28
Brendan Moyle
delayed initially by the erroneous belief that a large population persisted on the
west coast of New Zealand’s South Island (Williams 1986).
In the context of species recovery work, it is desirable to have management
tools that are objective and transparent. A failure to generate such tools may
perpetuate a bias toward well-studied or charismatic groups (Moyle 1998). However, the managers may simply lack the resources or time to engage in comprehensive analysis of extinction risk. In this setting, decision heuristics (rules of thumb)
may be adopted that are believed to yield satisfactory results.
Decision heuristics that have been adopted include rules based on genetic
theory and island biogeographic theory (Diamond 1975) and point-scoring systems used to establish priorities for species conservation activities (e.g., Molloy
and Davis 1992). Applications of criteria derived from genetic theory in the 1980s
used population size to assess population viability, with a consequence that some
populations were judged to be nonviable populations and were excluded from
consideration in recovery work (Hartley 1997). Applications based on island
biogeographic theory relied on island (and habitat patch) sizes and arrangements
to provide some guidance for target areas and management strategies for threatened species. Heuristics based on genetic theory and island biogeography ignored
a host of ecological considerations and the impact of various management
decisions.
More recently, decision rules (IUCN 1994) and point-scoring systems (e.g.,
Millsap et al. 1990) were developed in attempts to provide efficient means of
estimating extinction risks. For instance, the species priority ranking system in
New Zealand (Molloy and Davis 1992) incorporated a variety of management and
biological factors into a point-scoring system. These included the security of
existing ranges from disturbance, population size, threats, and success of management tools such as captive breeding or cultivation. Some factors were not directly
related to extinction risk, but the system was dominated by attributes expected to
be closely associated with it (Davis et al. 1992). Each factor was ranked ordinally
from 1 to 5, in which risk was inversely related to rank. The ranks for each
attribute were then summed to produce a cardinal number. However, summation
is not an appropriate way to deal with ordinal numbers, and the cardinal sum was,
unfortunately, nonsensical. Unweighted summation is sensitive to the number of
factors that are included in the system and may not reflect the threats that make the
greatest contribution to a species’ vulnerability to extinction.
Although a point-scoring system is more demanding of data than the genetic or
island biogeographic decision heuristics, it can still be attempted with rather
coarse data. The requirement to collect ordinal data means that some of the costs
of obtaining more accurate data can be avoided. If a conservation agency lacks the
time or resources to conduct comprehensive surveys, these types of decision
heuristics may improve the quality of the decisions made by managers. In this
regard, they may still be a useful model even if their grounding in ecological
realism is untested. To obtain better models with improved ecological realism
may simply be unachievable, given available resources.
Brendan Moyle
delayed initially by the erroneous belief that a large population persisted on the
west coast of New Zealand’s South Island (Williams 1986).
In the context of species recovery work, it is desirable to have management
tools that are objective and transparent. A failure to generate such tools may
perpetuate a bias toward well-studied or charismatic groups (Moyle 1998). However, the managers may simply lack the resources or time to engage in comprehensive analysis of extinction risk. In this setting, decision heuristics (rules of thumb)
may be adopted that are believed to yield satisfactory results.
Decision heuristics that have been adopted include rules based on genetic
theory and island biogeographic theory (Diamond 1975) and point-scoring systems used to establish priorities for species conservation activities (e.g., Molloy
and Davis 1992). Applications of criteria derived from genetic theory in the 1980s
used population size to assess population viability, with a consequence that some
populations were judged to be nonviable populations and were excluded from
consideration in recovery work (Hartley 1997). Applications based on island
biogeographic theory relied on island (and habitat patch) sizes and arrangements
to provide some guidance for target areas and management strategies for threatened species. Heuristics based on genetic theory and island biogeography ignored
a host of ecological considerations and the impact of various management
decisions.
More recently, decision rules (IUCN 1994) and point-scoring systems (e.g.,
Millsap et al. 1990) were developed in attempts to provide efficient means of
estimating extinction risks. For instance, the species priority ranking system in
New Zealand (Molloy and Davis 1992) incorporated a variety of management and
biological factors into a point-scoring system. These included the security of
existing ranges from disturbance, population size, threats, and success of management tools such as captive breeding or cultivation. Some factors were not directly
related to extinction risk, but the system was dominated by attributes expected to
be closely associated with it (Davis et al. 1992). Each factor was ranked ordinally
from 1 to 5, in which risk was inversely related to rank. The ranks for each
attribute were then summed to produce a cardinal number. However, summation
is not an appropriate way to deal with ordinal numbers, and the cardinal sum was,
unfortunately, nonsensical. Unweighted summation is sensitive to the number of
factors that are included in the system and may not reflect the threats that make the
greatest contribution to a species’ vulnerability to extinction.
Although a point-scoring system is more demanding of data than the genetic or
island biogeographic decision heuristics, it can still be attempted with rather
coarse data. The requirement to collect ordinal data means that some of the costs
of obtaining more accurate data can be avoided. If a conservation agency lacks the
time or resources to conduct comprehensive surveys, these types of decision
heuristics may improve the quality of the decisions made by managers. In this
regard, they may still be a useful model even if their grounding in ecological
realism is untested. To obtain better models with improved ecological realism
may simply be unachievable, given available resources.