Pollen records are from cores taken at two Ocean Drilling Program (ODP) drill sites (Site 1019: 41°40.696´N, 124°55.981´W, 989 meters below seafloor [mbsf]; Site 1020: 41°0.051´N, 126°26.065´W, 3050 mbsf) and from piston Core EW9504-17PC (42°14.55´N, 125°53.28´W, 2671 mbsf; Fig. 1). The siliciclastic clays and silt from the upper 60 m (meters composite depth [mcd]) of Holes 1020C, 1020D, 1019C, and 1019E were routinely sampled at 20-cm intervals (Lyle, Koizumi, Richter, et al., 1997). Core EW9504-17PC, composed predominantly of hemipelagic clay (Lund and Mix, 1998), was sampled at 5-cm intervals. Standard processing procedures that included the addition of known amounts of an exotic tracer to calculate pollen concentration were preceded and succeeded by sieving through 7-µm nylon screening. Taxonomic identification of pollen was based on comparison with modern pollen reference collections from western North America. Specific epithets are indicated for grains that were clearly identified; otherwise, pollen and spores are assigned to genera or higher rank. Other than the papillate grains of Sequoia (redwood) and the large inaperturate grains assumed to represent P. menziesii, inaperturate pollen of other genera in the Taxodiaceae, Cupressaceae, and Taxaceae that cannot be satisfactorily separated using light microscopy (e.g., Juniperus, Torreya, Cupressus, Libocedrus, Chamaecyparis, and Thuja) are here referred to as cedar type. Other synthetic pollen groups include chaparral (sclerophylous shrubs and other members of the Anacardiaceae, Rhamnaceae, and Rosaceae) and herbs (Gramineae, Cyperaceae, and Compositae, including Artemisia or sage). A mininum of 300 pollen grains were identified in each sample from Core EW9504-17PC; the initial pollen counts from Sites 1019 and 1020 presented here averaged ~110 and ~150 pollen grains, respectively. Pollen percentages were based on the sum of terrestrial pollen (excluding fern spores), and pollen concentration was calculated on the number of pollen grains per gram dry weight of sediment (gdws).
Age models were constructed by correlating the CaCO3 and Corg records to the best radiocarbon-dated sections and oxygen isotope records. Methods are reported in detail in Lyle et al. (Chap. 32, this volume). Age control for Core EW9504-17PC was achieved by (1) correlating the CaCO3 and Corg records from Lyle et al. (Chap. 32, this volume) with Core W8709-13PC (time scale of Lund and Mix, 1998) for 0-50 ka, (2) correlating the oxygen isotope record from 50 to 140 ka in the core (Lund and Mix, 1998) with the Martinson et al. (1987) age model, (3) allowing minor shifts in the record to minimize sedimentation rate changes without losing significant correlation, and (4) comparing the oxygen isotope record on the final age model to the Martinson et al. (1987) age model to see if the two age models remained consistent.
To construct an age model for Site 1020, we correlated the dated carbonate and organic carbon records from Core EW9504-17PC with the carbonate and organic carbon records of nearby Site 1020 and hence were able to transfer the age model from Core EW9504-17PC to Site 1020 (Lyle et al., Chap. 32, this volume). Below 140 ka, our age model for Site 1020 is based upon reconnaissance-scale oxygen isotope stratigraphy and biostratigraphic datums.
Development of an age model for Site 1019 was more problematic (Lyle et al., Chap. 32, this volume). The preliminary age model used here is based on reconnaissance-scale oxygen isotope stratigraphy supplemented by radiocarbon age control from 6 to 24 ka. We note that these preliminary age models will be modified when more age control (radiocarbon and stable isotope data) is available. Using the age models described above, the average sampling interval in Core EW9504-17PC was ~500 yr, the average sampling interval in the upper 57 mcd of Holes 1020C and 1020D was ~1700 yr, and the average sampling interval in the upper 40 m of Holes 1019C and 1019E was ~800 yr (Lyle, Koizumi, Richter, et al., 1997).