Food Choice in Frugivorous Bats Lauren Riegler Department of Biology, Trinity University ABSTRACT Frugivorous bats are important dispersers for many tropical plants and their conservation depends on furthering knowledge in their foraging behaviors and food preferences Brosset et al. 1996. This study investigated a possible fruit preference of five frugivorous bat species Carollia brevicauda , Artibeus lituratus, Artibeus jamaicensis, Artibeus toltecus and Platyrrhinus vittatus found in Monteverde, Co sta Rica among three wild fruit species Solanum umbellatum, Solanum aphyodendron and Ficus pertusa and two cultivated fruit species Musa accuminata and Carica papaya . Fruits were presented to the bats in the Bat Jungle of Monteverde, where the foraging of bats can be closely observed. Artibeus toltecus showed a slight trend of preference for Solanum umbellatum over Solanum aphyodendron. However, due to small sample size and pseudoreplication there was no significant preference for any of the fruits by a ny of the bat species . RESUME N Los murciÃ©lagos frugÃvoros son dispersores importante s de muchas plantas tropicales y su conservaciÃ³n depende del incremento de nuestro conocimiento de su comportamiento de forrajeo y preferencias dietÃ©ticas Brosset et al. 1996. Este estudio investigÃ³ la posible predilecciÃ³n de cinco murciÃ©lagos frugÃvoros Carollia brevicauda , Artibeus lituratus, Artibeus jamaicensis, Artibeus toltecus y Platyrrhinus vittatus por tres especies silvestres de frutas Solanum umbellatum, Sol anum aphyodendron y Ficus pertusa y dos especies de frutas cultivadas Musa accuminata y Carica papaya en Monteverde, Costa Rica. Las fruta s fueron presentadas a los murciÃ©lagos en la Jungla de MurciÃ©lagos de Monteverde, donde fue posible observar deteni damente el compartimiento de los murciÃ©lagos. Artibeus toltecus mo strÃ³ una tendencia leve de preferencia p or Solanum umbellatum. Sin embargo, debido a l pequeÃ±o tamaÃ±o de la muestra y a la seudo r replica ciÃ³ n , no se demostrÃ³ ninguna preferencia significativa por cualquiera de las fruta s de parte de ninguna de las especies de murciÃ©lagos . INTRODUCTION Tropical plants appear to have specialized on particular groups of animals for dissemination of their seeds. Janson 1983 reported that two thirds of the 258 species of fruit he examined in a Peruvian moist tropical forest could be classified as "bird" or "mammal" fruits. Bats have diverse roles in the ecology of tropical forests; each genus of fruit eating bat has an important part to play in a different group of plants. The optimal foraging theory states that different species exhibit a variety of strategies along the continuum from generalist to specialist to maximize net energy intake Pyke et al. 1977. Eating large fruits and eating fruits with high amount s of desirable nutrient contents are two ways bats maximize fitness. Many species of fruit bats h ave shown specialist strategies, such as Carollia spp. , by concentrating on certain types of fruits in order to forage more efficiently. An example is Carollia spp., who prefers fruit high in protein and carbohydrates Piper spp., but avoid fruits high in fiber figs Fleming 1987 . Also, Carollia spp. concentrate s on plants that produce fruits often Cecropia spp., a pattern that is characteristic of certain pioneer species Fleming 1987 . Because of these
specializations, tropical plants will suffer reduced dispersal success and survival whenever their particular frugivore are reduced or eliminated. Bats are very important dispersers of tropical plants and are vital for forest regeneration through seed dispersal Brosset et al. 1996. Bat dispersed plants such as Cecropia spp. , Piper spp. , and Solanum spp. in the Neotropics are some of the first and most abundant plants to invade clearings. They often hav e large seed banks present before clearings appear and are dispersed by defecation during flight by bats commuting between their daytime roosts and feeding grounds soon after the disturbance occurs Fleming 1987. These pioneer plants attract bats that may bring successional plant species into a clearing, leading to continued forest growth Fleming 1987. Seed dispersal by animals is advantageous in tropical habitats because it allows seeds to escape from density dependent predation, reduces competition fro m a parent plant or siblings and allows the plant to colonize new habitats Howe and Smallwood 