In this study we developed a methodology that associated land cover categories to aerial photographs. We selected 4, 42-mile (10 km2) study sites within metropolitan Tucson, Arizona for our pilot study. Land uses within these sites covered the range of landscapes found within Pima County, Arizona, although not necessarily in proportions representative of the metropolitan area. We designated land cover categories within the pilot sites based on recent (March, 1990) aerial photographs and developed a geographical information system (GIS) database of these land cover categories. We measured vegetation attributes on randomly selected samples within each land cover category and classified vegetation occurring on specific land cover categories according to the Brown et al. (1979) system. Our land cover classification system was nominal with a hierarchical structure, facilitating organization and providing flexibility for adding new categories.
Our results showed that although neighborhood parks contained the greatest vegetated areas within our four study plots, very low density housing (≥4 acres/house), rivers whose banks were partially stabilized, naturally occurring washes (with no bank stabilization), and natural open space contained the highest percentage of native vegetation. Within our four study sites, low density housing (≥4 acres/house), rivers with partially stabilized banks, naturally occurring washes (no bank stabilization), and natural open space land cover categories contained the most area that was covered with vegetation providing escape cover.