1 Crop size, ripening, and f ruit removal in Solanum umbellatum (Solanaceae) by Phyllostomid bats Eleanor Caves Department of Biology, Pomona College ABSTRACT Effective seed dispersal is of primary importance to most plants. Maxim um dispersal should res ult from a larger fruit crop, since large fruit crops are more obvious to frugivores; however, producing too many fruits can be detrimental to dispersal as it satiates the frugivore. Thus, plants must strike the proper balance to maximize dispersal. I ex amine the effects of ripe fruit density , ripe fruit number, and fruit crop size within infructescences on fruit removal probabilities (FRPs) of wild Solanum umbellatum in Monteverde, Costa Rica . Further, I examine preferences of frugivorous Phyllostomid b ats for various proportions of ripe fruits. Each day, the ripeness status of 1537 S. umbellatum fruits was recorded. Simultaneously, flight cage experiments were performed with five species of frugivorous Phyllostomid bat, in which the bats were offered arrays of three different proportions of ripe of S. umbellatum fruits. Results showed that FRP was independent of proportion of ripe fruits per infructescence on wild S. umbellatum , but was negatively correlated with total and ripe fruit crop size. In ca ptivity, bats exhibited a significant preference toward fruit arrays with higher proportions of ripe fruit; however, over time the bats learned to eat from all arrays equally. These results suggest that, because frugivorous bats in nature are not naÃ¯ve to S. umbellatum fruits (as they were in captivity), they can learn locations of fruiting plants; thus, S. umbellatum does not need to produce a certain proportion of ripe fruits to attract dispersers. Rather, frugivore selection may be driven by some unobs ervable factor, such as fruit nutritional quality. RESUMEN La dispersiÃ³n de semillas efect iva es de importancia primaria para la mayorÃa de las plantas. La dispersiÃ³n mÃ¡xima debe resultar de una cosecha de frutas mÃ¡s grande, puesto que cosechas de fruta s grandes son mÃ¡s obvias a frugÃvoros; sin embargo, la producciÃ³n de demasiadas frutas puede ser prejudicial para la disper siÃ³n porque satura a el dispersor . Por consiguiente, las plantas tienen que encontrar un equilibrio para m aximizar la dispersiÃ³n. E xaminÃ© los afecto s de la densidad de frutas maduras, los nÃºmeros de frutas maduras, y el tamaÃ±o de la cosecha de frutas dentro de las infrutescencias en las probabilidades de removimiento de las frutas (PRF) de Solanum umbellatum salvaje en Montever de, Cos ta Rica. AdemÃ¡s, examinÃ© las preferencias de murciÃ©lagos frugÃvoros en la familia Phyllostomidae para proporciones de varias frutas maduras. Cada dÃa, la categorÃa de madurez de 1537 frutas de S. umbellatum fue tomada . SimultÃ¡neamente, experimentos en u na j aula de vuelo se realizaron con cinco especies de murciÃ©lago frugÃvoro, en que a los murciÃ©lagos se les ofrecieron muestrarios de tres proporciones diferentes de frutas maduras de S. umbellatum . Los resultados mostraron que PRF era independiente de la proporciÃ³n de frutas maduras por infrutescencia en S. umbellatum salvaje, pero estaba relacionado negativamente con el tamaÃ±o de la cosecha de frutas to tal y madura. En cautiverio , los murciÃ©lagos mostraron una preferencia significante a muestrarios con proporciones de frutas maduras mÃ¡s grandes; sin embargo, con el tiempo, los murciÃ©lagos aprendieron a comer de todos los muestrarios igualmente. Estos resultados sugieren que, porque los murciÃ©lagos frugÃvoros en la naturaleza no son ingenuos a las frutas de S. umbellatum (como estaban en cautiverio) pueden aprender la s ubicaciones de las plantas con frutas; ademÃ¡s S. umbellatum no necesita producir una proporciÃ³n determinado de frutas madu ras para atraer los dispersores . En lugar de eso, la selecciÃ³n p or los frugÃvoros puede ser empujado por un factor que no se puede obs ervar, como la calidad nutritiva de una fruta.
