III. Management Goals and Criteria
defense-of-life, and incidental take. Indirect remov-
als, such as those that occur as a result of habitat
degradation, are captured either in the survival rate
or the carrying capacity. This distinction is largely
in deference to the ability to monitor direct, but not
indirect, removals.) If the rate of human-caused
removals is less than this upper bound, the demands
for the other demographic criteria can be reduced,
provided the persistence criterion is met. It is also
possible to meet ESA Fundamental Criterion 2
(90% probability of persistence over 100 years)
without meeting ESA Demographic Criterion 4
(human-caused removal rate less than h), but this
would require even higher survival and reproduc-
tive rates than specified by the second and third
demographic criteria (see discussion, below, of ESA
Threats-based Criterion 2). Thus, while the fourth
demographic criterion is not strictly necessary for
recovery, we have included it as a recovery criterion
because it reiterates MMPA Demographic Criterion
2, the combination of non-anthropogenic mortality
and anthropogenic mortality is critical, and the other
demographic criteria can only be set in the context
of the anthropogenic mortality.
There are three particular challenges in developing
and evaluating these demographic criteria: climate
change effects, density-dependence, and harvest.
First, sea-ice loss related to climate change is a
long-term threat that will present changing condi-
tions for ice-dependent Arctic species like the polar
bear. All of these demographic criteria are likely met
currently for the Polar Basin Divergent Ecoregion,
as well as for others; the concern is that they will
not continue to be met as climate-driven sea-ice
loss increases, which is why polar bears were listed.
Thus, the evaluation of the demographic criteria
needs to assess whether they will continue to be
met over the next 100 years. Second, survival and
recruitment (the first two demographic criteria)
may be density-dependent, that is, they naturally
decrease as the population size approaches carrying
capacity. Thus, a threshold value for those rates is
meaningless unless it is associated with a particular
population density. Here, we have chosen to estab-
lish these criteria in reference to the mnpl, which is
the population size, relative to the carrying capacity
at a point in time, that produces the highest net
annual production, assuming removals are unbiased
with regard to age and sex. This is a particularly
practical reference point because for polar bear
populations that are managed to be near mnpl,
the observed survival and recruitment rates can
be compared directly to the criteria. Third, for any
populations that are subject to direct human-caused
removals, the survival rate will be the product of
both the survival rate in the absence of anthropo-
genic take and the survival rate associated with
those removals, taking into account the sex and age
composition of the population and of the removals.
The survival rate in Demographic Criterion 1 refers
to the survival rate in the absence of removals, and
hence encompasses non-anthropogenic mortality;
the total take rate in Demographic Criterion 4 refers
to anthropogenic mortality.
The demographic criteria listed above are stated
in terms of average values of the true underlying
rates, not annual rates. Annual variation around
these mean values is expected; the criteria require
that the mean values of those stochastic processes
be above the indicated thresholds. Using average
values assumes that potential future change in how
much the rates vary from year-to-year will not, in
itself, have a meaningful effect on persistence. Also
the demographic criteria were derived assuming
a perfect ability to estimate them; the empirical
precision needed has not yet been developed. If the
demographic rates are measured or forecast with
considerable error, then it is possible to think that
the criteria have been achieved when the true values
do not, in fact, meet the criteria or, vice versa, to
think that the criteria have not been achieved when,
in fact, they have. The risk due to sampling error
has not been directly incorporated into the interpre-
tation of these criteria, but that consideration should
be evaluated carefully whenever a population status
assessment is made, and could be incorporated into
a future revision of this Plan.
The estimation of annual and mean rates for three
of the four demographic parameters (survival,
recruitment, and take rates) can be conducted with
monitoring programs that are already in place in
several polar bear subpopulations, including the
Southern Beaufort Sea. These programs currently
involve the marking and recapturing of individual
bears over time. Note, however, that the existing
monitoring programs are focused at the subpopula-
tion level but the ESA demographic criteria are
focused at the recovery unit level; research will be
needed to understand how to make inference at the
recovery unit level from data at the subpopulation
level (Regehr et al. 2015). The estimation of the
fourth demographic parameter, carrying capacity,
is notoriously challenging, because the link between
habitat variables and population responses is often
poorly understood. Modern statistical methods
(known as “hierarchical models”) provide a way to
estimate “latent” parameters like carrying capacity,
by integrating survival, recruitment, harvest,
habitat, and population size data into a single
statistical framework (Royle and Dorazio 2008). If
such a statistical model is developed for polar bears,
it can then be linked to forecasts of the habitat
variables (Durner et al. 2009) to provide the current
and projected estimates of carrying capacity needed
for Demographic Criterion 3.
As noted above, these demographic criteria should
be subject to periodic revision as new information
becomes available to inform their derivation.
28 Polar Bear Conservation Management Plan