¹Northern Arizona University, Flagstaff, Arizona; ²Forestry Research Institute of Ghana, Ghana, West Africa.

Abstract

Milicia excelsa is a valuable but threatened tropical hardwood. Frugivory and seed dispersal play an important role in tropical forest. We measured Milicia seed rain and conducted focal-tree observations and seed germination and predation trials to determine the importance of bats to Milicia seed dispersal. We conducted 300 hours of focal-tree observations at five fruiting Milicia trees. We measured seed rain using 56, 2-6 m² seed traps totaling 200 m², for 5,800 m² trap nights. Traps were placed in canopy gaps in a 2 ha circular array at 3 distances from the focal tree. We conducted 7 germination trials using seeds that had passed through Eidolon's digestive tract, been spit out in rejecta pellets, extracted from fruit with pulp attached, and extracted from fruit, rinsed and dried. We recorded temporal and total germination rates. To test if Milicia seed predation rates differed significantly by density, distance to parent tree, or treatment category, we conducted 48 seed predation trials placing groups of 12 and 2 seeds treated as in the germination trials in small-medium canopy gaps at two distances from two focal trees. We recorded diurnal and nocturnal seed removal rates.

Bats accounted for >98% of Milicia seed dispersal and Eidolon visitation increased seed rain by more than 200%. No avian generated seed rain was recorded. Eidolon dispersed seeds appear to germinate faster and escaped seed predation longer than un-dispersed seeds. Bat dispersal may be essential to the long-term survival of Milicia excelsa.

1. Introduction

It is well known among tropical ecologists that seed dispersal and pollination by bats plays an integral role in tropical forest succession, distribution, and community composition (Fleming and Heithaus 1981, Fleming 1982). Furthermore, many of these bat dispersed and pollinated plants have great economic and cultural significance (Howe 1986). Unfortunately, many of the people responsible for the management of these forests and those that rely on them for their livelihood are not aware of the importance of bats, and as such, bats rank among the most persecuted and threatened animals worldwide.

In 1996 I conducted a review on the role of bats in tropical ecosystems while working at Bat Conservation International (BCI). I came across two references (Omaston 1965, Thomas 1982) which suggested that Milicia excelsa, a valuable and threatened tropical hardwood, may rely solely on bats, especially the straw-colored fruit bat (Eidolon helvum), for effective seed dispersal. While close mutualisms between plants and their pollinators have been well-documented, only a handful of similarly close relationships have been demonstrated between plants and their seed dispersers (Howe 1984). Providing documentation of this relationship would provide a strong economic argument likely the most effective approach given the importance of Milicia and timber to West Africa's economy to change the current status of Eidolon from persecuted to protected.

Coincidentally, I was aware of research being conducted on Milicia regeneration in Ghana by Dr. Michael Wagner at Northern Arizona University (NAU). After discussing my findings with Dr. Wagner and Dr. Tuttle of BCI, I decided to pursue graduate research on this important ecological relationship. We decided that to really gain an understanding of the relationship between bats and Milicia we would need an approach not attempted before; we would build platforms in the very tops of the Milicia trees to observe the interactions between bats and Milicia. In June of 1997, with funding by BCI and NAU, I made an initial three-week trip to Ghana to evaluate potential study sites, meet collaborators, and begin logistical arrangements. The Forestry Research Institute of Ghana (FORIG), Dr. Wagner's cooperators on the Milicia research, agreed to provide logistical support and a counterpart through the government of Ghana's National Service internship. FORIG felt that this project might provide an opportunity to add their first staff wildlife scientist.

Ghana's foremost wildlife authority, Dr. William Oduro of Kumasi's University of Science and Technology (UST) introduced us to a prospective intern, Mr. Bright Kankam. Bright had recently graduated at the top of UST's Wildlife Conservation Program. Bright's enthusiasm and devotion to wildlife conservation was apparent immediately. We determined that the Afram Headwaters Forest Reserve, two hours drive north of Kumasi, would be an ideal location due to it's accessibility and the availability of mature Milicia. Previous research at the reserve also provided information on fruiting phenology for several large Milicia. We also felt the reserve had a high likelihood of being visited by bats from the Kumasi Eidolon colony whose numbers swelled to more than 300,000 noisy members just prior to their annual 1700 km migration.