2 INTRODUCTION S INCE SEEDS FIRST EVO LVED IN THE LATE D EVONIAN AND M ISSISSIPPIAN PERIODS , their dispersal has been a driving force in the evolution of plants (Fleming 1991). Near the parent, density dependent mortality results in the death of undispersed seeds and seedlings (Janzen 1969). To move seeds away, many plants evolved fruits with morphologies, chemistries, and phonologies favoring different animal dispersers (McKey 1975). The Fruit Crop Size Hypothesis states that more seeds will be dispersed from a plant with a large fruit crop. Plants that produce large fruit crops are more conspicuous, and as a result are easier for fr ugivores to discover. This should result in an increase in the fitness of the individual plant for two reasons. First, plants with a larger fruit crop will be visited by more disperser species, as well as more individual dispersers, causing more seeds to be dispersed. Second, large fruit crops require less searching and handling time, and are preferred by frugivores over plants with smaller fruit crops (Ortiz Pulido 2007). However, production of too many fruits decreases both the quantity and quality of seed dispersal further dispersers that arrive are likely to be opportunists that do not deliver seeds to safe or appropriate sites. Satiated dispersers often remain in the canopy of a single plant, dropping many seeds directly underneath the parent tree. In addition, satiated dispersers may also leave ripe fruits on the tree to rot, or allow ripe fruits to fall to the ground to rot. Thus, it is important for a plant to strike the proper balance (producing neither too many nor too few fruits) to maximize dispersal, which in turn maximizes fitness. Fruit crop size (defined as either total number of fruits per infructescence or per plant), proportion of ripe f ruits per infructescence, and number of ripe fruits per infructescence are important to FRP in many disperser taxa, including birds, arboreal mammals, and bats. Positive correlations between fruit crop size per infructescence and removal patterns have bee n observed in temperate bird dispersed plants (Davidar & Morton 1986, Denslow 1987, Jordano 1989, Wilson & Whelan 1993). Bird dispersed plants in a Mediterranean scrubland were found to show removal patterns that were independent of fruit nutritional qual ity, ripening rate, ripening time, and within plant fruit density, but which were significantly affected by fruit crop size per plant (Herrera 1984). In monkey dispersed Panamanian rainforest trees, seed dispersal and FRP were found to increase dramatical ly with increased crop size (Howe 1980). In bats, however, the trend is less clear. Larger fruit crop sizes per plant in bat dispersed fig trees (Moraceae: Urostigma) in Panama were shown to be positively correlated with FRP (Korine et al. 2000). Howeve r, work by Fleming (1981) examined the bat dispersed fruits of Piper amalgo (Piperaceae) and found that nightly FRPs were high, but that those probabilities were independent of nightly and seasonal ripe fruit crop size per infructescence and per plant, pro portion of ripe fruits per infructescence, and number of ripe fruits per plant. The tropical plant Solanum umbellatum (Solanaceae) provides a model system for examining several fruiting characteristics and their effect on FRP. S. umbellatum produces mult iple infructescences, each with anywhere between 10 and 100 fruits, and ripens variable numbers of fruits per night over a period of several weeks (Dinerstein 1986). In Costa Rica, S. umbellatum is primarily dispersed by frugivorous Phyllostomid (Chiropte ra: Phyllostomidae) bats that perch briefly on a fruiting tree to remove a fruit and subsequently return to a roost to consume the fruit (Charles Dominique 1991). Because Phyllostomid bats are able to determine fruit ripeness (using traits such as odor an d color) number of ripe fruits and their spatial arrangement may be important to their FRP (Janzen 1983, Korine & Kalko 2005).
