No abstract

-Secondary rainforest remnants might contribute to biodiversity conservation and preservation of healthy interspecific interactions with ongoing fragmentation. We studied the vertical distribution of trap-nesting bees along a fragmentation gradient of secondary forest remnants in Costa Rica. Fragment size did not affect bee abundance, diversity, and parasitism and mortality rates. However, height and edge effects influenced bee communities. Bees were more abundant in the canopy and the understory compared to an intermediate height, and bee diversity was higher in the canopy. Tree location (forest edge, intermediate distance, forest center) did not affect abundance but did affect bee diversity since most species preferred the forest interior. The cuckoo bees Aglaomelissa duckei and Coelioxys sp. 1 only partly followed their hosts' patterns, two Centris species. We conclude that an increasing amount of edge habitat will have negative consequences for bee communities and will reduce the conservation value of secondary forest fragments. fragmentation effects / parasitism rate / mortality rate / secondary forest fragments / vertical stratification

Foragers of several species of stingless bees deposit pheromone spots in the vegetation to guide recruited nestmates to a rich food source. Recent studies have shown that Trigona and Scaptotrigona workers secrete these pheromones from their labial glands. An earlier report stated that species within the genus Geotrigona use citral from their mandibular glands for scent marking. Since convincing experimental proof for this conjecture is lacking, we studied the glandular origin of the trail pheromone of Geotrigona mombuca. In field bioassays, newly recruited bees were diverted by artificial scent trails that branched off from the natural scent trail deposited by their nestmates only when they were baited with extracts from the foragers' labial glands. Compounds extracted from the mandibular glands, however, did not release trail following behavior. This demonstrates that the trail pheromone of G. mombuca is produced in the labial glands, as in Trigona and Scaptotrigona. Furthermore, in chemical analyses citral was identified exclusively in the foragers' mandibular glands, which disproves its supposed role as a trail pheromone. The labial glands contained a series of terpene-and wax type esters, with farnesyl butanoate as major constituent. We, therefore, postulate that the trail pheromone of G. mombuca is composed of a blend of esters.