IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.339.204.2018

Results

The revised splice (Fig. F2) covers the complete interval between Cores 339-U1387B-18X and 339-U1387C-8R, which is the last core in the (shipboard) splice and the first core in the single-cored section of Hole U1387C (see the Site U1387 chapter [Expedition 339 Scientists, 2013b]). The revised splice has a corrected mcd (c-mcd) depth scale (i.e., cores are shifted relative to each other without any correction for compression/expansion). Deeper than 171.6 mcd (older than MIS 16), the splice includes 41 transitions, of which only 6 did not need to be adjusted (Table T1). Shifts at the other transitions range from 0.01 to 5.56 m, with just a few having a negative value. For the interval between Cores 339-U1387B-26X and 339-U1387A-29X, the relative shifts published in Voelker et al. (2015) remain the same; the absolute c-mcd values, however, changed due to additional corrections above. Table T2 lists the correction needed to convert mcd to c-mcd values for each core within the revised splice. For Core 339-U1387C-8R, this value exceeds 50 m, highlighting how strongly the splice was modified from the shipboard version.

The revisions of the splice transitions are based on signals in the ln(Fe/Ca) and stable isotope records (Figs. F2, F3), although the weight percent sand data and, in one case, the alkenone-derived sea-surface temperate (SST) data (Voelker et al., 2017) were also considered, if necessary (Table T1). In general, the combined ln(Fe/Ca) and isotopic signals are taken into account. However, if only one such record (or additional data) is used, this is noted in Table T1. In a few cases, the revised splice tie point is based on the benthic foraminifer isotope signal because XRF and δ18O G. bulloides data are missing (not measured/not possible to measure) or inconclusive (e.g., Fig. F2F versus F3E). The benthic foraminifer isotope records are particularly important in the cases where a core needs to be appended to the core catcher of the previous core (Fig. F3). In these cases, it was impossible to generate proxy records that overlap, which is indicated by the comment “short coring gap” in Table T1. Many of these cases are associated with a pronounced benthic δ13C minimum (Fig. F3A–F3D), which is often related to an insolation maximum (e.g., Voelker et al., 2015). The clear trends visible in the benthic δ13C records allow the estimation that most of the coring gaps are minimal (~1–2 samples/12–25 cm/<600 y). An exception is the gap at the Section 339-U1387A-27X-CC to 28X-1 transition (Fig. F3B), which encompasses 2.25 ky (Voelker et al., 2015). Contrary to the shipboard splice, the revised splice also includes core catcher samples, not only in the cases where cores are appended but also as regular tie points (Table T1; e.g., Sections 339-U1387B-28X-CC through 339-U1387A-29X-1).

The shipboard splice incorporates a tie point from Core 339-U1387C-5R to Core 339-U1387B-36X. Neither the XRF nor the stable isotope records of both cores revealed any overlap, although core catcher samples of Core 339-U1387C-5R were included (Fig. F2G). To solve this problem, sections of Core 339-U1387A-36X were analyzed and subsequently incorporated into the splice. Both cores include a prominent contourite layer; therefore, it was easy to splice the two records based on their weight percent records (Fig. F4A), whereby the uppermost level of the weight percent sand maximum (i.e., 408.27 c-mcd) was chosen as a tie point. The planktonic and benthic foraminifer stable isotope records of both cores are also very similar (Fig. F4B, F4C) but highlight, like the weight percent sand data, the apparent lower sedimentation rate in Core 339-U1387C-5R, which is most likely an artifact caused by RCB coring (i.e., some sediment being washed out). With the core catcher of Core 339-U1387A-36X being included in the revised splice, this record reaches a little bit further back in time than the Core 339-U1387C-5R data, allowing the minimization of the coring gap that exists at the Core 339-U1387A-36X to Core 339-U1387B-36X transition (Fig. F3F).

After the study of Site U1385 by Birner et al. (2016) was published, it was obvious that the existing δ18O G. bulloides record of Site U1387 lacked several of the millennial-scale oscillations during MIS 40. In the shipboard splice of Site U1387, the MIS 40 interval is covered by Core 339-U1387C-2R. Because of the previous experience of RCB-drilled cores potentially losing some material (e.g., Core 339-U1387A-36X versus Core 339-U1387C-5R) and thus climate signals, Core 339-U1387A-33X was analyzed completely to verify if this core should be included in the splice instead. The two cores cover the same time interval, but as expected the Core 339-U1387A-33X record shows much more variability than that of Core 339-U1387C-2R (Fig. F2E). So, Core 339-U1387C-2R was replaced by Core 339-U1387A-33X in the revised splice, which is the preferred version of the revised splice. The splice tie point from Core 339-U1387B-32X to Core 339-U1387A-33X is based on perfectly matching benthic isotope records (not shown) because G. bulloides was rare/absent in most samples of the top half of Section 339-U1387A-33X-1 (generating the apparent gap in the splice record shown in Fig. F2E). Because a previous version of the revised splice contained corrections for the two tie points related to Core 339-U1387C-2R, Tables T1 and T2 also provide information on an optional revised splice that includes this core.