3 The purpose of this study is to test the effects (per infructescence) of proportion of ripe fruits, number of ripe fruits, and f ruit crop size on FRPs in wild S. umbellatum plants in Monteverde, Costa Rica. Simultaneously, I perform flight cage experiments with the primary dispersers of S. umbellatum , frugivorous Phyllostomid bats, to examine if bats favor certain proportions of r ipe fruit, as indicated by preferential eating of fruits, as well as differences in discovery time, when given artificial infructescences with varying proportions of ripe fruits. MATERIALS AND METHODS Study Site The field portion of this experiment w as conducted in Monteverde, Costa Rica, on the Pacific slope of the Cordillera de TilarÃ¡n. The flight cage portion of this experiment was conducted at the Bat Jungle, in Monteverde, Costa Rica. The Bat Jungle is a 209m 2 flight cage, designed to mimic the Monteverde Cloud Forest at night. Study Organisms Nine S. umbellatum plants at four study sites around the Monteverde area were used in this experiment (Figure 1); all were located in gardens or on forest edge. Data were collected near the peak of the S. umbellatum fruiting season, from April 5 to May 1, 2010; however, plants were at varying stages in their fruiting cycle when the experiment began. Figure 1. Map of study sites in the Monteverde area, Costa Rica. A total of nine study plants were used in this experime nt: six at the Centro EducaciÃ³n Creativa, one at the Bat Jungle, one at CASEM Co op, and one at the Hotel Belmar. All study plants were located between 1450 and 1600m, in Lower Montane Wet Forest. Map Legend: = Study Site = Ba t Jungle
4 The bats in the bat jungle have been acclimated to a 12 hour photoperiod in whic h 8:30 am to 8:30 pm is the night cycle. They are normally fed on a diet of banana, papaya, mango, watermelon, cantaloupe, apple, and available wild fruits three times daily at 8:30 am, 12:00 pm, and 4:30 pm. Five species of frugivorous bats in the family Phyllostomidae live in the Bat Jungle and were used in the flight cage experiment (number in parentheses indicates number of individuals): Artibeus lituratus (3) , Artibeus toltecus (36) , Artibeus jamaicensis (5) , Carollia brevicauda (9) , and Platyrrhinus vitatus (3). However, the majority of bats observed feeding on S. umbellatum fruits were A. toltecus . Field Experiment 50 infructescences of S. umbellatum fruits were used in this study, and were selected to be a representative sample of the infructes cences that I could reach without the aide of a ladder. Each of the 1537 fruits was marked with a unique identifying number, and each day the ripeness of each fruit was recorded as either unripe, ripening, ripe, or taken (Table 1). It should be noted tha t fruits disappeared for two reasons: either they were removed by a frugivore or they were dropped by the plant, as very ripe fruits are only weakly connected to the plant. Because it was not possible to distinguish between fruits that were taken by frugi vores and those that were of this paper. Plants were observed a total of 19 days, between 6:30 am and 9:30 am. TABLE 1: Characteristics used to classify r ipeness of S. umbellatum fruits in the field Ranking Fruit Color Fruit Size Fruit Texture Unripe Purely Green Small Hard Ripening Yellow Green Medium Slightly Squishy Ripe Purely Yellow Full Size Squishy Additional Observations Number of rotten fruits on each infructescence was recorded, and was included in counts of total numbers of fruits per infructescence. In addition, though no quantitative data were taken, qualitative information on the presence of ripe fruits on the ground underneath the parent tree was recorded. Flight Cage Experiment Three S. umbellatum fruit arrays (Figure 2) were designed to represent various proportions of ripe fruits that might be found on S. umbellatum in nature (Table 2). Fruits were collected from the field on the same morning that they were given to the bats. Each day, new fruits were placed on each array, in no particular orientation, and arrays were then placed in a feeding platform with three bowls; to ensure that results were not simply a product of fruit or a rray placement, the two different feeding platforms were used equally, and array types were shuffled between the three bowls. The bats were presented with three types of arrays of S. umbellatum fruits each day at 8:30 am; observation lasted for 15 minutes . When a fruit was removed during the observation period, time, ripeness status, and array type were recorded.
5 T ABLE 2: Numbers and proportions of ripe and unripe S. umbellatum fruits on three types of arrays offered to frugivorous Phyllostomid bats Ar ray # of Ripe Fruits # of Unripe Fruits Proportion of Ripe Fruits Proportion of Unripe Fruits Low 6 36 0.14 0.86 Medium 6 8 0.42 0.58 High 6 1 0.85 0.15 RESULTS Field Experiment 50 infructescences of fruit on nine S. umbellatum trees were examined, with a total of 1537 fruits. The average number of days to ripening was 1.95 0.30 (N = 825), while average number of days that a ripe fruit remained on an infructescence before being taken was 2.45 0.79 (N = 825). A total of 1211 fruits were taken du ring the course of the experiment, 825 of which were ripe, 212 of which were ripening, and 174 of which were unripe (Figure 3). The average number of ripe fruits that were taken from each infructescence (mean s.d. = 17.94 8.23) was higher than either ripening fruits taken (4.60 3.92) or unripe fruits taken (3.78 3.53). The difference in total number of fruits taken during the course of the experiment was significant between ripe and ripening (Friedman Rank test, df = 45, t = 13.326, p < 0.001) and between ripe and unripe (Friedman Rank test, df = 45, t = 14.152, p < 0.001). FRP was calculated using proportion of ripe fruits per infructescence that disappeared each night. FRP was found to be related to both number of ripe fruits on an infru ctescence and total number of fruits on an infructescence. However, FRP was independent of proportion of ripe fruits on a given infructescence (Figure 4). Total number of fruits on an infructescence, FIGURE 3. Mean numbe r of S. umbellatum fruits of each ripeness status removed by frugivores or dropped from plants. Error bars indicate standard deviation. Numbers over error bars indicate sample size. Bars with a dashed line above them were not significantly different (p > 0.05). 825 174 212
