Standardized methodologies across research programs are critical for developing a general understanding of spatiotemporal ecological processes. Urban ecology is among the fastest growing sub-disciplines in ecology has been successful in describing local and regional patterns of ecological responses to urbanization. However, the development of broad understandings of ecosystem response to urbanization has been impeded by inconsistencies in how the urban landscapes is defined. In a recent review of the literature, a ‘full-disclosure’ approach to quantifying urban landscape gradients is advocated for. This requires explicit reporting of three unifying decisions made in the process: defining the landscape’s thematic content, selecting representative data to represent content, and the choice of spatial scale at which it is analyzed.
In accordance with this framework, we present a general methodological workflow for generating objective and repeatable landscape gradients that requires explicit definition and justification of spatial scale, landscape variable selection, and data sources. This method integrates kernel density smoothing of data at ecologically relevant scales with a multivariate ordination of data to create a dual-axis continuous, numeric gradient representation of human mediated changes in landscape.
Results/Conclusions
Our method describes two dominant axes of variation, a trend which is observed across a range of landscapes: 1) a gradient of anthropogenic intensification describing variation from un-modified to modified landscapes, and 2) a gradient of anthropogenic transition describing a shift from predominantly agricultural to urban landscapes. By applying our method to a variety of urban / non-urban systems across the U.S., we establish the reliability and ecological utility of our dual-axis gradient approach, and demonstrate its potential to aid in standardizing methodology in defining urban landscape gradients.