Ecological understanding is limited by two interdisciplinary problems, scaling and linkages between ecosystems and human societies. Societies interact with ecosystems at broadly varying extents, thus scaling approaches are required to account for the multiple relationships. Similarly, all regions on the globe have been impacted by human activities; a robust scaling theory needs to incorporate the interactions between human activities and ecosystem processes. The research presented in this dissertation develops a concept of landscape rescaling, a landscape ecological framework that facilitates the investigation of the linked problems of scaling and social-ecological interactions within an individual city and linking a city with its surrounding hinterland. This approach highlights the identification of spatial organization as a means to integrate the diverse drivers of ecosystem processes occurring at multiple scales.

Landscape rescaling was implemented and refined in a series of investigations of the Phoenix, AZ, USA metropolitan region: (1) two spatially explicit simulation modeling studies, (2) analyses of linked vegetation-surface temperature-social patterns, and (3) analyses of soil heterogeneity at patch and regional scales. First, I examined influences of scale on models of urban land-use change and landscape interactions between two vegetation functional groups and their modification of biogeochemical processes within native desert habitat. In both of these models, the scale of analysis had a strong effect on model dynamics and the conclusions drawn from the models. Second, I examined the joint spatial heterogeneity in socio-economic status, vegetation, and surface temperature. Socio-economic information was obtained from the 2000 United States Census. Vegetation patterns were quantified using the Soil-adjusted vegetation index (SAVI) derived from remotely-sensed Landsat satellite data. Surface temperature patterns were quantified using a black-body inversion of the thermal band from the same Landsat data. These data were obtained at differing resolutions, thus requiring data rescaling to analyze the linked spatial patterns in these three sources of information. Vegetation was identified as a central link between surface temperature and social patterns within neighborhoods of the Phoenix, AZ metropolitan region with Path analysis, a multivariate statistical model. Due to differences in vegetation, surface temperature decreased by .28˚C for every $10,000 increase in median household income. These changes cascade to regional influences through modification of urban footprints. Third, I examined the distribution of soil organic matter, nitrogen, and nitrogen stable isotope ratios within individual patches, throughout the entire Phoenix, AZ metropolitan region, and examined differences in heterogeneity using a hierarchical scaling ladder approach. From the patch and regional data I estimated the effects of urbanization on soil storage of organic matter and total nitrogen and nitrogen stable isotope ratios using three scaling techniques. Overall, urbanization in this region has increased soil organic matter by 44%, total nitrogen by 48%, and has elevated δ15N by 21%. The relationships between patch and regional heterogeneity were further examined. By explicitly quantifying the uncertainty associated with scaling from points to patches and patches to land-use types, the influence of urbanization on soil scaling relations was identified. Intensive human management increased homogeneity of soil properties within individual patches and decrease homogeneity between patches compared to native desert.

These three studies of landscape rescaling theory provide the context for a new multiple-scale landscape ecological perspective and examples of its implementation. Together they document the strong scale dependencies of ecosystem patterns and processes within the Phoenix, AZ metropolitan region, the tight coupling between social and ecological processes within this urbanized arid region, and the influence of social-ecological coupling on scaling relationships within this region.