Translator Disclaimer
1 February 2010 Modeling the human invader in the United States
Author Affiliations +
Modern biogeographers recognize that humans are seen as constituents of ecosystems, drivers of significant change, and perhaps, the most invasive species on earth. We found it instructive to model humans as invasive organisms with the same environmental factors. We present a preliminary model of the spread of modern humans in the conterminous United States between 1992 and 2001 based on a subset of National Land Cover Data (NLCD), a time series LANDSAT product. We relied on the commonly used Maxent model, a species-environmental matching model, to map urbanization. Results: Urban areas represented 5.1% of the lower 48 states in 2001, an increase of 7.5% (18,112 km2) in the nine year period. At this rate, an area the size of Massachusetts is converted to urban land use every ten years. We used accepted models commonly used for mapping plant and animal distributions and found that climatic and environmental factors can strongly predict our spread (i.e., the conversion of forests, shrub/grass, and wetland areas into urban areas), with a 92.5% success rate (Area Under the Curve). Adding a roads layer in the model improved predictions to a 95.5% success rate. 8.8% of the 1-km2 cells in the conterminous U.S. now have a major road in them. In 2001, 0.8% of 1-km2 cells in the U.S. had an urbanness value of > 800, (>89% of a 1-km2 cell is urban), while we predict that 24.5% of 1-km2 cells in the conterminous U.S. will be > 800 eventually. Main conclusion: Humans have a highly predictable pattern of urbanization based on climatic and topographic variables. Conservation strategies may benefit from that predictability.
Thomas J. Stohlgren, Catherine Jarnevich, and Chandra P. Giri "Modeling the human invader in the United States," Journal of Applied Remote Sensing 4(1), 043509 (1 February 2010).
Published: 1 February 2010

Back to Top