42.869. G. commissarisi vs. H. underwoodi v = 704.115, t value = 32.907. A. geoffroyi vs. H. underwoodi v = 639.379, t value = 7.255. A Mann Whitney U test reveals that individual bats of A. geoffroyi and G. commissarisi differ in diversity of pollen grains carried U=84, U'=240, p=0.006. G. commissarisi individuals tend to hold higher pollen diversity mean H'=0.317 than those of A. geoffroyi mean H'=0.007. Overlap A Morista Horn index of overlap Table 6 reveals that there is a high degree of niche overlap between the two morphologically similar species, G. commissarisi and H. underwoodi Index value=0.592. There is no niche overlap between the larger and calcar lacking A. geoffroyi and H underwoodi and very little overlap between A. geoffroyi and G.

PAGE 6

commissarisi Index value=0.001. No species of plant is used by all three species and the two plants that are shared by A. geoffroyi and G. commissarisi are not common food sources for either bat Table 7. Four plant species are visited only by G. commissarisi morphospecies a, c, e, and k and two species are unique to A. geoffroyi morphospecies f and g. All plants visited by H. underwoodi are visited by G commissarisi, but a larger sample size is needed to determine if H. underwoodi is visiting more flowers than those found in this study. Additional Observations I observed that most of the pollen present on G. commissarisi and H. underwoodi is on the calcar. A smaller proportion of individuals of these two species carry pollen on the neck or head region. Members of the larger and calcar lacking species A. geoffroyi, tend to carry pollen on top of the head. From the pollen sampled from local bat po llinated flowers, only that of two plants occurred on bats captured for this study. Morphospecies c, Mucuna urens Papilionaceae, was being carried by G. commissarisi Morphospecies f, Vriesea gladioliflora Bromeliaceae, was being carried by A. geoffroy i Discussion This study illustrates an interesting situation where niche differentiation is apparently occurring between some species within the Monteverde bat and bat flower community, but overlap between others may be high. A. geoffroyi and G. commissarisi appear to partition resources in a way that limits interspecific competition, but from the results of this study, resource partitioning is not apparent between G. commissarisi and H. underwoodi. Both similar morphology between the two species and low sample size of H. underwoodi, could explain this discrepancy. Perhaps the low capture rate is evidence that they do not solely prefer nectar and they may in fact be feeding on insects or fruit. Though plausible, this explanation is doubtful because of high 100% pollen incidence rates of individuals caught and the high rates of capture success in the past Braun and Shelley 2002. A more likely explanation of low capture rates is that they are nectar specialists and exhibit small scale migrations i n response to flower availability LaVal pers. comm.. Low flower availability at the time of study would explain their absence and may also explain how they partition resources with G. commissarisi. G. commissarisi, known to feed on insects and fruit LaVal and Rodriguez H 2002; Nowak 1994 is probably more sedentary due to this generalist foraging strategy. This species does not need to move to areas with higher flower densities and it may actually benefit from low f lower density, because of less competition with morphologically similar H. underwoodi. As H. underwoodi migrate out of these common grounds, G. commissarisi may actually begin focusing on the plants that are in flower. This might explain the high pollen in cidence rate found on this species at the time of this study. When flower availability is high, G. commissarisi may switch its search image back to insects or fruit because of increased competition with H. underwoodi. These patterns of resource switching and availability might explain why I found high overlap between both species, low numbers of H. underwoodi, and high incidence of pollen on G. commissarisi. The small amount of overlap between A. geoffroyi and both H. underwo odi and G.