After returning to the U.S. I continued to refine the study design and solicit assistance for conducting the research. Meanwhile back in Ghana, Bright was conducting a survey of public perceptions surrounding the large Eidolon colony in the capitol of Accra and conducting test Milicia seed predation trials. In January of 1998, with the Milicia in bloom and support from the Columbus Zoo, The Conservation and Research Foundation, The Chicago Zoological Society's Conservation and Research Fund, and The Milwaukee Zoological Society's Foundation for Wildlife Conservation, I returned to Ghana with two trunks of tree climbing gear and field equipment to initiate the research.

2. Research

2.1 Research Objectives

The overall research objective was to determine the relative importance of bats, particularly Eidolon, to seed dispersal in Milicia. Importance was interpreted as the actual quantity of seeds dispersed relative to other animal taxa and the quality or effectiveness of dispersal. Quality or effectiveness would be the effects (if any) that dispersal has on survival and germination of the 2 mm elliptical seeds. Specifically, how many seeds are bats dispersing relative to other animals and seeds that fall beneath the parent tree and are bats dispersing seeds to places where they can survive and germinate?

2.2 Research Methods

We attempted to quantify the actual amount of seed dispersed by bats and Eidolon relative to other animal taxa. We also wanted to determine what proportion of the total seed crop of a tree is dispersed relative to what is not dispersed (falls beneath the parent tree). Because of the tremendous variation in our initial estimates of pre-dispersal mast crop for focal trees, we relied on estimates determined by adding the total amount of fruit removed by frugivores to the amount of fruit-falling beneath the parent tree. We employed seed traps to record seed-rain from volant frugivores. We conducted seed predation trials to determine the fate of dispersed and non-dispersed seeds and we conducted seed germination trials to evaluate the effects of bat dispersal on temporal and total seed germination rates.

2.2.1 Focal tree observations

We conducted almost 300 hours of diurnal and nocturnal observations in staggered 4 hour blocks at 5 fruiting Milicia trees (Heithaus et al. 1975, Fleming and Heithaus 1981). At three of these trees ground observations were made from palm thatch blinds. We conducted canopy observations from 2.5 m by 1.5 m wooden platforms with palm thatch blinds constructed in the crown of the focal trees. Two American and one Ghanian carpenter and climbers assisted in the construction of the platforms, nearly 35 meters above the forest floor. The main platform supports (joists) were fastened at the point of first trunk branching using heavy-duty webbing to minimize any long-term damage to the tree. Planks were then fastened to these joists with nails. Each platform took approximately 3 days to construct.

A combination of caving, rock climbing, and tree climbing ropes and technical equipment were used to haul materials up to the crown and gain initial and subsequent access (Lowman and Nadkarni 1995). We used binoculars, spotting scopes, and night-vision equipment to record the species or taxa and number of all animals observed visiting or feeding on fruits in or below the focal tree crowns. When possible we determined the feeding rate and recorded information on fruit handling (Chapman and Chapman 1996). Three trees were observed from the initiation of ripening until the cessation of frugivore visits and two focal trees were observed for a portion of their fruiting period.

2.2.2 Fruit-fall plots

We sampled 10% of the total area beneath the crown of 3 focal Milicia trees for fruit-fall by randomly placing 1.5 m² plots on the ground under the crown. We checked the plots daily and assigned the fallen fruits or fruit parts to one of five categories based on appearance; 1) Bat rejecta (fibrous fruit remains spit out by bats), 2) Parrot, fruit remains from feeding by parrots, 3) Anomalure, remains of fruit eaten by anomalures, 4) Other: remains eaten by other (unknown) animals, and 5) Knockdown/Fall down, fruits either falling on their own or knocked down by frugivores (indistinguishable). Plots were read from first fruit ripening to the cessation of fruiting.

