Table 1.
Initial body composition of bears used in the fasting simulations.
Table 2.
Properties used to parameterize the biophysical model.
Figure 1.
Polar bear surface areas calculated by Niche Mapper compared to expected surface areas.
Niche Mapper surface area calculations are for bears modeled in average body condition (total mass = 2.25× structural mass) with straight-line lengths of 1.8–2.5 m (see Text S1 for conversion of body length and condition to total body mass). Best [55] measured surface areas on 18 bears ranging from 11–375 kg and developed predictive equations for skin (surface area = 0.09*mass0.67) and fur (surface area = 0.11*mass0.67) surface areas (m2) as a function of animal mass (kg).
Figure 2.
Polar bear torso circumference calculated by Niche Mapper compared to expected axillary girth.
Torso cylinder circumferences modeled by Niche Mapper for bears of average body condition (total mass = 2.25× structural mass) compared to expected axillary girth for polar bears of different body lengths based on predictive equations from field study measurements of wild polar bears [66]. The error bars for the Thiemann et al. (1980–1996 data) represent the average 5.84% difference between predicted and actual measurements reported in the study.
Figure 3.
Radiant fur temperature comparison.
Comparison of radiant fur temperatures as measured by Best [48] using captive bears and Niche Mapper predictions for a model bears of the size under the same ambient temperatures for no wind and 7 m/s wind speed. To simulate the treadmill activity in the Best study, Niche Mapper targeted a metabolic rate 2.5× expected basal metabolic rate from the mouse to elephant curve. Comparisons were similarly close at 4 m/s wind speed (not shown).
Figure 4.
Denning polar bear metabolic rate comparison.
Average (± S.E.) standard metabolic rate (SMR) and lowest observed metabolic rate (LOMR) for polar bears under simulated denning conditions [56], compared to Niche Mapper metabolic calculations for simulations of a bear of the same size (204 kg) and core temperature (36.5°C) at steady state with the reported den environmental conditions (−3°C). Niche Mapper attempted to reach the average standard metabolic rate reported in all simulations only by varying flesh thermal conductivity. Bear fur depth was not reported, so the impact of fur depth on Niche Mapper metabolic predictions is shown, as is the impact of curled (legs tucked into torso) and uncurled posture.
Figure 5.
Comparison of predicted average polar bear weight loss to field measurements.
Average weight losses (±S.E.) for bears of all sizes and body condition predicted by Niche Mapper simulations compared to weight losses from field studies of bears in the western Hudson Bay population during the ice-free period. The Derocher & Stirling data [67] is for adult males and females. The Polischuk et al. [60] and Atkinson et al. study [59] male data presented include both adult and subadult males, and for Polischuk et al., females with yearlings.
Figure 6.
Comparing predicted mass-specific polar bear weight loss during a seasonal fast to field measurements.
Weight-specific average weight losses for males predicted by Niche Mapper compared to weight-specific mass losses male bears in the western Hudson Bay population, reported in Atkinson et al. [59].
Table 3.
Percentage of initial energy reserves remaining after simulated 120 day and 180 day fasts.
Table 4.
Percentage of hours Niche Mapper predicted panting would be required for polar bears to maintain a minimum metabolic rate during different fasting scenarios.
Figure 7.
Estimated total daily metabolic requirements for fasting polar bears.
Estimated total daily metabolic requirements relative to the target daily metabolic rate (55*mass0.78 kcal/d; from [26]) for average sized male and female bears (body lengths of 2.3 m and 2.0 m, respectively) during the current ice-free period in the western Hudson Bay, Canada. The dashed lines show the range of values Niche Mapper would accept; values outside the dashed lines indicate that Niche Mapper exhausted thermoregulatory options before reaching a heat balance at ±5% of the target metabolic rate. These results assume that fat is used to supply 90.85% of energy reserves during the fast.