Endocrine profiling of reproductive status and evidence of pseudopregnancy in the Pacific walrus (Odobenus rosmarus divergens).

Abstract

Endocrine profiling is an increasingly utilized tool for detecting pregnancies in wild populations of mammals. Given the difficulty in calculating reproductive rates of Pacific walruses (Odobenus rosmarus divergens) the use of endocrine techniques for determining pregnancy rates could be particularly useful for management of the population. The goals of this study were to 1) determine if progesterone and total estrogen concentrations in ovarian tissues of female walruses could be used to determine reproductive state and 2) determine if walruses undergo a functional postpartum estrus, as is seen in other pinnipeds. Ovaries were collected from female walruses (n = 13) hunted in subsistence hunts by Alaska Native communities. Females were categorized as postpartum, full-term pregnant, pregnant diapause or unbred. Total estrogen concentrations were greatest in unbred (n = 2) and pregnant (n = 2) females. Progesterone concentrations were also nominally larger in unbred (n = 2) than pregnant (n = 2) and postpartum (n = 9) animals. Small samples sizes precluded the use of statistical comparisons among groups. Corpora lutea tissue samples in this study did not reflect the presence of a postpartum estrus in the month of May as postpartum females yielded lower total estrogen concentrations than unbred or pregnant animals. Both unbred animals were in a state of pseudopregnancy, which has not been physiologically described for this species before. The progesterone profiles in late (59 ng/g) and early (140 ng/g) pregnancy were lower than expected and fell within the range of the postpartum females (36-210 ng/g), suggesting low production of the hormone by the corpus luteum during these phases of pregnancy. Profiling reproductive hormones in free-ranging walruses demonstrates that an endocrine approach may be a valuable tool for determining reproductive status of females, however increased sample sizes and time of year must be considered to accurately separate pregnant versus pseudopregnant individuals.