Comparative Structure and Organization of Canopy Bird Assemblages in Honduras and Brazil
Bird communities of tropical forests are famously diverse, with 250 species co-occurring at single 100-ha sites in Amazonia and over 180 species at single 100-ha sites in southern Middle America (Terborgh et al. 1990, Robinson et al. 2000). Although birds that frequent the forest canopy often constitute 40–50% of the species richness in these communities (Terborgh et al. 1990, Cohn-Haft et al. 1997), the difficulty of accessing the canopy has hindered studies of upper forest levels, so few published studies have focused directly on canopy bird assemblages and their ecology (Greenberg 1981, Loiselle 1988, Walther 2002, Naka 2004).
Despite the limited work on canopy birds, we know that they are an important component of forest bird communities and of the forest ecosystem. Canopy birds include important functional groups such as top predators, seed dispersers, and pollinators (Nadkarni and Matelson 1989, Howe 1996, Blake and Loiselle 2000, Holbrook and Smith 2000, Anderson 2001), and it has been argued that the loss of species in these groups following forest disturbance can have severe consequences for the forest ecosystem (da Silva et al. 1996, Loiselle and Blake 2002, Laurance et al. 2006).
As biodiversity is eroded through the continued effect of humanity in lowland neotropical rainforests, it will be essential to understand the processes that maintain and structure biological communities of forest ecosystems so that current levels of biodiversity can be preserved as much as possible. One way to elucidate patterns of diversity and the processes that create and maintain high levels of diversity in the tropics is through a comparison of similar systems at geographically distant locations (Pitman et al. 2001, Stevens and Willig 2002). Despite the use of this approach to examine bird communities in lowland neotropical rainforests (Karr et al. 1990, Robinson et al. 2000), no comparative study that focuses specifically on canopy birds has been published. Ground-based methods alone are insufficient for the study of canopy birds (Anderson 2009).
To date, only three studies have used canopy-based methods to describe entire bird assemblages in neotropical forest canopies: two in southern Middle America, at La Selva, Costa Rica (Loiselle 1988), and Barro Colorado Island, Panama (Greenberg 1981), and one in central Amazonian Brazil (Naka 2004). Although these studies have allowed us a preliminary understanding of canopy bird assemblages, some issues remain unresolved. One key question is whether canopy bird assemblages are dominated by forest birds (Loiselle 1988, Naka 2004) or by species associated with open habitats such as edges or clearings (Greenberg 1981). The harsh environment of the two-dimensional forest canopy is similar to open habitats in that it receives more direct sunlight and is subject to greater diurnal fluctuations of temperature and humidity, greater seasonal variation in water potential, and greater overall wind turbulence than is the forest interior (Endler 1993, Koch et al. 2004, Madigosky 2004).
As a consequence, we may expect canopy bird assemblages to be dominated by species that tend to occur in open habitats (Walther 2002, Burney and Brumfield 2009). Second, no consensus has been reached as to the trophic organization of canopy assemblages in lowland neotropical forests (Greenberg 1981, Loiselle 1988, Naka 2004). Because food resources in the forest canopy are highly variable over space and time (Frankie et al. 1974, Levey et al. 1994, Foster 1996, Leigh 1999), we may expect a high proportion of diet generalists, vagile species, and migrants in the canopy, able to exist on or track a variable and unpredictable diet. Finally, identifying the constituent vertebrate species of any given habitat is a fundamental step in field ecology used to characterize habitats and the ecosystems they constitute. The characterization of the core constituent species of the lowland neotropical forest canopy, and differentiating this group from visitors from other forest strata or neighboring habitats, have remained elusive and unquantified.
The major goal of our study was to present a broad characterization of canopy bird assemblages in lowland neotropical rainforests. We begin with the first description of a canopy bird assemblage from northern Middle America and use these data as a basis for comparisons with a canopy bird assemblage in central Amazonia similarly censused by means of canopybased methods. In particular, we address the following questions:
(1) What are the similarities or differences in the structure and composition of canopy bird assemblages in Honduras and Brazil, in particular as related to species richness, species abundances, composition of dietary guilds, and predominance of edge-living species and long-distance migrants at the respective sites?
(2) Does species richness of habitat and diet generalists and of migrants in canopy bird assemblages differ from random expectations drawn from regional pools of species?
(3) Which genera and species may be considered the core constituents of the assemblages in forest canopies in the humid lowlands of the neotropics?