Assessing the influence of spatial scale on the relationship between avian nesting success and forest fragmentation

Abstract

Ecological processes are dependent on the spatial and temporal scale at which they are viewed, and a process at any one scale may be influenced by factors at other scales. Thus, an ecological process at a broader scale may act to constrain processes at finer scales (Allen and Starr 1982, Thompson et al. 2000). Developing a full understanding of the spatial scales at which habitat conditions impinge on ecological processes therefore demands a multi-scale approach (Wiens 1989). The continuum of possible spatial scales can be broken into: (1) the space occupied by an individual, (2) the patch scale ? the habitat patch occupied by many individuals and species, (3) the landscape scale ? the collection of different habitat patches occupied by local populations, and (4) a biogeographic scale that encompasses different climates, vegetation formations, and assemblages of species (adapted from Wiens et al. 1986). Habitat fragmentation alters the spatial arrangement, shape and relative proportions of different habitat patches. These changes have a profound influence on ecological processes that are sensitive to alteration of the composition of environments, particularly at spatial scales 2-4 above. Two ecological processes, nest predation and brood parasitism by the Brownheaded Cowbird (Molothrus ater), are the primary influences on nesting success of most North American land birds (Martin 1992). Many studies that have investigated the relationship between habitat and nesting success have focused on the question of how vegetation characteristics of the nest micro-environment influence nest success (e.g., Martin 1992, Larison et al. 2001). This focus is at the scale of the space occupied by an individual bird (Wiens et al. 1986). In contrast, studies at broader scales are mostly concerned with how variation in predator/parasite density or movement pattern associated with coarser-scale habitat features has an impact on nesting success. For example, at the scale of a local patch, nests situated closer to habitat edges often experience higher predation or parasitism rates, largely due to elevated predator/parasite densities at habitat edges (Andren and Angelstam 1988, Burke and Nol 2000). The relative density of predators and parasites within edge habitat may further depend on the relative proportions of different habitats at broader, landscape scales. Thus, a handful of studies have examined how nesting success varies with degree of habitat fragmentation within 1-10 km radii of study sites (Robinson et al. 1995, Tewksbury et al. 1999). Understanding the ecological basis for scale dependence in nest predation and parasitism, requires an understanding of the distribution and abundance of predators and parasites in relation to edge, area, and biogeographic effects. These have been examined in greatest detail for the Brown-headed Cowbird. Cowbirds generally forage in open, short-grass habitats, particularly agricultural and other humanmodified habitats (Lowther 1993). However, they often parasitize hosts breeding in forested habitats. Female cowbirds therefore commute daily between their foraging areas and nearby forested habitats to parasitize hosts (Donovan et al. 2000). This gives rise to an "edge effect" where parasitism is greater along the edges of forests that are closer to cowbird feeding habitat. Individual cowbirds are capable of commuting up to 15 km between foraging and breeding resources (Curson et al. 2000). In the eastern United States, however, average commuting distances are more commonly 1-3 km (Thompson 1994, Gates and Evans 1998, Raim 2000, Thompson and Dijak 2000). The distance that cowbirds will penetrate forest interiors is correlated with the local population size of cowbirds in suitable habitat surrounding the forest (Donovan et al. 1997). This local abundance of cowbirds is, to a large extent, limited by the availability of suitable feeding areas, an area effect. Thus, local cowbird abundance increases as the relative area of human-transformed, usually agricultural habitats increases (Donovan et al. 1997). Thus, edge effects at the patch scale are expected to be constrained by variation in local cowbird abundance associated with area effects at local landscape scales within as much as a 10 km radius. At a biogeographic scale, the relative abundance of cowbirds is greatest within the Great Plains (incorporating portions of southern Canada, North and South Dakota, Nebraska, Kansas, and Oklahoma) and becomes progressively less towards the eastern and western edges of their range (Peterjohn et al. 2000). This pattern of relative abundance reflects the historical distribution of cowbirds in North America. They are believed to have been originally concentrated in the Great Plains of central North America but to have spread eastward and westward during the 19th and 20th centuries (Mayfield 1965, Rothstein 1994). Consequently, cowbird abundance decreases with increasing distance from the center of abundance in the Midwest, independently of continental patterns in land cover (Thompson et al. 2000). This is consistent with a general ecological pattern of spatial variation in abundance: density declines from the region of peak density towards the boundaries of the range (Brown 1984). Predator abundance and the risk of nest predation are hypothesized to exhibit similar scale dependence (Donovan et al. 1997, Tewksbury et al. 1999, Thompson et al. 2002), although biogeographic variation has yet to be documented in sufficient detail. Most studies have investigated the relationship between habitat features and nesting success at small and medium scales for individual species at usually single sites, but very few have investigated larger-scale influences. A notable exception is the study of Robinson et al. (1995), which quantified levels of nest predation and cowbird parasitism for nine species across a gradient of forest fragmentation (quantified within a 10 km radius of study sites) spanning nine landscapes within six mid-western states of the United States. Hochachka et al. (1999) also found that increased forest cover within a 10 km radius resulted in lower rates of cowbird parasitism of hosts in general. Using nest success data collected on the Ovenbird (Seiurus aurocapillus) collated in the Breeding Biology Research and Monitoring Database (BBIRD), we extend this approach to ask: what are the relationships between forest fragmentation and each of nest parasitism and nest predation? how do these relationships vary with the scale at which habitat fragmentation is assessed? The Ovenbird is a Neotropical migrant songbird that nests on the ground in the interior of mature forests in the eastern United States and Canada. It is known to be sensitive to forest edges, incurring higher nest mortality in edge habitat, and is a preferred host of the Brown-headed Cowbird (Van Horn and Donovan 1994). It is therefore expected to be highly sensitive to forest fragmentation. © 2006 Springer.