Abstract

Neutral landscape models are computer-generated representations of the spatial pattern of habitat. Most are based on a grid of cells, although network (graph theory) representations are increasing. Representation of the landscape varies from binary (habitat – non-habitat) to nearly continuous. The motivations for this approach to the study of landscapes include concepts derived from island biogeography as well as the idea that computer experiments might be substituted for real experiments at such spatial extents. Different spatial patterns can be generated using the three primary algorithms for grid-based neutral landscapes (random, hierarchical, and fractal), while network representations are usually variations on networks derived from observed landscapes. Scale, encompassing both resolution and extent in cellular models as it is in the network models, is an underlying issue that needs to be addressed more explicitly, but it is not a fatal flaw. Numerous applications of neutral landscapes have been developed in assessing actual landscapes, evaluating spatial metrics, and studying landscape connectivity – especially in relation to organism movement or dispersal. Neutral landscapes offer geographers opportunities to pursue spatial questions and to influence questions in landscape ecology.

DOI: 10.1111/j.1749-8198.2008.00090.x