Going Batty: Response of foraging glossophagine bats to auditorily perceived predator presence Hannah Kim Frank Department of Organismic and Evolutio nary Biology, Harvard University ABSTRACT Predator presence is known to alter the foraging beha vior of many species lead ing to decreased resource acquisition and lowered reproductive ra te (Werner 1994). Previous studies have examined the effect of predator presence on the emergence of roosting bats and the effect of predator models on foraging bats (Baxter et al. 2006). Only one study, however, has examined the importance of auditory cues in alerting foraging bats to the presence of predators (in this case, owls; Baxter et al. 2006). The study concluded that bats not only avoid perceived owls but also the call of a diurnal woodpecker, leading the authors to question whether the latter finding was the result of no vel sound avoidance or acoustic interference. This study, conducted in Monteverde, Costa Rica, investigated whether the findings of the earlier study were due to the effects of a novel sound or acoustic interference. Predator avoidance was quantified by counting number of visits to a hummingbird feeder after playing an owl call, a frog call or no call (a silent control). Significantly fewer visits were detected after the owl call than either the frog ca ll or silence (ANOVA, F = 11.0203, p = 0.0001, df = 2) indicating the results of the previous study were due to novel sound avoidance. This study also reaffirmed the importa nce of auditory cues as a technique for predator detection in foraging bats, demonstrating a globally significant trait of bats. RESUMEN Se sabe que la presencia de depredadores puede cambiar el comportamiento de forrajeo de muchas especies, llevando a una reduccin en la adquisicin de recursos y bajando la tasa reproductiva (Werner 1994). Estudios anteriores examinaron el efecto de la presencia de depredadores afuera de los sitios de descanso de murcilagos y el efecto de modelos de depredadores en los murcilagos que forrajean (Baxter et al. 2006). Sin embargo, solamente un estudio examin la importancia de seales auditivas como alerta ante depredadores sobre murcilagos que forrajean (bhos en este caso; Baxter et al. 2006). En este estudio se concluy que los murcilagos no solo evitan los bhos que perciben sino que tambin el canto de un carpintero diurno, lo que llevo a los autores a cuestionar si lo encontrado es el resultado de evasin de un nuevo sonido o de evita interferencia acstica. En este estudio, realizado en Monteverde, Costa Rica, se investig si los resultados del estudio anterior se deben a los efectos de un sonido nuevo o a interferencia acstica. La evasin de depredadores se cuantific co ntando el nmero de visitas a un comedero artificial despus del canto (playback) de un bho, una rana y s ilencio (como control). Haba menos visitas despus del canto del bho que despus del canto de la rana y el silencio, indicando que los resultados de Baxter et al. (2006) fueron debido a la evasin de un sonido nuevo. Este estudio tambin reafirma la importancia de seales auditivas como una tcnica para la detecci n de depredadores por parte de murcilagos que forrajean. INTRODUCTION Glossophagine bats are in the family Phyllo stomidae, also known as the American leafnosed bats (Nowak 1994). Members of the s ubfamily Glossophaginae, these bats exhibit 1
similar behaviors to hummingbirds, pollinating many tropical plants. They are specialized to extract nectar from flowers with elongate d tongues, a narrow snout with sensory hairs and the ability to hover (LaVal and Rodrgue z-H 2002). Still, many species have more generalized diets that include fruit and insects (Findley 1993). Because of this nectarivorous diet, glossophagine bats are attracted to hummi ngbird feeders at night, where they can be observed in large numbers (Heuer 2004). The presence or absence of predators near the feeding site is very important since predation is a major factor in bat mortality. In fact, predator avoidance has been offered as the reason for why bats roost colonially, emer ge from these roosts in groups and are nocturnal (Baxter et al. 2006). Risk of pred ation is especially high during foraging since with increased activity bats ar e more likely to encounter pr edators and because predators are often better at detecting m oving prey. Still, level of activ ity is positively related with resource acquisition and it is this tradeoff between obtaining food and the risk of mortality that defines the behavior of many species. Bats generally avoid predators and will avoid food if the predator is nearby (Baxter et al. 2006). This predator presence therefore hinders the foraging success of many species, leading to a lower reproductive rate, making the organism less evolutionarily fit (Werner 1994). Therefore predator presence is a powerful factor in determining not only a bats short-term feeding but also it s lifetime fitness. A variety of studies have been conducted to determine how roosting bats respond to the presence of predators. These studies have used methods that include using actual predators, visual models of predators and mode ls with playbacks of predator calls near bat roosts with mixed results. However, few researchers