To assess the potential for competition between fruit-eating birds and
bats, we quantified overlap in their use of fruit in a matrix of dist
urbed and undisturbed tropical forest. Seeds were obtained from 897 ba
ts (31 species) and 200 birds (29 species) mist-netted over 2.5 yr. We
calculated pairwise overlap in fruit diet by the simplified Morisita
index for those frugivore species with the most diet data (11 bats and
7 birds). Overlap was 0 for 70% of the bird-bat species pairs, and le
ss than 0.05 for the remainder. Two tanagers, Tachyphonus surinamus an
d Tangara schrankii, had greater dietary overlap with bats than did ot
her bird species. Cluster analysis based on diet overlap separated bir
ds and bats into different clusters. Correspondence analysis was used
to reduce the dimensionality of the frugivore x fruit diet matrix. The
first axis separated bats and the fruits on which they primarily fed
from birds and their major fruits and was related to fruit length, typ
e, and color. The second axis separated the frugivorous bats by size a
nd was correlated with fruit length and plant growth form. Six fruits
were eaten by both bats and birds: Phytolacca rivinoides, Marcgravia s
p., Souroubea guianensis, Philodendron cuneatum, Coussapoa villosa, an
d Havetiopsis flavida. As a group, these shared fruits comprised a min
or fraction (< 4%) of most bat diets, but a larger fraction (> 10%) of
most bird diets. These shared fruits were morphologically similar to
those eaten by birds (arillate or berry-like, brightly colored exterio
r or aril), suggesting bats may be sampling ''bird fruits''.