164
Selina S. Heppell, Deborah T. Crouse, and Larry B. Crowder
management policies on those life history stages that are most likely to increase
population growth.
The value of predicting the likely population-level effects of different management alternatives without expending either the time or money necessary to test
them in the field is obvious, provided the input parameters realistically reflect the
population’s vital rates under the circumstances. Fully testing different management alternatives in the field could also be risky for endangered species. It is
important to consider real management options, wherever possible, as alternative
hypotheses. In the Loggerhead Sea Turtle model, it is theoretically possible to
increase fecundity to a point at which population declines are arrested. But such a
management alternative is meaningless, as we have no way to affect fecundity in
sea turtles. Similarly, although small juvenile survival is an important vital rate to
the model, to date, we do not have access to large numbers of small juvenile
Loggerheads (it is believed they spend multiple years in a pelagic phase before
returning to coastal waters at about 50-cm carapace length). So, when simulating
the effects of TED requirements, we restricted the rate changes to large juveniles,
subadults, and adults, a meaningful management alternative to consider.
By considering multiple management approaches on a relative basis, we can
rank them by likely effect on population responses. For example, enhancing the
transition from solitary males to breeders by translocation of females is unlikely to
aid population growth based on the RCW model (because there are so few solitary
males) but enhancing the transition from helper to breeder (via drilling artificial
cavities, particularly in unoccupied habitat) should be successful. Similarly, with
a reasonable size limit, it appears that it would be relatively difficult to reduce
population sizes of Brook Trout via harvesting, but small changes in juvenile
growth and survival related to Rainbow Trout competition or reduced pH could
have dramatic effects. Again, the elasticity analysis approach should be useful to
managers who must decide how to allocate limited resources to a suite of alternative management approaches.
The advantages to be gained from matrix model analysis come from the flexibility of the model format. In each case we presented above, the matrix modeling
approach was modified to incorporate the particulars of the natural history of the
organism to be modeled and the data available for calculating parameters. The
complexity of each model reflects our current level of understanding of the
population’s vital rates. The sea turtle models are simple because the database to
support a more complex model simply is not available at this point. By contrast,
the RCW model is rather complex and reflects some of the natural complexity in
the woodpecker life history. The Brook Trout model includes density dependence
of vital rates when appropriate and examines both size and age classes. Matrix
models allow ample flexibility to examine a variety of life histories, while permitting a number of well-developed mathematical procedures (elasticity calculations
and others; cf. Caswell 1989) to aid analysis.
The deterministic linear modeling approach described in this chapter uses
sensitivity analyses to qualitatively compare the impact of stage-specific survival,
Selina S. Heppell, Deborah T. Crouse, and Larry B. Crowder
management policies on those life history stages that are most likely to increase
population growth.
The value of predicting the likely population-level effects of different management alternatives without expending either the time or money necessary to test
them in the field is obvious, provided the input parameters realistically reflect the
population’s vital rates under the circumstances. Fully testing different management alternatives in the field could also be risky for endangered species. It is
important to consider real management options, wherever possible, as alternative
hypotheses. In the Loggerhead Sea Turtle model, it is theoretically possible to
increase fecundity to a point at which population declines are arrested. But such a
management alternative is meaningless, as we have no way to affect fecundity in
sea turtles. Similarly, although small juvenile survival is an important vital rate to
the model, to date, we do not have access to large numbers of small juvenile
Loggerheads (it is believed they spend multiple years in a pelagic phase before
returning to coastal waters at about 50-cm carapace length). So, when simulating
the effects of TED requirements, we restricted the rate changes to large juveniles,
subadults, and adults, a meaningful management alternative to consider.
By considering multiple management approaches on a relative basis, we can
rank them by likely effect on population responses. For example, enhancing the
transition from solitary males to breeders by translocation of females is unlikely to
aid population growth based on the RCW model (because there are so few solitary
males) but enhancing the transition from helper to breeder (via drilling artificial
cavities, particularly in unoccupied habitat) should be successful. Similarly, with
a reasonable size limit, it appears that it would be relatively difficult to reduce
population sizes of Brook Trout via harvesting, but small changes in juvenile
growth and survival related to Rainbow Trout competition or reduced pH could
have dramatic effects. Again, the elasticity analysis approach should be useful to
managers who must decide how to allocate limited resources to a suite of alternative management approaches.
The advantages to be gained from matrix model analysis come from the flexibility of the model format. In each case we presented above, the matrix modeling
approach was modified to incorporate the particulars of the natural history of the
organism to be modeled and the data available for calculating parameters. The
complexity of each model reflects our current level of understanding of the
population’s vital rates. The sea turtle models are simple because the database to
support a more complex model simply is not available at this point. By contrast,
the RCW model is rather complex and reflects some of the natural complexity in
the woodpecker life history. The Brook Trout model includes density dependence
of vital rates when appropriate and examines both size and age classes. Matrix
models allow ample flexibility to examine a variety of life histories, while permitting a number of well-developed mathematical procedures (elasticity calculations
and others; cf. Caswell 1989) to aid analysis.
The deterministic linear modeling approach described in this chapter uses
sensitivity analyses to qualitatively compare the impact of stage-specific survival,