Use of Amazonian forest fragments by understory insectivorous birds

Ecology, Dec, 1995 by Philip C. Stouffer, Richard O. Bierregaard, Jr.

INTRODUCTION

Fragmentation of forest habitat affects the diversity and abundance of plants and animals in various forest types (e.g., Galli et al. 1976, Gottfried 1979, Shreeve and Mason 1980, Weaver and Kellman 1981, Main 1987, Klein 1989). For birds, studies in North America have generally indicated that forest birds, especially forest interior species, are vulnerable to fragmentation of breeding habitat (reviewed by Askins et al. 1990). The effects of fragmentation on tropical birds have been less thoroughly studied (but see Leck 1979, Willis 1979a, Bierregaard and Lovejoy 1989, Kattan et al. 1994). Available data suggest that some tropical species may be even more vulnerable to fragmentation than temperate species. Neotropical forest birds, especially in the Amazon basin, are generally more specialized in their foraging techniques, use narrower habitats and microhabitats, are more sedentary, and have larger territories than temperate forest birds (Orians 1969, Willis 1974, Remsen and Parker 1984, Karr et al. 1990, Terborgh et al. 1990).

To understand the effects of fragmentation, it has become increasingly important to examine fragments as part of a larger matrix that includes fragments of various sizes, continuous forest, and intervening disturbed areas (e.g., Askins et al. 1987, Keller and Anderson 1992, Fahrig and Merriam 1994). A related aspect of fragmentation that has received little attention is the effect of changing land use in the matrix surrounding fragments (e.g., Askins and Philbrick 1987). These landscape and temporal effects influence fragment use by birds. Moreover, interspecific variation in response to landscape effects are likely to be significant, especially among the sedentary, specialized birds of the Neotropics.

Part of the reason temporal effects are poorly known is related to the design of most studies of forest fragmentation. In most cases, response to fragmentation has been inferred from comparisons among fragments of various sizes sampled at the same time. These comparisons of fragments are useful for examining habitat associations (e.g., Ambuel and Temple 1983, Freemark and Merriam 1986, Blake and Karr 1987) or how fragment size influences species richness (e.g., Bolger et al. 1991, McCoy and Mushinsky 1994). The only direct measure of the effects of fragmentation, however, is to sample the same plot before and after it is isolated (e.g., Lovejoy et al. 1986). Repeated sampling also allows examination of successional processes, structural deterioration, and other gradual changes in fragments (Lovejoy et al. 1986, Robinson et al. 1992, Stouffer and Bierregaard 1995).

Here we examine the effects of fragmentation on the diversity and abundance of understory insectivorous birds at a site in Amazonian Brazil. These fragments were sampled from before isolation through 9 yr after isolation. We consider fragments of two sizes, 1 and 10 ha, that are surrounded by two distinct types of secondary growth. Patterns of human disturbance in the Brazilian Amazon often result in small forest fragments and abandoned fields like the ones we studied (e.g., Uhl 1987, Dale et al. 1994). Completely isolated fragments of this size can support no more than one or two pairs of even the commonest understory birds (Terborgh et al.. 1990), so fairly rapid extinction is inevitable (e.g., Diamond 1984). Because of rapid extinction, small fragments provide the opportunity to study recolonization on a relatively short temporal scale (see discussion in Rosenzweig and Clark 1994). Our emphasis on recolonization stresses the interaction of the fragments with the temporally variable matrix in which they are imbedded.

We describe the sequence of species loss and subsequent recolonization, how social organization and foraging behavior affect vulnerability to isolation, and the extent to which loss of forest insectivores is compensated for by increased abundance of nonforest species or previously uncommon species. Based on these results, we predict the species that are most likely to be threatened in a fragmented Amazonian landscape. We also suggest research that could further assess habitat requirements for persistence and recolonization of vulnerable birds.

STUDY SITE

The biological dynamics of forest fragments project study site

This study was conducted in terra firme forest 80 km north of Manaus, Brazil (2 [degrees] 30' S, 60 [degrees] W). Here we include data for five fragments of I ha and four fragments of 10 ha (see Lovejoy et al. 1986, Lovejoy and Bierregaard 1990, Bierregaard et al. 1992 for more detailed descriptions of the site). These fragments are located on three adjacent cattle ranches, with [approximately equal to]40 km separating the most distant fragments (Table 1). Although some parts of the area have been cleared for cattle ranching, these local disturbances are within an area that is mostly unbroken forest for hundreds of kilometres (see Vitousek 1994 for a LANDSAT image of the site).

TABLE 1. Characteristics of 1- and 10-ha forest fragments at the
BDFFP site. Number refers to the reserve code used in other
publications (see Lovejoy et al. 1986 for detailed descriptions of
the fragments). Code refers to the codes used here, and indicates
fragment size, type of secondary growth, and the number of the
replicate.