Arctic sea-ice proxies: Comparisons between biogeochemical and micropalaeontological reconstructions in a sediment archive from Arctic Canada

Abstract

Boxcore 99LSSL-001 from the southwest Canadian Arctic Archipelago (68.095°N, 114.186°W), studied by multiproxy approaches (sea-ice diatom biomarker IP25, phytoplankton-based biomarker brassicasterol, biogenic silica, total organic carbon, dinoflagellate cysts = dinocysts, diatoms) and their applications (sea-ice index PBIP25, modern analogue technique (MAT) transfer functions), provides a chronologically constrained (210Pb, 137Cs, two 14C dates) palaeoenvironmental archive spanning AD 1625–1999 with which to compare and evaluate proxies frequently used in sea-ice reconstructions. Whereas diatoms are rare, PBIP25, biogenic silica and qualitative dinocyst approaches show good agreement, suggesting that palaeo sea-ice histories based on biomarker and microfossil techniques are robust in this region. These combined approaches show fluctuating long open water to marginal ice zone conditions (AD 1625–1740), followed by high-amplitude oscillations between long open water and extended spring/summer sea ice (AD 1740–1870). Greater ice cover (AD 1870–1970) precedes recent reductions in seasonal sea ice (AD 1970–1999). Dinocyst-based MAT, however, produces a low-amplitude signal lacking the nuances of other proxies, with most probable sea-ice reconstructions poorly correlating with biomarker-based histories. Explanations for this disagreement may include limited spatial coverage in the modern dinocyst distribution database for MAT and the broad environmental tolerances of polar dinocysts. Overall, PBIP25 provides the most detailed palaeo sea-ice signal, although its use in a shallow polar archipelago downcore setting poses methodological challenges. This proxy comparison demonstrates the limitations of palaeo sea-ice reconstructions and emphasizes the need for calibration studies tying modern microfossil and biogeochemical proxies to directly measured oceanographic parameters, as a springboard for robust quantitative palaeo studies.