Determining the Spatial Scale of Species' Response to Habitat

E attributes such as population abundance and species richness depend not only on patch characteristics but also on the characteristics of the landscape surrounding the patch, known as patch context (Åberg et al. 1995, Gascon et al. 1999, Saab 1999, Szacki 1999, Fahrig 2001). Focal patch studies are one approach to studying the effects of patch context. In such studies, the data on species abundance or richness are collected in a number of patches or sites. The landscape predictor variables (e.g., habitat amount or fragmentation) are measured in areas that are centered on the patch or site locations (figure 1). Each patch (and its associated landscape) becomes a single data point in the data analysis (Brennan et al. 2002). In this way, researchers can examine the influence of habitat variables, measured at a large scale, on species abundance or richness. In this article, scale refers to the area or radius within which habitat predictor variables are measured. Therefore, we are referring only to the extent, and not the grain, of the predictor variable. Focal patch studies have been used to study the effects of road density and the amount of surrounding forest on wetland species richness (Findlay and Houlahan 1997), landscape habitat diversity on alfalfa insect richness (Jonsen and Fahrig 1997), amount of forest on raccoon density (Pedlar et al. 1997), amount of wooded border on alfalfa insect richness and density (Holland and Fahrig 2000), and amount of summer habitat and breeding pond density on leopard frog abundance (Pope et al. 2000). In addition to these focal patch studies, there are probably hundreds of existing data sets in which researchers have studied the effects of local or patch habitat variables on population abundance or species richness in a number of patches. Can these data sets be reanalyzed to study the effects of landscape context on population abundance or species richness? Given the relative ease of obtaining remotely sensed habitat data, these data sets represent a mine of potential information on the effects of landscape context. The main problem with using these data sets, however, is that in such studies the patches or sites are often rather closely spaced. This can result in data points that are not spatially independent because the landscape areas overlap (figure 1), possibly leading to pseudoreplication. These overlapping sites may constrain the number of data points that can be used for examining the relationship between species abundance or species richness and the measures of landscape context. Nevertheless, because the collection of field data is time-consuming and expensive, it would be beneficial if there were a way to use this data to examine questions related to the larger-scale landscape context. In this article, we present a randomization method and computer program (Focus; www.carleton.ca/lands-ecol/) that permit analysis of effects of landscape context in this situation. Version 2 of this software is available as a free download to researchers and will remain so for the indefinite future.