PAGE 7

commissarisi is most likely due to differing morphology. A. geoffroyi, a significantly larger species than the other two, is probably visiting larger flowers. Muchhala and Jarr in V 2002 showed that A. geoffroyi was visiting larger flowers than the smaller A. caudifera in Ecuadorian cloud forest. The two pollen types found on bats that were positively identified in this study evidence this same discrepancy in size. On A. geoffroyi, I found pollen from Vriesea glad ioliflora Bromeliaceae, a much larger flower than that of Mucuna ur ens Papilionaceae, which I found on G. commissarisi. Another more interesting explanation of how resource partitioning is occurring between G. commissarisi and A. geoffroyi, may relate to their differing foraging styles. Comparatively lower amounts of species richness of pollen on A. geoffroyi, suggests that this species has more specific feeding habits than G. commissarisi. When it is visiting flowers, A. geoffroyi is commonl y visiting two plants in particular type f Vr iesea gladioflora and type g TABLE 7 although visitation rates overall are much lower than those of the other two bat species. In addition, pollen abundances vary widely, suggesting that it is not always visiting multiple flowers in a night. As a pollinator, A. geoffroyi can thus be described as visiting relatively few species of flowers, and with little regularity. Lower incidence rates of pollen on A. geoffroyi might also be the result of feeder dependence, which would also explain high capture rates at the Gallery. The high proportion of reproductive males might suggest that they are spending less time foraging and are focusing on courtship, harem maintenance, a nd reproduction. Their search image may have shifted to the feeders because of increased foraging efficiency, which would allow more time for reproductive activities. Although feeder dependence may be an explanation of why relatively few A. geoffroyi indiv iduals are carrying pollen, it probably does not have a large effect because of the small amount of feeders and high levels of competition around them. Higher pollen species richness overall and more pollen species per individual, suggest that G. commissar isi occupies a broader niche than A. geoffroyi. In addition, this species often carries diverse pollen loads and carries pollen much more often than A. geoffroyi. When it is acting as a pollinator, G. commissarisi can thus be described as less specialized but visiting flowers with greater regularity and consistency. The differing foraging strategies of A. geoffroyi and G. commissarisi present an interesting dichotomy that may help to understand plant pollinator relationships in the Monteverde area. From a p lant's perspective, A. geoffroyi might be a good pollinator because it is less likely to be carrying foreign pollen. However, it is also fairly unreliable because of erratic foraging behavior and irregular visitation rates. On the other hand, a plant using G. commissarisi as a pollinator has more consistent visitation but may also have to deal with large amounts of foreign pollen coming in contact with the stigma. In responses to these differing selective pressures plants may have evolved to cope with the shortcomings that come as a result of these differing pollinator feeding strategies. Further studies within this plant pollinator community could examine the possibility that coevolution is not only morphological but also physiological. Perhaps p lants being visited by A. geoffroyi have not only evolved to match the relatively large body type of this species as shown by this study and that of Muchhala and Jarrin V 2002, but also to match its foraging strategy. Mechanisms to deal with erratic visi tation rates, such as increased stigma longevity, may explain the relationship in greater detail. Similarly, G. commissarisi

PAGE 8

pollinated plants may have coevolved on a morphological level. Perhaps stigmas of these plants have evolved to deal with large amou nts of heterotypic pollen. Further floral studies of the plants that A. geoffroyi and G. commissarisi are pollinating may reveal the extent of specialization within these relationships and the mechanisms by which this specialization occurs. Such studies mi ght show that pollinator behavioral traits can drive flower evolution and lead to greater specialization. The result of these specializations is greater diversity and finer niche partitioning. Continuing research might also attempt to quantify niche differ entiation among all seven species of nectarivorous bats in the Monteverde area. These studies might increase understanding of how resource partitioning mechanisms elevate diversity within communities and ecosystems as well as shed more light on coevolution theories presented in this paper. Acknowledgements I would first like to thank Richard LaVal for his inspiration, vast bank of knowledge, and guidance. Thank you to Michael and Patricia Fogden for permitting me to use the Monteverde Hummingbird Gallery. Thank you to Bill Haber and Willow Zuchowski for plant identification and help with methods. Thank you to the staff of Estacin Biol gica, the Monteverde community, and especially the Arguedas Villalobos family for their hos pitality and support. Thank you to the University of Wisconsin Madison Department of Wildlife Ecology. Thank you to Evan Adams, Sarah Bachman, and Catherine Ayers for their help in the field. Thank you to James Dugan for help in the field and critical rev iew of proposal. Much gratitude and special thanks goes out to Alan Masters, Rick F.X. Smith X, and R. Andrew Rodstrom for whom without, this project would not have been possible. Thank you Alan for your invaluable advice, critical review, and really for k eeping this project on the ground. Thank you Rick for your work servicing the nets, digital imagery, statistical help, entertainment, and the immense time and attentiveness you put into this project. Thank you Andrew for your experience, innovative ideas, handling of bats, innate composure in the field, and wide spanning general knowledge. F in almente, gracias a los Anoura geoffroyi para su cooperacin y su espritu tranquilo.

