Model results should be used to support and guide decisions rather than to
dictate decisions.
Models typically have significant uncertainties associated with their
results and output. Uncertainties in the input data are sometimes explicit
and obvious. Often, however, model uncertainties are not recorded or are
unknown and unstated. Sidebar 14.1 gives an example of how information
266
Thomas P. Maxwell et al.
Figure 14.1. The RAMAS Red List dialogue for entering an uncertain value.
In this example, the extent of occurrence for Grevilla caleyi (an Australian
shrub) is entered as a best estimate of 6 km
2 and a plausible range of 4 to
7 km
2 . This uncertainty is represented as a triangular fuzzy number [Data from
Akçakaya et al. (2000)].
Sidebar 14.1
Propagating uncertainty with RAMAS Red List
RAMAS Red List version 2.0 implements threatened-species criteria
of the International Union for the Conservation of Nature (IUCN)
(IUCN Species Survival Commission 2001). Those criteria constitute
dictate decisions.
Models typically have significant uncertainties associated with their
results and output. Uncertainties in the input data are sometimes explicit
and obvious. Often, however, model uncertainties are not recorded or are
unknown and unstated. Sidebar 14.1 gives an example of how information
266
Thomas P. Maxwell et al.
Figure 14.1. The RAMAS Red List dialogue for entering an uncertain value.
In this example, the extent of occurrence for Grevilla caleyi (an Australian
shrub) is entered as a best estimate of 6 km
2 and a plausible range of 4 to
7 km
2 . This uncertainty is represented as a triangular fuzzy number [Data from
Akçakaya et al. (2000)].
Sidebar 14.1
Propagating uncertainty with RAMAS Red List
RAMAS Red List version 2.0 implements threatened-species criteria
of the International Union for the Conservation of Nature (IUCN)
(IUCN Species Survival Commission 2001). Those criteria constitute