2.2.3 Seed-rain traps

We measured seed-rain on 1% of 20,000 m² of forest surrounding one fruiting Milicia tree. We used 56 seed traps of 2, 4, and 6 m² for a total of 200 m² . The traps were made of heavy plastic stretched between four wooden legs and averaged 1.0 meters in height. We placed the traps in small to medium canopy gaps in a circular array consisting of three 6,666 m² zones at three distances from the focal tree. Traps were checked at sunrise and sunset to distinguish between avian and chiropteran seed-rain.

2.2.4 Seed Predation Trials

We conducted a total of 40 seed predation trials in canopy gaps at two different distances from two focal trees to determine how seed density, distance dispersed from the parent tree, and treatment by bats, effected seed predation. In each gap we placed four groups of 12 seeds each that had been: 1) passed through the gut of Eidolon, 2) extracted from fresh fruit with pulp attached, 3) extracted from rejecta pellets spit out by bats, and 4) extracted from fruit, rinsed and dried. In the center of each plot we placed a group of 2 seeds that were treated as in 1 and 2 above. Trials were initiated at dusk and dawn and seed plots were checked each hour all day or night to determine seed removal rates.

2.2.5 Germination Trials

We conducted 7 germination trials placing 50 seeds each that had been treated as in the predation trials above in petri dishes. Seeds were kept moist and at room temperature. Seeds were checked daily and germinating seeds were recorded and removed. We recorded temporal and total germination rates for each treatment (Utzurrum and Heideman 1991).

2.3. Research Results

Bats, especially Eidolon, are by far the most important seed dispersal agents for Milicia, both quantitatively and qualitatively. The most conservative estimate suggests the Kumasi colony Eidolon disperse more than 140 million Milicia seeds in a single night and that bat dispersal appears to increase both seed survival and germination rates.

2.3.1 Focal tree observations

We conducted just under 300 hours of diurnal and nocturnal focal tree observations of 5 fruiting trees (0=60, range 15-90 hrs.). The following animals were observed taking or eating Milicia fruit: a minimum of four species of bat (including Eidolon and the hammerhead bat, Hypsognathus monstrasus); two arboreal mammals, a squirrel (Paraxerus sp.) and Pel's anomalure (Anomaluris peli); one terrestrial mammal, the brush-tailed porcupine (Atherurus africanus); and one avian species, the red crowned parrot (an allied subspecies of Policephalus robustus).

Several species of avian frugivores were observed commonly or occasionally in or near the focal trees but were never observed taking fruit. These included francolins, hornbills, touracos, bulbuls, and drongos. Heavy hunting pressure in the study area undoubtedly reduced populations of at least three mammalian frugivore; bushbuck (Tragelaphus scriptus) duikers (Cephalophus sp.) and the mona monkey (Cercopithecus mona). On at least one occasion a duiker was flushed near a focal tree. Bushbuck tracks were recorded once near another focal tree, and a troop of mona monkeys habitually passed within 50 meters of a third focal tree but we never recorded sign of their presence at that tree.

Bats were the most frequent mammalian visitor and were observed taking fruit at all 5 of the focal trees. They were recorded visiting the focal trees on more than 85 % of the nocturnal observation periods and were recorded actively feeding for more than 70% of the nocturnal observation time during the fruiting season. Eidolon visited all but one focal tree. During peak fruiting, non-Eidolon bats removed an average of 7.5 fruits/10 minutes. During an Eidolon feeding bout this increased to an average of 99 fruits/10 minutes (n=80, range 21-241).

The red-crowned parrot was the next most frequent visitor and was observed taking fruit from at least 4 of the 5 focal trees. They visited the focal trees on more than 50 % of the diurnal observation periods during peak fruiting and were recorded actively feeding for approximately 50% of the observation time (the remaining time was spent resting, preening, etc.). During active feeding parrots ate an average of 3.5 fruits/10 minutes (n=20, range; 1:30-4:00 minutes/fruit).

Anomalures were the third most frequent visitor to all but one of the focal trees and were observed eating fruit as well as twigs and bark during more than 40% of the observation periods. An accurate feeding rate could not be determined for the nocturnal anomalures due to their sedentary and secretive nature and the size of the focal tree crowns. However, limited observations, fruit-fall plots, and time spent eating non-fruit plant parts all suggest that anomalures consumed only a small fraction of the fruit crop relative to bats and parrots.

