The concentration of methane found by headspace (HS) analysis in sediments at Site 1144 was generally high, varying by more than an order of magnitude between 3,000 and >60,000 ppmv. Traces of ethane (<2 ppmv) were found in just three samples between 375 and 420 mcd, associated with the highest methane concentrations at depth. Carbonate concentrations were generally low, increasing at 180 mcd from 10-15 wt% in the upper interval to between 10 and 23 wt% in the lower interval, accompanied by greater amplitude fluctuations. Total organic carbon (TOC) obtained by difference (total carbon [TC] - inorganic carbon [IC]) decreases systematically from >1% above 50 mcd to <0.5% toward the base (below 420 mcd). Most sediments exhibit low C/N values, suggesting a predominantly marine organic source. Total chlorin abundance was measured to 316 mcd. Bulk concentrations exhibit strong fluctuations indicative of glacial-interglacial changes.
Headspace gas analysis was performed on every core from Hole 1144A, usually taken midcore. Analysis was conducted as described in "Organic Geochemistry" in the "Explanatory Notes" chapter. Concentrations of methane rise rapidly in the first three cores (to 20 mcd) (see Table T10; Fig. F18); this is related to the depletion of dissolved sulfate to essentially zero (<0.5 mM) in interstitial waters within the first core (see "Inorganic Geochemistry"). Such a relationship has been observed frequently (e.g., Kvenvolden and Kastner, 1990; Stein et al., 1995) and is attributed predominantly to microbial generation of methane below the interval of sulfate reduction. Sediment methane concentrations decline fairly uniformly from 20 to 200 mcd (Fig. F18), apparently reflecting decreasing organic matter (OM) availability (Fig. F19). A marked increase in methane is observed between 220 and 420 mcd, especially below 370 mcd. This is accompanied by traces of ethane in Cores 184-1144A-37X, 39X, and 40X. However, the C1/C2 ratio never drops below 50,000 because of the high methane concentrations. Methane rapidly decreases below 420 mcd. Without a strong thermal gradient, it is hard to explain the minimum and secondary peak in CH4 with depth. Dissolved sulfate measurements indicate that sulfate reduction is complete below 10 mcd. Density shows a marked increase and porosity a corresponding decrease at 420 mcd (see "Physical Properties"), indicating a change in the sediment. This reduction in pore space reduces the volume available for hydrocarbons and may account for the low methane levels in headspace samples at the base; any free gas generated may have escaped during XCB coring. Organic carbon abundance is also lowest in this deep zone. However, the increase in methane with depth between 220 and 420 mcd remains unexplained; a conclusion implicating increasing organic maturity awaits further postcruise analysis.
Soon after core recovery, the sediment was inspected through the transparent core liner for the presence of gas voids, from which gas was extracted using a steel piercing tool and gas-tight syringe. Voids were sampled in Cores 184-1144A-3H through 35X; below this level, voids were too scarce to routinely sample. Samples were analyzed as described in "Organic Geochemistry" in the "Explanatory Notes" chapter. Such void-space (VS) concentrations of methane as analyzed by the natural gas analyzer are approximately an order of magnitude higher than those in HS samples (Table T10; Fig. F18). Concentrations range from 23%-79% by volume with one anomalously low value from Core 184-1144A-23H, which was apparently contaminated by air during collection. The voids also contain 0.5% to 1.5% CO2 and as much as several percent N2 (after correction for air N2 using the O2 values). The methane data has little trend, although higher values between 190 and 290 mcd might indicate more complete sediment degassing before sampling for headspace analysis (Fig. F18). This is possibly the result of a change in drilling techniques from APC to XCB drilling, which routinely recovers more compacted sediment and less in situ free gas.
Three samples per core were analyzed for inorganic carbon, and at least one sample per core for total carbon, total nitrogen (TN), and total sulfur from Hole 1144A (Table T11). Carbonate content varies from 8.6% to 23.5 wt%, with lower average values and amplitudes above 175 mcd (average [AV] of 12.0% and standard deviation [SD] of 1.4), increasing to an AV of 15.1% and SD of 2.9 between 175 and 310 mcd. Values again decrease to ~10%-17% (AV of 13.7%; SD of 2.0) between 310 and 485 mcd until a final further increase below ~425 mcd to >20% (AV of 18.9%; SD of 3.3) (Fig. F19). The marked fluctuations in carbonate concentration between 175 and 310 mcd suggest a cyclic input of carbonate. Further, coulometer reaction times for >99.9% (complete) reaction of carbonate increased over this interval. This may indicate a lithologic change related to increased proportion of dolomite or increased biogenic carbonate.
The TOC concentration by difference (TC - IC) was determined for three samples per core to Core 184-1144A-12H and one sample per core below that. TOC values are higher (0.5%-1.3%; AV of 0.90%; SD of 0.17) above 195 mcd than below (AV of 0.56%; SD of 0.15%). Marked peaks (Samples 184-1144A-2H-5, 107-108 cm; and 12J-3, 107-108 cm) appear to correlate with a tentative assignment of oxygen isotope stage boundaries 2/1 and 6/5e (i.e., Termination Events I and II) and may well relate to pulsed input from allocthonous organic carbon remobilized by transgression of the continental shelf and/or autochthonous phytoplankton production (e.g., Harris et al., 1996) (see "Chlorin Analysis"). Sample resolution below 180 mcd is insufficient to explore this hypothesis. Concentrations decline steadily below 190 mcd and are especially low below 420 mcd (<0.5%) (Fig. F19). Rock-Eval TOC values are systematically 10% to 30% lower than those calculated by difference (Fig. F20). The decreasing TOC trend with depth downhole is similar to that seen in benthic foraminifers (see "Biostratigraphy").