PAGE 9

Literature Cited Bawa, K.S. 1990. Plant Pollinator Interactions in Tropical Rain Forests. Ann. Rev. Ecol. Syst. 21: 399 422. Braun, E. and D. Shelley 2002. Influence of Musa acuminata Musaceae On Bat Pollination Systems in the Monteverde Area. CIEE Tropical Ecology and Conservation Fall. Cunningham, S.lA. 1995. Ecolo gical Constraints on Fruit Initiation by Calyptrogyne ghiesbreightiana Arecaceae: Floral Herbivory, Pollen Availability, and Visitation by Pollinating Bats. American Journal of Botany 82 12: 1527 1536. Feinsinger, P. 1983. Coevolution and Pollination. In D.J. Futuyma, M. Slatkin eds.. Coevolution. Sinauer Press: Mass. Futuyma, D.J. 1997. The Evolution and Importance of Species Interactions. In Meffe, G. and Ronald Corrol eds.. Principals of Conservation Biology, 2 nd Ed. Sinauer Press: Mass. Haber, W 2000. Vascular Plants of Monteverde. In N.M. Nadkarni and N.T. Wheelwright eds.. Monteverde : The Ecology and Conservation of a Tropical Cloudforest. Oxford University Press, Inc., New York. Heithaus, R.E. 1982. Coevolution between bats and plants. In T .H. Kunz ed.. Ecology of Bats. Plenum Press, New York, New York. Hutchinson, G.E. 1959. Homage to Santa Rosalia or Why Are There So Many Kinds of Animals? The American Naturalist. XCIII: 145 159. LaVal, R.K. and B. Rodriguez H. 2002. Murcilagos de Costa Rica. Instituto Nacional de Biodiversidad, Costa Rica. Lemke, T.O. 1985. Pollen Carrying by the Nectar feeding Bat Glossphaga soricina in a Suburban Environment. Biotropica 172: 107 111 MacArthur, R.H. 1958. Population ecology of some warblers of northe astern coniferous forests. In L.A. Real and S.H. Brown eds.. Foundations of Ecology. University of Chicago Press, Chicago, IL Muchhala N. and P. Jarrin V. 2002. Flower Visitation by Bats in Cloud Forests Western Ecuador. Biotropica 343: 387 395 Nowak, R. M. 1991. Walkers Bats of the World. The John Hopkins University Press, London, England. Terborgh, J. 1992. Diversity and the Rainforest. Scientific American Library, New York, New York.

PAGE 10

Tables ____________________________________________________________________________________ TABLE 1 Reproductive condition of three phylostomid bats Males considered ‚activeƒ with enlarged testicles, females with visible teats _________________________________ _____________________________________ Species and sex Reproductively active No sign of Activity Anoura geoffroyi males 42 21 Anoura geoffroyi females 0 30 Glossophaga commissarisi males 1 47 Glossophaga commissarisi females 1 6 Hylonycteris underwoodi males 2 3 ______________________________________________________________________ TABLE 2. Incidence of pollen on two species of nectarivorous phylostomid bats. Number of individuals with pollen vs. number of individuals without pollen. X = 73.044 ______________________________________________________________________ Species and sex With Pollen Without Pollen Anoura geoffroyi males 13 53 Anoura geoffroyi females 11 21 Glossophaga commissarisi males 55 5 Glossophaga commissarisi females 7 0 ______________________________________________________________________ TABLE 3. Pollen abundances of two species of nectarivorous phylostomid bats. Males have significantly higher pollen abundance levels than females in both A. geoffroyi X = 627.038 and G. commissarisi X = 190.078. Between species there is not a significant difference in pollen load abundance X = .174. ______________________________________________________________________ Species and sex N Number of pollen grains Anoura geoffroyi males 8 2157 Anoura geoffroyi females 10 1099 Glossophaga commissarisi males 11 2396 Glossophaga commissarisi females 7 899

PAGE 11

______________________________________________________________________ TABLE 4. Pollen richness of sexes of two species of nectarivorous phylostomid bats. Number of pollen types carried are not significantly different between sex in both A. geoffroyi X = .143 and G. commissarisi X = 1.923. _______________________________________ _______________________________ Species and Sex N Number of pollen types Anoura geoffroyi males 8 3 Anoura geoffroyi females 10 4 Glossophaga commissarisi males 11 9 Glossophaga commissarisi females 7 4 ______________________________________________________________________ TABLE 5 Pollen richness of three species of nectarivorous phylostomid bats. Pollen load richness does not differ between species X = 1.143. _________________________________________ _____________________________ Species and sex N Number of pollen types Anoura geoffroyi 18 5 Glossophaga commissarisi 18 9 Hylonycteris underwoodi 5 2 ______________________________________________________________________ TABLE 6 Morista Horn Index of Overlap between three nectarivorous phylostomid bats. ______________________________________________________________________ Species Overlapping Morista Horn Index Of Overlap Value A. Geoffroyi vs. H. underwoodi 0 A. Geoffroyi vs. G. commissarisi 0.001382205 G. commissarisi vs. H. underwoodi 0.592444724 ___________________________________________________________________ TABLE 7 Pollen morphospecies richness and abundance on three species of nectarivorous bats. ____________________ __________________________________________________ Bat Species and Sex Pollen Types a b c d e f g h i j k Abund. H. underwoodi males 2 0 0 0 89 0 0 0 0 359 0 0 448 G. commissarisi males 9 8 5 778 852 6 0 0 91 395 180 81 2396 G. commissarisi females 4 0 0 4 403 0 0 0 35 457 0 0 899 A.geoffroyi males 3 0 11 0 0 0 210 1936 0 0 0 0 2157 A.g eoffroyi females 4 0 0 0 0 0 799 262 15 0 23 0 1099 Totals 11 8 16 782 1344 6 1009 2198 141 1211 203 81 6999