ABSTRACT: Haul-out behaviour is a central component in the life history of pinnipeds; seasonal patterns in this behaviour reveal insights into annual energy budgets and basic biology. In this study, custom-designed satellite-relay data loggers were deployed on 17 male walruses Odobenus rosmarus in Svalbard, Norway. Individual animals transmitted data for a mean ± SD of 255 ± 132 d (range = 54-471 d), performing on average 110 ± 65 haul-out events (range = 30-247 events); the data records for 5 animals were longer than 1 yr. Clear seasonal patterns occurred in the percentage of time hauled out and average haul-out duration; both variables reached maxima in summer and minima in winter. Time between haul-out events reached a maximum during the breeding season in winter. The walruses moved away from shore-based haul-out sites and onto sea ice in November and December, and returned to land again in June. Analyses using generalized additive mixed models and Cox proportional hazard models demonstrated that wind chill and time in the water prior to a haul-out event had the largest impacts on haul-out probability and duration, although wind chill had little impact during the summer when temperatures were not challenging, and also had reduced influence during the breeding period. Long periods at sea were followed by long periods hauled out, resulting in a relatively constant proportion of time spent hauled out over time scales of weeks. Our results indicate the importance of breeding in dictating male haul-out patterns during winter, and provide a baseline to assess impacts of future climate change.
KEY WORDS: Arctic · Biotelemetry · Cox proportional hazard models · Generalized additive mixed models · Marine mammals · Pinnipeds · Svalbard
Full text in pdf format Supplementary material | Cite this article as: Hamilton CD, Kovacs KM, Lydersen C
(2015) Year-round haul-out behaviour of male walruses Odobenus rosmarus in the Northern Barents Sea. Mar Ecol Prog Ser 519:251-263. https://doi.org/10.3354/meps11089
Export citation Share: Facebook - - linkedIn |
Previous article Next article |