Sulfur values are listed in Table T11 and range from 0.75% to 0.1%. We assume this sulfur occurs as pyrite, which is frequently observed in the cores (see "Lithostratigraphy"). Ten samples were noted to have visible pyrite in split-core sections during sampling or during carbonate sample preparation. Of these observed sulfide occurrences, approximately half had elevated S contents compared with adjacent samples without visible sulfide. Unlike OM, which is present as ubiquitous, fine-sized material, the sulfide occurs as discrete, irregular concentrations that may or may not be present in the CARB sample analyzed for total S. Six samples gave 0% S values after the detector began to yield unreliable S results (see "Organic Geochemistry" in the "Site 1143" chapter) and were subsequently re-analyzed (Table T11). The high values for the samples above Core 184-1144A-15H also correspond to the samples with higher organic carbon as expected from syngenetic sulfate reduction and organic carbon-sulfur systematics (Berner, 1984).
Our analytical methods allow us to chemically characterize the organic matter in several ways. Ratios of TOC to TN (C/N ratio) range from ~3 to ~9 at Site 1144 (Table T11; Fig. F19). However, these values appear too low and may result from residual NH4+ adsorbed on to, or absorbed within, the clay minerals (see "Organic Geochemistry" in the "Site 1145" chapter). Figure F21 shows the results for TN vs. TOC by difference. Average C/N values of marine plankton are between 5 and 8, whereas higher land plants have ratios >20. These values, however, may change during diagenesis and thermal maturation (Emerson and Hedges, 1988). Therefore, it is unclear whether a correction for clay-bound nitrogen is applicable. A woody fragment, recovered from Core 184-1144B-24H, can be used as a point of reference for terrestrial organic matter (Tables T11, T12). Overall, the C/N values exhibit no downhole trend, except that the amplitude increases downcore. Two higher C/N values (Sections 184-1144A-38X-1 and 40X-1) do not appear to correlate with any significant changes in TOC or any unusual core observations.
Examination of Rock-Eval S3 and oxygen index (OI) data (Table T12) indicates that they are not reliable for the first set of analyses. Thus, although the OI data are mainly indicative of type III terrestrial OM, we do not consider them to be definitive. The OI values are always a problem at low TOC and with immature OM. We have rerun several samples and find that the OI is indeed variable, but the results for Tmax (temperature of maximum release; see "Organic Geochemistry" in the "Explanatory Notes" chapter), TOC, and HI (hydrogen index) are repeatable and consistent. An additional problem arises in sediments containing such relatively low TOC content because of adsorption of pyrolyzates (S2 values leading to low HI values) to clays (Peters, 1986).
Total chlorin abundance was measured between 0 and 316 mcd at a sampling interval of two samples per section (Table T13; Fig. F22A). Analytical methods are discussed in "Organic Geochemistry" in the "Explanatory Notes" chapter. The concentration of pigments determined by solvent extraction and fluorescence spectrophotometry could not be quantified precisely because of the absence of a known standard but is expressed in relative absorbance units per gram dry weight of sediment. Total chlorin abundance varies by more than an order of magnitude (0.34-4.27 units) and exhibits a general decreasing trend downhole. This trend is consistent with that seen in TOC measurements (Fig. F22B) at various sites during Leg 184 and may reflect a dilution signal seen as a corresponding increase in carbonate abundance related to the evolution and deepening of the South China Sea (Fig. F19). However, an alternative explanation requiring diagenetic transformation of chlorins over time awaits further shore-based research.
The distribution of chlorins downhole shows some marked similarities to that of TOC. This confirms the conclusion of previous studies relating chlorin abundance to organic input to sediments (e.g., Harris et al., 1996). However, there are some notable discrepancies within the two data sets. Several marked maxima in TOC abundance do not correspond to high chlorin concentrations (e.g., 95-105 mcd and 150-180 mcd). It is suggested that these represent periods of increased nonmarine OM enrichment and coincide with elevated C/N ratio values. Comparison with NGR data indicates that maxima in NGR counts, indicative of higher clay content and (by inference) increased allocthonous input, coincide with such increases in TOC (Fig. F22C). Discrepancies also occur in the exact timing of absolute maxima in chlorins and TOC, which we tentatively conclude coincide with Termination Events (Termination I, 5 mcd; Termination II, 117 mcd; Termination III, 209 mcd; Termination IV, 252 mcd; and Termination V, 309 mcd). The resolution of the TOC data in the upper 180 mcd permits us to conclude that during Terminations I and II, maximum TOC abundance occurs before that in chlorins. This may result from the input of allocthonous OM reworked from the continental shelf during rising sea levels before a climatic and nutrient-related effect on marine phytoplankton productivity. The general pattern observed in NGR data suggests a reverse correlation with chlorin abundance, prompting our conclusion that chlorins are a proxy for autochthonous marine production at Site 1144.