6 including ripe and unripe , was negatively correlated with proportion of fruits taken (linear regression, df = 1, f = 7.746, r 2 = 0.161, p = 0.005). Therefore, larger infructescences had proportionately fewer fruits taken. The impact was slight, however, as the regression coefficient suggests that the total number of fruits on an infructescence accounts for only 1.6 percent of the variation in proportion of fruits removed. Number of ripe fruits on an infructescence was also slightly negatively correlated with FRP (linear regression, df =1, f = 7.059, r 2 = 0.0147, p = 0.008). However, FRP was not significantly correlated with proportion of ripe fruits on a given infructescence (linear regression, df = 1, f = 0.821, r2 = 0.0017, p = 0.365). Thus, larger infructescences resulted in decreased fruit removal, a s did higher numbers of ripe fruits per infructescence. Proportion of ripe fruits per infructescence, however, had no effect on FRP. a) b) c) FIGURE 4. The effects of number of ripe fruits per infructescence, proportion of ripe fruits per infructescence, and total number of fruits on an infructescence on the proportion of ripe fruits taken in S. umbellatum in Mont everde, Costa Rica. a. Number of ripe fruits per infructescence had a significant, negative effect on the proportion of ripe fruits taken overnight (p = 0.008). b. The proportion of ripe fruits on an infructescence did not have a significant effect on th e proportion of ripe fruits taken overnight (p = 0.365). c. Total number of fruits on an infructescence had a significant, negative impact on the proportion of ripe fruits taken overnight (p = 0.005).
7 Additional Observations A total of 24 rotten fruits were observed on ten infructescences, but only on four different tr ees. The four trees that had fruits rotting were the same four trees underneath which fallen ripe fruits could be found each day. Few or no ripe fruits were ever found on the ground underneath the other five study trees. On average, ripe fruits remained on the tree longer on trees that had rotten fruits (mean s.d. = 2.77 0.487) than on trees on which no rotten fruits were observed (2.36 0.834), and that difference was statistically significant (one tailed t test, df = 24, t = 1.909, p = 0.034). Tr ees that contained rotting fruits appeared to be located further from other potential bat dispersed plants than those that did not. All study plants were located in gardens, having been gardens. Trees with rotting fruits were not located near any specialized garden; rather they were planted by themselves and were surrounded primarily by grass. Flight Cage Experiment Over the course of 15 trials, bats ate a total of 225 ripe fruits and only five unripe fruits, out of a total of 270 ripe and 675 unripe fruits that were offered, and thus bats significantly preferred ripe fruits to unripe fruits (chi square test, df = 1, = 208.3, p = 3.043E 46) (Figure 5). Bats took fruits from all array types, but preferentially from Medium and High, as shown by three indicators of preference: number of ripe fruits eaten during the 15 minute experimental period, number of ripe fruits eaten during the first three minutes of observation, and time to discovery, defined as number of seconds until the first fruit was eaten from each type of array (Figure 6). On average during each 15 minute observational period, bats ate 5.26 0.96 ripe fruits from array type High and 5.46 1.06 from array type Medium. However, on average, fewer ripe fruits were eaten from array type Low (4.26 1.75), and this difference was significant (ANOVA, df = 2, f = 3.636, p = 0.034). Average ripe fruits consumed within the first three minutes of o bservation was also significantly greater from array type High (3.66 1.54) and array type Medium (4.066 1.624) than from array type Low (2.066 1.709) (ANOVA, df FIGURE 5. Total number of ripe and unripe fruits eaten by frugivo rous Phyllostomid bats in the Bat Jungle, Monteverde, Costa Rica. n = 226. Over the course of the experiment, significantly more ripe fruits than unripe fruits were eaten by the bats (p = 3.04E 46).
8 = 2, f = 6.345, p = 0.0039). Finally, average discovery time (in seconds) for array type High was 24.86 15.73, which was shorter than for both array type Medium (39.66 89.45) and array type Low (179.4 247.3). Differences in discovery time were significant between array Low and Medium as well as between array Low and High (ANOVA, df = 2, f = 4.715, p = 0. 0 142). Despite the significant preference exhibited by the bats for array types Medium and High, preferences seem to have changed over the course of the experiment (Figure 7). Within 15 days, number of ripe fruits eaten from each t ype of array converged on six, the total number of ripe fruits available on each array; during the last two days of observation, all six ripe fruits were eaten from all three types of arrays. No significant differences were found in rate of change of a) b) c) FIGURE 6. Responses of Frugivorous Phyllostomid bats to three arrays of Solanum umbellatum fruits in the Bat Jungle, Monteverde, Costa Rica. n = 15. Error bars represent Standard Deviation. Bars with dashed lines over them wer e not significantly different (p > 0.05). a Average total number of ripe fruits eaten during the course of a single 15 minute trial. b Average number of ripe fruits eaten within the first three minutes of a trial. c Average time to discover each of the three types of arrays, defined as number of seconds until the first fruit was eaten from each array.
