The resulting layer was filtered to remove any patches of less than 4 pixels in size, and to fill in any holes of 4 pixels or less. Water was removed, unless it was completely surrounded by other natural land cover. A 30 meter buffer of the roads was also removed. Any patches in the resulting data layer that were less than 20 acres in size were removed. The remaining data layer consisted of all the Potential Conservation Areas.
Prior to GIS, each record of a species occurrence was mapped on USGS 7.5 minute quadrangle maps. A dot was placed on the map at the estimated lat/long center of the occurrence location and the lat/long recorded in the database. Each record was also given a mapping precision based on the known location. Second precision records meant the location was known precisely. Minute precision records were known to occur within a mile and a half of the lat/long point. General precision records were known to the township or quadrangle name level.
The MNFI database has now been incorporated into a GIS. Because the MNFI database precedes GIS technology, older records are given a spatial extent by taking the lat/long point of the record and applying a buffer based on a mapping precision. Second precision records are given a 100 meter buffer, minute precision records are given a 2,000 meter buffer, and general records are given a 8,000 meter buffer.
Newer records are entered as polygons with a digitized spatial extent. Occurrences best represented by a point, (i.e. small plant populations or nest sites) are represented by a small (12 meter) polygon.
To deal with these potential shortcomings, three aspects of the data, species habitat requirements, spatial extent of the occurrence, and the age of the record, are combined to create a probability layer. The layer represents the likelihood of encountering a sensitive species or natural community, based on known sightings, in a given area.
Species were grouped into habitat classes. The appropriate habit was extracted from the IFMAP land use coverage. Then the spatial extent of each occurrence was used to define a boundary from which to extract the potential habitat.
After creating a bounded habitat layer, the age of a record was used to assign a low, medium, or high probability (0.25, 0.50, and 1.00) to the habitat parcel. The probability represents the likelihood that the occurrence still exists. Records prior to 1970 were assigned a low probability. Records after 1982 were given a high probability. Records between 1971 and 1982, by default, were given a medium probability.
Each occurrence is also given a biodiversity value, based on the species global status, state status, occurrence quality rank, and the probability value. The greater the threat of imperilment to the species, the higher the value assigned to the occurrence. In a similar vein, the higher the quality of each occurrence, the higher it's assigned value. The biodiversity value of each occurrence is then calculated by adding the values for the global status, state status, and the quality ranking, and then multiplying the sum by the probability value.
The end result is a vector layer of the habitat within each occurrence boundary, with attributes of probability and biodiversity value.
Core Area - the size of the PCA after a 300 foot buffer is removed. Stream Corridor - the presence or absence of a stream Landscape Connectivity Percent - the percent of a quarter mile buffer around each PCA that is made up of other PCAs or the transition zone. Landscape Connectivity Proximity - the number of other PCAs or transition zone areas within 100 feet. Restorability - the percent of a quarter mile buffer that is made up of restorable land. Conservation Value - the sum of the conservation values from the eo_habitat layer that intersect the PCA. Quality Percent - the percent of the PCA that contains vegetation who's class is unchanged from circa 1800 Quality Area - acres of the PCA that contain vegetation who's class is unchanged from circa 1800 Scores were given for each criteria result, and the individual scores were tallied resulting in one total score for each PCA.