Using geospatial data and principles of landscape ecology to identify field sites and characterize landscapes

Abstract

We explored the utility of using readily available geospatial datasets to assist in field site identification. Specifically, we used readily available geospatial data to locate potential field sites that represented specific points along an urban–rural gradient of development intensity at landscape and regional scales. We also incorporated development age (young versus mature) in our site selection. Using publicly available spatial datasets, we computed a set of variables that characterized the landscape at two spatial scales and combined these to develop a gradient of urbanization. We then used maps of these gradients, digital orthophotographs, and ancillary geospatial data layers to identify candidate field sites of the requisite landscape- and regional-scale development intensity. These field sites were then visited to verify the current condition, and acceptable sites were used in companion studies to survey avian communities. We were also interested in how land cover estimates derived from different data sources varied. We compared estimates of percent land cover at two spatial scales derived from coarse scale land cover maps and fine scale screen-digitized from aerial photographs. At finer spatial scales, it appears the added costs associated with screen-digitizing yield much more precise estimates of land cover, whereas at coarser scales, although satellite-based land cover classifications may be somewhat less accurate, they may be sufficiently correlated with aerial photo-interpreted classifications that the expenditure is not worthwhile.