have examined the behavior of foraging, not roosting, bats in response to predators (Baxter et al. 2006). Heuer (2004) examined the response of foraging nectarivorous bats to a model snake or owl, known and important predators of bats, positioned at th eir feeder sites (Altringham 1996). She noted that bats avoided feeders with either model present but not a control without a model. As a follow up, Daz (2005) tested the hypothesis that bats were av oiding a novel object and not a predator by comparing bats reactions to a snake model, a novel object (a shoe) and a control, which lacked any object. She found th at bats avoided the snake model but not the shoe and concluded that bats only respond to the threat of a recognizable predator. However, bats do not only use visual a nd olfactory signals. Bats also use echolocation, which involves the transmitta nce of sound (Nowak 1994). Additionally, studies with bats and other prey animals have demonstrated that auditory cues are very important in predator detection (Baxter et al. 2006; Berger 1999; Schnitzler and Kalko 1998). Only one study, however, has looked at foraging bats detecti on of predators using auditory cues. Baxter et al (2006) used playbacks of ow ls, woodpeckers and silence at spatially separated stations to determine whether bats avoid owl calls. The bats they studied did avoid owl calls but also avoi ded the woodpecker calls. The researchers theorized that the avoidance of woodpeckers was likely du e to an avoidance of novel sounds or acoustic interference since the bats of their study were insectivorous bats, which rely heavily on echolocation (Nowak 1994). The woodpecker would indeed be a novel sound because the bats of their study, genus Myotis tend to roost in cav es and very rarely in hollow trees where they would be likely to encounter the diur nal woodpecker (Nowak 1994). This study attempted to clarify the results of Baxter et al. (2006) and to determine whether bats avoid a novel sound or acoustic interference by using pl aybacks of owl and 2
frog calls. Also, if the bats avoid predator s ong (owl), it further rein forces the conclusion of Baxter et al. (2006) that acoustic cues are importan t in detecting pr edator presence during foraging. This study expected a difference in the number of bat visits to the feeder after the various calls. Also, it was predicte d that there would be fewer visits to the feeder immediately after the predator call than after either the familiar call or silence, as bats should avoid the perceived predator. Si nce the bats are familiar with the frogs and know they are not harmful, the bats would not avoid the feed er after the frog call unless it is indeed the noise that bothers them. MATERIALS AND METHODS Bats were observed on 11 nights between July 14 and July 31, 2007, between 6:40 pm and 9:00 pm. They were observed visiti ng a hummingbird feeder in the garden at Selvatura Adventure Park, located at 1612 m on the Caribbean slope of the Tilarn mountain range (Millard 2004). The garden cont ains six feeder stands, each with four feeders. For this experiment, all feeders but the one observed feeder were removed to concentrate the bats and accurate ly monitor all bat activity. Three treatments were used in the expe riment: the call of the Bare-shanked Screech Owl, the call of the Common Tink Frog ( Eleutherodactylus diastema ) and a silent control. These species were chosen because they are known to live in the region of the study site; this ensures bats would be fa miliar with both calls (Fogden 1993; Leenders 2001). The silent control trials replaced the call with silence but were otherwise identical. Treatments were performed in a random order by sampling without replacement to control for the variation in visits with time of night. All treatments were run an equal number of times per night, twice a night except for one night in which trials were only once, for nine nights (the first two nights utilized a different procedure). Trials were run a total of 17 times for each call type. Only 15 trials were us ed in data analysis, however, since the data from July 21 were excluded based on abnorma l activity possibly due to aberrant weather and the presence of a screaming animal. For each trial the call was played and then th e number of visits were recorded for a five-minute period following the call (or sile nce). Calls were played, from various locations, for roughly one minute each to allow the bats time to hear the call and react. The frog call lasted 52 seconds, the owl 55 seconds and silence for approximately one minute. A visit was defined as a pause in front of the feeder; repeated visits by the same bat were counted as separate visits as described in Heuer (2004) and Daz ( 2005). Visits were observed using a headlamp and small flash light, both covered in red cellophane to minimize disturbance to the bats. Trials were spaced 20 minutes apart to allow bats to forget the previous call. This period was chos en based on observations of the bats on the first two nights as being adequate recovery time. Also, based on the first two nights of observa tions, the most reliable predictor of bat activity appeared to be presen ce of food in the feeder. On the subsequent nights feeders were refilled with a 20 % sugar solution by volum e. Volume of sugar solution in the feeder was measured immediately before the call a nd after the five-minute period to determine volume eaten by the bats during each trial. Other data was recorded as well. The amount of moonlight was quantified as no moon, slight moon or nearly full. Weather c onditions were recorded in regards to the 3
