TESTING K-BENTONITE CORRELATIONS USING TRACE ELEMENTS IN APATITE PHENOCRYSTS FROM THE SALONA FORMATION OF CENTRAL PENNSYLVANIA

Volcanogenic apatite trace element chemistry from K-bentonites promises a method for high-precision correlation from the local to intercontinental scale. This tool is uniquely suited for testing the isochronous nature of biozones, stratigraphic sequences, and chemostratigraphic profiles. Previous studies of trace elements in volcanogenic apatite show distinct elemental concentrations, i.e. fingerprints, for characterizing K-bentonites. Yet, basic questions remain concerning this discriminating power that relate to the expected range, variability, and geologic meaning of apatite trace element values. We examined these issues with an example from the Hagan K-bentonite cluster in the Salona Formation of central Pennsylvania, a well-exposed portion of the Late Ordovician cratonic margin of Laurentia. This formation is dominated by thick calcilutites, thin wackestones, and lesser amounts of fissile black shale. The Salona Formation is unusual in the Upper Ordovician of North America for containing multiple thick and closely spaced K-bentonites within a deep subtidal carbonate-dominated facies. K-bentonite beds thicker than 5 mm were sampled from the basal Salona Formation at five locations (N = 42), forming a ~100 km transect. Trace elements (F, Cl, Mg, Mn, and Fe) where analyzed in single apatite phenocrysts from thirty-four samples using an electron microprobe (eight samples anomalously lack apatite phenocrysts). Initial results show that several beds display significant variation in phenocryst chemistry. Comparison with published data of Cenozoic volcanogenic apatite trace elements suggest that some of the K-bentonites may be composite beds, either from a heterogeneous source or multiple eruptions. Despite this, the beds can be characterized given enough analyses. On the basis of 10 to 20 analyses, three of the beds possess distinct apatite trace element profiles. Because of the relatively large number of beds, varied apatite trace element profiles, variable phenocryst content, and analytical precision of the electron microprobe, 20 to 50 analyses per sample are required for the remaining beds. Our interpretation of this new data suggests correlations that disagree slightly with those previously made on the basis of stratigraphic position, thickness, and whole rock chemistry.