untitled

1980s proved to be a false alarm, at least in this case, as one might expect in the

majority of cases (>90%). Nonetheless, the PVA approach gives us a formal starting

point in evaluating extinction risk.

Population viability analysis models are attractive because they supply hard numbers

for the kind of uncertain (stochastic) processes that threaten small populations. This

has led to widespread proliferation of PVA models, as discussed by several in-depth

reviews (Boyce 1992; Morris and Doak 2002). However, we should be cautious about

the reliability of projected extinction risks, for a number of reasons. First, we rarely

have precise, reliable estimates of growth rates in any population, let alone those that

are under threat. Minor errors in our estimates of demographic parameters (because

the number of years of data is too short, for example) multiply geometrically over

time, leading to inflated (or deflated) estimates of extinction risk (Ludwig 1999; Fieberg

and Ellner 2000). Projection of extinction risk beyond a short time ahead (10–20%

of the length of the time series) provides high uncertainty that the extinction risk

estimates have little value (Fieberg and Ellner 2000). In other words, 40 years of

304 Chapter 17

0.10

0.08

0.06

0.04

0.02

0

0 20 40 60 80 100

Time

Cumulative probability of extinction

Fig. 17.6The
cumulative probability
of extinction over a
given span of time for a
grizzly bear population
with demographic
parameters identical
to the Yellowstone
population studied
during 1959– 82.
σ=0.08.

90

80

70

60

50

40

30

1960 1965 1970 1975 1980 1985 1990 1995

Year

Number of female grizzles

Fig. 17.7Population
dynamics of female
Yellowstone grizzlies
from 1959 to 1995.
(Based on the combined
data of Eberhardt et al.
1986 and Haroldson
1999.)

17.7.3Strengths and
weaknesses of PVA