Landscape size affects the relative importance of habitat amount, habitat fragmentation, and matrix quality on forest birds

Authors

  • Adam C. Smith,

  • Lenore Fahrig,

  • Charles M. Francis


A. C. Smith (adam_c_smith@sympatico.ca) and L. Fahrig, Geomatics and Landscape Ecology Research Laboratory, Carleton Univ., 1125 Colonel By Drive, Ottawa, ON, Canada KIS 5B6. – C. M. Francis, Environment Canada, Canadian Wildlife Service, National Wildlife Research Center, Carleton Univ., Ottawa, ON, Canada KIA 0H3.

Abstract

It is important to understand the relative effects of landscape habitat loss, habitat fragmentation, and matrix quality on biodiversity, so that potential management options can be appropriately ranked. However, their effects and relative importance may change with the size of the landscape considered because the multiple (and potentially conflicting) ecological processes that are influenced by landscape structure occur at different spatial scales (e.g. dispersal, predation, foraging). We estimated the relative effects of habitat loss, habitat fragmentation, and matrix quality (measured as the amount of forest, the proportion of forest area contained in large core forests, and the density of roads respectively) on fragmentation-sensitive forest birds in southern Ontario, Canada using a range of landscape sizes (0.8–310 km2). We used three complementary statistical approaches to estimate relative effects of these correlated landscape factors – 1) multiple regression, 2) information theoretic (AIC) estimates of the most parsimonious model, and 3) multi-model inference to average effects across all supported models. We controlled for spatial autocorrelation, local habitat, roadside sampling bias, time of day, season, habitat heterogeneity, and the interaction between the effects of habitat amount and fragmentation. We found that relative effects of habitat amount and fragmentation were scale dependent; habitat amount had a consistently positive effect that was consistent over more than two orders of magnitude in landscape area (~1–300 km2). In contrast, the effects of habitat fragmentation depended on the size of the landscape considered. Indeed, for veery Catharus fuscescens, habitat fragmentation had positive effects at one scale and negative effects at another. The effects of matrix quality were generally weak and changed little with scale. For the number of fragmentation sensitive species and the presence of veery, habitat amount was most important in large landscapes and habitat fragmentation in small landscapes but for the presence of ovenbird Seiurus aurocapilla, habitat amount was most important at all scales.

Ancillary