degree of rain and were quantified as dry, drizzl e, rain and downpour. Number of visits per trial was the average of counts by two individuals. Number of visits as a percentage of the nightly average was used in data analysis to c ontrol for the fluctuations in numbers of visits each night (between 481 and 922 visits per trial) Visits as a percentage of the nightly average was then analyzed with an ANOVA te st to determine whether bats selectively avoid any treatment. ANOVA test s were also used to determine the effect of weather, moonlight and time of night on the number of visits. A regression was used to relate volume eaten with number of visits. On two ni ghts bats were netted wi th butterfly nets to identify species using a guide by LaVal and Rodrguez-H. (2002). RESULTS Three species of bats were found to visit the feeders. Anoura geoffroyi Glossophaga commissarisi and Glossophaga soricina were all caught as they visited the feeder. All three species are common in th e Monteverde area (LaVal and Rodrguez-H. 2002). There were significantly fewer visits to the feeder after the owl call than after either the frog call or the silent control (ANOVA, F = 11.0203, p = 0.0001, df = 2; Figure 1). Weather was shown not to affect the number of visits in this experiment (ANOVA, F = 1.5309, p = 0.2090, df = 4). Similarly, the time of ni ght did not affect the number of visits during a trial (ANOVA, F = 0.3316, p = 0.9593, df = 9). Moonlight was significantly related to number of visits within a trial, showing fewer bat visits with no moon (ANOVA, F = 5.2969, p = 0.0083, df = 2; Figure 2). In ad dition, volume eaten was compared with number of visits and no correlation was discovered (R2 = 0.0422, p = 0.1483, N = 51; Figure 3). 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 frog owl silence call type Figure 1. Effect of call type on visits (mean s). Data over all nights except July 21 shows a strong relationship with fewer visits after the ow l call (Means for frog, owl, silence:1.0212 0.1028; 0.8918 0.1209 and 1.0870 0.1228) % nightly average is the number of visits in a trial divi ded by the average number of visi ts per trial for that night. 4
0 200 400 600 800 1000 1200 almost full none slight Amount of Moonlight 0 10 20 30 40 50 60 70 80 020040060080010001200 Visits Figure 2. Effect of moonlight on number of visits. There were significantly less visits in the absence of the moon. Figure 3. Volume change as a factor number of visits. No significant relationship was found. (Fit line: Y = 25.3945 + 0.0140 X) DISCUSSION Trials demonstrated that there were fewer visits after the owl call than the frog call or silence, supporting the hypothesis and pred iction of this study that bats avoid an auditorily perceived predator. The response to the predator call, reduction in number of visits to the feeder, supports the findings of Heuer (2004), Daz (2005) and Baxter et al. (2006) that bats respond to th e detection of predators with avoidance and that predator presence is an important factor in foraging act ivities. This is vitally important since in altering foraging behavior to a void the predator, bats lose an opportunity to gain food and potentially reduce lifetime reproductive success (Werner 1994). The fact that bats accept these costs to avoid predators indicates the impor tance of predation to these animals. Also, this study confirms the importance of owls as a predator of bats as was found in Heuer (2004) and Baxter et al. (2006). Finally, the relationship conf irms the findings of Baxter et al. (2006), emphasizing that aud itory cues are important for bats in the detection of predators. This study also clarifies the conclusions of Baxter et al. (2006) by determining whether their findings of reduced bat activity near woodpecker calls was the result of avoidance of a novel sound or an effect of acous tic interference. Since bats in this study did not avoid the frog call, it can be inferred that the bats in Baxter et al. (2006) were likely avoiding a novel sound and not acoustic interf erence. There was no novel sound in this trial so if the bats had avoided the feeder af ter the frog call, it woul d have been to avoid acoustic interference. Still, since the calls we re not played during the five-minute periods in which visits were counted, the role of acous tic interference cannot totally be ruled out. Also, the bats observed in this study are nectar ivorous bats in the tr opics, visiting a known and consistent feeding source. These bats likely have very different behavioral ecologies than insectivorous bats in British Columbia, th e bats observed in Baxter et al. (2006), for which echolocation may play a more important role in foraging and therefore for which acoustic interference would be a larger concern. 5