On several occasions a single squirrel was observed in the late afternoon feeding at one of the focal trees. Approximately 5 feeding bouts (visits to the tree) were observed and none lasted more than 45 minutes. Squirrels ate an average of 1 fruit/3 minutes (n=15 observations). Two brush-tailed porcupines were observed on one occasion foraging beneath the crown of a focal tree. A feeding rate could not be determined.

2.3.2 Fruit-fall plots

The actual volume of fruit recorded by category has not yet been quantified. However certain patterns are evident at all three trees. At the two focal trees visited by Eidolon with fruit-fall plots, the number of rejecta pellets outnumbered both parrot-eaten fruit and fruits falling or knocked down (KD/FD) early in the fruit ripening period. The ratio then changed to reflect relatively equal numbers of rejecta, KD/FD, and parrot towards the middle of the fruiting period, and in late in the fruiting period a large KD/FD component followed by parrot and rejecta pellets. The same general pattern occurred at the tree not visited by Eidolon, however, the rejecta pellet and parrot component was relatively similar in the early to mid-fruiting period, shifting to a predominance of parrot and KD/FD late in the fruiting period.

2.3.3 Seed-rain traps

The seed-rain grid was centered on focal tree #1. Approximately 12 days after the first bat visits to tree #1, a second Milicia tree with ripe fruit and signs of feeding by bats was located approximately 200 meters to the northeast. The next closest known fruiting Milicia was approximately .75 km west. Seed traps were first checked on March 6 shortly after the first ripe fruit was detected. The first chiropteran seed-rain was recorded on March 12, and the last on April 13. A seed trap hit consisted of a portion to several defecation events (splats) recorded on one trap from one night. The average number of Milicia seeds per splat was 9, with a range of 0-40. A total of 287 seed trap hits were recorded. This extrapolates to a minimum of 257,400 seeds (12.87 seeds/m² ) recruited to the forest floor on the 2 ha. of area sampled in a single fruiting season. No avian generated Milicia seed-rain was recorded.

2.3.4 Seed Predation Trials

The preliminary results indicate that both fecal seeds and control seeds escaped initial detection and survived significantly longer than seeds with fruit pulp attached and rejecta seeds. When all treatments are combined the seed predation rate appears greater for trials closer to the focal tree than for the paired trials further away (n=26 trials). Groups of two seeds appear to escape predation longer than groups of 12. independent of seed treatment.

2.3.5 Germination Trials

The control seeds appear to have the most rapid germination rate followed closely by fecal seeds, seeds with fruit attached, and seeds from reject pellets. Total germinations rates appear to follow the same pattern, but they may be less pronounced.  

2.4 Discussion 

2.4.1 Focal tree observations

Four focal trees received greater than 90% of the observations, and a fifth was observed occasionally during peak fruiting. Bats and parrots were clearly the only frugivores to consistently consume Milicia fruit at these trees. As noted in the results, several other avian and mammalian frugivores were common or abundant in the study area, but were never observed taking fruit. Some of these species were even observed on several occasions in or under the focal trees and appeared to ignore the fruit, as was the case with hornbills and francolins.

It is important to note that the Milicia seeds are small and light-demanding (photoblastic) therefore they need to be scatter-dispersed in small numbers and land in a forest gap or other high-light situation for optimal seed survival and seedling growth. They will not germinate when buried in the soil and do not fare well with competition from other Milicia seedlings (Nyong'o et al. 1994). With this in mind, species such as the brush-tailed porcupine which defecates in latrines and other animals which defecate seeds in large clumps, such as duiker, elephant, or bushbuck, would not be high-quality dispersal agents regardless of the quantity of seeds they consume. This may also simplify the question of secondary dispersal by ants, or small rodents as cached seeds would not be expected to germinate.

The anomalures were also observed feeding on twigs and bark as well as Milicia fruit. Given their sedentary and obligate arboreal nature, anomalures would be expected to defecate the majority of seeds, regardless of viability, under the (Milicia seed-hostile) shade of the parent tree or neighboring trees. The same qualities that bestow poor disperser status on anomalures is exemplified in the red-crowned parrots. While parrots undoubtedly consumed a large quantity of Milicia fruit, they remained in the focal trees during their feeding bouts for long periods of time, defecating below the focal tree canopy.

