Introduction

By 2007 more than half of the world's population is expected to reside in cities (United Nations,2004). As urban populations and the number of cities expand, natural and agricultural lands are transformed into highly altered landscapes. These changes in demography and land use have contributed to the alteration of biogeochemical cycles at local, regional and global scales (Vitousek et al., 1997a; Pouyat et al., 2003). Yet we lack sufficient data with which to assess the underlying mechanisms of land-use change (Groffman et al., 2004), largely because of the difficulty encountered when applying established biogeochemical research methods such as large-scale field manipulations to urban and suburban ecosystems (Pouyat et al., 1995a). Moreover, current conceptual and quantitative biogeochemical models incorporate human effects only indirectly (Groffman and Likens, 1994).

As a result, most urban ecosystem studies have relied on a comparative approach or ‘natural experiments’ to investigate urban effects on biogeochemical cycles in ecological remnants characteristic of a particular area or region (Pickett et al., 2001). This approach takes advantage of remnant systems as ‘whole ecosystem’ manipulations by which the effects of multiple urban stress and disturbance factors are assessed with established statistical methods and modelling approaches (Pouyat et al., 1995a; Breitburg et al., 1998; Carreiro and Tripler, 2005; Carreiro et al., Chapter 19).