Additionally, this study examined the effect of time of night, weather and moonlight on the number of visits to the feed ers. Both time of night and weather did not affect the number of visits indicating that predator avoidance was likely responsible for differences in numbers of visits between tr ials. Daz (2005) also noted no significant effects in visitation as a result of weather or time of night; ho wever, she noted that bats are discouraged entirely from visiting the feeder by heavy rain. In contrast, I noted no significant difference in visits over all weather types, recording upwards of 500 visits in the five-minute periods in which th ere was torrential rain. The third effect examined was that of moonlight. Heuer (2004) found a significant relationship indicating th at with more moonlight there are fewer visits to the feeder. My study showed the opposite with significantly le ss bats on nights with no moon or obscured moon. This finding runs contrary to findings w ith many different specie s of bats including vampires and insectivorous species, which seem to avoid moonlight presumably to avoid predation (Altringham 1996). It also runs counter to the findings of Elangovan and Marimuthu (2001) who found that the number of f eeding visits of an Indian fruit bat to a tree was negatively correlated with percent moonlight. The significance of the current studys finding of fewer bats with less moonlight is that moonlight did not affect the number of visits to the feeder in this experiment. The bats in this experiment may be less affected by the amount of moonlight because of the light conditions of the site. Although the garden is not lit at night and is su rrounded by tall trees that block some of the surrounding light, there are buildings near the si te with lights that may acclimatize the bats that visit these feeders and make the e ffect of extra moonlight less relevant. Lastly, volume measurements were taken as another way to measure potentially the bat activity at the feeder. This variable did not relate to visits, likely the result of a variety of difficulties ascertaining the correct volume of solution in the feeders. As a result, volume measurements were not used in any of the data analysis other than to demonstrate that volume eaten (as well as could be measured) does not seem to correlate with number of visits. If one were able to ensure that they were measur ing the correct volume of sugar water consumed by the bats during the trial, it may tell a different story about foraging behavior and may be an interesting future study. Overall, this study reinforces the conclu sions of previous studies that predator avoidance is a major factor in Chiropteran foraging behavior. It also emphasizes the importance of auditory cues in predator de tection and their role in affecting foraging behavior, something only previously examin ed by one study. This previous study was carried out with insectivorous bats in Britis h Columbia. Because the current study shows similar conclusions, but was with nectarivorous bats in Cost a Rica, it demonstrates that avoidance of an auditorily perceived owl is a global trait of bats, emphasizing the importance of auditory cues and the role of ow ls as bat predators. Still, the high numbers of bats visiting after all treat ments demonstrates the importance of these feeders as a food resource in the region. Additionally, this study clarified the findings of Baxter et al. (2006), suggesting that confounding results we re the result of nove l sound avoidance and not acoustic interference. Fina lly, this study examined the effects of weather and time of night and determined that they were not responsible for the differences in number of visits per trial. The effect of moonlight was also ex amined with unanticipated results. This was interpreted to mean that, contrary to the lite rature, moonlight did not significantly affect visitation in this particular experiment. Future studies may consider playing the call 6
throughout the experimental period to look at the role of acoustic interference versus predator avoidance. This may also strengthe n the differences between treatments since in this study visitation tended to recover quickly with bats only avoiding feeders for a short period, approximately one minute, after the owl calls. Future studies may also wish to examine how much food is actually consumed by the bats during each trial as a different measure of risk taking. ACKNOWLEDGEMENTS I would like to extend a tremendous thank you to Arturo Cruz for accompanying me for nine nights, enduring extreme boredom and a lot of rain. I am also grateful to him for helping me net and identify the bats (or rather, for identifying the bats). I would also like to thank Tania Chavarra for her guidance and assistance throughout the entire process and especially with the statis tics. I am very grateful to the owners of Selvatura for allowing me to use their site. Thanks to Karen Masters for her advice on improving my project, help with statistics and general guidance. Thank you to Camryn Pennington and Pablo Allen for their help with statistics, general paper writing and, in Cams case, fo r allowing me to bug her with questions while she studies for the GREs. I am very grateful to Nick Fabeck, Jessica Walthew and Amy Elliott for help with Excel and editing. Lastly, Beth Schlimm has my eternal gr atitude for sitting in a thunderstorm to assist me on the wettest data collecting night of them all. LITERATURE CITED Altringham, J.D. 1996. Bats: biology and behavior. pp. 130, 219-221. Oxford University Press. New York, New York. Baxter, D. J. M., J. M. Psyllakis, M. P. Gillingham and E. L. OBrien. 