We were never able to determine if the parrots were passing whole seeds but our observations of their fruit handling suggested they had sufficient dexterity to extract and crush seeds with their bill tips. We found one avian defecation on our platform that contained crushed Milicia seed immediately after parrots were noted feeding. However, the fact that we recorded 286 bat seed-rain events and no avian seed-rain in 5,800 m² trap days/nights sufficiently eliminates parrots as high quality dispersers of Milicia seed. In fact, our fruit-fall data shows that the high number of fruits knocked down by parrots (partially eaten or rejected after one bite) may make them more of a seed predator regardless if they are able to pass viable seeds.

In contrast, both Eidolon and other bats constantly flew around and between the focal trees and several feed roosts while feeding, or even between multiple fruiting (Milicia) trees. Given this feeding behavior, their preference to defecate while in flight, and their rapid digestion time (Thomas 1982), Eidolon and other bats perform a high-quality dispersal service distributing millions of Milicia seeds in small groups across the forest each night.

Lastly, much of the ecological literature on frugivory refers to fruits knocked-down by frugivores as an opportunity for secondary dispersal (by terrestrial vertebrates or insects). In contrast, we believe that due to the high level of below-crown seed predation by insects and the photoblastic nature of Milicia seed, knockdown actually represents a lost opportunity cost for the seeds, i.e. they have lost an opportunity for high-quality bat dispersal.

2.4.2 Fruit-fall plots

 The fruit-fall plots were established to help determine what proportion of the fruit crop was removed by each frugivore taxa relative to what fell beneath the tree therefore providing a pre-dispersal estimate of total mast crop (Chapman and Chapman 1996). This information will not be available until more in-depth statistical analysis is completed (target date, July 1, 1999).

The reduced numbers of rejecta pellets in the fruit-fall plots when Eidolon were not present supports our nocturnal observations that non-Eidolon bat species feed primarily by what we call pluck and run, landing briefly to take a single fruit and flying to a nearby feeding roost to consume it, before returning for another fruit. Eidolon on the other hand arrive in a large flock (12-100 animals) with approximately half of the animals feeding by pluck and run and the other half remaining in the focal tree and crawling from fruit to fruit, with the rejecta pellets falling below the crown. While the bats remaining in the tree are likely less efficient dispersers for the same reasons noted for the parrots above, this appears to be more than compensated for by the shear number of Eidolon feeding at one time, and the movement of members of the flock between fruiting trees and feeding roosts. This is evidenced by the 200% increase in seed trap hits during Eidolon feeding bouts over.

The fruit-fall plots also allow us to correlate the number of fruits eaten by parrots determined from our observations with the evidence provided by hundreds of partially parrot- eaten fruits in our fruit-fall plots. The lack of anomalure-eaten fruit remains in our plots also supports our limited nocturnal observations and contention that anomalures are relatively unimportant Milicia seed dispersers.

2.4.3 Seed-rain traps

Due to time and labor constraints we were only able to follow seed-rain at one focal tree, although some of the seed-rain hitting our traps undoubtedly originated from other focal trees (and less-likely unknown fruiting trees). However, our observations at the other 4 focal trees indicate that our results would not have differed significantly if more trees had been monitored. All five of the focal trees appeared to be visited consistently by bats and parrots, with only one focal tree believed to have not been visited by Eidolon. This tree had the smallest fruit crop of the 5 monitored and also appeared to be more distant from any other fruiting Milicia (as determined by extensive area searches). These factors may explain why Eidolon never found this tree or ignored the tree as it provided an insignificant (fruit) return on their foraging time and energy investment.

Many of the bat splats hitting the seed traps contained small groups of seeds, with an average of 9 for all splats with seeds. Many splats contained only a single seed or two or three seeds. If the final statistical analysis on our seed predation trials supports what appears to be a survival advantage for groups of two versus groups of 12 seeds (independent of treatment) this will provide further evidence of the high quality of Milicia seed dispersal by bats. The design of our seed trap array will also allow us to calculate the amount of seed-rain as a function of distance from the focal tree.