9 ripe fruits eaten, as represented by the slopes of the three regression lines (f test, df = 24, t = 0.883, p > 0.05). Time to discovery also became more equal for the three types of arrays as time passed; the last three days of observation, less than fifteen seconds passed between discovery of first array (which occurred at 10 seconds, 10 seconds, and 12 seconds, respectively) and the discoveries of the second (at 13 seconds, 15 seconds, and 18 seconds) and third arrays (at 21 seconds, 23 seconds, and 23 secon ds). Using the methods outlined in Zar (1984) for comparing the slopes of three regression lines, significant differences in change in discovery time were found between the slopes of the line for arrays Low and Medium (f test, df = 24, t = 2.985, p < 0.05 ), as well as between the slopes of the lines for arrays Low and High (f test, df = 24, t = 2.246, p < 0.05). DISCUSSION My results showed that FRP was independent of proportion of ripe fruit on an infructescence. Significant, negative correlations were found between number of ripe fruits on an infructescence and total fruit crop per infructescence; but as evidenced by low R 2 values, each of these factors only accounts for approximately one percent of the observed trend, indicating that there are oth er factors that determine the probability of fruit removal from S. umbellatum infructescences. Interestingly, the results of the field experiment contradicted the flight cage experiment, which clearly showed that bats feed preferentially from infructescenc es with higher proportions of ripe fruit. a) b) FIGURE 7. Number of Ripe Fruits Eaten and Time to Discovery of fruits by frugivorous Phyllostomid bats over a 15 day period of flight cage ex periments. a. Total number of ripe fruits eaten each day increased toward 6 (the total number available) for each type of array. b. Time to discovery for each type of array converged over the course of the experiment, showing a general decreasing pattern . Significant differences in slope were found between Low and Medium (p < 0.05) as well as between Low and High (p < 0.05).
10 This observed preference could be due to the processes Phyllostomid bats use to discover and localize food. Using a combination of olfaction and low frequency, multi harmonic echolocation signals, Phyllostomid bats determine target range, target position in space, and size, shape, and texture of objects (Korine & Kalko 2005). However, olfaction and echolocation are not used interchangeably olfactory senses are primarily used in initial detection and localizati on of a fruit, while echolocation is principally used for the final, short range localization (Thies & Kalko 1998). When ripening, many fruits undergo chemical changes that attract frugivores via odor (Janzen 1983). Thus, because initial localization of fruit is odor based, it is possible that higher proportions of ripe fruits produce enough odor to overpower the scent of nearby unripe fruits. As a result, bats in the flight cage may have initially found it easier to localize greater proportions of ripe fruits, which are subsequently preferred by Optimal Foraging Theory. One unexpected result of this experiment was that bats learned within a 15 day period to locate and eat fruits equally from each type of array. Although in flight cage experiments it w as observed that proportion of ripe fruits had an impact on frugivore preference, that preference was only a significant factor initially, when the bats were naÃ¯ve. Presumably, the rapidity with which bats learn can be beneficial to S. umbellatum , because it limits the period of naivety. In general, are able to create and learn a topological representation of their environment, which they use for planning routes (Winter & Stich 2005). Thus, frugivorous bats in the Monteverde area may be aware of the location of several S. umbellatum individuals, which they visit regularly, regardless of proportion of ripe fruits, ripe fruit number, and fruit crop size. Frugivores find fruits through a number of location cues, including memory, odor, and color (Janzen 1983); it is possible that frugivorous bats utilize certain cue s initially and then later switch to another. In the flight cage experiment a ten to fifteen day lag was observed before bat learning fully manifested itself, suggesting that factors such as proportion of ripe fruit may be important at the beginning of a fruiting season. Early in fruiting season, frugivores may be naÃ¯ve; however, in the case of S. umbellatum , fruits are produced almost year round, in all months except June and July, with fruiting peaks occurring from September to October and again from April to May (Dinerstein 1983). Bats are able to hold up to 40 different feeding locations in working memory without any sign of decay (Winter & Stich 2005); perhaps for a tropical fruiting plant with near year round availability, naÃ¯ve frugivores may not be a significant issue. Further