Male orchid bees were attracted to chemical baits and collected in nine Atlantic Forest fragments in southeastern Brazil. Fragments differed in size and shape. Three additional sites were also sampled in a nearby large fragment. Three hypothetical core areas of each fragment were measured as the total area minus an area of 50, 100, and 200-m-wide perimeter. Abundance and richness were not correlated with either fragment size or ratio area/perimeter, but were positively correlated with the size of core areas. These results suggest that orchid bee conservation requires the preservation of the fragments with the largest possible core areas. Neither size nor shape alone (area/perimeter ratio) seemed to be good indicators of the value of a given fragment for sustaining diverse and abundant faunas of orchid bees.
Atlantic Forest; conservation; Euglossine; habitat fragmentation