1982. A study of forest regeneration along the Peruvian floodplains found that dispersal either by the wind or by bats was the most effective way for fast growi ng trees to colonize newly exposed soil Foster et al. 1986. If bat dispersal were to be eliminated, the diversity of new plants would be greatly reduced and this would in turn lead to support fewer species of animals. Past studies have required mist net ting and/or flight cages to obtain fecal samples from fruit eating bats in the Monteverde area in order to determine their diet, but this current study allows observation of the foraging behavior of the bats without handling Dinerstein 1983; LaFlamme 2000 . In this study, five different frugivorous bat species were observed Carollia brevicauda , Artibeus lituratus, A. jamaicensis, A. toltecus, Platyrrhinus vittatus first hand in an artificial setting carefully reconstructed to replicate the natural envir onment as close as possible while observing their preferences of different wild fruit species found in the cloud forest of Monteverde. Certain species or genera have been studied for their fruit preferences and diet width in Monteverde and other areas; ho wever, none of these studies was able to observe the bats forage in their own habitat or observe the frugivores without disturbing them Dinerstein 1983; Handley and Leight 1991; Wendeln et al. 2000. Presented with three species of fruit Solanum umbellat um , Solanum aphyodendron and Ficus pertusa one would expect Artibeus jamaicensis to prefer Ficus pertusa fruits and Artibeus toltecus to prefer the Solanum spp. fruits, specifically Solanum umbellatum due to the high numbers of this fruit species found in fecal samples in the wild Dinerstein 1983. Overall, the larger bats A. lituratus, P. vittatus, C. brevicauda would prefer large fruits, so they should not show an interest in these wild fruits preferred by the smaller frugivores Kalko 1994. Strong preferences for certain species of fruit supports the idea that these bats are important to the environment as seed dispersers through their specializing on resources for optimal foraging. Determining the preferences of wild fruits found in nature by frugi vorous bats helps to uncover specialization strategies, which assist scientists in the conservation of not only frugivorous bats, but the plants for which they disperse. METHODS The study was conducted in the Bat Jungle of Monteverde that houses seven sp ecies of bats two nectar feeding and five frugivorous found in the Monteverde area; the five species of frugivorous bats consist of 21 individuals Artibeus toltecus 10 individuals ,
Artibeus jamaicensis 6 ind. , Artibeus lituratus 1 ind. , Platyrrhin us vittatus 3 ind. , and Carollia brevicauda 1 ind.. This set up was not an exact replicate of the natural environment for several reasons: cultivated fruits such as banana and papaya were presented in bowls, only females were present and there was a t ime change forced on the bats made to behave as if daytime were night, but it was a unique opportunity that allowed for the researcher to observe the bats up close without disturbing them. Three species of wild fruits Ficus pertusa , Solanum umbellatum and Solanum aphyodendron were collected in areas where the bat species were found in nature. Ficus pertusa was collected from one tree in the Orchid Garden in Santa Elena, S. umbellatum was collected from one tree near the DoÃ±a Flory restaurant in Monteve rde, and S. aphyodendron along with S. umbellatum from one tree was collected from ten trees along the road past La Collina Lodge Fig. 1a. Ripe fruits were carefully selected fruits were determined to be ripe when they were squishy, sweet, and of a ce rtain color Â€ S. umbellatum , yellow; S. aphyodendron , yellow/green; F. pertusa , dark purple, picked with the help of a tree clipper for fruits that were too high up, stored in a cooler with ice until they could be placed in a refrigerator and then present ed to the bats within the next three days. FIGURE 1a. Map of the Monteverde and Santa Elena area with the collection sites that contained the three species of fruits offered to the frugivorous bats. The bat shows where the Bat Jungle was located, the c ircles show where Solanum umbellatum was collected, squares show where Solanum aphyodendron was collected, and the triangle shows where Ficus pertusa was collected. Monteverde, Costa Rica. April May 2006.
