Fluvial transport of weathering products of continental rocks and soils is generally agreed to be the main source of REE into the ocean (Holser, 1997). Relative to the REE composition of shale, REE content of seawater is enriched in heavy rare earth elements (Elderfield and Greaves, 1982; Piepgras and Jacobsen, 1992). The phenomenon is attributed by Elderfield and Greaves (1982) to the greater stability of HREE complexes because of their smaller ionic radii and their enhanced interactivity with anions. Whether this hypothesis helps account for the pattern evident in Site 1014 sediments is a moot point. The surprising fact that the REE composition of these samples roughly mirrors that of seawater (despite orders of magnitude differences in concentration of REE in the two media) may simply result from the bias in the sediments toward the biogenic origin of the major initial particulate constituents.
The presence of a distinct negative Ce anomaly in the Site 1014 sediments reflects an important characteristic feature of seawater. In modern seawater, the anomaly is attributed to the dominant oxidation to Ce4+ and its incorporation into Mn and/or Fe oxyhydrides (Shaw and Wasserburg, 1985) or to its enrichment in authigenic minerals. Rare earth elements are also incorporated into phosphate grains from seawater during the transportation and deposition stages (Ilyin, 1998). A prominent feature of various ancient and modern sedimentary environments, strongly negative Ce anomalies, and—less commonly—zero and positive Ce anomalies associated with local anoxia (the Japan Trench presumably serves as an example of the latter [Lerche and Nozaki, 1998]) used to be regarded as a reliable indicator of redox conditions in water masses and associated sediments (e.g., Wright-Clark and Holser, 1981; Wright et al., 1987). However, the REE composition in sedimentary rocks is now considered to be complicated not only by a large range of primary effects but also by diagenetic effects (Holser, 1997). At Site 1014, diagenetic effects include bioturbation, pyritization, incipient organic carbon maturation, biogenic opal dissolution, and possibly various authigenic mineral reactions. But to what extent has the REE signature been diagenetically overprinted?
One well-known constraint (Wright et al., 1987) on REE distribution is their tendency to accrete in phosphatic remains during early diagenesis. Thus, pelagic sediments of the equatorial Pacific are reported by Toyoda et al. (1990) to have high P content and correspondingly high REE. In Site 1014 sediments, phosphatic debris is sparse, and a low P content (0.002%-0.139%) is accompanied by low total REE. However, P content reportedly (Shipboard Scientific Party, 1997) decreases significantly at the base of Subunit IB (below 340 mbsf), whereas over this interval, REE content registers an increase and therefore cannot be strictly resident in phosphatic components.
The REE content is better accounted for by uptake in the seawater column by calcareous fossils and siliciclastic clays, organic matter, and/or by detritals (Sholkovitz et al., 1994). Under strictly diagenetic conditions, even the precipitation of calcite within microburrows (e.g., chondrites, etc.), and bioturbation generally, can be expected to have contributed to the trace element signature. Thus, acquisition of REE as well as other elements such as uranium from pore waters was probably an important factor. Plots of La/Lu and Ce* vs. depth for eight samples (Fig. 3) in Hole 1014A appear to vary sympathetically and reflect the general enrichment in LREE; however, these patterns reveal very little about possible changes in pore-water conditions (Palmer, 1985; German and Elderfield, 1990). At Site 1014, the uranium content ranges from 4.3 to 25 ppm, far in excess of the low abundances (~0.08 ppm) reported by Liu et al. (1988) for many marine carbonates.
A distinctly negative Eu anomaly present in the chondrite-normalized plot (not shown) of all samples is, as indicated earlier, far less evident in the NASC-normalized plot (Fig. 1). The reason for this is not clear. However, the Eu/Sm ratio provides a measure of the Eu anomaly and ranges from 0.19 to 0.46 in Site 1014 sediments. If diagenesis was a factor, this ratio perhaps ought to be of a more uniform value—closer, say, to the NASC value of 0.22 (Chaudhuri and Cullers, 1979).
Results of experimental work by Aja (1998) and studies conducted by Zhao et al. (1992) leave little doubt that REE mobilization may occur during diagenesis and that adsorption by clay minerals plays an important role in such fractionation processes. In this respect, the clays at Site 1014, although of primarily authigenic origin (and so, at best, only faintly representative of provenance), will have contributed to REE mobilization. Note, however, that the Quaternary clay-rich Subunit IA does not show the maximum enrichment in REE (see Fig. 2); there is also a distinct tendency for CaCO3 and total REE to vary antithetically. Thus, by process of elimination, the REE are linked to the siliciclastic clay component and probably also to biogenic opal dissolution. The latter is a classic diagenetic process common toward the base of the sequence at Site 1014 and likely elsewhere in the California Borderland at these depths.