Correspondence Graeme S. Cumming, DST/NRF Center of Excellence, Percy FitzPatrick Institute, University of Cape Town, Rondebosch, Cape Town 7701, South Africa. Email: firstname.lastname@example.org
Despite long-standing awareness of the potentially important ecological role of fruit bats, we know little about the ecology of the vast majority of species. Here we report the results of a pilot satellite tracking study aimed at establishing the scale of movement of the straw-coloured fruit bat Eidolon helvum. This was the first ever attempt to track African fruit bats using satellite telemetry. We tagged four bats with solar-charged 12 g satellite transmitters at Kasanka National Park in December 2005 and obtained a combined total of 104 different location fixes over a 149-day period. Before migrating, bats foraged as far as 59 km from the roost in a single evening; by contrast, one migrating individual moved 370 km in one night. Bats travelled an average 29 km day−1 over the period of study, with bats that appeared to be migrating moving north-west from Kasanka at an average 90 km day−1. The greatest cumulative distance travelled by a single bat was 2518 km in 149 days. The results show conclusively that the straw-coloured fruit bat E. helvum is capable of migrating thousands of kilometres across central Africa on an annual basis, implying that the fruit pulse in northern Zambia is richer than anything on offer in the Democratic Republic of the Congo at the same time of the year.
Fruit bats (Megachiroptera) occur in many African ecosystems in large numbers. Fruit bats play an important ecological role in the pollination of flowering plants, the dispersal and germination of seeds and the subsequent establishment of woody vegetation (Taylor & Kankam, 1999; Medellin & Gaona, 1999; Henryi & Jouard, 2007). Although many of the functions of bats are also performed by birds, some woody plants in Africa are almost entirely bat pollinated, and because the dispersal behaviours of fruit bats differ from those of birds, bats provide additional redundancy in these keystone ecological processes (Walker, 1992; Naeem, 1998; Duncan & Chapman, 1999).
An estimated 5–10 million straw-coloured fruit bats Eidolon helvum congregate between October and December each year at Kasanka National Park in north-central Zambia (Sorensen & Halberg, 2001; Richter & Cumming, 2006). The Kasanka colony (Fig. 1) is one of the largest known aggregations of fruit bats in the world. Straw-coloured fruit bats are found either year round or seasonally throughout much of sub-Saharan Africa, but there has been little previous research on their broad-scale movement patterns (or indeed, on those of any African fruit bat, Thomas, 1983). The broad-scale ecological role played by fruit bats in African savannas is largely unexplored, and as a result the potential contributions of fruit bats to ecosystem processes remain extremely difficult to assess.
We argued previously that straw-coloured fruit bats at Kasanka National Park migrate south from equatorial rainforests to take advantage of a seasonal pulse of fruit production in northern Zambia (Richter & Cumming, 2006). Our initial results showed that the timing of fruit bat arrival and departure matched closely the periods of increase and decrease in fruit availability at Kasanka National Park. However, this work (which was conducted solely at Kasanka National Park) was not sufficient on its own to determine the distances over which bats were responding to fruit availability.
In this follow-up study, we test the hypothesis that the scale of E. helvum movements in southern Africa is similar to that of broad-scale spatial variation in rainfall and consequently in fruit production. Broad-scale (latitudinal) variations in rainfall and plant phenology are well documented and relate closely to fruit production and to the reproductive strategies of bats (Hepburn & Radloff, 1995; Cumming & Bernard, 1997). If the mechanism driving E. helvum movements is a response to fruit availability, we expect that bats would move over large areas between rainfall regions to exploit spatial and temporal variations in fruit production. The alternative hypothesis suggests that the bats remain within the same rainfall region [i.e. are derived from the southern Democratic Republic of the Congo (DRC) and eastern Zambia] but aggregate annually at Kasanka National Park. This study constituted both a pilot project to test the application of satellite telemetry to straw-coloured fruit bats and a first empirical test of whether or not the bats were undertaking long-distance movements.
Alternative collar designs were tested and refined using a dummy transmitter on captive E. helvum at the Lubee Bat Conservancy. In December 2005, four wild adult male E. helvum were fitted with Microwave Telemetry's (Columbia, Maryland) 12 g, solar-charged satellite Platform Transmitter Terminals (PTTs) at Kasanka National Park (12°33′S 30°09′E). The original collar design was provided by Jon Epstein at the Consortium for Conservation Medicine. Our collars were constructed of a single piece of lightweight vegetan leather held together with brass rivets. A wool lining, cut to about 4 mm in length, was sewed on by hand to reduce chafing on the bat's upper back. The wool and transmitters were attached to the leather using Teflon fishing line (passed through the wire loops on the transmitter) and a quick set epoxy, which also covered the knots in the fishing line and ensured that they did not untie (Fig. 2).
Four bats were tagged. All were large males; their weights were 320, 290, 302 and 305.5 g, respectively. Tagged individuals were tracked using the Argos satellite system for periods that ranged from 27 to 149 days. We recorded both short- and long-distance movements, including nightly foraging flights in the park's vicinity and longer range migratory flights.