Ten fruits each of two wild species were presented to the bats during every treatment. The two fruit species were placed inside bowls and then set in the same stand along with the other cultivated fruits given to the bats every day by the Bat Jungle staff such as banana, papaya, mango, pineapple, watermel on, tomato, cantaloupe or avocado. During the treatments there were two bowls of wild fruits and ten to eleven bowls of cultivated fruits presented to the fruit bats in three different stands everyday Fig. 1b. The treatments consisted of 1 S. umbellat um vs. S. aphyodendron , 2 S. umbellatum vs. F. pertusa , and 3 S. aphyodendron vs. F. pertusa . Each treatment was repeated four times over a 16 day period from April 20 to May 6. The experiment was run two times before observations were recorded in orde r for the bats to become accustomed to having wild fruit offered to them. The fruits were set out 30 minutes before the bats were released by a staff member, through op ening the door of the Ânight area Â‚ where the bats slept at night, at 0900 and then obser vations were recorded for two hours. The number of visits that resulted in a piece of fruit from one to three cm in width being taken from each species of bat was noted for the two wild fruit species along with two bowls of the favorite cultivated fruits of the bats Musa accuminata banana and Carica papaya Â€ papaya in the same stand Richard Laval, personal communication. The recording of visits to the cultivated fruits was started at the fifth session and only eight sessions of cultivated fruit data were collected. Bats were identified using the MurciÃ©lagos de Costa Rica: BATS Laval and RodrÃguez 2002. The total observation period was 22.5 hours over 12 sessions one session was cut short by one hour and another by half an hour due to unpredictable circumstances, but these sessions lasted long enough for total removal of the wild fruits to occur. FIGURE 1b. Set up of the Bat Jungle where observations took place. The Night Area refers to the holding area for the bats at nighttime. A visitor ent ers the viewing area from the right into a darkened room, where bats have been habituated to amber lighting in order to observe bats. The Viewing Area and Bat Jungle are separated by glass. The entrance to the bat area is for setting out the food for the bats and for cleaning the Bat Jungle at night. Monteverde, Costa Rica. April May 2006.
RESULTS Artibeus toltecus was the only frugivorous bat species in the Bat Jungle that ate the wild fruits presented during the study and, since there were only ten in dividuals of this species present, there was a high level of pseudoreplication. None of the bats ate F. pertusa . Artibeus toltecus ate three times more papaya than banana fruits and ate almost every piece of Solanum spp. fruit presented to it over the stud y period Table 1. Three of the other four frugivorous bats were observed eating the banana and papaya during the treatments. A total amount of fruit pieces eaten over eight of the twelve sessions shows certain trends of A. lituratus, A. jamaicensis and P . vittatus Table 2. A. jamaicensis ate three times more banana pieces than papaya fruit, A. lituratus ate similar amounts of the cultivated fruits and P. vittatus ate twice as much banana as papaya fruits. TABLE 1. Total amount of each wild fruit specie s taken by A. toltecus during twelve sessions, along with the amounts and rates of removal pieces of fruit per minute of the M. accuminata and C. papaya by A. toltecus over eight sessions 15.5 hours. The frugivorous bat was presented with a total of 80 fruits each of S. umbellatum and S. aphyodendron and a large supply of cultivated fruits M. accuminata and C. papaya that was never exhausted during the observational periods. Fruit Species Presented Solanum Solanum Musa Carica Frugivorous Bat Species umbellatum aphyodendron accuminata papaya Artibeus toltecus 77 75 46 139 Rates of removal 0.0495 0.149 TABLE 2. Total amount of each fruit species taken by three of the frugivorous bat species that did not take wild fruit during eight sessions. The frugivorous bats were presented with a large supply of cultivated fruits M. accuminata and C. papaya that they never exhausted during the two hour observation periods these values are from a total of eight sessions over a period of 15.5 hours. Fruit Species Presented Frugivorous Bat Species M usa accuminata Carica papaya Artibeus jamaicensis 102 34 Artibeus lituratus 10 11 Platyrrhinus vittatus 40 24 Artibeus toltecus showed no preference between S. umbellatum and S. aphyodendron Fig. 2. There was no significant difference in the average time periods necessary for A. toltecus to eat fruits of either S. umbellatum or S. aphyodendron Mann Whitney U, tied p = 0.4539 Table 3.
