Introduction

Urbanisation provides an environmental gradient from a highly developed core to a rural or natural area (Numata,1976; McDonnell and Pickett, 1990). Urban–rural gradients provide a useful laboratory in which to examine environmental effects on communities. Analysing changes in a biological community along such a gradient provides a scientific basis for planning ecological cities, but also makes it easier to test hypotheses through management of urban landscapes.

Urban ecosystem studies in Japan started in 1971 as a part of a project on ‘human survival and environment’ (Numata, 1976). Numata defined an urban ecosystem as one consisting of man-modified and man-made ecosystems, and he proposed three approaches to urban ecosystems: (1) examining the flow of energy, matter, people and information; (2) studying the impact of an urban environment on terrestrial and aquatic ecosystems; and (3) studying reactions to the abiotic environment (in the Clementian sense). He and his colleagues reported a series of urban ecological studies of metropolitan Tokyo (see Numata, 1982).

Although urban ecological study in Japan covers many fields, such as pollution of air and water, urban climates and the water cycle, one important subject is the impact of urbanisation on flora, fauna and the biological community. Early studies focused on biological indicators of urbanisation; lichen (Taoda, 1973), alien plants (Hotta, 1977), soil arthropods (Aoki, 1979), and the retreat of wild animals (Chiba, 1973) were used as indicators. Through comprehensive study of urban ecology, Numata (1987) identified the importance of landscape ecology in urban ecosystem studies.