The above palaeo¯ux calculations suggest that water-column euxinia could have been achieved even when continued circulation supplied oxygen to deep waters. The trace-metal budget for the studied Pliocene sapropels strongly favours continued circulation during deposition, as this would have provided an extra input of trace metals, required to account for the observed amounts of trace metals in the sapropels. For example, the total amount of nickel sequestered in sapropel i-282 (average of sites 964, 967 and 969), assuming 80% coverage, is 8 3 10 13 g. The water column in a fully stagnant eastern Mediterranean, however, contains only ,2 3 10 12 g Ni (,3%) 13 , and rivers, assuming global average river composition and a present-day discharge, can supply only ,8 3 10 12 g Ni (that is, about 10%) during sapropel formation. Unfortunately, exact river discharges of trace metals during sapropel formation are unknown. A ten-times increase in river water discharge during sapropel formation, however, is highly unrealistic 2,3 . Consequently, it is unlikely that Ni discharge increased to such an extent. Budgets for other trace metals in sapropel i-282 and i-292 yield comparable results 13 . It therefore seems that stagnation of the water column was not as important as generally believed 2,3 . M . MethodsSediment samples (0.5±1 cm resolution) were freeze-dried and ground. Organic-carbon contents (C org ) were determined with an NCS-analyser after carbonate removal with 1M HCl (ref. 13). Sediments were dissolved in HClO 4 ±HNO 3 ±HF; dried residues were dissolved in 1M HCl for analysis with inductively coupled plasma atomic emission spectroscopy (Mo, Ni, V, Ba), Zeeman atomic absorption spectroscopy (Cd), and¯ow-injection analysis system (Se) 13 . Pyritic sulphur (S pyr ) was extracted by Cr(II) reduction 16 . Evolved sulphide was trapped in 1M NaOH and measured polarographically, or trapped as Ag 2 S and measured for sulphur stable-isotope composition (d 34 S pyr ) by combustion isotope ratio monitoring mass spectrometry relative to the V-CDT standard 16 : R 34 S= 32 S, d 34 S pyr ½ 1; 000 3 R pyr =R VÀCDT 2 1. Organic compounds were extracted with methanol/dichloromethane. Elemental sulphur was removed with activated Cu. Apolar fractions were isolated on activated Al 2 O 3 , and analysed by gas chromatography and gas chromatography/ mass spectrometry 30 .

No abstract

Ice observations and sediment collected in a summer transit through the Northwest Passage provide insights on suspension freezing, the most important sediment entrainment mechanism for the Arctic Ocean. No evidence was seen for entrainment by bottom adfreezing, bluff slumping, river flooding, dragging ice keels, or significant eolian transport from land to sea. Lack of eolian sediment loading in the Northwest Passage, together with that already reported for northern Alaska, eliminates wind as an important source for fine sediment in the pack of the Beaufort Gyre and related parts of the Transpolar Drift. Muddy sediment with pebbles and cobbles, algae with holdfasts, ostracodes with appendages, and well-preserved mollusks and sea urchins were collected from two sites in a 50 km long stretch of turbid ice. These materials indicate that suspension freezing reaching to a water depth of 25–30 m during the previous fall was responsible for entrainment. This mechanism requires rapid ice formation in open, shallow water during a freezing storm, when the ocean becomes supercooled, and frazil and anchor ice attach to and ultimately lift sediment and living organisms to the sea surface. The mechanism, already known to be important in the Beaufort Sea, probably also affects wide, shallow Siberian shelves and leads to cross-shelf transport of shallow-water organisms and dropstones with "glacial striations" toward deep basins. This makes distinguishing glacial–interglacial cycles more difficult.

No abstract