OOS 19-1
Green roof and biodiversity potential along the urban-rural gradient

Tuesday, August 11, 2015: 8:00 AM
328, Baltimore Convention Center
Michael McKinney, University of Tennessee, Knoxville
Nick Sisco, Environmental Studies, University of Tennessee, Knoxville, TN

Green roofs have great potential for increasing overall urban biodiversity. However, their biodiversity value will be strongly influenced by characteristics of these “habitat fragments”. Specifically, green roofs will act as “urban and suburban filters” where the area and connectedness of each roof (habitat fragment) will determine which groups of organisms can colonize and persist. Widely dispersing organisms, e.g., wind-dispersed spontaneous vegetation, birds, and insects, will likely be among the most important ecological components in green roof ecosystems. Quantitative data on the specific area and configuration of roofs in urban and suburban areas is typically lacking. This is essential information for measuring exactly how much biodiversity potential is available in and around cities. We analyzed GIS data obtained from the KGIS database produced by the City and County governments of the Knoxville, Tennessee area using ArcGIS software. These data record the areal “footprint” of all buildings in Knox County, ranging from downtown Knoxville through all outlying suburban and rural areas. This approximates the total rooftop cover and the specific configurations of all rooftops in the County.


Approximately 3.5% of the total surface area of Knox County (47.72 km2/1362 km2) is covered by rooftops. In terms of the urban-rural gradient, samples of 1 km x 1 km cells in the downtown urban core have the highest roof cover, with a maximum of 22.7% and an average for the urban core of roughly 16% roof cover. Cells in the suburban areas typically average around 11% roof cover but range from 4-17%, depending on the specific housing subdivisions analyzed. Cells sampled in rural parts of the County, far from the urban core, average just around 2% with a range from 0-7%. The frequency distribution of roof area in highly urban cells shows a strongly right-skewed distribution. This indicates that for highly urban areas, available roof area consists of many smaller habitat fragments and a few very large buildings. The mean distance between urban rooftops is around 400-500 m, with a roughly normal or slightly right-skewed distribution.  In suburban cells, the frequency distribution of roof area is less right-skewed reflecting less dominance of large buildings. Interestingly, the mean distance between suburban rooftops is around 200-300 m, less than for urban roofs, which is an important influence on dispersal potential.