Anthropogenic edges, treefall caps, and fruit-frugivore interactions in a neotropical montane forest

In a montane tropical forest in southwestern Colombia, we investigated how anthropogenic edges may alter bird-mediated seed dispersal from edge to for est interior as a function of edge age and presence of treefall gaps. We es timated fruit abundance and mist-netted birds at four distances from edge t o forest interior (0-10, 30-40, 60-70, and 190-200 m) in three young (<12 y r) and three old (>40 yr) edges. Fruit-sampling plots (50-m(2) plots) at each of the four distances were cla ssified into gap and intact forest. Fruit abundance and frugivore capture r ates varied from edge to forest interior, but such changes depended on edge age. At new edges, the total number of fruits was higher at the forest edg e than at the forest interior, whereas bird captures showed the opposite tr end. At old edges, the total number of fruits and bird capture rates did no t vary among the four distances. In a first group of 12 plant and four bird species, the distribution of individuals in fruit (7 species) and captures (3 species) from edge to forest interior differed between old and new edge s. In a second group of 18 plant and five bird species, which included thos e that were not amenable for a comparison between old and new edges and tho se that were not influenced by edge age, the distribution of individuals in fruit (12 species) and captures (3 species) was not uniform from forest ed ge to forest interior. Lastly, 124 plant and 19 bird species with <20 indiv iduals in fruit and captures, respectively, were classified into very spars e and sparse species. We found that all but the sparse frugivores were more abundant at the forest edge than in the forest interior. Because very spar se and sparse plant species showed such a clear trend, we used seeds retrie ved from mist-netted birds to assess potential seed movement of these speci es from edge to forest interior. Seeds of very sparse and sparse plant spec ies were found both at forest "edge" (0-10 m) and at forest "interior" (the three other distances combined). Our results suggest that birds are not responding to changes in fruit abund ance (resource-base-driven mechanism). Instead, they indicate that frugivor e capture rates reflect either a direct edge effect or a non-edge induced e ffect on birds. The apparent uncoupling of processes generating the observe d patterns in fruit and frugivore abundance may affect seed dispersal in im portant ways. Furthermore, our results indicate that, as edges age, "edge e ffects" (i.e., maximum distance at which changes induced by edge creation a re apparent within forest stands) change.