PATTERNS OF VARIATION IN MUDROCK GEOCHEMISTRY, CAPE ROBERTS PROJECT DRILLCORES, MCMURDO SOUND, ANTARCTICA

The three Cape Roberts Project (CRP) cores, drilled in 1997 - 99, penetrated a composite section of ~1500 m of Quaternary through lower Oligocene (possibly uppermost Eocene) sediments, at least half of which contains evidence of significant glacial influence. Evidence of significant glacial influence is less obvious, however, in the lower Oligocene/upper Eocene sediments.

Major, minor and trace element abundances in 150 CRP mudrock samples have been reported previously (Krissek and Kyle, 1998, 2000, 2001), and stratigraphic variations in elemental abundances have been interpreted as a record of sediment supply from four distinct sources: Precambrian/lower Paleozoic crystalline basement (CB), recycled Paleozoic/lower Mesozoic Beacon Supergroup sediments (BS), Mesozoic Ferrar Dolerite (FD), and Cenozoic McMurdo Volcanic Group rocks (MVG). The entire data set has now been analyzed using factor analysis to evaluate more pervasive patterns of variation.

Three major factors (87% of the total variance) were defined by analyzing the major element dataset. Each of these three factors distinguishes contributions from the relatively silica-rich sources (CB and BS) vs. those from the more mafic sources (FD and MVG). Three major factors (79% of the total variance) were defined by analyzing the minor and trace element dataset. The most important of these three (50% of the total variance) appears to distinguish contributions of MVG vs. FD materials. The two other factors identify variations within a compositionally heterogeneous MVG component.

Five major factors (82% of the total variance) were defined by analyzing the entire dataset. Only one of these factors (12% of the total variance) identifies a CB or BS component; the other factors distinguish MVG vs. FD contributions, or identify variations within compositionally heterogeneous MVG or FD components. This apparent heterogeneity within the MVG and/or FD components will complicate future efforts to quantify the contribution from each source area using chemical mass balance models.