Resource information
Heterogeneous land cover patterns contribute to unique ecological conditions in cities and little is known about the drivers of these patterns among cities. We studied tree cover patterns in relationship to urban morphology (for example, housing density, parcel size), socioeconomic factors (for example, education, income, lifestyle characteristics), and historical legacies in Baltimore, Maryland, and Raleigh, North Carolina. Utilizing a multimodel inference approach and bivariate analyses, we analyzed two primary datasets employed in previous research predicting urban tree cover—one comprising continuous data (US Census), and the other consisting of categorical variables (Claritas PRIZM) that incorporate consumer purchasing data. Continuous data revealed that urban morphological characteristics were better predictors of tree cover patterns than socioeconomic factors in Raleigh and Baltimore at the parcel and neighborhood scales. Although the categorical dataset provided some evidence for the importance of socioeconomic and lifestyle characteristics in predicting tree cover patterns, the hierarchical nature of these data preclude separating the impacts of these factors from levels of urbanization. Bivariate analyses of continuous and categorical variables revealed that the highest correlation coefficients were associated with variables describing urban morphology—parcel size, percent pervious area, and house age. In Baltimore, historical census data were better predictors of present-day tree cover than census data from recent years. Most notably, parcel size, a key predictor of tree cover, has decreased with time in Raleigh to sizes consistently seen in Baltimore. Our findings demonstrate that urban morphology, the main driver of tree cover patterns in these cities, may lead to the homogenization of tree canopy in Raleigh and Baltimore in the future.
