Ecological integrity of priority habitats was based on degree of local human development, amount and local connectivity of habitat, and quality of habit. Indicators were selected to reflect the needs of focal species, as well as other key ecological attributes of these habitats. Ecological integrity was estimated for floodplain forests, freshwater wetlands, human development, major rivers, rice agriculture, tallgrass prairie, and tidal wetlands.
Ratings of each indicator are based on a viability assessment conducted following the Open Standards framework. Indicators with ratings of Good or Very Good are within the acceptable range of natural variation (with Good suggesting some management may be needed to get the attribute back to an ideal condition). Indicators with ratings of Fair or Poor are outside of the acceptable range of natural variation, and require some level of restoration. EEMS is a tree-based, fuzzy logic modeling system, where output values range from -1 to +1, indicating whether a given output is entirely FALSE or TRUE, respectively. We translated this approach to equate to the Open Standards framework of rating indicators Poor, Fair, Good, or Very Good. To convert raw data to fuzzy variables, the threshold for +1 was set at the breakpoint between Good and Very Good for any given indicator, while the threshold for -1 was set at the breakpoint between Fair and Poor for any given indicator. Because of the hierarchical nature of EEMS and the fact that converting raw data to fuzzy variables truncates the range of the fuzzy variables that are used as input for higher level outputs, our final maps depict first-level indicators (i.e., direct results of converting raw variables to fuzzy variables) on a scale where Very Good >= 0.9 and Poor <= -0.9. For higher level indicators, Very Good >= 0.5 and Poor <= -0.5.