FIGURE 2. The mean time period re quired for Artibeus toltecus in treatment 1 to finish bowls containing ten pieces each of two different species of wild fruit. The Mann Whitney U test tied p value of 0.4539 shows no significant preference for one fruit over the other. Monteverde, Costa Rica. April May 2006. TABLE 3. Mean time periods required for Artibeus toltecus to finish ten pieces of either Solanum umbellatum or Solanum aphyodendron. a Refers to treatment 1 with S. umbellatum vs. S. aphyodendron, b refers to treatments 2 and 3 w ith either Solanum spp. vs. F. pertusa. Two choices combined the means of both Solanum spp. in treatment 1 in which the frugivorous bat ate both fruit species. One choice combined the means of both Solanum spp. in treatment 2 and 3 in which the frugivorous bat only ate one fruit species these are treatments involving F. pertusa . Mann Whitney U Test was used to determine preferences of A. toltecus over two species of wild fruit, S. umbellatum and S. aphyodendron , and differences between 1 wild fruit bein g presented with another wild fruit choice or 2 being the only wild fruit choice available . All tied p values show no significant differences in fruit choice. Specie s of fruit Mean Time min. Species of fruit Tied p value S. umbellatum a 7 2.5 S. u. a vs. S. a. a 0.4539 S. aphyodendron a 92.5 S. u. a vs. S. u. b 0.4539 S. umbellatum b 60 S. a. a vs. S. a. b >0.9999 S. aphyodendron b 87.5 S. u. b vs. S. a. b 0.5 516 Two choices S. spp. a together 82.5 Two choices vs. one choice 0.6259 One choice S. spp. b together 73.75 Additional Observations Artibeus toltecus was seen repeatedly returning to Solanum spp. bowls even once it had been emptied of fruit. This shows that certain individuals were returning many times to the Solanum spp. fruits pseudoreplication when foraging. An individual A. toltecus during one session was observed in an empty Solanum spp. bowl for over five seconds while searching for fruit. The taking of a piece of fruit usually occurs much more quickly 0 20 40 60 80 100 120 1 Fruit Species Mean Time minutes Solanum aphyodendron Solanum umbellatum
than this no more than one second. Artibeus toltecus individuals were also seen approaching bowls up to four times before entering and taking a piece of fruit. This shows that they c arefully select from the many choices of fruit offered. Artibeus lituratus was seen roosting in the same place every day during the study. Also, during observational sessions at stand 3, it was noticed that A. lituratus almost always visited the same bowl in stand 3 whenever foraging, regardless of which cultivated fruit was present in the bowl at the time. The bowl and stand were close to the roosting site, which suggests territoriality. Platyrrhinus vittatus was seen more often at stand 1 than the othe r stands and this might be due to the fact that 2 out of the 3 individuals roosted close to stand 1 everyday. This was more evidence that suggested the possibility of territoriality. Many of the females were discovered to have been pregnant when they gave birth during the study. Artibeus toltecus femal es had a total of seven babies . DISCUSSION Artibeus toltecus was shown to slightly prefer Solanum umbellatum over Solanum aphyodendron , as predicted, but due to small sample size and pseudoreplication, this preference was not shown to be statistically significant. Artibeus toltecus is known to feed on S. aphyodendron and S. umbellatum in nature and Solanum umbellatum is known to be a major part of the diet of this frugivorous bat during parts of the year Di nerstein 1983. Female bats, which are can be picky eaters and have higher energy demands than most males because they are either pregnant or lactating for ten months of the year, were the only sex present in the Bat Jungle at the start of the study Flemi ng 1987. Seven A. toltecus babies were born during the study and the preference of the reproductive mothers for S. umbellatum might be explained by the fact that the fruits eaten most frequently by reproductive Monteverde bats are those species that