IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.344.209.2018

Results and discussion

We present the measurements of dissolved pore fluid REEs in Tables T2, T3, and T4. To illustrate the downhole behavior of dissolved REEs as a series, we selected representative members of the light, middle, and heavy REE groups: La, Dy, and Yb, respectively. Downhole profiles of alkalinity, NH4+, and Ba2+ in the pore fluids are shown to contextualize the dissolved REE profiles with the major geochemical characteristics of the sediment column (see the Expedition 344 summary chapter [Harris et al., 2013a]). Dissolved concentrations of REEs are normalized to PAAS, and representative ratios of (Dy/La)N, (Yb/La)N, and (Yb/Dy)N, where N represents the normalized values, are shown to assess the degree the relative abundances of the light, middle, and heavy REEs to one another. Ratios of other representative members of the light, middle, and heavy REEs did not vary appreciably from the chosen ratios or affect interpretation. Ratios of a composite average of light, middle, and heavy REEs also did not diverge significantly from the interpreted results.

Sites U1378/U1380

REEs measured in pore fluids of the upper slope of Costa Rica show great variability in the upper 200 mbsf, with La ranging from 9 to 162 pM, Dy from 24 to 120 pM, and Yb from 29 to 145 pM (Fig. F2A; Table T2). REEs in this upper 200 mbsf interval also display relatively greater abundances of middle and heavy REEs over light REEs (Fig. F2B). The highest normalized ratios occur at 24 mbsf, where (Dy/La)N is 41, (Yb/La)N is 190, and (Yb/Dy)N is 5. This sample depth also corresponds to the lowest concentrations of La and Dy in the upper 200 mbsf of the sediments. Yb concentration at this depth is 97 pM, which is significantly higher than La and Tb, accounting for the elevated ratios. What accounts for the low concentrations of light and middle REEs in this sample, especially relative to the surrounding samples, is less clear, but we cannot discredit the validity of this signal that is above our detection limits and within the reported error. Comparatively, ratios in the upper 200 mbsf, not including the 24 mbsf sample, average ~8 for (Dy/La)N, ~18 for (Yb/La)N, and ~2 for (Yb/Dy)N. Overall, normalized patterns of dissolved REEs in the upper 200 mbsf interval are characterized as increasing from light to heavy REEs (Fig. F2C).

Deeper than 200 mbsf, REE concentrations distinctly decrease with depth in the core. In this interval, La decreases from ~100 to ~10 pM, Dy from ~10 to ~4 pM, and Yb from ~50 to ~2 pM (Fig. F2A). The relative abundances of middle and heavy REEs over light REEs are elevated but less so than in the upper 200 mbsf interval (Fig. F2B). Normalized ratios average ~5 for (Dy/La)N, ~9 for (Yb/La)N, and ~2 for (Yb/Dy)N. Compared to the REE patterns of the upper 200 mbsf, the lower depths in the sediment at the slope site are relatively flat (Fig. F2C).

Although there is no discernible correlation between REEs and NH4+ or Ba2+ throughout the entire core, a relatively strong relationship exists between REEs and alkalinity, particularly in the lower depths of the core (Fig. F3). For pore fluid samples deeper than 200 mbsf, the R2 values for a linear regression between REEs and alkalinity are 0.8886, 0.9807, and 0.9554 for La, Dy, and Yb, respectively. If we extrapolate these linear relationships to the interval shallower than 200 mbsf, the data show a net deficit relative to the alkalinity for La and Tb in a number of samples. By contrast, extrapolating the linear relationship for Yb shows a slight net surplus in the upper 200 mbsf.

