Comparing sea-ice sediment load with Beaufort Sea shelf deposits: Is entrainment selective?

Modern dispersal of sea-ice-rafted debris (IRD) is important for the Arctic Ocean sediment budget from sources to sinks. Sediment entrainment occurs mainly through the action of small ice crystals (frazil) attaching to sedimentary particles in shallow water, ct mechanism that could be selective. The principal source for entrainment of IRD by suspension freezing into the Beaufort Gyre, western Arctic Ocean, is the adjacent shallow (< 30 m) shelf, here called the source surface. The texture, clay-mineral composition, coarse sand (> 250 mu m) lithology, and carbon and carbonate content of IRD in the Beaufort Gyre were compared to sediments from the probable source surface, in order to determine whether preferential entrainment occurs with any of these sediment parameters. IRD is generally much finer grained than the source surface, showing that silt- and clay-size particles are preferentially entrained by frazil ice, although anchor ice can locally incorporate very high percentages of sand and coarser clasts, The coarsest IRD is also the most poorly sorted. The clay mineralogy of the < 2 mu m IRD fraction is very similar to that of the source surface, indicating no selective entrainment within the clay fraction. The lithology of sand in IRD also matches that of the source surface, although the number of coarse grains is too small (< 100) in most samples for a statistically meaningful count, The average organic-carbon content of IRD is three times higher than that of the source surface, but we attribute this to summer algal growth on ice flees rather than to selective entrainment. A relatively low carbonate content in IRD may be because much of the carbonate in the source is of silt size while about 50% of the IRD measured is clay size. The low carbonate content may also reflect solution under acidic summer conditions on sea ice. Selective export of silt- and clay-size particles by ice rafting from the shallow shelf with time should lead to the formation of a slightly coarser lag, even though same of the dirty ice drops its sediment load in the entrainment area. Further mineralogical and lithological analysis on IRD promises to become a useful tool for the study of sediment dispersal paths by drift ice in the Arctic today and in the past, and also for the study of sources of anthropogenic pollutants found in sea ice.