2.4.4 Seed predation trials

Several seminal seed dispersal studies in the tropics have supported the hypothesis that an optimal seed survival zone exists at a certain distance from the parent tree where the effects of high near-tree predation and high seed-rain rates are offset by lower seed-rain rates but correspondingly lower predation rates (Janzen 1970, Augsburger 1984). It was obvious that seed predation rates below the focal tree crowns were extremely high. Fallen fruits, newly dropped rejecta pellets, and seeds in bat splats were almost immediately attacked by seed bugs (order Hemiptera) and other invertebrates.

We conducted an experiment during one Eidolon feeding bout on 14 rejecta pellets falling beneath the focal tree. We counted the number of seeds remaining on and in the pellets upon hitting the ground and returned to count these again early in the morning. The mean number of seeds/pellet was 14 initially, and 3 the following morning. It also appears that seed predation rates were higher in the paired predation plots (trials) closer to the focal tree than the plots further away, independent of seed group size or treatment.

We predicted that the removal of the fruit pulp by the bats digestive system (as represented by the control and fecal seed groups) might reduce invertebrate seed predation if olfactory cues were important in seed discovery by seed predators. We also believed that seed group size (density) may be another variable affecting seed predation and survival for the same reason. Our preliminary results suggests that both seed density and seed treatment (removal of pulp) indeed decreased seed predation rates. If these preliminary results are supported by statistical analysis, it may help explain the mechanisms by which seed dispersal by bats increases Milicia reproductive fitness.

2.4.5 Germination trials

Passage through the gut of bats appears to enhance at least temporal, if not both temporal and total Milicia seed germination rates. This may be due to either the removal of an inhibitory chemical in the ovary (pulp) surrounding the seed, or the effects of some type of chemical scarification resulting from gut passage. The positive effect of passage through bat and bird guts on seed germination has been demonstrated in previous studies (Utzurrum and Heideman 1991, Izhaki 1995).

Another interesting aspect of the germination trials is an extremely poor germination rate among seeds extracted from rejecta pellets relative to both control seeds and fecal seeds. This may mean that the sucking action of a bat feeding on Milicia fruit is a strong selective force resulting in the differential ingestion of healthy (viable) versus less viable seeds.

2.5 Research Summary

In almost 300 hours of diurnal and nocturnal observations at five fruiting Milicia trees only bats and parrots were recorded eating significant quantities of Milicia fruit. Due to their sedentary feeding behavior, parrots were at best poor seed dispersers and may actually be Milicia seed predators. We recorded 287 bat generated seed trap hits and no avian seed-rain in 5,800 m² seed trap days-nights. In addition, parrots knocked down large quantities of fruit, making it unavailable for bat dispersal. 

We established fruit-fall plots beneath three Milicia trees to help quantify pre-dispersal mast crop size and the quantity of fruit dispersed by frugivore taxa. The plots also confirmed our observations that parrots knocked down large quantities of fruit and that non-Eidolon bats took fruit to nearby roost trees for consumption while Eidolon fed both in the focal trees and took fruits to feeding roosts. Due to their large flock size Eidolon dispersed more seeds than other bats. In contrast to other potential seed dispersers, bats scattered small groups of seeds over a large area, increasing the probability that they would land and survive in favorable sites.

Seed predation and seed germination trials indicate that bat-dispersed seeds may escape predation longer and germinate faster than non-bat dispersed seeds. This may be due to several factors including; the delivery of the seed away from areas of high seed predation, the removal of insect-attracting and germination-inhibiting pulp from the seed, and the dispersal of low density seed groups which are less likely to attract predators and result in less seed competition. Our research shows that the fruit-handling behavior of bats may also result in the differential ingestion of viable versus less-viable seeds.  

3. Conservation

3.1 Conservation Objectives

We believe the results of our research will help convince the Ghanaian government to provide protected status for the country's two largest Eidolon colonies, bringing a halt to the official and random persecution of these colonies. This will not only ensure that Eidolon continues to fulfill it's role as the primary dispersal agent for Milicia, but also preserve the integral role it plays in pollinating and dispersing the seeds of many other plants. Another goal is to use the information gained from this research as a foundation for developing an environmental education program on the interrelationship between humans, wildlife, and tropical forests. This program can then be used by Ghana's fledgling conservation organizations for school and civic presentations to promote bat and wildlife conservation, especially in the vicinity of the large Eidolon colonies and protected natural areas.

