High-resolution analyses of mineral and major element composition including org-C and carb-C were conducted for the latest Pleistocene to Holocene sediments from ODP Site 1017 covering the last 80 k.y. Compositional variations associated with glacial/interglacial as well as millennial-scale changes in paleoceanographic conditions are identified at the intermediate depth of the California margin off Point Conception. Significant variations in mineral and major element composition were detected in both time scales, which were semiquantified subjectively by Q-mode factor analysis of 11 major elements. Five factors were extracted, which together explain more than 99% of variance. To characterize these five factors, correlation was examined between the factors and mineral composition. We also conducted grain-size analysis of selected samples to characterize grain size of these factors.
Factor 1 has high scores of Fe2O3, MgO, CaO, and LOI and represents enrichment of fine-silt- to clay-size amorphous material (including organic matter), calcite, and smectite. Factor 2 has high scores of SiO2, Na2O, and K2O and represents enrichment of coarse-silt- (and sand) size quartz and feldspar. Factor 3 has high scores of SiO2, CaO, and P2O5 and represents enrichment of coarse-silt- to sand-size bioclasts and quartz in the case of turbidite sand, and fine-silt- to clay-size autochthonous biogenic carbonate in the case of hemipelagic sediments. It is likely that factor 3 in hemipelagic sediments represents the degree of carbonate preservation and/or productivity. Factor 4 has high scores of TiO2, MnO, and P2O5 and possibly represents enrichment of titanomagnetite-like mineral. This factor may represent a size-sorting effect during transportation in the case of turbidite sands and a possible northern source of detrital material in the case of hemipelagic sediments. Factor 5 has high scores of Fe2O3, MgO, and K2O and probably represents enrichment of glauconite. It is likely that glauconite in hemipelagic sediments is authigenic and could be related to the bottom-edge position of OMZ.
Grain-size distribution of hemipelagic sediments is well expressed by factors 1 and 2, which explains most of their compositional variations. Temporal variations of factor 1 (representing fine silt and clay) show millennial-scale as well as orbital-scale variations with higher loading (finer grain size) during warm periods and lower loading (coarser grain size) during cold periods. The intervals characterized by lower factor 1 loading coincide with faintly laminated intervals. Together with their principal grain sizes of 13-14 µm, these faintly laminated intervals are considered as contourites. These results suggest that bottom-current intensity at Site 1017 (water depth of 956 m) changes in concert with stadial/interstadial as well as glacial/interglacial climatic cycles during the last 80 k.y. with periods of stronger current intensities coinciding with cooler periods and vice versa. Relative bottom-water oxygen level inferred from DOPT suggests that bottom water was better ventilated during cold periods when contour current intensity was stronger. This evidence argues for the hypothesis by Kennett and Ingram (1995) and Behl and Kennett (1996), which suggested the possibility of increased ventilation of NPIW during the Younger Dryas and stadials during the last glacial period. This, in turn, suggests intensification of intermediate-water circulation in the North Pacific during stadials of the last glacial episode, possibly associated with Dansgaard-Oeschger cycles.
Our result also suggests that org-C and carb-C, which are occasionally used as indicators of surface productivity, are strongly influenced by grain-size sorting caused by contour current at least at Site 1017 and possibly at other sites in the California margin. Only after subtracting the influence of grain-size effect do signals of productivity, preservation, or input from terrestrial source become evident. Correlation between org-C and factor 1 (fine-grain-size indicator) suggests presence of additional organic carbon at several stratigraphic intervals. These are tentatively attributed to metabolizable organic matter still contained within the top 1.4 m of the sequence and an additional supply of terrestrial organic carbon from nearby rivers and/or enhanced surface productivity around 11, 43, 48, 55, and 72-81 ka. Carbonate preservation and/or productivity effects are represented by higher factor 3 loading. Factor 3 loading shows positive peaks approximately at 10, 25, 47, and 77 ka.