Application of geospatial models to map potential Ruditapes philippinarum habitat using remote sensing and GIS

Abstract

This study applied geographic information system (GIS)-based models to map the potential Ruditapes philippinarum (Korean littleneck clam) habitat area in the Geunso Bay tidal flat, Korea. Remote-sensing techniques were used to construct spatial datasets of ecological environments, and field observations were undertaken to determine the distribution of macrobenthos. The mapping of potential habitat was completed and eight controlling factors relating to the distribution of tidal macrobenthos were determined. These were the tidal flat digital elevation model, slope, aspect, tidal annual exposure duration, distance from tidal channels, tidal channel density, spectral reflectance of the near-infrared bands, and surface sedimentary types, which were all generated from satellite imagery. The spatial relationships between the distribution of R. philippinarum and each control factor were calculated using a frequency ratio, logistic regression, and artificial neural networks combined with GIS data. Individuals were randomly divided into a training set (50%) to analyse habitat potential using the frequency ratio, logistic regression, and artificial neural network models, and a test set (50%) to validate the predicted habitat potential map. The relationships were overlaid to produce a potential habitat map with a R. philippinarum habitat potential (RPHP) index value. These maps were validated by comparing them to surveyed habitat locations such as those in the validation data set. From the validation results, the frequency ratio model showed prediction accuracy of 82.88%, while the accuracy of the logistic regression and artificial neural networks models was 70.77% and 80.45%, respectively. Thus, the frequency ratio model provided a more accurate prediction than the other models. Our data demonstrate that frequency ratio, logistic regression, and artificial neural networks models based on GIS analysis are effective for generating potential habitat maps of R. philippinarum species in a tidal flat. The results of this study will be useful for conserving and managing the macrofauna of tidal flats.