Mapping the potential distribution of shorebirds in Japan: the importance of landscape-level coastal geomorphology

Abstract

Several recent studies have predicted potential habitats along coastal areas using large-scale physical environmental variables to identify target areas for conservation. However, no indices or methodologies for predicting tidal-flat habitats at a large spatial scale have been developed. Tidal flats supporting large populations of shorebirds have been identified in semi-enclosed seas. Thus, bays are probably important topographic units for evaluating the locations of shorebirds' non-breeding habitats. A GIS-based methodology was developed to extract ‘bay units’ at any scale from coastline data. Using three environment variables (the area of the bay units at three spatial scales, the percentage of shallow water area in each bay unit, and the spring-tide range), it was possible to predict tidal-flat habitats for six shorebird species with high accuracy (AUC > 0.95, sensitivity >90%). Results showed that the percentage of shallow water area in small- and medium-scale bays was the best predictor of tidal-flat habitats, followed by the area of bays at a large spatial scale. This indicates that the size (scale) of a bay and the percentage of shallow water present are highly related to the presence of tidal-flat habitats. The prediction maps for individual species of shorebirds clearly showed differences in the distribution patterns of species. These maps were overlaid to identify potentially species-rich areas and thus where conservation and restoration of the tidal flats in these bays would be important. The model, which uses simple coastal data, is a useful, resource-efficient method for identifying target conservation and restoration areas across broad scales. Copyright © 2011 John Wiley & Sons, Ltd.