Site U1381

On the incoming plate at Site U1381, the discernible characteristic of dissolved REEs in pore fluids is localized maxima in concentrations at ~20 mbsf for La, Dy, and Yb and increases, particularly in La concentrations, deeper than 80 mbsf (Fig. F4A; Table T3). In the uppermost pore fluid sample from the upper 2 mbsf of the sediment, concentrations of light and middle REEs are elevated, with La = 192 pM, Dy = 43 pM, and Yb = 36 pM. Comparatively, in samples from 8 and 14 mbsf dissolved concentrations of La, Dy, and Yb average ~60, ~28, and ~46 pM, respectively. These dissolved REE concentrations reach local maxima of 193, 76, and 74 for La, Tb, and Yb, respectively, at ~20 mbsf before decreasing to averages of ~60, ~17, and ~20 pM between 40 and 80 mbsf. Deeper than 80 mbsf, concentrations of light and middle REEs are particularly elevated, averaging 273, 38, and 26 pM for La, Tb, and Yb, respectively, for the three samples measured in this interval.

Normalized REE ratios generally show greater abundances of middle and heavy REEs relative to light REEs throughout the core (Fig. F4B). These ratios are relatively constant throughout the upper 80 mbsf, averaging ~3, ~8, and ~2 for (Dy/La)N, (Yb/La)N, and (Yb/Dy)N, respectively. An exception is the sample at 50 mbsf, which shows elevated ratios of 10, 41, and 4 for (Dy/La)N, (Yb/La)N, and (Yb/Dy)N, respectively. These elevated values may be an artifact of the relatively low concentrations of REEs because this particular sample records the lowest values for the entire core. For the three samples deeper than 80 mbsf, normalized ratios are the lowest in the core, averaging 2, 2, and 1 for (Dy/La)N, (Yb/La)N, and (Yb/Dy)N. These ratios are reflected in the overall normalized REE patterns, with the deepest samples plotting below samples from the upper depths of the sediment (Fig. F4C). By comparison, REEs measured in the framework wedge of Site U1381 exhibit concentrations that are orders of magnitude greater than in the pore fluids and show moderate anomalies in the Eu signal and a relatively greater abundance of heavy REEs over light REEs (see Yan and Shi, 2016) (Fig. F5). Unlike dissolved REEs in the pore fluids of slope Sites U1378/U1380, no observable correlation exists between REEs and alkalinity at the incoming plate Site U1381.

Site U1414

At the second site on the incoming plate, Site U1414, dissolved REEs in the pore fluids are characterized by a general peak in concentrations in the upper 75 mbsf of the sediment and relatively low, constant values deeper than 75 mbsf (Fig. F6A, Table T4). In this upper 75 mbsf, REEs reach maximum values of 834, 223, and 207 pM for La, Dy, and Yb, respectively. Deeper than 75 mbsf, REE profiles are fairly constant, averaging ~25, ~10, and ~20 pM for La, Dy, and Yb. The exception to this profile are two samples at ~200 and ~250 mbsf, where La is 197 and 108 pM, respectively, Dy is 37 and 20 pM, and Yb is 38 and 20 pM. Normalized REE ratios indicate a greater abundance of heavy REEs, particularly centered around 100 mbsf (Fig. F6B). In general, the deepest pore fluid samples in this core have the lowest normalized ratios and the flattest overall patterns, which is consistent with REEs at Sites U1381 and U1378/U1380 (Fig. F6C). REEs measured in the framework wedge underlying these sediments are consistent with the framework wedge at Site U1381, showing orders of magnitude greater concentrations than the pore fluids, with moderate anomalies in the Eu signal and relatively greater abundances of the heavy compared to the light REEs (see Yan and Shi, 2016) (Fig. F5).

REEs at Site U1414 show a general positive correlation with pore fluid alkalinity, although linear regressions are relatively weak, with R2 values of 0.418, 0.529, and 0.672 for alkalinity correlations with La, Dy, and Yb, respectively (Fig. F7). Pore fluid REEs also show a slight positive correlation with dissolved Ba2+, except in the Ba2+ peak deeper than 300 mbsf (Fig. F8). For the interval not including the Ba2+ peak, the R2 values for a linear regression between Ba2+ and REEs are 0.4874, 0.6068, and 0.5611 for La, Dy, and Yb, respectively. The concentrations of Ba2+ in this interval, however, is relatively low, with values between 0.5 and ~2 µM.