Volume 30
Issue 2

Tropical bats face a housing crisis. New World leaf-nosed bats, members of the remarkably diverse family Phyllostomidae, are especially hard hit. Among their favored roosts are foliage, hollow trees and caves, all of which are disappearing in the face of deforestation and human disturbance. And sometimes roosts are willfully destroyed due to the lingering misconception that all bats are vampires. As bat populations decline, so, too, do the essential ecosystem services from insect control to pollination and seed dispersal that well-housed bats provide.

Artificial roosts could play a key role in reducing this housing shortage in the tropics. But that option is beginning to look more challenging than it might seem. A variety of substitute roosts and strategies have been explored by researchers, many of them with support from BCI’s Student Research Scholarship program. But despite many promising results, these projects often also raise new questions that need to be explored if artificial roosts are to help keep tropical bat populations and ecosystems healthy.

In temperate climates, bat enthusiasts usually think of artificial roosts as the familiar wooden boxes divided into narrow vertical roosting chambers. Unfortunately, most leaf-nosed bats disdain these well-tested bat houses. Indeed, no single roost design will be appropriate for all phyllostomids. In addition to trees and caves, some phyllostomids sleep in mammal burrows, termite nests or leaf tents. Many species also use such structures as mines, tunnels, buildings, bridges and culverts. The gamut of natural roost types provides researchers with a diverse palette for designing and testing artificial ones.

Replicated hollow trees are the main artificial roost style that has been tested by neotropical scientists so far. BCI Scholar Detlev Kelm and colleagues at the University of Erlangen-Nuremburg pioneered this technique in Costa Rica (BATS, Summer 2008). They built simple, inexpensive boxes about 6 feet (2 meters) tall and 2 feet (90 centimeters) wide and installed them on the ground in forest fragments. Within weeks, the roosts attracted 10 species of leaf-nosed bats.

Kelm’s team found that when bats colonized these artificial roosts, seed dispersal increased in the nearby vicinity. And the bats mostly dispersed seeds of “pioneer plants” the trees and shrubs that grow first in denuded forest areas and provide shelter for other plants to take root.

Ecologists were excited by this result because a lack of seed dispersal into forest clearings (such as pastures) is a major barrier to regeneration: no seeds, no trees, no forest. It appeared these artificial roosts could work like seed magnets to promote forest recovery on unused farmlands. But would the roosts pay off in hot, open pastures as well as they had in fragmented forests?

To find out, one of us (Reid) built and monitored 48 artificial roosts in southern Costa Rica in 2009-10 with BCI Scholarship support. One-third of the roosts were placed within forest fragments, while two-thirds were in abandoned cow pastures. Half of the pasture roosts were placed in the open sun and half in the shade of small trees. The roosts were made of fibrolite, an inexpensive, concrete-based material that’s widely available in Latin America. To help bats find the roost boxes amid dense, tropical grasses, they were installed on poles or trees several meters above the ground. To gauge seed dispersal, seed traps fine netting stretched across wire hoops were set beneath and at a random point 33 feet (10 meters) from each roost.

Unfortunately, bats rarely used roosts in abandoned pastures. Instead, the bats sought out artificial roosts in adjacent forest fragments. During the two-year study, bats visited nearly 90 percent of the roosts in forests. Using infrared video cameras, Reid identified several species of phyllostomid bats. When fruit bats visited the roosts, increased amounts of pioneer-plant seeds were detected in the seed traps, supporting the notion that bat roosts may increase seed dispersal, at least in forests.

For dispersal to matter, however, seeds must also germinate and survive; after two years of monitoring, Reid’s team detected no significant change in the number of seedlings beneath active bat roosts. Perhaps seedlings would have been more abundant had seed-carrying bats visited the roosts more often, but ultimately seed dispersal is most important in abandoned pastures, which bats rarely visited.

