# Model sensitivity limits attribution of greenhouse gas emissions to polar bear demographic rates

## TL;DR

This study shows that the way greenhouse gas emissions are linked to polar bear population changes depends heavily on modeling choices, making the results uncertain.

## Contribution

The study reveals that the existing framework for linking emissions to polar bear demographics is overly sensitive to decision rules.

## Key findings

- The number of ice-free and fasting days varied significantly based on different decision rules.
- The threshold for reproductive failure was not consistently reached in the Southern Beaufort Sea subpopulation.
- Modeling assumptions strongly influence conclusions about the impact of greenhouse gas emissions on polar bears.

## Abstract

Greenhouse gas emissions continue to increase and negatively affect sea ice conditions that polar bears rely on. It is therefore important to better understand how specific emissions levels affect polar bear demography. A recent study proposed a framework to address this issue, but sensitivity to decisions rules of the approach may limit its utility. We tested how sensitive the approach is to decisions rules related to sea ice concentration, choice of subpopulation boundaries, and modeling choices for bears in the Chukchi Sea and Southern Beaufort Sea subpopulations. We found that the number of ice-free days, number of fasting days, and when 10% of reproductive females exhibited recruitment failure varied considerably depending on equally-valid decisions rules versus those used in the existing study. Whereas the previous study suggested that both subpopulations surpassed the critical number of ice-free days that negatively affect recruitment, we found this threshold was never reached by the Southern Beaufort Sea subpopulation and only once for the Chukchi Sea subpopulation for the decision rules we considered. Our results suggest that the previously published approach is too sensitive to modeling assumptions and choice of decision rules to accurately evaluate the impacts of GHG emissions on polar bear demographic rates.

The online version contains supplementary material available at 10.1038/s41598-025-89218-3.