are g enerally high in soluble carbohydrates but relatively low in protein, which is the case in S. umbellatum Dinerstein 1983. Artibeus toltecus ate most, but not all of the wild fruits presented in the Bat Jungle because the wild fruit removal was impeded b y the presence of cultivated fruits such as banana and papaya, which are preferred by the fruit bats. Contrary to the general bat preference for M. accuminata , Artibeus toltecus seemed to show a preference towards papaya over banana. This trend could be a result of the nutrients that they were suddenly receiving from the Solanum spp. fruits, which could have caused a shift in their cultivated fruit choices. None of the frugivorous bats of the Bat Jungle ate the F. pertusa fruits and none of the bats, oth er than A. toltecus , ate the Solanum spp. fruits. T he other frugivorous bats ignored them because they eat other species of fruits in the wild and had a constant source of cultivated fruit, which has been shown to be preferred by bats Richard Laval, perso nal communication. The other frugivorous bats in the Bat Jungle specialize in other species of fruits besides Solanum spp., such as Ficus velutina for A. jamaicensis and A. lituratus Dinerstein 1983. However, this researcher was unable to locate the rig ht species of Ficus spp. in time for the study and the species presented, F. pertusa another plant with fruits eaten by frugivorous bats, was not of interest to the specific species of bats found in the Bat Jungle.
The other frugivorous bats in the Bat Jungle either suggested a preference for banana or had no preference between banana and papaya. Artibeus jamaicensis and P. vittatus showed a slight preference for banana, but due to the small sample size and pseudoreplication it was not statistically supp orted; however, data from the Bat Jungle itself did support the idea that the bats prefer banana over the other fruits Richard Laval, personal communication. The amount of fruit given and eaten was recorded on a daily basis and banana was presented in th e highest number of bowls out of all the cultivated fruits and was eaten more often by the bats of the Bat Jungle. Artibeus lituratus showed no preference over the fruits, but due to the fact that data were collected from only one individual the results we re not significant. Data were not collected on Car o l lia brevicauda due to its similar morphology with the other bats within the jungle and due to the fact that there was only one individual present in the Bat Jungle during the study. Some of the results mi ght be skewed due to the fact that C. brevicauda could have been confused with another species of bat during observations. The fastest mean time for removal of the fruit occurred with S. umbellatum in treatment 2 60 minutes. This mean time was lower tha n S. umbellatum in treatment 1 72.5 min., possibly due to the fact that the fruits were not competing with any other wild fruit since A. toltecus was not interested in F. pertusa . Solanum umbellatum was removed slightly faster than S. aphyodendron 72. 5 min. vs. 92.5 min.. This trend was observed possibly because S. umbellatum is known to be a major part of the diet of A. toltecus found in higher numbers in the fecal samples of A. toltecus in the wild and due to the high carbohydrate level found in S . umbellatum, which reproduc tive bats have been shown to prefer Dinerstein 1983. Also, the mean times for both Solanum spp. in treatments 2 and 3 73.75 min. were lower than the mean times for the Solanum spp. in treatment 1 82.5 min.. This supports t he idea that the wild fruits were being eaten faster in treatments 2 and 3 because they were not competing with each other when they were placed in the same stand with F. pertusa . Destruction of populations of fruit eating bats, during misguided campaigns to eradicate vampires, has adversely affected the reproductive success of their food plants Fleming 1987. The absence of fruit eating bats will make the reforestation of cleared areas much more difficult. Without a healthy population of fruit bats, the long