Landscape Fragmentation Assessment Using a Single Measure
Wildlife Society Bulletin, 28 (4):
875--881 (2000)Abstract
Measurement of fragmentation is crucial for determining its consequences and to develop policy for nature conservation. We propose a fragmentation measure |φ| which combines, using a multidimensional Euclidean distance, 4 main characteristics of fragmented landscapes: total habitat area, total habitat perimeter, number of patches, and patch isolation. Its properties can be summarized as: 1) |φ| reflects the overall fragmentation status; 2) every component of |φ| is accepted as a measure of fragmentation; 3) every component of |φ| is a normalized variable; 4) every component of |φ| is easy to compute; 5) average patch size, interior habitat, and habitat connectedness are included indirectly in |φ|; 6) |φ| is independent of the land-use type; and 7) |φ| can be calculated for raster and vector data. We show that the normalized values composing |φ| prevent misinterpretation of features as fragment number or boundary length. A sensitivity analysis, based upon artificial patterns, showed that increasing fragmentation is correlated with smaller values of |φ|. Wildlife managers are encouraged to use |φ| for objective evaluation of fragmented landscapes.