Reid’s roosts were used mostly for nocturnal feeding: bats brought insects or fruits to the roost at night, but they slept elsewhere during the day. Kelm’s roosts, in contrast, typically were used as day roosts and colonized permanently. The difference in bat behavior between the two experiments might be due to differences in roost design. Reid’s roosts, at about 2 feet (60 centimeters) tall, were significantly smaller then Kelm’s and were open at the bottom. These may have let in too much light and caused bats to seek darker locales for day roosts.

Another way that bats could improve forest restoration is by depositing nutrient-rich guano that could stimulate plant growth on degraded, tropical soils. Graduate student Ellen Holste of Michigan State University is currently analyzing the nutrient contents of soil collected below active and inactive artificial roosts in Reid’s experiment to identify any impacts related to guano.

With the growing interest in artificial roosts in the Neotropics, several key questions must be answered. Most importantly, what roost designs would attract phyllostomid bats? Research so far has focused on replicating hollow trees, but colonization has been spotty with current designs.

One intrepid naturalist, Mauricio Garca, recently dug a 16-foot (5-meter) long bat cave into a hillside in Costa Rica in hopes of increasing bat populations on his farm. He reports that bats use it as a feeding roost but are not sleeping there.

Many other possible designs are still unexplored. One option might be to “garden” for tent-making bats by planting their preferred tent plants mostly broad-leaved understory palms and heliconias. These bats craft their own roosts by gnawing on the stems of appropriate, very large leaves until the leaves fold over to provide shelter. A 2009 study by Felipe Melo and colleagues found that even the smallest of the tent-makers disperse large tree seeds. Another option might be to enlarge simulated hollow trees into tower roosts, similar to those developed by BCI’s Bat House Program and its partners for some North American bats.

Since fruit bats seem to roost primarily in forests, researchers should investigate the impact on degraded habitat of artificial roosts installed in adjacent forest. Kelm’s team found increased seed rain around artificial roosts within small forest fragments. Do such roosts also increase seed dispersal in nearby abandoned pastures? Casallas-Pabn recently initiated such an experiment, with support from a BCI Scholarship, in Colombia. Artificial roosts were affixed to isolated, remnant trees and to smaller trees in forest fragments. Seed dispersal and seedling recruitment are being monitored at the roosts and in adjacent pastures.

Another important question concerns the impact of artificial roosts on pest control. We have observed that insectivorous phyllostomids, such as Micronycteris bats, are often the first to use a new artificial bat roost. Recent work by BCI Scholar Margareta Kalka and her colleagues in Panama found that the common big-eared bat (Micronycteris microtis) is a voracious predator of plant-eating insects, and its presence reduces damage to tree seedlings (BATS, Summer 2008). Could strategically placed artificial roosts enhance such predation in recovering forests?

The development of artificial roosts for New World leaf-nosed bats is creating important new opportunities for research and action, but their utility for tropical bat conservation or forest restoration is still largely undemonstrated. At this point, we know that roost boxes are suitable for several species of phyllostomids in forests. We also know that seed dispersal increases when fruit bats use these roosts. Now we need to learn whether these initial conclusions can be generalized outside of Costa Rica and in other tropical ecosystems; which roost designs are most effective for phyllostomids; how we can reconcile bats’ preference for roosting in forests with the need for seed dispersal beyond forest boundaries; and how artificial roosts affect other ecosystem functions, such as pest control.

The challenges are significant. Even after finding the most bat-friendly designs for artificial roosts, they must also be made of locally available materials that are inexpensive enough for farmers to purchase or build and lightweight enough to carry to remote areas. And the roosts must withstand ravages of a tropical climate and the ubiquitous termites.

Forest restoration is not a substitute for preserving intact, natural habitat. But the continued development of artificial roosts for the Neotropics may nonetheless help reduce the housing crisis and get tropical bats back to work.

J. LEIGHTON REID is a doctoral candidate in Environmental Studies at the University of California at Santa Cruz. His roost study is conducted in collaboration with Ellen Holste (Michigan State University) and Zak Zahawi (Organization for Tropical Studies).

DIEGO CASALLAS-PABN is a doctoral candidate in Biological Sciences at the Universidad Nacional de Colombia in Bogot. His study is conducted in collaboration with Rosario Rojas Robles.