3.2. Conservation Action

Although bats are considered keystone species in tropical forest ecosystems because of their role in maintaining biological diversity, they rarely receive the research or conservation attention garnered by more visible and popular wildlife species such as elephants or primates. This is true at both the international and individual country level.

We felt that the best way to win support for bat conservation in Ghana, a country whose struggling economy relies heavily on natural resource extraction, would be to clearly demonstrate the role of bats in maintaining economically valuable species such as Milicia. We feel that this approach, together with plans for promoting bat-based eco-tourism, will convince the appropriate authorities to provide protected status for Ghana's largest bat colonies.

3.2.1 Protected Status for Eidolon colonies

We are currently pursuing two strategies that we feel will be successful in having the Kumasi Eidolon colony (and possibly the Accra colony) designated with protected status. In collaboration with Ghana's three non-government conservation organizations, we will use the study results to demonstrate the reliance of Milicia on Eidolon to the directors of the Forestry Department and the National Forest Products Marketing Board, as we have already done successfully with FORIG and UST. We are counting on these agencies, all of which are influential in the Ghanian government, to join us in requesting that the Ministry of the Environment provide protected status for the Eidolon colonies.

Our second strategy relies on the emerging importance and enormous potential of eco-tourism to Ghana's economy. We are currently collaborating with the Ghana Wildlife Department's Officer-in-Charge at the Kumasi Zoo to design and build Africa's first Bat Observation Tower in the midst of the Kumasi Zoo colony. This 20 meter tall observation tower will provide visitors a unique opportunity to observe the bats at eye-level and provide a spectacular view of their evening flight. Together with an educational display on the interrelationships between bats, other frugivores, and tropical forests, thousands of Ghanians and foreign tourists will learn to appreciate and understand the importance of the bat colony. At the same time, the Tower will demonstrate the economic importance of the bats to generating revenue from eco-tourism.

3.2.2 Conservation education

 By helping those who manage the forests and those who depend on it for their livelihood better understand the mutual interdependence between forest plants and animals, we will have a much better chance of keeping all of the pieces of this intricate living puzzle. Using a set of slides supplied by Bat Conservation International, we have already given presentations on the role of bats in rainforest ecology to students, conservationists, and professional foresters at FORIG, UST at Kumasi, Kakum National Park, and the Ghana Association for the Conservation of Nature (GACON). We received very positive feedback from these presentations, especially from the classically-trained professional foresters. Many of these foresters said that they had no idea how important bats were to forest regeneration.

We decided that we would wait until we had slides from our Milicia and Eidolon project before giving presentations to the schools and villages in the vicinity of the Kumasi Eidolon roost and our study area. We believe this will make the concepts less abstract and our presentation will depict familiar locations and individuals. Bright Kankam recently received a full set of slides depicting our research. Bright will be working closely with GACON and the Nature Conservation Research Center (NCRC, an Accra-based non-government organization) using these slides and the BCI slide set to tailor an educational presentation for teachers, students, and civic leaders in Kumasi and Kwapanin, the village closest to our study site. We have already located a generator in Kwapanin, and FORIG will provide the slide projector. Bright will give the presentations in English and the local Twi dialect. We are expecting that the NCRC will collaborate with us to develop a similar program for the capitol, Accra.

4. Further Research Needs and Conservation Action Plan

4.1 Further research needs

Additional research is needed to tie seed dispersal by bats to plant demography, the ultimate measure of plant reproductive success. We have been contacted by a student from the University of Aberdeen regarding potential bat research projects in Ghana, and have proposed research along these lines. We also collaborated with the NCRC to conduct several simple seed-rain experiments at the Kumasi Zoo and Accra Eidolon colonies. We placed plastic sheets beneath day roost trees and recorded the amount and percent of seed-rain by plant species. This work could be easily implemented and expanded to gather additional information on the dispersal and pollination of other economically and ecologically important plants by Eidolon.