experimentation at the beginning of the S. umbellatum fruiting season is necessary in order to determine if early season preferences exist as a result of frugivore naivety, or if the long fruiting season of S. umbellatum d oes not allow sufficient time for bats to become naÃ¯ve. Given that FRP cannot be fully explained by any of the factors tested in this experiment, other factors must be considered. One factor might be placement of a fruit or infructescence on a plant. In S. umbellatum many infructescences are located above the canopy of the tree, pointing up in an umbrella like shape. Fruits may preferentially be removed from these infructescences, as they are the easiest to find and thus might be preferred. Bats also use memory to learn to identify and prefer certain fruits (Gerardo Herrera M. 1999). Plants in the genus Solanum are known to provide both soluble sucrose sugars and free amino acids that can be assimilated in the very short intestinal transit times of m ost bats (Charles Dominique 1991). Thus, nutritional (e.g. rewards in relation to costs) to frugivores, and these differences should subsequently drive
11 f rugivore preference (Fleming 1991, Martin 1995). Some of this difference in preference was observed in this study, in that fewer than half of the study plants were ever observed with rotten fruits or ripe fallen fruits. Apart from plant location (in that plants with rotten fruits were not located in specially planted bat or native gardens) there were no observable differences between trees that had rotten fruits and those that did not. It is possible, then, that they differed in ways that could not easily be observed, such as fruit quality. Within species variation in fruit nutritional content has been shown to be a result of numerous factors, including fruit size, differences in environmental conditions between plant microsites, and soil variation (Izhak i et al. 2002). Given that a limited number of ripe S. umbellatum fruits are available each night, the population of frugivorous bats in the Monteverde area ought to disperse all the fruits (Howe & Estabrook 1977). However, ripe fruits were found remaini ng on the plant for an average of more than two days, and both rotten fruits and dropped ripe fruits were seen. For a plant with a dribbling fruiting strategy, these observations were surprising, for it is assumed that this strategy evolved to avoid frugi vore satiation. As few as twenty years ago, it was nearly impossible to find rotting fruits on S. umbellatum plants, and it was extremely rare to observe ripe fruits on the ground or tree (N.T. Wheelwright, pers. comm.); the majority of ripe fruits were t aken at night, and were replaced by newly ripened fruits by late evening of the following day at the earliest (E. Dinerstein, pers. comm.). My experiment took place in the morning, and I found that many ripe fruits remained on the plant even after the pri mary frugivores had come and gone. This observation correlates with a Monteverde wide decline in bat abundance since 1973, evidenced by decreasing numbers of bats caught in mistnets per hour, as well as declines in activity recorded by bat detecting senso rs (R.K. LaVal, unpubl. data). A decline in frugivore population introduces an imbalance into this coevolved plant disperser system and causes many fruits to either not be dispersed maximally or to not be dispersed at all, falling under the parent tree to suffer rot or predation . The evolutionary reaction, if any, of S. umbellatum to the decline in frugivore dispersers remains to be seen, but currently minor dispersers such as birds (Wheelwright 1984) may become more important dispersers, as bat populat ions dwindle. Perhaps competition amongst S. umbellatum individuals for frugivores will cause unusually strong selection on whichever trait maximizes effective seed dispersal. In an unaltered system, S. umbellatum is likely utilizing the fact that Phyllo stomid bat frugivores are not naÃ¯ve; thanks to frugivore memory, learning, and other unexplored factors, fruits are dispersed regardless of proportion or numbers of ripe or total fruits. However, this paper shows that continued observation of S. umbellatu m in the Monteverde area may provide insight into the consequences of frugivore decline on the fruiting strategies, seed dispersal, and overall fitness and success of a fruiting plant. ACKNOWLEDGEMENTS I would first and foremost like to thank Dr. Alan Masters, my advisor, for help designing and carrying out this experiment. My deep appreciation goes to Richard LaVal, owner of the bat jungle, as well as the entire staff of the Bat Jungle, for their patience and assistance with my project. Thanks to th e entire CIEE staff, Anjali, Pablo, Moncho, and Yimen, for helping with statistical analyses, supplies, and everything that I needed. Finally, I would like to express my gratitude to the other students of CIEE TEC Spring 2010, for their support and encour agement.
