RESULTS AND DISCUSSION

Disseminated organic matter from Leg 184 samples had TOC contents of 0.2-1.1 wt% and ¹³C stable isotope values ranging -19 to -24. One sample from Hole 1143C yielded a ¹³C value of -19.5, indicating that the OM source for this sample is mainly marine. There were 17 samples from Site 1148 that showed ¹³C values between -22.6 and -24.6, in the range between normal marine-sourced and land-sourced OM. Three samples from Hole 1146A, two from Hole 1144A, and one from Hole 1143C yielded similar values (-21.5 to -23.9). Therefore, most of the samples contain some (25%-75%) land-derived OM. Table T1 lists these data. Percent carbon is organic carbon content determined as part of the isotope analysis (i.e., after removal of carbonate). Therefore, the original TOC content of the sample was one-third to one-half as much percent carbon, depending on the carbonate content. Figure F1 shows these data plotted vs. age.

These ¹³C isotope values (-22.6 to -24.6) are in the range reported for marine-derived sediment OM. However, because "pure" marine-sourced OM is usually thought to be in the range of -18 to -21 for tropical oceans, it appears that ~25%-75% of the OM may be land derived (¹³C = -26). Opsahl and Zepp (2001) report riverine dissolved organic carbon (DOC) and Raymond and Bauer (2001) report riverine DOC and particulate organic carbon (POC) that show ¹³C isotope values of -26 to -31, in the same range as most of our "woody" material.

The presence of mixed marine-terrestrial OM in these sediments is supported by the shipboard Rock-Eval data (Shipboard Scientific Party, 2000), which shows borderline type II (marine-sourced OM)/type III (land-plant-sourced OM) values on hydrogen index (HI) vs. oxygen index (OI) plots, as well as by C/N values that often exceed 10. (HI = S₂/TOC, where S₂ = mg of pyrolysis hydrocarbons/g sample; OI = S₃/TOC, where S₃ = mg pyrolysis CO₂/g sample.) Some caution must be used with all these results because of the very low TOC values (and high carbonate carbon values), which make Rock-Eval results less definitive. The very low total N values make the organic C/N values less certain, and it is possible that some N is not organic (i.e., ammonia nitrogen, probably fixed to clay minerals).

There appears to be a trend in the data with age; ¹³C values of older samples are slightly more negative (isotopically lighter). This can be explained by the greater amounts of land-derived organic matter (with lighter isotopic value) present in the older sediment samples. This explanation is reasonable because the amount of land material transported and preserved in the sediment would probably have been greater in the past when the SCS was not as deep or the hole locations were closer to shore or during times of greater weathering and erosion. In addition, microbial sulfate reduction preferentially uses marine OM (Westrich and Berner, 1984); this could in part explain the high amounts of land-sourced OM indicated by the ¹³C isotope (light values) estimate.

Woody Samples

The presence of wood samples in the Leg 184 sediments is rare, as it is in most open-marine deep-sea sediments. It requires unusual events to transport the wood fragments from land via rivers to the ocean, further transport (floating, wind, and ocean currents) 100 km or more offshore, to sink to the seafloor, and be buried in the sediment. Indeed, most Leg 184 shipboard scientists were skeptical that any wood fragments could be found. Several collections of black sediment masses turned out to be sulfides (pyrite and troilite) rather than the sought-after land-sourced OM. After the first find at Site 1143, the search continued and organic-rich concentrations were recovered from sediments from Sites 1144, 1145, 1146, and 1148. Interestingly, the (now) more offshore southern Site 1143 yielded a 5-Ma sample from deep (230 meters below seafloor [mbsf]) sediments, northern SCS Site 1148 yielded a deep (515 mbsf) sample (27 Ma), and Site 1144 yielded a deep (235 mbsf) sample that was very young (0.2 Ma). Basin position, age, and depth do not appear to exclude the presence of woody fragments. However, four of the samples came from <130 mbsf and were young (<1.2 Ma), although only three of the sites had older sediments. The shallowest and youngest samples appeared more peaty, whereas the deeper, older samples appeared more like lignite (brown coal) (N. Bostick, pers. comm., 2000), as expected from organic maturation studies (Bostick, 1979).

Seven samples containing centimeter-size concentrations (wood fragments) of organic matter were measured for ¹³C stable isotope values on the bulk sample, the extracted humic acid, and the residue after extraction. Carbon isotope values of the concentrated OM and woody OM were generally isotopically lighter (depleted in ¹³C) than the bulk sediment OM. Carbon isotope values of the bulk and woody OM ranged from -23.9 to -30.6. Most woody OM showed ¹³C isotope values from -23.5 to -26.5, so some of these samples are isotopically lighter (more negative) than expected for normal land wood. This can be explained by the results of Leavitt and Long (1982, 1986), who report that leaf and twig material are 2-4 lighter (more negative) than wood of junipers and pines. Benner et al. (1987) reported that lignin from leaves and wood is 3-6 lighter than cellulose.

Humic Extracts

Extraction of humic materials was performed on three samples to remove any possible marine-sourced soluble OM that may have been absorbed into or adsorbed onto the wood during its transport from land to the deep sea or during burial in the sediment. For the three samples from which humic acid was extracted/recovered, the ¹³C carbon isotope values are within 0.2 of the bulk sample value (and within the range of replicate analyses). In one case (Section 184-1146C-7H-2), the humic extract was 0.7 lighter (more negative) after extraction than the lignite from which it was extracted (or 1.0 lighter than the unleached sample). The results are listed on Table T2. Figure F2 shows these data plotted vs. age.

Because the age of all but two of the woody samples is <1.5 Ma, it is not possible to give great significance to any isotopic trends with age. Probably no trends would be expected, and, if found, they would be questionable because of the presumably episodic occurrence of the woody materials, which represents unusual occurrences of transport from land, probably during severe ocean storm and/or land rainfall or some other events. It is interesting that woody materials are mainly concentrated/observed during the last 1.5 m.y. (but only three sites contained older sediments).

For samples that have intermediate values between marine and terrestrial OM, the relative amounts of each type can be estimated. Taking -20 for the marine organic matter ¹³C source value (the heaviest SCS value is actually -19.5) and using -26 as the terrigenous end-member:

¹³C of -21.5 = ~75% marine and 25% terrigenous OM.

¹³C of -23.0 = ~50% marine and 50% terrigenous OM.

¹³C of -24.5 = ~25% marine and 75% terrigenous OM.

Of course, it is not this simple because some of the woody samples are lighter than -26 yet can be no more than 100% terrigenous OM. However, as shown by our terrestrial/woody samples, the end-member is probably between -25 and -29. Any assumptions about these end-member values will result in errors in the proportions of the source of the OM. Likewise, samples visually identified as lignite that have values of -24 and -25 are not partly marine OM, as shown by the laboratory chemical extraction of sorbed soluble OM (humic material), which has essentially the same isotope value as the bulk woody sample.

Furthermore, when estimating actual input of sedimentary OM rather than preserved (measured data) sedimentary OM, syngenetic and diagenetic microbial sulfate reduction, which results in FeS and pyrite, must be considered. This mechanism mainly uses the more metabolizable marine OM (Berner and Westrich, 1985). Therefore, the originally deposited OM probably had contained at least 30% more marine OM than now measured, based on the amount of pyrite sulfur present (0.1-0.3 wt%) in most samples (organic C/S = 3; Wang, Prell, Blum, et al., 2000). Thus, the original OM deposited in the sediments was 50%-90% marine OM.