A simple fuzzy logic approach to analyze integrated interdisciplinary 'four-component' conceptual resource management models

Derivation of a GIS-based watershed-scale conceptual model for the St. Jones River Delaware from habitat-scale conceptual models

“…Each habitat model has its own set of VECs and services, but draws from the same set of drivers and stressors (See Reiter, 2004 andReiter et al, 2006 for more details on the model and basic methodology).…”

“…Although habitat models are useful for issues that have primary impacts at the habitat scale, they are less functional for addressing issues at larger scales such as retracing the full path of common drivers of interest, finding the most sensitive stressors in the watershed for focusing research attention, making management decisions that would impact multiple habitats at once, or simply having an idea of how the VECs react to the stressors at a watershed scale (Reiter, 2004). While the question is of major interest to planners and managers, there is limited existing procedure or literature on the question given that the four-component methods themselves are fairly new.…”

“…The narrowed habitat-scale conceptual models and associated matrices developed by Reiter (2004) were used as the basis for the construction of the watershed-scale conceptual model. Using the 2002 land use/land cover (LULC) GIS-based map of Kent County, Delaware (Kent County, 2002), we approximated the weight of each habitat within the watershed by combining related LULC attributes into our model classification scheme and creating a shapefile for each habitat using ArcView GIS 3.3.…”

“…We then created a numerical representation of the watershed model which incorporated the calculated weight of each habitat in the matrices. In each matrix, the qualitative description from the stakeholder workshop was replaced by relative numerical equivalents (not applicable ¼ 0, low ¼ 0.25, medium ¼ 0.5, high ¼ 0.75; Reiter, 2004). To create the numerical model, we incorporated the weight of each habitat within the driver-loading matrix, and the sum of the weighted driver across all habitats was considered to be the driver's weight for the watershed model (Table 2).…”

“…For example, to calculate the weight for the driver ''Development'' at the watershed scale, we multiplied the area of each habitat in percent by the ''Development'' driver value (0-5) from the habitat models: (0.1499*5) þ (0.0185*5) þ (0.1608*5) þ (0.4129*5) þ (0.0881*5) þ (0.0001*4) þ (0.003*2) þ (0.0138*5) þ (0.0001*1) þ (0.0122*2) ¼ 4.2509. Then, using the method of Reiter (2004), we created the social, ecological, flow-component, and feedback matrices for the watershed. With the numerical values for the drivers, we developed a set of equations for each parameter in the diagrammatic watershed model using the methodology of Reiter (2004).…”