IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.339.204.2018

Materials and methods

The analytical measurements were done on the working and archive halves of Site U1387 core. Samples for discrete analyses were taken from the working halves, including core catcher sections, with great care taken to avoid sampling the fillings between the biscuits in the XCB cores of Holes U1387A and U1387B. The average sample resolution is 12–13 cm, although the actual sample-to-sample distance can be between 9 and 17 cm or increase to ~6 cm in the higher resolution interval across Termination X (MIS 21–22 transition). The samples were prepared in the Laboratory for Sedimentology and Micropaleontology of the Marine Geology Division of Instituto Português do Mar e da Atmosfera (IPMA) (formerly at Laboratório Nacional de Energia e Geologia [LNEG]) following the established procedure (i.e., weighing, freeze drying, washing through 63 µm mesh, and drying and weighing of the >63 µm fraction) (Voelker et al., 2015). The weight percent of sand >63 µm was calculated by dividing the dry weight of the washed sample by the weight of the >63 µm fraction.

For stable isotope analysis of Globigerina bulloides shells, 8–12 clean specimens were collected from the >250 µm fraction. The samples were analyzed with Finnigan MAT-251 and MAT-252 mass spectrometers, each coupled to an automated Kiel I carbonate preparation system, at the Center for Marine Environmental Sciences (MARUM; University Bremen, Germany). The mass spectrometers’ long-term precision is ±0.07‰ for δ18O based on repeated analyses of internal (Solnhofen limestone) and external (NBS-19) carbonate standards. Stable isotopes in benthic foraminifer shells of Planulina ariminensis or Cibicidoides pachyderma were analyzed. Samples were either measured at MARUM (majority of samples) or at the Geozentrum Nordbayern (Erlangen, Germany). For further details, see Voelker et al. (2015) because benthic isotope data are used here mostly to evaluate the short coring gaps.

The X-ray fluorescence (XRF) core scanner data were collected every 3 cm downcore over a 1.2 cm2 area with a downcore slit size of 10 mm and three separate runs using generator settings of 10, 30, and 50 kV and a current of 0.2, 1.0, and 1.0 mA, respectively (Bahr et al., 2014; Voelker et al., 2015). Sampling time was 30 s directly at the split-core surface of the archive half using XRF Core Scanner II (AVAATECH Serial Number 2) at MARUM. The split-core surface was covered with a 4 µm thin SPEXCerti Prep ultralene foil to avoid contamination of the XRF measurement unit and desiccation of the sediment. The majority of data reported here have been acquired by a Canberra X-PIPS silicon drift detector (SDD; Model SXD15C-150-500) with 150eV X-ray resolution, the Canberra DAS 1000 digital spectrum analyzer, and an Oxford Instruments 100W Neptune X-ray tube with rhodium (Rh) target material. In January 2015, this tube was replaced by an Oxford Instruments XTF5011 X-ray tube 93057, which affected the analysis of some of the off-shipboard splice sections. The tube substitution had, however, no effect on the ln(Fe/Ca) results shown here (no offsets observed between sections analyzed prior to or after January 2015). The XRF scanner’s raw data spectra were processed by X-ray spectra analysis by a iterative least-square software (WIN AXIL) package from Canberra Eurisys. Following Bahr et al. (2014, 2015), we are presenting the XRF data as the natural logarithm of the Fe/Ca ratio. Small gaps in the ln(Fe/Ca) splice are caused by “missing” sections (i.e., sections that were added into the splice during late stages of the revisions, when XRF measurements had been “completed”).