most important management issues and environmental factors (Rule
No. 2). For example, in the cumulative effect model, we did not include the
mechanisms or human actions that drive the demographic variables in the
model (Rule No. 7; see also Figure 2.2). Hypothetical scenarios focusing
on a limited set of presumed, key factors were a useful way to limit
complexity while still exploring a full range of parameter values. Our results
indicated that only some of the innumerable environmental and anthropogenic conditions that could be linked to the key factors of wolf population
trends merit more detailed investigation.
The Voyageurs Park cumulative effect projects would have benefitted
from even greater interaction between park staff and modelers (Rules Nos.
1 and 2). Numerous conditions resulted in initially vague project objectives
and priorities: a project mandated by an agency outside the park, a long
lead time between project instigation and modeling, staff turnover, and
political pressures on park management. Further, we proposed a novel
approach to cumulative effect analysis to a staff with limited experience
with either modeling or wolves. In retrospect, it would have been helpful
to develop some initial analyses or model exercises to connect the new
managers to the project and establish more clear objectives for the project
from the start.
One of the barriers we experienced with managers was their expectation
that the model would “solve their problem” or at least convince constituents
that managers were doing the right thing (Rule No. 3). Strategic modeling
helps management by revealing the relative importance of different factors
and the conditions under which the population is most vulnerable and
secure. It may also help identify thresholds for rapidly increasing risk that
suggest management criteria. However, modeling does not relieve managers from establishing clear objectives under diverse political pressures or
making judgments under uncertainty. Stochastic modeling can provide
important insights, but does not tell managers whether or not to prohibit
specific human actions or even which management approach is “best” under
conflicting societal demands. We had to help managers understand that
stochastic population modeling is experimental, not prescriptive. Further,
modeling is a process not a product, an interactive, adaptive activity that
evolves with the management objectives.
2.5 Conclusions
We illustrated a pragmatic approach to modeling that involved working
with expert biologists and managers to construct a simple population
model that addressed specific management-oriented questions. The model
included the basic processes of wolf demography and social structures
necessary to make accurate predictions. Simple simulation experiments
were used to determine the population impacts of changes in demographic
42
Jean Fitts Cochrane et al.
No. 2). For example, in the cumulative effect model, we did not include the
mechanisms or human actions that drive the demographic variables in the
model (Rule No. 7; see also Figure 2.2). Hypothetical scenarios focusing
on a limited set of presumed, key factors were a useful way to limit
complexity while still exploring a full range of parameter values. Our results
indicated that only some of the innumerable environmental and anthropogenic conditions that could be linked to the key factors of wolf population
trends merit more detailed investigation.
The Voyageurs Park cumulative effect projects would have benefitted
from even greater interaction between park staff and modelers (Rules Nos.
1 and 2). Numerous conditions resulted in initially vague project objectives
and priorities: a project mandated by an agency outside the park, a long
lead time between project instigation and modeling, staff turnover, and
political pressures on park management. Further, we proposed a novel
approach to cumulative effect analysis to a staff with limited experience
with either modeling or wolves. In retrospect, it would have been helpful
to develop some initial analyses or model exercises to connect the new
managers to the project and establish more clear objectives for the project
from the start.
One of the barriers we experienced with managers was their expectation
that the model would “solve their problem” or at least convince constituents
that managers were doing the right thing (Rule No. 3). Strategic modeling
helps management by revealing the relative importance of different factors
and the conditions under which the population is most vulnerable and
secure. It may also help identify thresholds for rapidly increasing risk that
suggest management criteria. However, modeling does not relieve managers from establishing clear objectives under diverse political pressures or
making judgments under uncertainty. Stochastic modeling can provide
important insights, but does not tell managers whether or not to prohibit
specific human actions or even which management approach is “best” under
conflicting societal demands. We had to help managers understand that
stochastic population modeling is experimental, not prescriptive. Further,
modeling is a process not a product, an interactive, adaptive activity that
evolves with the management objectives.
2.5 Conclusions
We illustrated a pragmatic approach to modeling that involved working
with expert biologists and managers to construct a simple population
model that addressed specific management-oriented questions. The model
included the basic processes of wolf demography and social structures
necessary to make accurate predictions. Simple simulation experiments
were used to determine the population impacts of changes in demographic
42
Jean Fitts Cochrane et al.