Urban sprawl typically consists of low-density urban development dominated by single-family housing and automobile-oriented land use patterns. Sprawl impacts landscape structure and composition, especially along the urban periphery. However, few studies have simultaneously examined sprawl at the building level and by building type (e.g. single family, multi-family) and its relationship to forest landscapes within an urbanizing region.
(1) To map and quantify 30-years of sprawl and assess its impacts on forest landscapes in southeast Michigan, a seven-county region centered on the City of Detroit (2) to investigate how different building types, densities, and distances affect forest structure.
We used the Random Forests algorithm to analyze high resolution remote-sensing imagery and computed three landscape metrics of forest fragmentation and cohesion, incorporating data on built types and densities. Finally, we investigated the relationship between single-family housing sprawl and forest landscape functionality.
The built-up expansion was correlated with an increase in overall tree canopy in the region. However, multilevel analysis revealed these same forest landscapes became less cohesive and more fragmented over time as a result of urban sprawl. Additional correlation tests revealed an increase in patch density and decrease in effective mesh size (meff) and patch cohesion in areas proximate to low-density single-family housing.
The analysis documents how urban sprawl negatively impacts forested landscapes. Single-family housing in particular had a detrimental impact on the functionality of adjacent forested landscapes. High thematic resolution enables policy-makers and planners to identify specific policies and interventions to increase landscape functionality.