Urban forests are valuable resources, and urban forest inventories are valuable management tools. The forest of Cary, North Carolina was inventoried using aerial photographs and Geographic Information Systems (GIS). The inventory began by creating a classification scheme compatible with the town's vegetation management program. The classification scheme included fifteen categories. Six categories were developed and undeveloped forests, one was farm land, two were water bodies, and six were other features of development. The scheme was applied to aerial photographs, and the results were mapped into a geographic information system (GIS) to create an urban forest cover map.
Ground sampling was used to determine forest composition and mapping accuracy. Sampling points were assigned using a systematic allocation, which produced 157 sampling points distributed over 22,943 hectares. Each sampling point was located on the ground, and stand composition data was collected using plotless sampling procedures.
Change and ecological analyses were performed on the cover map. The change analyses concentrated on a 1,045 hectare area in the southwestern section of the town. Aerial photographs from 1988 were classified, digitized, then overlaid on the vegetation map. Ecological analyses of the forest used two methods. The first identified ecologically significant regions based on the amount of the undeveloped forest cores that could be linked by floodzones. The second method involved using forest patch shape, size, and distribution to describe the landscape of Cary, and to compare developed and undeveloped areas.
Sampling revealed differences in basal area in developed and undeveloped forest types. In deciduous forests, the mean basal area of undeveloped stands was 9.7 m2, and 5.9m2 in developed stands. This represents a reduction of 3.7 m2 (39%). In coniferous forests, the mean basal area of undeveloped stands was 10.9 m2, and 4.6 m2 in developed stands, for a reduction of 6.2 m2(57%). In mixed forests, the mean basal area of undeveloped stands was 11.9 m2, and 7.8m2 in developed stands, for a reduction of 4.1 m2 (34%). The dominant species of the forest types were the same in the developed and undeveloped stands, although species less common in the undeveloped stands were not found in developed stands. When size distributions were examined, the overall diameter distributions of developed deciduous and mixed stands matched the distributions of undeveloped stands. In coniferous forests, the distribution of undeveloped stands was significantly different from developed stands.
Change analyses revealed 16 regions that had been undeveloped forest in 1988, but had been developed by 1993. These stands represented 107 hectares, or 10%, of the total area of the study region. These areas have reduced basal areas and canopy cover, and increased disturbance due to human activities.
Due to the high concentration of linkable 5 hectare forest interiors, the most ecologically significant area surrounded a lake in the northern section of town. Undeveloped forests comprised the greatest portion of the landscape. Developed categories, such as developed forests, recreational facilities, and impervious surfaces, were more frequent in the southeastern section of town, while undeveloped forests and farms were more frequent in the northwestern section.
This information has many implications. Water quality can be degraded due to increased runoff. Wildlife is impacted because of reduced habitat area and lack of connected forest patches. Urban climates will be affected due to decreased canopy cover. Municipal planning and urban forest management can mitigate the ecological impact of development by using this information to specify species for tree plantings, setting species replacement ratios, and setting guidelines for contiguous forest preservation. Utilization of this data can ensure that the urban forest is a self- sustaining ecosystem that serves and shelters all of its inhabitants.