Document Type



Master of Science



Date of Defense


Graduate Advisor

John G. Blake


Bette A. Loiselle

Robert J. Marquis


Foraging specialization is one mechanism that has been hypothesized to promote species coexistence and, thus, higher bird species diversity in tropical forests. Dead-leaves suspended in vegetation above the forest floor constitute a focal substrate that may promote specialization (Rosenberg 1990a, b). In a study locality in Amazonia, Remsen and Parker (1984) reported 16 species of birds closely associated with this system. Therefore, it represents a microcosm of high diversity in tropical forests. As such, one natural question is how this substrate can support so many coexisting species, and at what level does specialization occur? In the present study, I address these questions using two approaches. First, I compare composition and abundance of arthropods (grouped by taxa and trophic group) in dead-leaf clumps from two different habitats (river edge and terrace forests) and from clumps composed of small, medium, or large leaves (Chapter I). Second, I compare diets of Automolus rufipileatus and A. ochrolaemus, two species that specialize on this substrate, that occur in river edge and terrace forests, respectively. Diets of the specialists are compared to that of A. rubioinosus, a non-specialist that occurs in terrace forest (Chapter II). Samples of arthropods from dead-leaf clumps in terrace and river edge forests were collected (n= 12) during the dry season at Pakitza (Manu National Park, Dpt. of Madre de Dios, Peru). I sorted arthropods to OTU (Operational Taxonomic Units) (Vandermeer 1972) and counted and measured individuals for each unit. Each taxon was assigned to one trophic group based on known primary food habits (predators, blood suckers, parasitoids, scavengers, herbivores, fungivores, detritivores, omnivores, and non-feeding). Individuals also were divided into size categories (1 =0.1-4.99 mm; 2=5.0-9.99 mm; 3=>_10.0 mm). Composition of arthropod taxa, trophic groups and size classes available in dead-leaf clumps from river edge and terrace forest did not vary greatly, although more arthropods/m3 were available in river edge forest. I gathered stomach contents (n = 7, A. rufipileatus; n = 24, A. ochrolaemus; n= 10, A. rubioinosus) from birds collected during the dry season in the study area from museum collections (MUSM and FMNH). Arthropod fragments were sorted, mounted and identified to OTU. They also were measured and assigned to two size categories (1 = 0.1 -4.9mm; 2= >5mm). Jacob's Electivity Index (1974) was used to evaluate the use of arthropod taxa by the two specialists relative to what was available from the samples from river edge and terrace forest. In general, the two specialists overlapped more in diet than either did with the non-specialist. Diets of the two specialists were similar in arthropod taxa and size of prey items they contained but they differed from the non-specialist in the number of Orthoptera consumed and size of prey. Anecdotal data on aggressive interactions between the two specialists when they come in contact at the edges of their habitats suggest that segregation might have developed to avoid interactions with each other in exploiting the same substrate, and same prey items. This study, although limited in scope as it summarizes data only from the dry season, contributes to our knowledge of the level at which specialization occurs in the Automolus-dead leaf system. Thus it is helpful in understanding general patterns of how species coexist, addressing the main question of why there are more species in the tropics.