12 LITERATURE CITED CHARLES DOMINIQUE, P. 1991. Feeding Strategy and Activity Budget of the Frugivorous Bat Carollia persplicillata . Journal of Tropical Ecology 7: 243 256. DAVIDAR, P., AND E.S. MORTON. 1986. The Relationship Between Frui t Crop Sizes and Fruit Removal Rates by Birds. Ecology 67: 262 265. DENSLOW, J.S. 1987. Fruit Removal Rates from Aggregated and Isolated Bushes of Red Elderberry, Sambucus pubens . Canadian Journal of Botany 65: 1229 1235. DINERSTEIN, E. Reprodu ctive ecology of fruit bats and seasonality of fruit production in a Costa Rican cloud forest. Ph.D. Dissertation. University of Washington College of Forest Resources, 1983. DINERSTEIN, E. 1986. Reproductive Ecology of Fruit Bats and the Seasonality of Fruit Production in a Costa Rican Cloud Forest. Biotropica 18: 307 318. FLEMING, T.H. 1981. Fecundity, Fruiting Pattern, and Seed Dispersal in Piper amalgo (Piperaceae), a Bat Dispersed Tropical Shrub. Oecologia 51: 42 46. ----. 1991. Frui ting Plant Frugivore Mutualism: The Evolutionary Theater and the Ecological Play. Pages 119 114 in P.W. Price, T.M. Lewhinsohn, G.W. Fernandes, and W.W. Benson, eds. Plant Animal interactions. John Wiley and Sons, Inc., New York. GERARDO HERRERA M., L. 1999. Preferences for Different Sugars in Neotropical Nectarivorous and Frugivorous Bats. Journal of Mammology 80: 683 688. HERRERA, C.M. 1984. A Study of Avian Frugivores, Bird Dispersed Plants, and Their Interaction in Mediterranean Scrublands. Ecological Monographs 54: 2 23. HOWE, H.F. AND G.F. ESTABROOK. 1977. On intraspecific competition for avian dispersers in tropical trees. The American Naturalist 111: 817 832. HOWE, H.F. 1980. Monkey dispersal and waste of a neotropical fruit. Ecology 61: 944 959. IZHAKI, I., E. TSAHAR, I. PALUY, AND J. FRIEDMAN. 2002. Within population variation and interrelationships between morphology, nutritional content, and secondary compounds of Rhamnus alaternus fruits. New Phytologist 156: 217 223. JANZEN, D.H. 1969. Seed Eaters Versus Seed Size, Number, Toxicity, and Dispersal. International Journal of Organic Evolution 23: 1 27. ----. 1983. Dispersal of Seeds by Vertebrate Guts. Pages 232 262 in D.J. Futuyma and M. Slatkin, eds. Coevo lution: Sinauer Associates Publishing, Sunderland MA. JORDANO, P. 1989. Pre Dispersal Biology of Pistacia lentiscus (Anacardiaceae): Cumulative Effects of Seed Removal by Birds. Oikos 55: 375 386. KORINE, C., AND E.K.V. KALKO. 2005. Fruit Detect ion and Discrimination by Small Fruit Eating Bats (Phyllostomidae): Echolocation Call Design and Olfaction. Behavioral Ecology and Sociobiology 59: 12 23. KORINE, C., E.K.V. KALKO, AND E.A. HERRE. 2000. Fruit characteristics and factors affecting fru it removal in a Panamanian community of stranger figs. Oecologia 123: 560 568.
