Abstract
Determining the characteristic spatial scales at which species respond most strongly to the amount of available habitat is crucial for developing cross-species scaling rules and for predicting species' responses to landscape modification. A Biologist's Toolbox article by Holland and colleagues ("Determining the Spatial Scale of Species' Response to Habitat," BioScience 54: 227-233) presents a multiscale approach and computer program (Focus) for detecting characteristic scales using a resampling procedure, species-habitat regression models, and nonoverlapping sampling sites. Holland and colleagues refer to these nonoverlapping areas as "spatially independent sites," when in fact spatial independence includes additional concerns not addressed by the Focus approach. Here I discuss issues of spatial heterogeneity-spatial autocorrelation and spatial dependence-as they relate to measuring the spatial scaling of organism-environment relationships with regression models. I present an empirical example with cactus bugs (Chelinidea vittiger), demonstrating how spatial heterogeneity complicates the task of determining characteristic scales of species-habitat relationships. Finally, I provide some cautions and suggestions for researchers who are considering using Focus to examine scaling patterns with existing data sets.
Original language | English (US) |
---|---|
Pages (from-to) | 533-537 |
Number of pages | 5 |
Journal | BioScience |
Volume | 56 |
Issue number | 6 |
DOIs | |
State | Published - Jun 2006 |
Keywords
- Autocorrelation
- Landscape heterogeneity
- Patch context
- Spatial regression
- Spatial scale
ASJC Scopus subject areas
- General Agricultural and Biological Sciences