habitat-based model envisioned by park biologists, Cochrane (2000) used
the demographic wolf population model to predict the relative effects of
four environmental factors (prey availability, human-caused mortality,
immigration, and disease mortality) on the persistence of wolves in the
park.
To predict the relative impacts, Cochrane (2000) employed a fullfactorial experimental design with the four environmental factors at
five levels each. The wolf population in the model was assumed to occupy
a maximum of 15 territories, 3 inside the park and 12 surrounding the park.
The response variable was the likelihood that wolf population size inside
the park fell below specified thresholds in any year before the 30-year time
horizon. Ten response variables were measured with population sizes from
0 to 18 wolves in increments of 2. The level for each primary environmental
factor was specified in terms of the levels of one or more demographic parameters in the simulation model. The level of prey availability affected mean
litter size and the rates of dispersal and winter mortality; human-caused
mortality affected the rate of winter mortality in territories outside the
park; and disease mortality affected the rate of summer pup mortality. The
levels of immigration were 0 to 24 immigrants per year in increments of 6.
The results of this factorial analysis (Figure 2.3) suggested that disease
mortality is the most important factor affecting whether or not the park
wolf population would remain near its initial population size. Immigration
had the most impact on the likelihood that the wolf population in the park
would fall below a threshold of seven wolves. Human-caused mortality in
wolf territories outside the park had little effect on the number of wolves
in the park, except when the population was already very small under
extreme conditions of no immigration, very low prey biomass, and high
disease mortality. While changes in the demographic parameters associated
with alternative levels of prey availability had little effect on wolf population size, prey availability had stronger effects in experiments where prey
determined territory spacing or sizes (results not shown). Thus, prey availability within foreseeable ranges had an effect on total wolf numbers
(through the number of territories that can “fit” within the small park) but
very little relationship with the likelihood of extirpation. It was easier to
maintain breeding pairs rather than a large population in the park, and
these breeding pairs were highly resilient to extirpation because of the
readily available pool of replacement breeders.
The results of this factorial analysis helped ameliorate concerns about
human-caused mortality of wolves outside the park while focusing new
attention on the spread of disease from dogs to wolves. In addition, the
model results indicated which environmental conditions would likely
enhance the security of the population. Those environmental conditions,
which supported the population’s reproductive capacity more than a
constant, large population size, could be monitored in lieu of intensive
wolf population sampling or trend interpretation.
2. Modeling for Endangered-Species Recovery
33
the demographic wolf population model to predict the relative effects of
four environmental factors (prey availability, human-caused mortality,
immigration, and disease mortality) on the persistence of wolves in the
park.
To predict the relative impacts, Cochrane (2000) employed a fullfactorial experimental design with the four environmental factors at
five levels each. The wolf population in the model was assumed to occupy
a maximum of 15 territories, 3 inside the park and 12 surrounding the park.
The response variable was the likelihood that wolf population size inside
the park fell below specified thresholds in any year before the 30-year time
horizon. Ten response variables were measured with population sizes from
0 to 18 wolves in increments of 2. The level for each primary environmental
factor was specified in terms of the levels of one or more demographic parameters in the simulation model. The level of prey availability affected mean
litter size and the rates of dispersal and winter mortality; human-caused
mortality affected the rate of winter mortality in territories outside the
park; and disease mortality affected the rate of summer pup mortality. The
levels of immigration were 0 to 24 immigrants per year in increments of 6.
The results of this factorial analysis (Figure 2.3) suggested that disease
mortality is the most important factor affecting whether or not the park
wolf population would remain near its initial population size. Immigration
had the most impact on the likelihood that the wolf population in the park
would fall below a threshold of seven wolves. Human-caused mortality in
wolf territories outside the park had little effect on the number of wolves
in the park, except when the population was already very small under
extreme conditions of no immigration, very low prey biomass, and high
disease mortality. While changes in the demographic parameters associated
with alternative levels of prey availability had little effect on wolf population size, prey availability had stronger effects in experiments where prey
determined territory spacing or sizes (results not shown). Thus, prey availability within foreseeable ranges had an effect on total wolf numbers
(through the number of territories that can “fit” within the small park) but
very little relationship with the likelihood of extirpation. It was easier to
maintain breeding pairs rather than a large population in the park, and
these breeding pairs were highly resilient to extirpation because of the
readily available pool of replacement breeders.
The results of this factorial analysis helped ameliorate concerns about
human-caused mortality of wolves outside the park while focusing new
attention on the spread of disease from dogs to wolves. In addition, the
model results indicated which environmental conditions would likely
enhance the security of the population. Those environmental conditions,
which supported the population’s reproductive capacity more than a
constant, large population size, could be monitored in lieu of intensive
wolf population sampling or trend interpretation.
2. Modeling for Endangered-Species Recovery
33