The Analysis of Fauna Composition for Predicted Habitat of Amphibians.

The significance of biodiversity has been widely accepted in recent years. In this context ecological information are indispensable for regional landscape planning in both urban and rural areas. In the present study, we propose a procedure to estimate species distribution in a rural area using amphibians as materials. Estimation model of potential habitats was constructed based on the relationship between amphibian species composition and land attributes. The investigation of amphibians in 24 sites was performed in and around Mito city, central part of Ibaraki prefecture as a case study. The percentage of inhabitation of each species was calculated by the current situation of the amphibians in the investigation sites, and the 8 types of fauna composition were identified using TWINSPAN. Then, landform types and vegetation types were classified by GIS. And the discriminant function which classifies each fauna composition type was computed. In the first stage, it was divided into hilly areas and plateau/lowland areas by landform classification. Hilly areas are characterized by habitation of three amphibian species; Hynobius tokyoensis, Rana ornativentris and Rana rugosa. In the following stage, the more detailed division of each fauna composition type, which was explained mostly by the present situation of Hyla japonica, Rhacophorus schlegelii and Bufo japonicus formosus was possible by using vegetation classification. The discrimination ratios of predicted inhabitation were between 65 - 86 %. We calculated the potential habitation possibility of amphibian species in the non-investigated area using the obtained distinction function formula. The map of the predicted habitat of amphibians was created about 840 km2 of the case study area with a grid cell of about 1km around. In addition, verifying investigation was conducted at 13 sites and the validity of the predicted potential habitat map was accepted. By establishing a series of research flows for the prediction technique, the application of a broader-scale landscape planning was attained.