Analysing forest fragmentation using spatial autocorrelation, graphs and GIS

This paper presents a replicable methodology to assess the extent of fragmentation in forests induced by natural and man-made features and disturbances. The analysis is presented in the context of conservation ecology research and its relevance to the development of environmental protection policy is discussed. Methodologically, the approach utilizes Geographical Information Systems (GIS), Relational Database Management Systems (DBMS), symbolic mathematics software, spatial autocorrelation statistics and graph data structures. The intersection of these technologies and techniques offers a potentially fertile area to extend the meaningful definition of natural forested areas and their management as well as our understanding of how these areas can be preserved through conservation policy. The paper concludes with suggestions for further extensions of the research.