Jonathan Sumrall

The rocks that make up the stratigraphy of the Upper Peninsula of Michigan are extremely difficult to differentiate in hand sample. The three students that helped perform the field and lab work for this project found out firsthand how difficult this task can be. We worked closely with a colleague of mine from Shawnee State University and the US Forest Service to perform this project. The Engadine Group of the middle Silurian was the focus of this study. The Engadine Group has been broken down into smaller subdivisions in other areas; however, it has not been differentiated in the Hiawatha National Forest. The smaller subdivisions include (oldest to youngest): Rockview Dolostone, Rapson Creek Dolostone, and the Bush Bay Dolostone. The Bush Bay Dolostone is further subdivided into three members (Oldest to youngest): Prentiss Creek Member, Swede Road Member, and McKay Bay Member.

Methods: Field measurements were made to determine the thickness of each body of exposed rock. Descriptions of each section of rock were made by the students. Core samples were drilled for each unit measured totaling 65 samples. Core samples were brought back to SHSU to be processed. This involved splitting each core in half. One half of the core was used for analyses, while the other half was catalogued for long term storage. From the half used for analyses, petrographic thin sections and powders for geochemical analyses were created. Powdered samples were dissolved and analyzed at TRIES laboratory at SHSU for trace element composition. Duplicate powders were sent to the University of Alabama Stable Isotope Laboratory for carbon and oxygen stable isotope analyses. Duplicate cores were used to measure the magnetic susceptibility of each unit to determine if there were differences among rock units. Finally, thin section petrography was performed using polarizing light microscopes and cathodoluminescence microscopes.

Field Data: A total of eight sites were examined to determine the stratigraphy present and measure the thickness of these exposures. The sites consisted of two quarries and six bedrock exposures. The most significant amount of rock was located at Scott’s Quarry where ~30 meters of rocks were exposed. The most significant observation from the field work was the identification of a single unit of rock that could be traced over three miles between multiple sites. This single unit allowed for the differentiation of several members of these rocks. Other significant observations was the presence/absence of fossils, chert content, and degree of karst development.

Petrographic Data: From the eight sites, thin section analysis varied slightly from the initial field observations. The most significant variance included the ability to identify fossil fragments in thin section, especially for units that were noted as having little to no fossils present. All three formations were identified and have specific characteristics to each. In Scott’s Quarry, the oldest units of this study were present. The base of Scott’s Quarry contained rocks of the Manistique Group, which is the group of rocks deposited directly before the Engadine Group. The Manistique Group in Scott’s Quarry consisted

of chert-rich dolostones that contained crinoids and moderate silica replacement of fossils. In addition, the Manistique Group contained minor iron oxides and detrital quartz grains. The Rockview Formation was located directly on top of the Manistique Group in Scott’s Quarry. The Rockview Formation in Scott’s Quarry consisted of brachiopodal, crinoidal, and algal dolostones with micro-stylolites and chalcedony (silica cement). It also consisted of doloboundstones to dolowackestones, representing alternating microbial mats and low energy, lagoonal environments. Both the stylolites and chalcedony suggest shallow burial resulting in these diagenetic features. The top of Scott’s Quarry (~2 meters) transitions into the Rapson Creek Dolostone. This transition is marked in the field by a color transition, and large voids filled with dolomite crystals are found in thin section. Rapson Creek was also identified at two other sampling locations and had a very similar appearance in thin section. The Bush Bay

Dolostone was more difficult to differentiate. The Prentiss Creek Member and Swede Road Member were grouped because there was little to no petrographic difference between these two. One sampling location was confirmed as these two members. The key identifying petrographic characteristic appears to be the presence of pressure solution of echinoderm fragments. This produces a distinct microtexture when examined under a polarizing light microscope. The McKay Bay Member was sampled at one location and had a distinct petrographic appearance in thin section. This member contained large euhedral dolomite crystals, red (iron oxide) stained stylolites, and abundant moldic porosity. In the field, the McKay Bay Member had the highest development of karst features.

Trace Element Data: The trace elemental dataset reinforces the petrographic classification and further refines the differentiation of these rocks. Few studies have concentrated on the trace element concentration of these rocks. The trace elemental concentrations are reported as weight oxide percentage. The Manistique Group of Scott’s Quarry has an average SiO2 of 0.12%, Fe2O3 of 0.39%, Al2O3 of 0.37%, and a total trace concentration of 1.14%. The Rockview Formation has an average SiO2 of 0.0.62%, Fe2O3 of 0.27%, Al2O3 of 0.30%, and a total trace concentration of 1.50%. The Rapson Creek Formation has an average SiO2 below detection limit, Fe2O3 of 0.09%, Al2O3 of 0.04%, and a total trace concentration of 0.51%. The Lower Bush Bay Members (Prentiss Creek and Swede Road) has an average SiO2 below detection limit, Fe2O3 of 0.04%, Al2O3 below detection limit, and a total trace concentration of 0.45%. The McKay Bay Member has an average SiO2 below detection limit, Fe2O3 of 0.10%, Al2O3 of 0.04%, and a total trace concentration of 0.52%.

Stable Isotope Data: The stable isotope dataset shows little variance between all samples collected. Average δ13C values are +1.7 per mil +/- 0.9 per mil, and average δ 18O values are -5.0 per mil +/- 0.7 per mil. While the values do not vary much, spatial variance among vertical sections (e.g. Scott’s Quarry) show interesting trends that follow previous differentiation trends. For example, the Manistique rocks of Scott’s Quarry generally increase in both 13C and 18O values moving upward. Both the Rockview Formation and Rapson Creek Formation of Scott’s Quarry show a decrease in both 13C and 18O values moving upward. Carbon stable isotopes between Scott’s Quarry and other sites correlate similar to other datasets; however, oxygen stable isotopes display higher variance and do not correlate as well.

This may be due to diagenetic effects and variation in stable isotope variations of oxygen isotopes in meteoric, lake, and pore waters.

Magnetic Susceptibility Data: The magnetic susceptibility dataset has shown to be the least useful for differentiating the Engadine Group mainly due to the extremely low magnetic signature of these carbonate rocks. Trends do exist with this dataset. In Scott’s Quarry, the Manistique Group shows a general decrease in magnetic susceptibility, the Rockview Formation shows a general increase in susceptibility, and the Rapson Creek shows a general decrease in susceptibility. The changes in trends occur at the same interval interpreted by other datasets to represent the transition between these same units. In this regard, magnetic susceptibility may show some use in future studies; however, other more robust measures were identified in this study.

Conclusion: Coupling multiple geochemical and petrographic techniques with field observations achieved a basic differentiation between the units of the Engadine Group within the Hiawatha National Forest. Petrographic, trace element, and stable isotope datasets proved to be the most robust, while magnetic susceptibility appears to be the least useful for this task. Future studies will build on this differentiation and expand the number of proxies for differentiation.