Diatom abundances in sediment core CESAR_83-006

The modern Arctic Ocean is regarded as a barometer of global change and amplifier of global warming (Graversen et al., 2008, doi:10.1038/nature06502) and therefore records of past Arctic change are critical for palaeoclimate reconstruction. Little is known of the state of the Arctic Ocean in the greenhouse period of the Late Cretaceous epoch (65-99 million years ago), yet records from such times may yield important clues to Arctic Ocean behaviour in near-future warmer climates. Here we present a seasonally resolved Cretaceous sedimentary record from the Alpha ridge of the Arctic Ocean. This palaeo-sediment trap provides new insight into the workings of the Cretaceous marine biological carbon pump. Seasonal primary production was dominated by diatom algae but was not related to upwelling as was previously hypothesized (Kitchell and Clark, 1982, doi:10.1016/0031-0182(82)90087-6). Rather, production occurred within a stratified water column, involving specially adapted species in blooms resembling those of the modern North Pacific subtropical gyre (Dore et al., 2008, doi:10.1016/j.pocean.2007.10.002), or those indicated for the Mediterranean sapropels (Kemp et al., 1999, doi:10.1038/18001). With increased CO2 levels and warming currently driving increased stratification in the global ocean (Sarmiento et al., 1998, doi:10.1038/30455), this style of production that is adapted to stratification may become more widespread. Our evidence for seasonal diatom production and flux testify to an ice-free summer, but thin accumulations of terrigenous sediment within the diatom ooze are consistent with the presence of intermittent sea ice in the winter, supporting a wide body of evidence for low temperatures in the Late Cretaceous Arctic Ocean (Falcon-Lang et al., 2004, doi:10.1016/j.palaeo.2004.05.016; Amiot et al., 2004, doi:10.1016/j.epsl.2004.07.015; Otto-Bliesner et al., 2002, doi:10.1029/2001JD000821), rather than recent suggestions of a 15 °C mean annual temperature at this time (Jenkyns et al., 2004, doi:10.1038/nature03143).

Lithostatigraphy of the CESAR cores

Sixteen piston cores and twelve gravity cores were successfully recovered from northern and southern crests of the eastern Alpha Ridge andfrom the Alpha Ridge graben. All but one core contain late Cenozoic muds with variable amounts of sand- to pebble-sized clastic material that probably reflects transport by ice during the past 4-5 Ma. Sixteen Cenozoic-Holocene lithostratigraphic units have been delimited on the basis of sediment texture, structure, colour, detrital carbonate and authigenicferromanganese content. The composition of the upper 13 units in the CESAR cores is similar to the Fletcher's Ice Island cores; hence most units can be broadly correlated over most of the Central Arctic Ocean. Three new lithostratigraphic units (A1-A3) occur at the base of CESAR cores from the northern Alpha Ridge crest. Paleomagnetic and palynological data indicate a Late Miocene-Early Pliocene age for unit A3, which confirms previous reports of a slow sedimentation rate during the Cenozoic. CESAR core 6 was obtained from an erosional surface on top of a fault block at the north edge of the Alpha Ridge graben. This core contains ca. 2m of laminated diatom ooze of Campanian-Maastrichtian age and two ?Paleogene volcanic ash units below a brown mud unit which probably corresponds to units A2 and A3. The biosiliceous ooze contains no foraminifera or silicoflagellates and only few dinoflagellates. There is little difference in biogenic or clastic sediment content between light and dark laminae and the rhythmites do not appear to be annual varves produced in an upwelling environment. The microstructure and fluctuating mineral composition of the laminae most closely resemble those of lami- nated chert beds in the Triassic forearc basins of Japan.

(Table 1) Oxygen isotopes in minerals at DSDP Hole 83-504B basalts

The D/H, 18O/16O and 87Sr/86Sr ratios of the basaltic basement from the Leg 83 section of DSDP Hole 504B show that in that area the oceanic crust has experienced intensive but not pervasive alteration. Isotope ratios of the basalts are very heterogeneous because of an input of oxygen, hydrogen, and strontium from seawater. The hydrogen isotopic composition of many samples displays the complete thermal history of the water-rock interactions. High-temperature mineral formations (actinolites, epidotes, and chlorites) were overgrown by a mineralization at lower temperatures (mixedlayer smectites, iddingsites, and smectites) during successive stages of cooling of the oceanic crust by cold seawater. From 87Sr/86Sr data bulk water/rock ratios up to 5:1 have been calculated. There is evidence that some primary minerals like high-An plagioclases contain oxygen from altered basalts. We have discussed the probability that there existed a seawater/crust interface, now at a depth of 620 m sub-basement, during the high-temperature water/rock interactions. This interface was covered during later magmatism by thick flows, pillow lavas, and intrusives.