Carbon, nitrogen, and vegetation along an urbanization gradient: a Boston case study integrating field, remotely sensed and socioeconomic data
Understanding the role humans play in modifying ecosystems through urban development is central to addressing our current and emerging environmental challenges. Urbanization can drastically modify carbon (C) and nitrogen (N) cycling in terrestrial ecosystems. However, spatiotemporal distribution of these modifications and their impact on ecosystems are not well-quantified. In this dissertation, I combined field and remotely sensed data, models and laboratory analysis, and socioeconomic data to understand the variations in ecosystem characteristics and their socioeconomic covariates along a 100-km urbanization gradient in the Boston region. Vegetation and soil C and N chemistry from 139 field plots show that C and N content increased in soil and decreased in vegetation with urbanization for forest, residential and other-developed land use classes. Landsat normalized difference vegetation index correlated positively with aboveground biomass and foliar N content (but not N concentration), and negatively with impervious surface area (ISA) fraction. Patterns in foliar N concentration are associated more strongly with changes in species composition than with phenotypic plasticity. My results demonstrate the need to account for ISA fraction when scaling vegetation and soil data across urban landscapes. Measured atmospheric inorganic N inputs at nine sites along the gradient correlated significantly with proximity to urban core and modeled on-road CO2 emissions. N leaching rates correlated positively with atmospheric N input rates. A regional model underestimated atmospheric N inputs at urban sites and overestimated it at rural sites, thus highlighting the need to incorporate the effects of urbanization in N deposition models. Analysis of census variables, forestland owner surveys, and biomass highlighted the scale-dependent relationships between socioeconomic variables and vegetation biomass. Owner occupancy showed the strongest and most consistent relationship with biomass across different scales. Combined with either housing age or educational attainment, owner occupancy explained ~80% variance in biomass in different spatial extents of the gradient. Conservation awareness among landowners was higher near the urban core and correlated positively with educational attainment and landholding size. My results demonstrate the complex spatial variations in urban biogeochemistry and help develop a mechanistic understanding of urban ecosystem function and its socioeconomic covariates.