Last Glacial Maximum to Holocene Sea Ice Decline in the Bering and Chukchi Seas: a Comparison of Diatom-, IP25-, and Grain Size-Based Sea Ice Proxies

Abstract:

With the rate of Arctic sea ice decline increasing over the past decade, and ice-free conditions in the central Arctic predicted for the second half of this century, it is increasingly important to put sea-ice decline in the context of natural variability. Unfortunately, satellite observations of sea ice only extend back to 1979, and direct observations before then are spotty at best. Recently, there have been several advances in sea ice proxies. Specifically improvements in diatom-based proxies and the molecular biomarker, IP25, have allowed quantitative reconstructions of Arctic sea ice.

Here we apply our own IP25 and diatom assemblage sea ice calibrations for the Bering and Chukchi seas to two cores in the Bering Sea that extend to the Last Glacial Maximum in order to explore differences in the proxy records. In addition, we present high-resolution grain size analyses as a way to test the presence or absence of sea ice.

Before 14.5 ka, the sediments are composed of up to 20% clay by volume, reflecting transport by sea ice or from a proximal coastline during this time of lowered sea level. IP25 indicates extensive sea ice and the diatom assemblages are dominated by Thalassiosira antarctica and pennate benthic diatoms. Today, T. antarctica is associated with heavy winter and spring sea ice and forms resting spores during an abrupt advance of sea ice in the fall. The pennate benthic taxa include rare instances of IP25-producing species as well as other Naviculoid species that are commonly found attached to spring sea ice.

Between 14.5 and 12 ka, clay sized particles make up only 8% of the sediments and IP25 indicates that sea ice was absent from both locations. However, diatoms commonly found in the marginal ice zone today such as Fragilariopsis spp. Thalassiosira hyalina, and Thalassiosira nordenskioeldii are common.

Sea ice was extensive (up to 60% annual coverage) at both sites during the Last Glacial Maximum and it declined during the deglaciation. IP25 indicates that sea ice was absent at both sites after 14.5 ka, except for a minor advance during the Younger Dryas. However, in contrast, diatom assemblages indicate that sea ice was present until 11 ka. We hypothesize that two different ice conditions existed, with the more recent ice unsuitable for IP25-producers.