Steroid hormones in Pacific walrus bones collected over three millennia indicate physiological responses to changes in estimated population size and the environment

Abstract

Abstract The Pacific walrus (Odobenus rosmarus divergens) is an iconic Arctic marine mammal and an important resource to many Alaska Natives. A decrease in sea ice habitat and unknown population numbers has led to concern of the long-term future health of the walrus population. There is currently no clear understanding of how walrus physiology might be affected by a changing Arctic ecosystem. In this study, steroid hormone concentrations (progesterone, testosterone, cortisol and estradiol) were analysed in walrus bones collected during archaeological [3585–200 calendar years before present (BP)], historical [1880–2006 common era (CE)] and modern (2014–2016 CE) time periods, representing ~ 3651 years, to track changes in reproductive activity and cortisol concentrations (biomarker of stress) over time. Our results show that modern walrus samples have similar cortisol concentrations (median = 43.97 ± standard deviation 904.38 ng/g lipid) to archaeological walruses (38.94 ± 296.17 ng/g lipid, P = 0.75). Cortisol concentrations were weakly correlated with a 15-year average September Chukchi Sea ice cover (P = 0.002, 0.02, r2 = 0.09, 0.04, for females and males, respectively), indicating a possible physiological resiliency to sea ice recession in the Arctic. All steroid hormones had significant negative correlations with mean walrus population estimates from 1960 to 2016 (P < 0.001). Progesterone in females and testosterone in males exhibited significant correlations with average September Chukchi Sea ice cover for years 1880–2016 (P < 0.001 for both, r2 = 0.34, 0.22, respectively). Modern walruses had significantly lower (P = < 0.001) reproductive hormone concentrations compared with historic walruses during times of rapid population increase, indicative of a population possibly at carrying capacity. This is the first study to apply bone as a tool to monitor long-term changes in hormones that may be associated with changes in walrus population size and sea ice cover.