13 MARTIN, T.E. 1985. Resource Selection by Tropical Frugivorous Birds: Integrating Multiple Interactions. Oecologia 66: 563 573. MCKEY, D. 1975. The Ecology of Coevolved Seed Dispersal Systems. Pages 159 191 in L.E. Gilbert and P.H. Raven, eds. Coevolution of Animals and Plants. University of Texas Press, Austin. ORTIZ PULIDO, R., Y.V. ALBORES BARAJAS, S.A. DIAZ. 2006. Fruit Removal Efficiency and Success: Influenc e of Crop Size in a Neotropical Treelet. Plant Ecology 189: 147 154. THIES, W., AND E.K.V. KALKO. 1998. The roles of echolocation and olfaction in two Neotropical fruit eating bats, Carollia perspicillata and C. castanea , feeding on Piper . Behav Eco l Sociobiol 42: 397 409. WHEELWRIGHT, N.T., W.A. HABER, K.G. MURRAY, AND C. GUINDON. 1984. Tropical Fruit Eating Birds and Their Food Plants: A Survey of a Costa Rican Lower Montane Forest. Biotropica 16: 173 192. WILSON, M.F., AND C.J. WHELAN. 1 993. Variation of Dispersal Phenology in a Bird Dispersed Shrub, Cornus drummondii . Ecological Monographs 63: 151 172. WINTER, Y., AND K.P. STICH. 2005. Foraging in a complex naturalistic environment: capacity of spatial working memory in flower bat s. Journal of Experimental Biology 208: 539 548. ZAR, J.H. 1984. Biostatistical Analysis: Second Edition. Prentice Hall Publishing, New Jersey
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 M39-00313
Tamao de la parcela, maduracin y eliminacin de la fruta en Solanum umbellatum (Solanaceae) por los murcilagos Phyllostomid
Crop size, ripening, and fruit removal in Solanum umbellatum (Solanaceae) by Phyllostomid bats
Effective seed dispersal is of primary importance to most plants. Maximum dispersal should result from a larger fruit crop, since large fruit crops are more obvious to frugivores; however, producing too many fruits can be detrimental to dispersal as it satiates the frugivore. Thus, plants must strike the proper balance to maximize
dispersal. I examine the effects of ripe fruit density, ripe fruit number, and fruit crop size within infructescences on fruit removal probabilities (FRPs) of wild Solanum umbellatum in Monteverde, Costa Rica. Further, I examine preferences of frugivorous Phyllostomid bats for various proportions of ripe fruits. Each day, the ripeness status of 1537 S. umbellatum fruits was recorded. Simultaneously, flight-cage experiments were performed with five species of frugivorous Phyllostomid bat, in which the bats were offered arrays of three different proportions of ripe of S. umbellatum fruits. Results showed that FRP was independent of proportion of ripe fruits per infructescence on wild
S. umbellatum, but was negatively correlated with total and ripe fruit crop size. In captivity, bats exhibited a significant preference toward fruit arrays with higher proportions of ripe fruit; however, over time the bats learned to eat from all arrays equally. These results suggest that, because frugivorous bats in nature are not nave to S. umbellatum fruits (as they were in captivity), they can learn locations of fruiting plants; thus, S. umbellatum does not
need to produce a certain proportion of ripe fruits to attract dispersers. Rather, frugivore selection may be driven by some unobservable factor, such as fruit nutritional quality.
La dispersin de semillas efectiva es de importancia primaria para la mayora de las plantas. La dispersin mxima debe resultar de una cosecha de frutas ms grande, puesto que las cosechas de las frutas grandes son ms obvias a los frugvoros; sin embargo, la produccin de demasiadas frutas puede ser perjudicial para la dispersin porque satura al dispersor. Por consiguiente, las plantas tienen que encontrar un equilibrio para maximizar la dispersin. Examin los efectos de la densidad de las frutas maduras, los nmeros de frutas maduras, y el tamao de la cosecha de frutas dentro de las infrutescencias en las probabilidades de removimiento de las frutas (PRF) de Solanum umbellatum salvaje en Monteverde, Costa Rica. Adems, examin las preferencias de los murcilagos frugvoros en la familia Phyllostomidae para las proporciones de varias frutas maduras. Cada da, la categora de madurez de 1537 frutas de S. umbellatum fue tomada. Simultneamente, los experimentos en una jaula de vuelo se realizaron con cinco especies de murcilago frugvoro, en que a los murcilagos se les ofrecieron muestrarios de tres proporciones diferentes de frutas maduras de S. umbellatum. Los resultados mostraron que PRF era independiente de la proporcin de frutas maduras por infrutescencia en S. umbellatum salvaje, pero estaba relacionado negativamente con el tamao de la cosecha de frutas total y madura. En cautiverio, los murcilagos mostraron una preferencia significante a los muestrarios con proporciones de frutas maduras ms grandes; sin embargo, con el tiempo, los murcilagos aprendieron a comer de todos los muestrarios igualmente. Estos resultados sugieren que, porque los murcilagos frugvoros en la naturaleza no son ingenuos a las frutas de S. umbellatum (como estaban en cautiverio) pueden aprender las ubicaciones de las plantas con frutas; adems S. umbellatum no necesita producir una proporcin determinado de frutas maduras para atraer a los dispersores. En lugar de eso, la seleccin por los frugvoros puede ser empujada por un factor que no se puede observar, como la calidad nutritiva de una fruta.
Text in English.
Seed dispersal by bats
Costa Rica--Puntarenas--Monteverde Zone
Dispersion de semillas por murcielagos
Costa Rica--Puntarenas--Zona de Monteverde
Tropical Ecology Spring 2010
Ecologa Tropical Primavera 2010
t Monteverde Institute : Tropical Ecology