A Bayesian Network Modeling Approach to Forecasting the 21st Century Worldwide Status of Polar Bears - Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications - page 240

Table A1. (continued)

Node
Name

Node Title Node Description States

In essence, this is the tolerance of bears to live in more crowded conditions than those
at which they presently live. And, it is a function of food availability.
I believe that bears have a reasonable tolerance of crowding if food is abundant or if
they are in good condition while waiting for sea ice to return etc. Examples of these
situations include (1) portions of the high Arctic-like near Resolute, where bear densi-
ties on the sea ice in spring are apparently much higher than they are in most of the
polar basin, and (2) the high densities at which polar bears occur on land in hudson
Bay in summer when they are loafing and waiting for the sea ice to return.
I assumed that crowding tolerance has little or no effect on outcome likelihoods until
habitat quantity was reduced substantially requiring bears from one area to either per-
ish or find some place else to go on at least a seasonal basis.
Thereafter, if relocations of members of some subpopulations meant invading the areas
occupied by other bears, crowding tolerance entered an assessment of whether or not
relocation was a practical solution.

G Relocation

Possible

Is it likely that polar bears displaced from one region could either seasonally or perma-
nently relocate to another region in order to persist.
This is a function of foraging effects (e.g., prey availability) in the alternative area
(here I am specifically focusing on prey availability in the alternative area rather than
the area from which the bears may have been displaced) crowding tolerance, and
contiguity of habitats.

yes
no

A Foraging

habitat Value

This node expresses the sum total of things which may work to alter the quality of
habitats available to polar bears in the future. The idea here is that sea ice is retreat-
ing spatially and temporally, but is the ice that remains of comparable, better or worse
quality as polar bear habitat. Our RSF values are projected into the future with the
assumption that a piece of ice in 2090 that looks the same as piece of ice in 1985 has
the same value to a polar bear. Perhaps because of responses we cannot foresee, it may
be better seal habitat, or it may be habitat for an alternate prey. Conversely, it may be
worse because of atmospheric and oceanic processes (e.g., the epontic community is
less vibrant because of thinner ice which is not around for as long each year). Or it
may be worse habitat because of oil and gas development, tourism, shipping etc.

better
same_as_now
worse

D Change in

Foraging
habitat
Distribution

This node expresses the combination of the quantitative ways the retreat of sea ice may
affect use of continental shelf habitats.
Our analyses indicate, in addition to reductions of total ice (and RSF Optimum ice)
extent (expressed under habitat quantity), we will see seasonal retreats of the sea ice
away from coastal areas now preferred by polar bears, and these retreats are projected
to progressively become longer.

improved_availability
same_as_now
reduced_availability
Gr_reduced_avail-
ability
Unavailable

These changes will affect polar bears by reducing the total availability of ice substrate
for bears. They also will make ice unavailable for extended periods in some regions
where bears now occur year-round. This will result in the opportunity for seasonal oc-
cupancy but not year-round occupancy as they have had in the past.
Note that in the PBCE because it includes the North Beaufort and Queen Elizabeth
and East Greenland each of which has different starting points, the values in the CPT
express kind of an average. Similarly, in the Seasonal region, there is a huge difference
between Hudson Bay and Foxe Basin or Baffin Bay. So, again the CPT values are a
sort of an average, trying to reflect these differences.

l2 Vital Rates This expresses the combined effect of changes in survival of adult females and of

young and reproductive patterns. The probabilities assigned each of the states reflects
the relative importance to polar bear population dynamics of each of these vital rates to
the growth of the population.
This node does not reflect human influences on population growth such as hunting, or
mortalities resulting from bear-human interactions. Those things, along with effects of
parasites, contaminants, etc. are brought in as modifiers at the level of the next node.

improve
same_as_now
decline