Male straw-coloured fruit bats can and do migrate at least 2000 km over a 3-month period. Summing each successive hop, the total distances moved by individuals 5915, 5916, 5917 and 5918 were 1875, 2518, 2046 and 1328 km, respectively. These movements were accomplished in 27, 149, 81 and 45 days, respectively. As the crow flies, the distances travelled by each bat were 878, 1975, 1492 and 1104 km. All four individuals moved from Kasanka towards the central and northern DRC (Fig. 3). Habitat use by individuals included Zambezian miombo woodland, forest-savanna mosaic, lowland forest and riverine forest. Although the bats at Kasanka National Park roost in a protected area, only two of our four tagged individuals moved across current conservation areas in the DRC.
Differences in reproductive timing among the females in the Zambian colony suggest that the large colony at Kasanka is an aggregation of multiple smaller colonies. In support of this argument, the timing of departure of each of the four tagged bats as well as the exact path of their migration differed. Although all individuals migrated to the north-west of Kasanka, their final locations were hundreds of kilometres apart.
The distances that were travelled over different legs of an individual bat's migration route varied widely, with one individual travelling as far as 370 km in a single night. A more typical mean distance for migrating bats is around 90 km day−1 (Fig. 4). The mean distance travelled per day between all data points was 29 km, with a standard deviation of 53.5 km (Fig. 5). These data include a short period of residency while bats were at Kasanka as well as movements during the migratory period. Data from pre-migration movements indicated that E. helvum will forage as far as 59 km from the central roost location at Kasanka.
Although our sample size is miniscule in comparison with the total number of bats present in the Kasanka colony, our results show unambiguously that straw-coloured fruit bats can move thousands of kilometres on an annual basis. The Kasanka colony is completely vacated each year. The long distances moved by some of our tagged individuals, together with differences in reproductive timing within the colony, suggest that the Kasanka National Park colony is composed of individuals drawn from a large area of central Africa. This result supports our previously published conjecture that the bats move south from the equatorial rainforests to northern Zambia to take advantage of a seasonal pulse in fruit production (Richter & Cumming, 2006). Given that central Africa is an area in which considerable spatial variation in competition and predation would be inevitable, and that long-distance movement leads to an increased risk of mortality, the results imply that the annual pulse of fruit production in northern Zambia is a much richer resource than what is available at the same time of year in the equatorial rainforest.
The length, direction and duration of recorded flights indicate that this species has the potential to disperse seeds and nutrients widely. Our results constitute the furthest recorded migration by any African mammal. Straw-coloured fruit bats are thus a potential broad-scale vector for pathogens; the last known location for one of our tracked individuals was within the range of an Ebola outbreak in 2007 and within 500 km of a site where Ebola antibodies were found in fruit bats (Leroy et al., 2005). We have no evidence to suggest that the last known location from this bat was the terminal migration site, and our results indicate that 500 km is a realistic distance for these animals to travel in a matter of days, making E. helvum a potentially fast-moving vector of infectious diseases.
Although we were able to obtain some useful data on fruit bat movements through the use of satellite PTTs, we would not strongly recommend further studies using the same (currently available) technology. The primary problems with this application of satellite telemetry are the relatively high expense (both of transmitters and of the ARGOS satellite downloads) and the need to use a solar-powered transmitter to reduce transmitter weight. Our impression was that once the bats entered the rainforests, the transmitters seldom received sufficient sunlight to charge fully. As a result, we were able to obtain only 5 months of data and even these data were somewhat patchy. Unfortunately, this particular problem has no obvious solution; lightweight VHF transmitters are cheaper, but following bats through the DRC in a microlight would be costly and dangerous. In captive trials, an 18 g transmitter was too heavy for captive bats to fly with; our research was only made possible through a technological advance that reduced the weight of the lightest available solar-charged satellite transmitter from 18 to 12 g. A 9.5 g model that was unavailable to us is now on the market, and a 5 g model is in its testing phase. Further studies of this kind on bats would ideally use a sub-15 g battery-powered satellite Global Positioning System transmitter. Eidolon helvum is an unusually good test species for future tracking studies of bats, because it is a clumsy flier and tends to fly well above the canopy when it commutes (making good satellite fixes relatively easy to obtain).
Lastly, our findings have some important implications for the long-term sustainability of forests and human livelihoods in one of the financially poorest parts of the world. The potential contribution of the Kasanka colony to fruit tree regeneration is particularly important in northern Zambia, where local communities are highly dependent on fruit, wood and other forest products. Only a small fraction of the migratory route of the Kasanka colony is currently protected, and forested areas throughout much of central Africa are under threat from climate change and expanding human populations. Eidolon helvum is a classic example of a species that needs a functional network of roosting and foraging sites (Poiani et al., 2000); preserving only one site on its migration route will not protect this particular population (Fleming & Eby, 2003). Although many of the same ecological functions may be performed by birds, the sheer number of fruit bats congregating at Kasanka suggests that their ecological impact will be high. Conservation of this population and the ecological functions that they perform will, in the long term, depend on international cooperation to maintain fruit-producing woodlands in northern Zambia and the northern DRC as well as a network of suitable stopover sites along E. helvum migratory routes.
We are grateful to Edmund and Kim Farmer and the Kasanka scouts for their assistance at Kasanka, the Zambia Wildlife Authority for issuing the necessary permit, Allyson Walsh for allowing us to test harness designs on captive fruit bats at the Lubee Bat Conservancy and Paul Racey for his valuable comments on an earlier draft of this manuscript. This research was supported by an NSF SGER grant to G.S. Cumming. All procedures were approved by the University of Florida's Animal Care and Use committee.