term survival of some of Costa RicaÂƒs plants may be seriously put at risk, and a significant number of tropical trees and shrubs will lose their seed dispersers Fleming 1987. More studies on the interactions of frugivorous bats and the plants they d isperse must be conducted for the conservation of fruit bats and the forests they inhabit. Future studies could add to the knowledge of frugivorous bat food preferences by testing more species of each bat in order to diminish the effect of pseudoreplicatio n, using mist nesting as a supplement to the Bat Jungle, removing the cultivated fruit in order to test the preferences of wild fruit only, using only one stand to lessen effects of location preference or territoriality and presenting wild fruits that have been picked all at same time to avoid fruit aging as a confounding artifact. ACKNOWLEDGMENTS Thanks to my advisor, Javier MÃ©ndez, for all the productive arguments and laughter throughout this study and the writing of this paper, to Richard Laval for his vast knowledge and incredible generosity in letting
me take advantage of his amazing Bat Jungle, to the Bat Jungle staff for their kindness and UNO games, to Ollie Hyman and Maria Jost for assisting me throughout my study and for helping in making sense o f these statistics, to Karen Masters and Alan Masters for getting me started on my path towards bats and lastly to the bats for making this experience in Costa Rica even more incredible. Ã°J LITERATURE CITED Bonaccorso, F. J. & T. J. Gush. 1987. Feeding behavior and foraging strategies of captive phyllostomid fruit bats; an experimental study. J. Animal Ecol. 56: 907 920. Brosset, A., Charles Dominique, P., Cockle, A., Cosson, J.F., and Masson, D. 1996. Bat communities and deforestation in French Guia na. Canadian Journal of zoology, 74 11: 1974 1982. In: Genova, M. 2000. Costa Rican Bat Diversity in Primary and Secondary Forests and Banana Plantations. CIEE Tropical Ecology and Conservation. Dinerstein, E. 1983. Reproductive ecology of fruit bats and the seasonality of fruit production in a Costa Rican cloud forest. Doctoral Dissertation, University of Washington. pp. 107 136 Fleming, T. H. 1987. Fruit bats: prime movers of tropical seeds. Bats 53: 3 5. Foster, R. B., J. Arce, and T. S. Wachte r. Dispersal and the sequential plant communities in Amazonian Peru floodplain. pp. 357 370. In: A. Estrada and T.H. Fleming eds., Frugivores and seed dispersal. Dr. W. Junk Publs., Dordrecht, Netherlands; 1986. Handley, C. O., Jr. & E. G. Leight, Jr. 1991. Demography and natural history of the common fruit bat Artibeus jamaicensis on Barro Colorado Island, Panama. C. O. Handley, D. E. Wilson and A. L. Gardner: 141 149. Howe, H. F., and J. Smallwood. 1982. Ecology of seed dispersal. Ann. Rev. Ecol. Syst. 13:201 228. Janson, C. H. 1983. Adaptation of fruit morphology to dispersal agents in a neotropical forest. Science 219:187 189. Kalko, E. 1994. Diversity in tropical bats. In: Tropical Biodiversity and Systematics , H. Ulrich, ed. Proceedings of the Interation Symposium on Biodiversity and Systematics in Tropical Ecosystems. Bonn. In: Genova, M. 2000. Costa Rican Bat Diversity in Primary and Secondary Forests and Banana Plantations. CIEE Tropical Ecology and Conservation. LaFlamme, R. 2000. Ni che partitioning and overlap in two frugivorous bat species: Artibeus toltecus and Sturnira ludovici . CIEE Tropical Ecology and Conservation. pp. 237 248. Laval, R. K. 2006. Personal communication. Laval, R. K., and B. RodrÃguez H. 2002. MurciÃ©lagos de Costa Rica BATS. Instituto Nacional de Biodiversidad. Santo Domingo de Heredia. pp. 160 205. Pyke, G. H., H. R. Pulliam, and E. L. Charnov. 1977. Optimal foraging: a selective review of theory and tests. The Quarterly Review of Biology, 522: 137 154. In: Johnson, R. M. 2005. Niche partitioning by frugivorous bats in the San Luis Valley, Costa Rica. CIEE Tropical Ecology and Conservation. Wendeln, M. C., J. R. Runkle & E. K. V. Kalko. 2000. Nutritional values of 14 fig species and bat feeding prefer ences in Panama. Biotropica 323: 489 501.