1 Introduction

The quantification of spatial heterogeneity is necessary to elucidate relationships between ecological processes and spatial patterns (Turner, 1990; Turner et al., 2003*). Therefore the measurement, analysis and interpretation of spatial patterns receive much attention in landscape ecology (Haines-Young and Chopping, 1996*). A great variety of metrics for landscape composition (e.g., the number and amount of different habitat types) and configuration (the spatial arrangement of those classes) were developed for categorical data. Software packages are widely used (e.g., FRAGSTATS, see McGarigal and Marks, 1995; McGarigal et al., 2002*), and many metrics have also been integrated into existing geographic information system (GIS) software (e.g., Patch Analyst in ArcView; and module Pattern in IDRISI).

Although spatial pattern analysis should be considered to be a tool rather than a goal in and of itself, and the objectives or questions driving any analysis must include the qualities of the pattern to be represented and why, there are many examples in the literature when this has not been correctly followed (Li and Wu, 2004*). Several issues associated with the interpretation of landscape metrics are widely used by practitioners (Gustafson, 1998; Haines-Young and Chopping, 1996; Li and Wu, 2004; Turner et al., 2003*). Many metrics are sensitive to changes in the spatial resolution (grain size) of the data or the area (extent) of the landscape (Wickham and Riitters, 1995*), and numerous correlations occur among landscape indices (Riitters et al., 1995*; Cain et al., 1997). The downscaling and upscaling of landscape metrics as functional and structural landscape indicators at different scales still remains a challenge (Mander et al., 2005).

On the other hand, the common usage of the term “landscape metrics” mostly refers to indices developed for categorical map patterns (McGarigal et al., 2002*), but it is sometimes also used for topographic measures (Iampietro et al., 2005; Vivoni et al., 2005) that characterize landscape, or it may also just refer to some combination of several characteristics that are important to a particular species (Schils, 2006; Fernández et al., 2007).

The main aim of this paper is to give an overview of the development and state of the art of the applications for landscape metrics, one of the classical landscape ecological tools, and study objects that help us better understand the relationships between landscape pattern and processes.