Geoscience Reference
In-Depth Information
Figure 11.2
General data flow of the two main categories of
GIS
species distribution models identi-
fied in this chapter.
falling within the adequate temperature range are to be included as suitable
environments). No attempt is made to define confidence intervals to the indi-
vidual estimate, nor is any information provided on the relative importance of
one variable over another (e.g., vegetation types vs. temperature). Moreover,
no estimate of the degree of association or its variability is provided with the
relationship.
On the other hand, models that fall into the analytical group introduce
variability in the sense that advice from different specialists is combined to
define species-environment relationships, thus introducing variability in
terms of different opinions of the experts (deductive-analytical), or that the
species observation data are analyzed in a way that takes into account the range
of acceptability of all environmental variables measured, their confidence lim-
its, and their correlation. Both the deductive-analytical and the inductive-
analytical approaches tend to estimate the relative importance of the different
environmental layers considered in the analysis, thus moving toward an objec-
tive combination of environmental variable layers.
Examples of deductive-analytical models are based on techniques such as
multi-criteria decision-making (
MCDM
) (Pereira and Duckstein 1993), Delphi
(Crance 1987), and nominal group technique (NGT) (Allen et al. 1987).
Generally speaking, these techniques use the advice of more than one special-
