| Abstract: | ABSTRACT Wildlife biologists are using land-characteristics data sets for a variety of applications. Many kinds of landscape variables have been characterized and the resultant data sets or maps are readily accessible. Often, too little consideration is given to the accuracy or traits of these data sets, most likely because biologists do not know how such data are compiled and rendered, or the potential pitfalls that can be encountered when applying these data. To increase understanding of the nature of land-characteristics data sets, I introduce aspects of source information and data-handling methodology that include the following: ambiguity of land characteristics; temporal considerations and the dynamic nature of the landscape; type of source data versus landscape features of interest; data resolution, scale, and geographic extent; data entry and positional problems; rare landscape features; and interpreter variation. I also include guidance for determining the quality of land-characteristics data sets through metadata or published documentation, visual clues, and independent information. The quality or suitability of the data sets for wildlife applications may be improved with thematic or spatial generalization, avoidance of transitional areas on maps, and merging of multiple data sources. Knowledge of the underlying challenges in compiling such data sets will help wildlife biologists to better assess the strengths and limitations and determine how best to use these data. |
