Sulfur-isotope analyses of gypsium samples from the Southeast Greenland margin, southern Alaska and Gulf of Alaska (Table 1)

Gypsum grains were identified in Miocene-Pleistocene sediment cores from two deep-water ODP sites, Site 918 off the SE Greenland margin and Site 887 in the Gulf of Alaska, and in Holocene sediment cores from shallow-water localities in Disenchantment Bay and Muir Inlet in southern Alaska. Although initial morphologic and textural observations suggested a complex system in which the gypsum may have had more than one origin, quantitative sulfur isotope analyses of the gypsum provide evidence of its detrital nature. d34S values in gypsum from southern Alaska range between +0.0 and +7.1 per mil. Gypsum has d34S values between -27.1 and -27.5 per mil in the Gulf of Alaska and values between -28.5 and +0.2 per mil off the SE Greenland margin. All of these isotopic signatures are too highly depleted in d34S to have precipitated from seawater, present or past. In addition there is no significant change in d34S values for gypsum crystals with differing physical characteristics (abraded vs. unabraded) from the same stratigraphic horizon, suggesting all the gypsum is detrital regardless of the degree of abrasion. The isotopic and physical evidence, in combination with the onshore geology the environmental setting, and site characteristics of the gypsum-bearing marine localities, lead us to propose that the ultimate source of the gypsum is precipitation from freeze-induced terrestrial sediment or soil brines. Furthermore the combined evidence suggests that the subsequent occurrence of gypsum in glacimarine sediments results from ice-rafting (by icebergs or sea ice) of the frozen regolith and/or, in the proximal glacimarine setting of southern Alaska, very rapid burial via turbidity currents.

Sedimentology of the Reykjanes Ridge, Southeast Greenland

Marine records from the Reykjanes Ridge indicate ice sheet variations and abrupt climate changes. One of these records, ice-rafted detritus (IRD), serves as a proxy for iceberg discharges that probably indicates ice sheet fluctuations. The IRD records suggest that iceberg discharge 68,000-10,000 yr B.P. happened more frequently than the 7000- to 10,000-yr spacing of the Heinrich events. An IRD peak 67,000 to 63,000 yr B.P. further suggests that the Middle Weichselian glaciation started about 12,000 yr earlier in the North Atlantic than in the Norwegian Sea. Several later IRD events, in contrast, correlate with Norwegian Sea IRD-rich layers and imply coeval ice sheet advances in the North Atlantic and the Norwegian Sea. Coccoliths in a core from the Reykjanes Ridge show distinct peaks in species that record occasional inflow of warm surface water during the last glaciation, as previously reported from the eastern Labrador Sea. High abundances of coccoliths, together with a decrease ofNeogloboquadrina pachydermasin. and relatively low delta18O values, imply enhanced advection of the North Atlantic Current 69,000-67,000 yr B.P., 56,000-54,000 yr B.P., 35,000-33,000 yr B.P., and 26,000-23,000 yr B.P. This advection provided a regional moisture source for extension of ice sheets onto the shelf. In contrast, most of the IRD events are characterized by cold polar surface water masses indicating rapid variations in ocean surface conditions.

Growth rates of Fe-Mn nodules and age and sedimentation rates of host sediments from the Southeast Pacific

Growth rates of nine ferromanganese nodules collected from the Southeast Pacific were estimated using the alpha radiogpaphic technigue. Growth rates range from 1 to 16 mm per million years. In three nodules measurements were made on two opposite sides; two of them showed no growth in one of measured directions during the last 300 ky, whereas in the third nodule growth rates on the opposite sides differ by factor 2. Average sedimentation rate of deposits underlying the nodules estimated by the radiocarbon and excess 230Th methods, were 4 mm/1000 years with rather minor variations. Difference between sedimentation rates and nodule growth rates is caused by activity of benthic fauna, as suggested by inversion of radiocarbon dates with depth.