(Appendix 1-4) Geochemistry, desorption methods, TOC, TC, and mineral content of different settings

Sorption of volatile hydrocarbon gases (VHCs) to marine sediments is a recognized phenomenon that has been investigated in the context of petroleum exploration. However, little is known about the biogeochemistry of sorbed methane and higher VHCs in environments that are not influenced by thermogenic processes. This study evaluated two different extraction protocols for sorbed VHCs, used high pressure equipment to investigate the sorption of methane to pure clay mineral phases, and conducted a geochemical and mineralogical survey of sediment samples from different oceanographic settings and geochemical regimes that are not significantly influenced by thermogenic gas. Extraction of sediments under alkaline conditions yielded higher concentrations of sorbed methane than the established protocol for acidic extraction. Application of alkaline extraction in the environmental survey revealed the presence of substantial amounts of sorbed methane in 374 out of 411 samples (91%). Particularly high amounts, up to 2.1 mmol kg**-1 dry sediment, were recovered from methanogenic sediments. Carbon isotopic compositions of sorbed methane suggested substantial contributions from biogenic sources, both in sulfate-depleted and sulfate-reducing sediments. Carbon isotopic relationships between sorbed and dissolved methane indicate a coupling of the two pools. While our sorption experiments and extraction conditions point to an important role for clay minerals as sorbents, mineralogical analyses of marine sediments suggest that variations in mineral composition are not controlling variations in quantities of sorbed methane. We conclude that the distribution of sorbed methane in sediments is strongly influenced by in situ production.

Description of lithofacies associations and their interpreted settings from sequences of sediment core CRP-1 (Table 2)

Cape Roberts Project drillcore 1 was obtained from Roberts Ridge, a sea-floor high located at 77°S, 16 km offshore from Cape Roberts in western McMurdo Sound, Antarctica. The recovered core is about 147 m long with the upper 43.15 metres below the sea floor (revised figure) being dated as Quarternary and the older part of the sequence being Miocene. The core includes nine facies: sandy diamict, muddy diamict, gravel/conglomerate, mud(stone), clay(stone) and carbonate. These facies occure in associations that are repeated in particulare sequences throughout the core, and are interpreted as representing different depositional environments through time. Seven lithofacies associations are interpreted as representing offshore shelf, ice protected/below wave-base; prodeltaic/offshore shelf; delta front/sandy shelf; ice system; subglacial till/rainout diamict/debris flow diamicts singly or in combination; and a carbonate-rich shelf bank. The facies associations are used to infer that the Quaternary section represents deposition on a polar shelf with perhaps two or three glacial fluctuations. The Quaternary carbonate unit indicates a period of ice sheet retreat, but local glacial activity may have increased with an increase in costal precipitation. The Miocene section represents polythermal glacial systems. The older Miocene section is glacially dominated whereas the younger section is much less so. The glacially dominated section may provide evidence for a major glacial advance thar resulted un a low stand of global eustatic sea level at that time. After the low stand, eustatic sea level was gradually rising during deposition of the younger section dominated more by non-glacial processes.

dD values of C31 n-alkane (dD wax) from sediment core GeoB9501-4 offshore the Northwest African Sahel

Various studies have demonstrated that the stable hydrogen isotopic composition (dD) of terrestrial leaf waxes tracks that of precipitation (dDprecip) both spatially across climate gradients and over a range of different timescales. Yet, reconstructed estimates of dDprecip and corresponding rainfall typically remain largely qualitative, due mainly to uncertainties in plant ecosystem net fractionation, relative humidity, and the stability of the amount effect through time. Here we present dD values of the C31n-alkane (dDwax) from a marine sediment core offshore the Northwest (NW) African Sahel covering the past 100 years and overlapping with the instrumental record of rainfall. We use this record to investigate whether accurate, quantitative estimates of past rainfall can be derived from our dDwax time series. We infer the composition of vegetation (C3/C4) within the continental catchment area by analysis of the stable carbon isotopic composition of the same compounds (d13Cwax), calculated a net ecosystem fractionation factor, and corrected the dDwax time series accordingly to derive dDprecip. Using the present-day relationship between dDprecip and the amount of precipitation in the tropics, we derive quantitative estimates of past precipitation amounts. Our data show that (a) vegetation composition can be inferred from d13Cwax, (b) the calculated net ecosystem fractionation represents a reasonable estimate, and (c) estimated total amounts of rainfall based on dDwax correspond to instrumental records of rainfall. Our study has important implications for future studies aiming to reconstruct rainfall based on dDwax; the combined data presented here demonstrate that it is feasible to infer absolute rainfall amounts from sedimentary dDwax in tandem with d13Cwax in specific depositional settings.

(Table 2) Chlorine geochemistry and mineralogy of altered oceanic crust samples

Bulk chlorine concentrations and chlorine stable isotope compositions were determined for hydrothermally altered basalt (extrusive lavas and sheeted dikes) and gabbro samples (n = 50) from seven DSDP/ODP/IODP drill sites. These altered oceanic crust (AOC) samples span a range of crustal ages, tectonic settings, alteration type, and crustal depth. Bulk chlorine concentrations range from < 0.01 wt.% to 0.09 wt.%. In general, higher chlorine concentrations coincide with an increase in temperature of alteration and amphibole content. d37Cl values of whole rock AOC samples range from -1.4 to +1.8 per mil. High d37Cl values (>=~0.5 per mil) are associated with areas of higher amphibole content. This observation is consistent with theoretical calculations that estimate amphibole should be enriched in 37Cl compared to co-existing fluid. Negative to near zero d37Cl values are found in areas dominated by clay minerals. Chlorine geochemistry is a rough indicator of metamorphic grade and mineralogy. AOC is a major Cl host in the subducting oceanic lithospheric slab. Here we show that bulk chlorine concentrations are ~3 times higher than previous estimates resulting in a greater contribution of Cl to the mantle.