ODP Site 1229 is located on the Peru shelf at ~150 m water depth (Fig. F1) and in the immediate vicinity of ODP Leg 112 Site 681 (Suess and von Huene, 1988). During Leg 201, sediments from Hole 1229E were sampled at 10-cm intervals from 0 to ~700 cm below seafloor (cmbsf). Lithologic Subunit IA (0–40.7 meters below seafloor [mbsf]) is characterized by alternations of olive-green well-laminated diatom ooze and clay-rich silt and clay- and silt-rich diatom ooze (Shipboard Scientific Party, 2003). Shipboard chemical measurements of pore water showed a pronounced anomaly between 100 and 300 cmbsf (Shipboard Scientific Party, 2003) consisting of brief maxima in alkalinity, dissolved inorganic carbon (DIC), ammonium, and sulfide coinciding with a brief negative excursion in dissolved sulfate. The same anomaly is also apparent in the ammonium and alkalinity profiles of ODP Site 681 (Suess, von Huene, et al., 1988). Based on radiocarbon analyses, Fink et al. (2006) reconstructed the sedimentation chronology for the uppermost section of Hole 1229E (~400 cm). Their results suggest a ~2-m section at the top of the hole that represents rapid and continuous accumulation at this site during the late Holocene (~100 cm/k.y.). Deposition during the middle part of the Holocene (2–8 ka), however, seems to have been very slow (~10–20 cm/k.y.) and results in a reduction of the thickness between the uppermost Holocene sediments (0–200 cmbsf) and the Pleistocene/Holocene boundary (~380 cmbsf). To assist our interpretations in characterizing the processes responsible for the observed fluctuations in geochemical parameters, the 14C dates from these authors (Fink et al., 2006) have been incorporated on all vertical profiles used in the present study.
Immediately after shipboard sampling, sediments were frozen and shipped under ice to Corpus Christi, Texas (USA), where they were freeze-dried and subsequently homogenized. Weight percentage of organic carbon (Corg) and total nitrogen (TN) were determined using an automated ANCA-SL elemental analyzer coupled with a PDZ-Europa 20-20 continuous-flow isotope mass spectrometer (EA-CFMS). Carbonate was removed prior to analysis by vapor-phase acidification with HCl for 24 hr followed by drying at 40°C for 24 hr (Hedges and Stern, 1984; Harris et al., 2001). During all sample runs, two to three quantitative standards (glucosamine) were included after every four to five samples. The average precision of elemental analyses determined from replicate analyses of selected samples is ±2%–5% for both Corg and TN.
Stable isotopic signatures of organic carbon (
13Corg) were determined on preacidified samples using the same automated EA-CFMS mentioned above. Details of the procedure are provided in Fry et al. (1992). The continuous-flow mode minimizes the sample size needed for analysis, and samples as small as 2 mM C can be analyzed with a precision of 0.1
, thus greatly reducing analytical effort. Isotopic compositions are reported in 13C notation and are referenced to Peedee belemnite (PDB) standard. The 13C notation is defined as follows:
13C = {[(13C – 12C)sample/(13C – 12C)standard] – 1} x 1000. (1)
Alkaline CuO oxidation (Hedges and Ertel, 1982) was performed on intervals of Hole 1229E to obtain a suite of biomarkers, which include phenolic compounds derived from the lignin biopolymer (an unambiguous biomarker for vascular plants). Because of time and cost constraints, CuO oxidation analyses were only performed on samples from the surface to ~450 cmbsf. The deepest section of this record covers the Pleistocene–Holocene transgression and offers information on potential changes in OM source fluctuations across this boundary. The CuO oxidation method has undergone revisions and adaptations since its initial conception in the early 1980s (Hedges and Ertel, 1982), which have led to "cleaner" chemistry, change and decrease in solvent utilization, and increased sample throughput (Goñi and Hedges, 1992; Louchouarn et al., 2000; Dalzell et al., 2005). Briefly, a sediment amount providing 2–4 mg Corg (Louchouarn et al., 2000) is oxidized under alkaline conditions with CuO at 155°C for 3 hr in pressurized stainless steel mini-reaction vessels (3 mL; Prime Focus Inc.). The aqueous solution is then acidified with 6-N HCl and extracted three times with ethyl acetate. Extracts are dried with Na2SO4 and then evaporated to dryness using a LabConco solvent concentrator. The CuO reaction products are redissolved in a small volume of pyridine (200–500 µL) and a subsample is derivatized with bis-trimethylsilyl trifluorocaetamide (BSTFA) containing 1% trimethylchlorosilane (TMCS).
Separation and quantification of trimethylsilyl derivatives of CuO oxidation byproducts were performed by gas chromatography–mass spectrometry (GC-MS) on a Varian quadrupole GC-MS system (3800/1200 L) fitted with a fused capillary column (CP-Sil 8 CB/MS, 60 m x 0.25 mm inner diameter; Varian Inc.). Each sample was injected, under splitless injection mode, into a straight glass liner inserted into the GC injection port, and He was used as the carrier gas (~1.3 mL/min). The GC oven was temperature-programmed from 100°C, with no initial delay, to 300°C at 4°C/min and held constant at the upper temperature for 10 min. The GC injector and GC/MS interface were both maintained at 300°C. The mass spectrometer was operated in the electron impact mode (EI, 70 eV), in scan mode. Compound identification was performed using column retention times and by comparing the full spectra of each sample to those produced by commercially available standards. Quantification was performed using relative response factors adjusted to trans-cinnamic acid (CnAd) as the internal standard. Replicate analyses of standard estuarine sediments (i.e., National Institute of Standards [NIST] standard reference material [SRM] 1944; N = 6) showed that the analytical precision of the major CuO oxidation products and related parameters averaged ~5%. The yields and composition parameters produced in this study match those obtained with both microwave digestion (Goñi and Montgomery, 2000; Houel et al., 2006) and with larger reaction vessels with conventional oven procedure (Louchouarn et al., 2000). The average standard deviation obtained from replicate analyses of ODP samples was somewhat higher (15% ± 10%; range = 3%–30%). This is likely caused by very low levels of terrigenous organic matter (see below), which generate higher variability during quantification (cf. Visser et al., 2004).