2006. Behavioural response of bats to perceived predation risk while foraging. Ethology 112: 977-983. Berger, J. 1999. Anthropogenic extinction of top carnivores and interspecific animal behaviour: implications of the rapid decoupling of a web involving wolves, bears, moose and ravens. Proceedings of the Royal Societ y of London 266: 2261-2267. Daz, L. 2005. Effects of novel object versus predator presence on nectarivorous bat foraging behavior. pp. 123-129. Tropical Ecology and Conservation. CIEE, Summer. Elangovan, V. and G. Marimuthu. 2001. Effect of moonlight on the foraging behaviour of a megachiropteran bat Cynopterus sphinx Journal of Zoology 253: 347-350. Findley, J. S. 1993. Bats: a community perspective. pp. 25-26. Cambridge University Press. Cambridge, Great Britain. Fogden, M. 1993. An annotated checklist of the birds of Monteverde and Peas Blancas. pp. 1730. Michael Fogden. Monteverde, Costa Rica. Heuer, R. 2004. The effects of predator presence on nectarivorous bat foraging behavior. pp. 88-95. Tropical Ecology and Cons ervation. CIEE, Summer. LaVal, R. K. and B. Rodrguez-H. 2002. Murcilagos de Costa Rica/ bats. pp. 134, 140-141, 144-145, 148149. Editorial INBio. Costa Rica. Leenders, T. 2001. A guide to amphibians and reptile s of Costa Rica. pp. 102. Zona Tropical. Miami, Florida. Millard, E. 2004. Foraging strategies and body partitioning of glossophagine bats. pp. 1-11. Tropical Ecology and Conservation. CIEE, Spring. Nowak, R. M. 1994. Walkers bats of the world. pp. 125-129, 186-190. Johns Hopkins University Press. Baltimore, Maryland. Schnitzler, H.-U. and E.K.V Kalko. 1998. How echolocating bats search and find food. In H. T. Kunz and P. A. Racey (Eds.). Bat biology and conservati on. pp. 183-196. Smithsonian Institution Press. Washington, DC. 7
8Werner, E. E. 1994. Individual behavior and higher-order species interactions. In L. A. Real (Ed.). Behavioral mechanisms in evolutionary ecology, pp. 300-301. University of Chicago Press. Chicago, Illinois.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00148
Frank, Hannah Kim
Ir batty: respuesta de forrajeo de los murcilagos con la presencia de depredadores percibidos auditivamente
Going batty: response of foraging glossophagine bats to auditorily perceived predator presence
Predator presence is known to alter the foraging behavior of many species leading to decreased resource acquisition and lowered reproductive rate (Werner 1994). Previous studies have examined the effect of predator presence on the emergence of roosting bats and the effect of predator models on foraging bats (Baxter et al. 2006). Only one study, however, has examined the importance of auditory cues in alerting foraging bats to the presence of predators (in this case, owls; Baxter et al. 2006). The study concluded that bats not only avoid perceived owls but also the call of a diurnal woodpecker, leading the authors to question whether the latter finding was the result of novel sound avoidance or acoustic interference. This study, conducted in Monteverde, Costa Rica, investigated whether the findings of the earlier study were due to the effects of a novel sound or acoustic interference. Predator avoidance was quantified by counting number of visits to a hummingbird feeder after playing an owl call, a frog call or no call (a silent control). Significantly fewer visits were detected after the owl call than either the frog call or silence (ANOVA, F = 11.0203, p = 0.0001, df = 2) indicating the results of the previous study were due to novel sound avoidance. This study also reaffirmed the importance of auditory cues as a technique for predator detection in foraging bats, demonstrating a globally significant trait of bats.
Se sabe que la presencia de depredadores puede cambiar el comportamiento de forrajeo de muchas especies, llevando a una reduccin en la adquisicin de recursos y bajando la tasa reproductiva (Werner 1994). Estudios anteriores examinaron el efecto de la presencia de depredadores afuera de los sitios de descanso de los murcilagos y el efecto de los modelos de los depredadores en los murcilagos que forrajean (Baxter et al. 2006). Sin embargo, solamente un estudio examino la importancia de las seales auditivas como alerta ante depredadores sobre los murcilagos que forrajean (bhos en este caso; Baxter et al 2006). En este estudio se concluyo que los murcilagos no solo evitan a los bhos que perciben sino que tambin el canto de un carpintero diurno, lo que llevo a los autores a cuestionar si lo encontrado es el resultado de evasin de un nuevo sonido o de evitar la interferencia acstica. En este estudio, realizado en Monteverde, Costa Rica se investigo si los resultados del estudio anterior se deben a los efectos de un sonido nuevo o la interferencia acstica.
Text in English.
Tropical Ecology 2007
Ecologa Tropical 2007
t Monteverde Institute : Tropical Ecology