Introduction

Landscape pattern analysis (LPA) has been a major part of landscape ecological research for the last two decades (Romme 1982, O'Neill et al. 1988, Turner 1989, 1990, Turner and Gardner 1991, Pickett and Cadenasso 1995, Gustafson 1998, Wu and Hobbs 2002). The ultimate goal of LPA is to link spatial patterns to ecological processes at different scales. The importance of LPA lies in the needs to: (1) monitor, quantify, and project the change of a given landscape; (2) compare and contrast patterns between different landscapes; and (3) help understand processes underlying observed patterns, so that landscape dynamics may be better understood and predicted (Turner et al. 2001, Wu 2004). Thus, appropriate and effective use of LPA methods is vital to the development of landscape ecology.

After two decades of rapid development, landscape ecology has begun to mature. However, many problems still persist in the application of LPA (Li and Reynolds 1995, Tischendorf 2001, Fortin et al. 2003, Li and Wu 2004). Li and Wu (2004) have called for a serious rethinking of why and how landscape pattern analysis should be used, with an intent to discourage the rampant and blind use of LPA methods. They argued that theoretical guidance should be sought in the practice of LPA. Fortin et al. (2003) stated that methodological developments often undergo four phases: (1) the introduction phase with key papers describing a new methodology, (2) the testing phase with many papers applying the new methodology, (3) the critical review phase with limitations of the methodology identified and with rethinking of its fundamental purposes, assumptions, and formulations, and (4) the standardization phase with the most effective methods being selected as the norm.