Environmental Engineering Reference
In-Depth Information
It was obvious that fractal geometry was both interesting and useful for land-
scape studies and that we had only scratched the surface of what might be done.
Simulated by this and by Mandelbrot's work (e.g. Mandelbrot 1967, 1983) I
attended a Gordon Conference on fractals in the summer of 1986. A broad spectrum
of issues and applications were discussed, but the most interesting was the presen-
tation by Dietrich Stauffer on percolation theory. The images he presented (which
were random 2D structures) looked remarkably like landscapes. Discussions with
Stauffer at the conference and subsequent study of his first topic on percolation
theory (Stauffer 1985) made it clear that this theoretical framework would provide a
rich set of tools for landscape ecology.
Yet how does one use randommaps and percolation theory in landscape ecology?
Caswell's paper on neutral models and community theory (Caswell 1976) was well
known and widely discussed, but had not been of particular interest to problems in
landscape ecology. However, it was clear that combining Caswell's approach with
Stauffer's percolation theory would provide a neutral model for landscape ecology.
In the fall of 1986 we began writing the programme RULE (Gardner 1999) and
running a series of simulations to describe this approach. In the spring of 1987, with
the completed manuscript in hand, we presented the results during the 2nd Annual
Meeting of the U.S. chapter of the International Association of Landscape Ecology
at the University of Virginia. Frank Golley, who was instrumental in the develop-
ment of landscape ecology in the U.S. had just established the new journal appro-
priately titled “Landscape Ecology”. Frank requested at the Charlotte meeting that
we submit this manuscript for publication in this new journal and by the end of that
summer our paper (Gardner et al. 1987) had appeared in print.
The diversity of issues for which neutral models may be useful has been
significant, yet the full range of applications has yet to be fully explored (Gardner
and Urban 2007). Although new methods have been developed (e.g. Hagen-Zanker
and Lajoie 2008; Li et al. 2009; Wang and Malanson 2007), existing methods for
analysis and comparison of maps available within the original software (now called
“Qrule”,
http://www.al.umces.edu/Qrule.htm
) have rarely been fully employed.
This chapter will explore these two areas by first developing a new neutral for the
analysis of landscape change within the Piedmont of Maryland, a large area that is
being rapidly altered by urban development. Analysis methods will then be used to
test the hypothesis that the contagion processes associated with human develop-
ment have been responsible for non-random changes in the patterns of forest
patches. Specifically, there may be an insignificant reduction in the total amount
of forest cover, but significant changes in the adjacency of land cover types and the
probability that distribution of the size of forest patches has occurred.
15.2 A Programming Philosophy
Documentation and practical examples illustrating the use of Qrule are available in
a number of publications (Gardner 1999; Gardner and Walters 2001; Pearson and
Gardner 1997). Recently added revisions have also been described (Gardner and
