because three interacting factors collectively undermined the acceptability
of FORPLAN solutions.
The first factor was the rapidly increasing public interest and participation in natural resource management decision making (Behan 1990; Knopp
and Caldbeck 1988; Wondolleck 1988). The second factor was the agency’s
strategic mistake of reducing all major aspects of the problem to a single
LP solution, making the models very large and often requiring dubious
transformations of information in the process. But the third factor, the difficulty of explaining the derivations of FORPLAN solutions, was perhaps the
most problematic (O’Toole 1983). With enormous public interest in the
management implications of model solutions, this final factor was a fatal
flaw.The lesson for modelers with a stake in resource management is simple:
scientifically sound models are a necessary, but not sufficient, condition
for successful model application in the modern public arena of resource
management. Increasingly, models are expected to explain themselves in
convincing and intuitive ways.
8.6.2 Habitat Suitability Index Models
Another example of a modeling approach that fell short of expectations
is Habitat Suitability Index (HSI) modeling. Such models have been developed for a wide variety of wildlife species as part of a formal habitatevaluation procedure that was extensively applied by the U.S. Fish and
Wildlife Service (Verner et al. 1986). These models focus on providing a
simple, formalized method for assessing impacts on wildlife habitat. The
HSI models attempt to provide information useful to managers on the site
characteristics that affect the use of particular habitats by a species. The
models typically consist of simple relationships among habitat quality and
multiple characteristics, such as canopy cover, diameter classes of trees and
shrubs, tree stem densities, area of open water, and distance to forest cover.
The objective is to combine these variables to provide an overall index of
suitability.
The HSIs are based on local habitat variables, ignoring species interactions except those caused by the indirect effects of related habitat variables.
Early HSI models ignored most landscape characteristics, making the
models inappropriate for situations where the sizes, shapes, edge effects, and
neighborhood relationships of habitats have a greater effect on habitat preference than local forest composition and structure. Because they are based
only upon habitat variables, they cannot take account of historical factors
driving local abundances, such as demography. Nor can they deal with the
absence of species resulting from interactions not described by the given
habitat variables, such as restrictions caused by pathogens. Considerable
effort to develop new methods to ameliorate some of these limitations have
been developed recently, making extensive use of remote-sensing methods
(Scott et al. 2001). Though inherently static entities, HSIs can also be
extended to include the dynamics of underlying environmental factors,
8. Evolving Approaches and Technologies
159
of FORPLAN solutions.
The first factor was the rapidly increasing public interest and participation in natural resource management decision making (Behan 1990; Knopp
and Caldbeck 1988; Wondolleck 1988). The second factor was the agency’s
strategic mistake of reducing all major aspects of the problem to a single
LP solution, making the models very large and often requiring dubious
transformations of information in the process. But the third factor, the difficulty of explaining the derivations of FORPLAN solutions, was perhaps the
most problematic (O’Toole 1983). With enormous public interest in the
management implications of model solutions, this final factor was a fatal
flaw.The lesson for modelers with a stake in resource management is simple:
scientifically sound models are a necessary, but not sufficient, condition
for successful model application in the modern public arena of resource
management. Increasingly, models are expected to explain themselves in
convincing and intuitive ways.
8.6.2 Habitat Suitability Index Models
Another example of a modeling approach that fell short of expectations
is Habitat Suitability Index (HSI) modeling. Such models have been developed for a wide variety of wildlife species as part of a formal habitatevaluation procedure that was extensively applied by the U.S. Fish and
Wildlife Service (Verner et al. 1986). These models focus on providing a
simple, formalized method for assessing impacts on wildlife habitat. The
HSI models attempt to provide information useful to managers on the site
characteristics that affect the use of particular habitats by a species. The
models typically consist of simple relationships among habitat quality and
multiple characteristics, such as canopy cover, diameter classes of trees and
shrubs, tree stem densities, area of open water, and distance to forest cover.
The objective is to combine these variables to provide an overall index of
suitability.
The HSIs are based on local habitat variables, ignoring species interactions except those caused by the indirect effects of related habitat variables.
Early HSI models ignored most landscape characteristics, making the
models inappropriate for situations where the sizes, shapes, edge effects, and
neighborhood relationships of habitats have a greater effect on habitat preference than local forest composition and structure. Because they are based
only upon habitat variables, they cannot take account of historical factors
driving local abundances, such as demography. Nor can they deal with the
absence of species resulting from interactions not described by the given
habitat variables, such as restrictions caused by pathogens. Considerable
effort to develop new methods to ameliorate some of these limitations have
been developed recently, making extensive use of remote-sensing methods
(Scott et al. 2001). Though inherently static entities, HSIs can also be
extended to include the dynamics of underlying environmental factors,
8. Evolving Approaches and Technologies
159