The success of this research project has already focused attention on other plant-animal relationships relative to forest management. The director of FORIG has asked Bright Kankam to research the role of forest elephants in seed dispersal of another valuable and endangered timber tree in Ghana, baku (Tieghemella hecklii). The relationship between forest elephants and baku may be very similar to that of Eidolon and Milicia, however elephant populations are significantly lower than those of Eidolon.

4.2 Conservation action plan

I am planning on returning to Ghana in August of 1999 to facilitate the construction of the Kumasi Zoo Bat Tower. With support from FORIG, GACON, and the NCRC, Bright and I will give presentations to Forestry Department, National Timber Marketing Board, and Ghana Wildlife Department (GWD) officials to gather support for protected status for the Kumasi Eidolon colony. A retired structural engineer and docent with Prescott's zoo (Arizona) is currently working on a design for the bat tower and we are now looking for an architect to draw up the plans. Then we will price materials and estimate the amount of labor necessary for construction.

The director of the Kumasi Zoo will act as our liaison with the GWD to obtain permission and permits necessary for construction. We estimate that we will need approximately 2-3,000.00 dollars (U.S.) for tower construction by a local Ghanaian contractor. We will also solicit ideas and funding for the tower's interpretive display component. The tower will have two observation levels, one for the physically challenged if possible. A small access fee will be charged for locals and a larger fee for foreign tourists with the proceeds to benefit the Zoo and the GWD.

LITERATURE CITED

 Augsburger, C. K. 1984. Seedling survival of tropical tree species: interactions of dispersal distance, light-gaps, and pathogens. Ecology 65:1705-1712.

Chapman, C. A. and L. J. Chapman. 1996. Frugivory and the fate of dispersed and non-dispersed seeds of six African tree species. J. of Trop. Ecology. 12:491-504.

Fleming, T. H. 1982. Foraging strategies of plant visiting bats.pp. 287-325. In: The Ecology of Bats. Kunz, T.H. ed., Plenum Press, New York. 425 p.

Fleming, T. H., and Heithaus, E. R. 1981. Frugivorous bats, seed shadows, and the structure of tropical forests. Repr. Bot. 45-53.

Heithaus, E. R., T. H. Fleming, and P. A. Opler. 1975. Foraging patterns and resource utilization in seven species of bats in a seasonal tropical forest. Ecology 56: 841-854.

Howe, H. F. 1984. Constraints on the evolution of mutualism. Am. Nat. 123:764-777.

Howe, H. F. 1986. Seed Dispersal by Fruit-Eating Birds and Mammals. pp. 123-189. in: Seed Dispersa, D. R. Murray, ed. Academic Press. San Diego. 322 p.

Izhaki, I., Korine, C. and Z. Arad. 1995. The effect of bat (Rousettus aegyptiacus) dispersal on seed germination in eastern Mediterranean habitats. Oecologia. 101:335-342.

Janzen, D. H. 1970. Herbivores and the number of tree species in tropical forests. Am. Nat. 104:501-528.

Lowman, M. D., and N. M. Nadkarni. 1995. Forest Canopies. Academic Press, San Diego. 624pp.

Nyong'o, R.N., Cobbinah, J. R., and J. Appiah-Kwarteng. 1994. Flowering and fruiting patterns in Milicia excelsa and Milicia regia Welw. Ghana J. of For. Vol. 1, 1994.

Omaston, H. A. 1965. Pollen and seed dispersal in Chlorophora excelsa and other Moraceae, and in Parkii filicoidea (Mimosaceae), with special reference to the role of the fruit bat, Eidolon helvum. Commonwealth For. Rev. 44:96-104.

Thomas, D. W. 1982. The ecology of an African savanna fruit bat community: resource partitioning and role in seed dispersal. PhD thesis. University of Aberdeen, Scotland. 205pp.

Utzurrum, R. C. B. and P. D. Heideman. 1991. Differential ingestion of viable vs. nonviable Ficus seeds by fruit bats. Biotropica 23(3): 311-312.

Last Updated ( Thursday, 10 July 2008 )