Sc. Latta et al., AN EXPERIMENTAL-STUDY OF NEST PREDATION IN A SUBTROPICAL WET FOREST FOLLOWING HURRICANE DISTURBANCE, The Wilson bulletin, 107(4), 1995, pp. 590-602
We used artificial nest experiments to study rates and patterns of egg
predation in subtropical wet and lower montane wet forest on the isla
nd of Puerto Rico. Levels of depredation were higher for ground nests
than for elevated nests, but we found no difference in rates of egg pr
edation between forest edge (5-50 m) and forest interior (100-250 m) n
ests. We quantified 25 forest vegetation variables surrounding ground
and elevated, forest interior nests, and correlated nest success with
these vegetation parameters. Utilizing discriminant function analysis,
nest success was correctly classified in 59% of cases (for ground nes
ts) using two Vegetation variables (shrub density and vegetation in th
e 2.5-3.0 m foliage height class), and in 71% of cases (for elevated n
ests) using six vegetation variables (canopy cover, vegetation in the
0-0.5, 0.5-1.0, 1.0-1.5, and 3.0-4.0 m foliage height classes, and Cec
ropia tree density). Our data suggest that the absence of an induced e
dge effect is the result of the creation of canopy openings and early
successional vegetative associations in a matrix across the forest fol
lowing hurricane disturbance in 1989. Comparative data suggest that th
e high rates of egg predation recorded here may he typical for insular
forest habitats.