Fruit-eating bats (Phyllostomidae) catalyze forest recovery by dispersing seeds of early-successional trees and shrubs into disturbed habitats. We tested whether artificial bat roosts can accelerate forest recovery by increasing seed rain and soil nutrients in abandoned pastures in southern Costa Rica. We installed nine artificial roosts in each of five sites (N = 45 roosts). Three roosts per site were installed on wooden or steel posts in abandoned pasture and exposed to complete sunlight; three were installed in pasture on giant stakes of Erythrina poeppigiana; and three were installed on trees in adjacent closed-canopy forest. Giant stakes were four-meter tree limbs planted in the ground which quickly sprouted canopies and provided a minimal amount of shade for bat roosts. We hypothesized that: (1) fruit-eating bats would colonize roosts in forest more often than abandoned pasture; (2) colonization would be greater in roosts attached to giant stakes than to posts; and (3) colonization by bats would increase seed rain and soil nutrients (N, P) but not seedling recruitment. Roosts were installed in 2009-2010 and monitored twice per month for two years. Bats were identified by their droppings and with infrared video cameras.
Results/Conclusions
Bats rarely used roosts in abandoned pastures, even when the roosts were attached to giant stakes. Most forest roosts, in contrast, were used during at least one two-week monitoring period. The difference in roost usage between habitats may have been due to higher temperatures in pasture roosts. Rather than colonizing roost boxes as day roosts, bats generally used them as feeding roosts for consuming fruit or insects. Most bat detections were of arthropodivores (Phyllostominae) rather than fruit-eating bats (Carolliinae). When fruit-eating bats used roosts, abundance of bat-dispersed seeds increased below the roost relative to a control 10 m away. Seedling abundance and species richness, however, were not affected by roost usage. Given that (1) bats rarely used pasture roosts; (2) most of those that did were arthropodivores; and (3) increased seed dispersal did not result in a change in seedling abundance or species richness, we conclude that artificial bat roosts are not likely to accelerate tropical forest recovery in abandoned pastures. Future studies could test alternative roost box designs or assess the impact of roost use by arthropodivores on herbivore abundance and tree growth in established tree plantations with closed canopy.