Geographic Information Systems (GIS) provide a powerful tool for the
investigation of species-habitat relationships and the development of
wildlife management and conservation programs. However, the relative
ease of data manipulation and analysis using GIS, associated landscape
metrics packages, and sophisticated statistical tests may sometimes
cause investigators to overlook important species-habitat functional
relationships. Additionally, underlying assumptions of the study design
or technology may have unrecognized consequences. This volume examines
how initial researcher choices of image resolution, scale(s) of
analysis, response and explanatory variables, and location and area of
samples can influence analysis results, interpretation, predictive
capability, and study-derived management prescriptions. Overall, most
studies in this realm employ relatively low resolution imagery that
allows neither identification nor accurate classification of habitat
components. Additionally, the landscape metrics typically employed do
not adequately quantify component spatial arrangement associated with
species occupation. To address this latter issue, the authors introduce
two novel landscape metrics that measure the functional size and
location in the landscape of taxon-specific 'solid' and 'edge' habitat
types. Keller and Smith conclude that investigators conducting GIS-based
analyses of species-habitat relationships should more carefully 1) match
the resolution of remotely sensed imagery to the scale of habitat
functional relationships of the focal taxon, 2) identify attributes
(explanatory variables) of habitat architecture, size, configuration,
quality, and context that reflect the way the focal taxon uses the
subset of the landscape it occupies, and 3) match the location and scale
of habitat samples, whether GIS- or ground-based, to corresponding
species' detection locations and scales of habitat use.
