Identifying reliable indicators of fitness in polar bears

Abstract

Animal structural body size and condition are often measured to evaluate individual health, identify responses to environmental change and food availability, and relate food availability to effects on reproduction and survival. A variety of condition metrics have been developed but relationships between these metrics and vital rates are rarely validated. Identifying an optimal approach to estimate the body condition of polar bears is needed to improve monitoring of their response to decline in sea ice habitat. Therefore, we examined relationships between several commonly used condition indices (CI), body mass, and size with female reproductive success and cub survival among polar bears (Ursus maritimus) measured in two subpopulations over three decades. To improve measurement and application of morphometrics and CIs, we also examined whether CIs are independent of age and structural size-an important assumption for monitoring temporal trends-and factors affecting measurement precision and accuracy. Maternal CIs and mass measured the fall prior to denning were related to cub production. Similarly, maternal CIs, mass, and length were related to the mass of cubs or yearlings that accompanied her. However, maternal body mass, but not CIs, measured in the spring was related to cub production and only maternal mass and length were related to the probability of cub survival. These results suggest that CIs may not be better indicators of fitness than body mass in part because CIs remove variation associated with body size that is important in affecting fitness. Further, CIs exhibited variable relationships with age for growing bears and were lower for longer bears despite body length being related to cub survival and female reproductive success. These results are consistent with findings from other species indicating that body mass is a useful metric to link environmental conditions and population dynamics.

Fig 1

Plot of the percent of females age 5 and older observed with spring cubs (year t + 1) grouped by their body mass measured the prior spring (year t)(a) and the modeled probability (with 95% confidence intervals as dashed lines) of cub production relative to maternal body mass from a logistic regression (b).

Fig 2

Relationships between maternal mass and the mass of first year cubs (i.e., COY) and yearlings in one and two cub litters of adult female polar bears captured in the spring in the Chukchi Sea 1986–1994 (a) and yearlings captured in the Chukchi Sea 2008–2017 (b).

Fig 3

Relationship between the percent of cubs surviving and the mass of their mothers for raw data summarized among ranges of body mass (a) and for modeled probabilities (with 95% confidence intervals as dashed lines) based on a logistic regression (b). Cub survival was determined during the time between initial spring capture following den emergence to the following spring when their mothers were recaptured. Maternal mass was measured at initial capture of the cubs and was the only maternal metric related to cub survival.

Fig 4. Values of three condition metrics (BCI, BMI, and energy density) relative to age for male and female polar bears captured in the Chukchi Sea.

Fig 5. Values of three condition metrics (BCI, BMI, and energy density) relative to length for male and female polar bears captured in the Chukchi Sea.

Data are shown for bears age 2 and older.