GSA 2020 Connects Online

Paper No. 2-4
Presentation Time: 2:35 PM

VISUAL DERIVATIVE SPECTROSCOPY AS A QUANTITATIVE APPROACH TO ANALYZE THE NON-SILICA COMPONENT OF HIGH SILICA CHERT STONE TOOLS


LEWIS, Angela R.1, BEBBER, Michelle R.2, EREN, Metin2, WILLIAMS, Jeremy C.3 and ORTIZ, Joseph D.4, (1)Department of Geology, Kent State University, PO Box 5190, Kent, OH 44242-0001, (2)Department of Anthropology, Kent State University, PO Box 5190, Kent, OH 44242-0001, (3)Geology, Kent State University, PO Box 5190, Kent, OH 44242-0001, (4)Department of Geology, Kent State University, 221 McGilvrey Hall, 325 S. Lincoln St, Kent, OH 44242

Chert color or visible reflectance (400 – 700 nm wavelengths) is the result of minor and trace elemental composition. Visual derivative spectroscopy (VDS) have been used extensively to identify minerals in marine sediment cores. Principle component analysis (PCA) of chert reflectance can aid provenance determination through separation of potential formation conditions and removal of quartz from high SiO2 (majority > 90 wt.%) cherts. Stone tools recovered from Northern Ohio and the Welling Site in Central Ohio were analyzed with VDS and compared to chert outcrops. The Welling Site (33-Co-2) is in Coshocton County near thick chert beds of the Upper Mercer member of the Pottsville formation (Middle Pennsylvanian 315 Ma) that was used by Clovis people to make stone tools. Mineralogy of Welling Site tools was measured with traditional X-ray diffraction (XRD) and calculated with the Ortiz (2011) varimax rotated PCA method that removed the dominant signature (e.g., quartz). For verification, scanning electron microscope energy dispersive spectroscopy (SEM EDS) was used on tools with > 2% loss on ignition and quartz was removed from XRD spectra. Stone tool reflectance was significant for tool type (e.g., projectile points) and recovery location (p = 0.03 and p = 0.00, respectively). Drills were the only significant tool type and all tools were similar to Welling except tools from Lorain and Erie Counties. Chert tool variance was from precipitation in iron ore rich areas (PC 1 73%) then weathered clays (PC 2 13%). Three Northern Ohio tools were not related to PC 1, did not cluster with the other tools or outcrops and had Euclidean distances closest to the Zaleski formation. Those tools could have been originally made from Zaleski chert or from outside Ohio. Welling XRD bulk mineralogy was mostly quartz with some siderite, calcite, pyrite and organics. More clays and pyrite were found when quartz was removed from the XRD spectra. VDS determined minerals were PC 1 74% clays, talc and gypsum then PC 2 20% opal. Welling tools were better quality (greater SiO2 wt.%, lower loss on ignition) than Northern Ohio tools but had more clays, presumably from more weathering. Welling tools had less Fe minerals but that result could change with improvement of the reference spectral library.