Pilot Study using Elemental Geochemistry as a Means of Discriminating Chert Stratigraphy and Provenance in the Roberts Mountains Allochthon, Nevada

Biogenic cherts in the Roberts Mountain Allochthon (RMA) are analyzed for major and trace element geochemistry with the goal of developing a geochemical means
to differentiate stratigraphic units. This methodology allows for the identification of
cherts that cannot be confidently assigned to stratigraphic units because of structural
complexities, poor exposure, or lack of fossil data. Samples were collected from 7 localities in 5 ranges, all with known ages based on radiolarian biostratigraphy. These
samples are assigned to one of three formations, the Slaven Chert, Cherry Spring
chert within the Elder Formation, or porcellanite in the upper Vinini Formation. A
total of 82 hand samples were analyzed by portable X-ray fluorescence (pXRF) on
the rock surface for a broad suite of elements. Powders of 66 of these samples were
also analyzed by pXRF in the lab at both 120s and 240s run intervals to determine
whether either powering or run duration could improve results. The 66 powdered
samples were then analyzed by Inductively Coupled Plasma— Mass Spectrometry
(ICP-MS for trace elements) and Inductively Coupled Plasma—Optical Emission
Spectrometry (ICP-OES for major oxides) to compare whether pXRF could perform
as well as ICP in the classification process. Multivariate linear discriminant analysis
(LDA) was used to develop a series of classification models on each of the analytical
datasets. Each model was then tested using quantitative methods. Using the LOOCV
method of model testing, pXRF analyses of hand sample surfaces, while the fastest
and cheapest method, produced the least reliable classification model that predicted
unknowns only 62% of the time. Classification models based on pXRF analyses of
powdered samples increased the model’s predictive power to 75%. Increasing pXRF
run time from 120s to 240s on powdered samples decreased the number of analyses
below the limit of detection (LOD), but did not increase the model’s predictive power.
Trace elements by ICP-MS (including REE’s) had only a slightly increased predictive
power (79%) and are more expensive to run. Suggestions for improving overall model
performance, regardless of the analytical method, include using a larger dataset to
‘train’ the model that may incorporate a fuller range of geochemical variability in
each of the units across the RMA. Alternatively, development and application of a
model in a smaller region, such as a single mountain range or project area, may also
reduce lateral variations and increase model performance. Major oxide compositions
(ICP-OES) were not a useful predictor of stratigraphic unit but provided insights re-
*E-mail: noblepj@unr.edu
1078 Elizabeth M. Benge, Paula. Noble, Richard W. Murray, and John Muntean
garding provenance. Together, major oxide and REE data rule out an open ocean setting for the RMA cherts and support previous work on RMA detrital units that indicate the RMA basin formed proximal to the Laurentian margin. REE patterns show
an Eu anomaly but no Ce anomaly and, together with La-Th-Sc analysis, indicate a
continental margin depositional environment with continental arc and/or cratonal
influence on the pelagic sequence. Collectively, these data show that it is possible to
discriminate between the Ordovician, Silurian, and Devonian chert units in the RMA
using geochemistry, even regionally across multiple mountain ranges. Tradeoffs exist
between cost and reliability of pXRF versus ICP-MS data. These geochemical data
also are useful in evaluating geologic provenance of chert.
Key Words: Roberts Mountains allochthon, chert, geochemistry, Niton