How has the state-of-the-art for quantification of landscape pattern advanced in the twenty-first century?

Abstract

Landscape ecology was founded on the idea that there is a reciprocal relationship between spatial pattern and ecological processes. I provide a retrospective look at how the state-of-the-art of landscape pattern analysis has changed since 1998. My objective is to show how pattern analysis techniques have evolved and identify some of the key lessons learned. The state-of-the-art in 1998 was derived from information theory, fractal geometry, percolation theory, hierarchy theory and graph theory, relying heavily on the island-patch conceptual model using categorical maps, although point-data analysis methods were actively being explored. We have gradually winnowed down the list of fundamental components of spatial pattern, and have clarified the appropriate and inappropriate use of landscape metrics for research and application. We have learned to let the objectives choose the metric, guided by the scale and nature of the ecological process of interest. The use of alternatives to the binary patch model (such as gradient analysis) shows great promise to advance landscape ecological knowledge. The patch paradigm is often of limited usefulness, and other ways to represent the pattern of landscape properties may reveal deeper insights. The field continues to advance as illustrated by papers in this special issue.