Frugivory And Seed Dispersal By Two Species Of Fruit Bats

Cynopterus sphinx and Rousettus leschenaulti are two species of abundant and widely distributed pteropodid bats in Xishuangbanna, Southwest China. Both are frugivorous and frequently found coexisting in the same areas. We studied the diet of them and investigated their roles as seed dispersers of their dietary plants (Musa acuminate, Morus macroura and Syzygium oblatum) and the effect of manipulation by the two fruit bat species on seed germination of Ficus racemosa and Ficus hispida. The roles of odor and vision in foraging behavior of Rousettus leschenaulti to lychee were evaluated in capture conditions. Home range and activity of Rousettus leschenaulti in winter were studied by radio-tracking in the suburb of Haikou city of Hainan Province. The results were following:Fruits of 20 plant species of 11 families and leaves of 2 plant species of 2 families were used by C. sphinx and fruits of 14 plant species of 8 families and leaf of 1 plant species of 1 families were used by R. leschenaulti. They mainly fed on fruits, including plant fruits such as lychee and longan, only leaves of 2 plant species were found. Figs were steady food resources for the fruit bats. These two species were sympatric and dietary overlap between them was relatively high in rain season (June ~ October), and 64% of the plant species recorded were used by these two species of bats in this season.The foraging behavior of Cynopterus sphinx on fruits of wild banana (Musa acuminata) and subsequent seed dispersal were studied. We found Cynopterus sphinx began to visit M. acuminata about 20–40 min after sunset. In the first half of the night (19:00-24:00), bat activity peaked twice, at 19:30-20:00, and 22:00-22:30. The number of individuals of C. sphinx captured by mist net were (mean±SE) 2.2±0.33/day and 1.4±0.32/day in the rainy season (September to October) and dry season (December to November), respectively; the difference was not significant. The number of seed pellets expelled was 9.0±1.12/day and 7.2±1.37/day in the rainy and dry seasons respectively; again the difference was not significant. The removal curves for marked fruits were similar for 10 focal trees. Fruits were consumed heavily within two weeks after ripening and all the marked fruit were removed within one month. The difference in seed dispersal was significant between different feeding roosts indicating that patterns of seed dispersal may not be uniform. We found the seeds of M. acuminata can be dispersed by C. sphinx to a distance of about 200 m, and C. sphinx can be considered as an effective seed disperser of M. acuminata.Observations on the foraging behavior of the two frugivorous bats at M. macroura fruiting trees indicated that they seldom fed on the fruit at the parent tree, but left with the fruit and flew to roosts in other trees to consume it. Their foraging activity peaked at 22: 00-22: 30 and 23:00-23:30. Mesua ferrea and Alstonia scholaris were favored feeding sites for bats, another tree species such as Baccaurea ramiflora, Mayodendron igneum, Caryodaphnopsis tonkinensis, Celtis tetrandra etc. also been used as feeding sites by bats. Feeding roost density decreased with increasing distance from the parent, but no correlation between distance and seed deposition. The effect of bat digestion on seed germination was investigated, and we found that germination percentage of all treatments involving ingestion by bats was significantly lower than control seeds and some germination parameters of seeds from different treatments were changed.There was no significant peak in foraging of the two frugivorous bats on fruits of Syzygium oblatum. Mean germination percentage of seed without pulp spat out by fruit bat in 3 different habitats (Non-roost, roost and under parent tree) were 91.7 %, 86.7%, 78.3 %, respectively. Seed germination percentage in non-roost was significantly high than under parent trees. The germination curves of seeds without/with pulp spat out by fruit bats and seeds in fruits were similar and final seed germination percentage were 88.0%, 92.0% and 86.7%, respectively. There were no significant differences among them. Rodents were important seed predators of Syzygium oblatum, including two species of squirrels and mouse, respectively. Seed and fruit predation were both distance-independent for both low and high density.We found most seeds of F. racemosa germinated during the first 6 d for all the treatments and both bat species. For F. racemosa, manipulation by Rousettus leschenaulti did not significantly affect GP. Tmin and GS of the F. racemosa seeds ingested by R. leschenaulti (including seeds from scats and ejecta) were same with those of control seeds. T50 was 4 d for control seeds but 5 d for seeds from scats and ejecta. The GP for F. racemosa seeds from C. sphinx scats and ejecta did not differ significantly from the controls. The treatment received by F. racemosa seeds ingested by C. sphinx (including seeds from scats and ejecta) did not change Tmin and GS. T50 was 4 d for all treatments. Germination in F. hispida occurred within 1 wk and GP reached about 99% in all treatments after 7 d. The GP of F. hispida was not significantly affect by the manipulation by R. leschenaulti. Compared with control seeds, the treatment received by the F. hispida seeds in the digestive tracts of R. leschenaulti resulted in an increase in Tmin and GS of 1 d, which was also true for the seeds from ejecta. T50 was 6 d for control seeds and seed from scats, but 7 d seeds from ejecta.Night feeding activity began at a similar time for both sexes of Rousettus leschenaulti, about 90-110 minutes after sunset. Foraging activity time of male was 391.8 min and that of female was 533.7 min at night. Based on 93 radiotelemetry positions, mean home range for tracked individuals was 2289 ha, The home range of different individuals showed a little overlap, R. leschenaulti often fly along the same flight route several days to foraging areas. The distance form day roost to foraging areas were 7.45-11.7 km and foraging areas of female were more than that of male. Feeding behavior of Rousettus leschenaulti toward lychee, a cultivated and preferred food of this bat, was investigated under captive conditions. We found that feeding activities comprised 25-30% of the total activity of these animals, and feeding and rest durations were not significantly different between two sexes. To evaluate the role of odor and vision in foraging behavior, we provided animals with mimetic lychee, real lychee and mimetic lychee soaked in the juice of lychee and recorded the number of approaches to different"fruit"types. The results indicated that bats approached real fruit significantly more than mimetic fruit, and the number of approaches to the soaked mimetic fruit was also significantly more than to the unsoaked mimetic fruit. The role of different sensory cues in the foraging behavior of these bats were discussed. We emphasize that the olfactory cue is important in locating food resources and discriminating food items.