Evaluating the effectiveness of ArcView 2.1 as a tool for predicting rare natural communities at the Stoddard Properties, Stoddard New Hampshire

Duffy, Michael
Rick Van de Poll
Department of Environmental Studies
Geographic Information Systems (GIS) technology can enable property managers to effectively track the occurrence of natural communities’ boundaries and species ranges; it can also produced detailed, customized maps to assist with management goals. Can GIS be used to predict the potential habitat for rare species or natural communities of a given property, making their inventory more efficient and cost-effective? This project evaluated that potential by using existing electronic data such as soil, elevation and vegetation cover maps, and manipulating them in an ArcView 2.1 Macintosh software environment. Digital maps were obtained from a number of federal, state and private sources to study four contiguous parcels covering approximately 1902 hectares (4700 acres) in Stoddard, New Hampshire. This land, referred to as the Stoddard properties, is a significant natural area in the southwest part of the state. Management of the properties calls for an extensive inventory of plant, animal and natural community resources, including rare species and natural communities. Four natural communities (northern acidic rocky summit, boreal circumneutral talus woodland, hardwood-conifer seepage swamp, and the level bog acidic fen complex) that had documented occurrences on the Stoddard properties were chosen to test the utility of ArcView software. Ecological profiles were compiled for each one. These general profiles were fined to produce specific quantitative data on physical and biological attributes such as soil type, expected pH range, aspect, dominant species and canopy closure. These attribute values were then used to query the digital maps layers in order to produce limited sets of appropriate physical and biological features for each natural community type. These new data layers were then brought together in one map. Intersections of the various attributes produced a final map layer of predicted habitat. This was then compared to the actual occurrences for each natural community. In three out of four cases, the predicted habitat contained the documented occurrence – only the hardwood-conifer seepage swamp was not successfully predicted.

Read Full Text