The results of our analyses are presented in Table T2 along with additional geochemical data (Shipboard Scientific Party, 2002) for parameters discussed in this paper. Here we focus on three important results, which are briefly summarized. First, the measured Corg concentrations of Eocene sediments are an order of magnitude lower than modern marine pelagic sediments despite similar sedimentation rates. Second, anomalously high Corg concentrations relative to average values are evident for middle Eocene sections at Site 1218. Third, the downcore trends in organic carbon are not "flat" but display significant variations, both on a weight percentage and mass accumulation rate basis. This observation holds even when anomalously high values are excluded.
In contrast to modern surface sediments of the Pacific equatorial region, organic carbon concentrations throughout the Eocene sediments are lower by an order of magnitude. Corg values of modern sediments range between 0.2 and 0.4 wt% Corg (Lyle, 1988; Lyle, 1992; references in Hammond et al., 1996). Average values at Sites 1218 and 1219 equal 0.03 wt% Corg, with a standard deviation of ±0.03 and ±0.01 wt%, respectively, for each site. Values range between 0.01 and 0.08 wt% Corg for Site 1219 and 0.01 and 0.43 wt% Corg for Site 1218. The median value for both sites equals the mean value of 0.03 wt% Corg. More than 85% (584 of 668) of the samples from Site 1218 contain organic carbon values of 0.05 wt% or less. At Site 1219, 99% of all samples are in this range (n = 352). These average Corg values typify all of the Paleogene samples we analyzed from Leg 199, including Sites 1220 and 1221, reported elsewhere in this volume, and are consistent with the generally low shipboard values reported in the Leg 199 Preliminary Report (Shipboard Scientific Party, 2002).
About 30 samples, or 3% of all samples, exhibit relatively high organic carbon values ranging between 0.10 and 0.28 wt% Corg. Values at the low end of this range (0.10 wt% Corg) are more than three standard deviations greater than the average value for both sites. Values at the high end of the range (0.28 wt% Corg) are an order of magnitude greater than the average value for both sites. Put into perspective, the Corg concentration of this "anomalous" subset of samples is typical modern pelagic sediments. We also observe a higher frequency of anomalies (0.10 wt% Corg) at Site 1218 compared to Site 1219, both in terms of absolute numbers and proportionate to the number of samples from each site. At Site 1218, 25 of 668 samples were above this value vs. 4 of 355 samples at Site 1219. (Note: only 0.005% of all samples we analyzed [5 of 1020] would have been above the detection limit of 0.2% using the shipboard method or Corg analysis!)
We note that the generally low Corg values at both sites still exhibit variability; downcore profiles are not flat, even when the anomalously high values are removed from consideration. For example, Figure F3 shows the changes in organic carbon, calcium carbonate, and biogenic opal on a weight percent basis. Note that Corg is not in phase with calcium carbonate or biogenic silica, and that it tends to lead calcium carbonate. This pattern persists when presented on a carbonate-free or opal-free basis (not shown). However, maxima and minima for all three parameters are coeval when presented on a mass accumulation rate-basis (i.e., the Corg "lead" disappears) (see below).
Positive correlations were found between bulk sedimentation rate and organic carbon weight percent, suggesting that organic carbon preservation is favored by "rapid" burial, hence its removal from the sediment/water interface. The Corg content is independent of bulk MAR at Site 1219 (not shown). In contrast, the inverse trend between these parameters (bulk sediment MAR and wt% Corg) is found for Site 1218 as shown in Figure F4. The maximum Corg value at Site 1218 is 0.45 wt% (vs. 0.08 wt% at Site 1219) and the total MAR is three times greater than at Site 1219 (3.5 g/[cm2 x k.y.]). At Site 1218, high organic carbon values (>0.1 wt%) are associated with low total mass accumulation rates (<0.5 g/[cm2 x k.y]). For MAR values greater than 0.5 g/(cm2 x k.y), samples display low Corg (<0.10 wt%), opposite the relationship observed in modern sediments. An important implication of this graph is that the processes which produced low Corg (0.1 wt%) were operative over the full range of modern mass accumulation rates (i.e., we cannot invoke low MAR to explain the low Corg, especially because we find just the opposite relationship at Site 1218 [low MAR and high Corg]). In Figure F5, the component MARs of organic carbon, calcium carbonate, and biogenic opal are shown for the Eocene and suggest strong coherence between the two productivity signals and the burial flux of Corg. Recent work by Francois et al. (2004) suggests that syndepositional redistribution of sediment by bottom currents (sediment focusing) can significantly affect the mass accumulation rates. Newer work by Lyle et al. (2005) suggests that "sediment focusing" has not obliterated the primary productivity signal in the eastern equatorial Pacific for sediments of the past 18 k.y. years. They demonstrate coherence for the sedimentation parameters both within and between the sediment columns at every scale (i.e., within the sediment column at the centimeter scale down to 500 m depth, and from tens of meters to 10,000 km at the horizontal scale). They also show that the maximum credible redistribution of sediments transported by the fastest deep currents is ~150 km, about 1.5 map degrees. Although sediment focusing may occur as a natural process, it is probably not relevant to our study.
As with the sedimentation rate, the surface area:mass ratio of host phases is recognized as an important factor that controls the organic carbon content of marine sediments (e.g., Weliky, 1983; Hedges and Keil, 1995). As the amount of clay, and clay-sized particles, increases in a sediment assemblage, so does the overall surface area:mass ratio and organic carbon content. To estimate the clay content, or nonbiogenic fraction, we used the simple relationship:
We report that no trend was found for either site between the organic carbon content and the fractional amount of the clay component.