CONODONT APATITE δ18O INDICATES CLIMATIC CONTROL ON SALINITY GRADIENT ACROSS THE US MIDCONTINENT SEA DURING PENNSYLVANIAN GLACIO-EUSTATIC HIGHSTANDS

Pennsylvanian phosphatic “core shales” are interpreted to have been deposited in waters deep enough to produce a pycnocline that may have developed as a consequence of increased riverine runoff into the largely landlocked Midcontinent Sea during the more humid interglacial intervals (Algeo et al. 2008). We tested whether a gradient in surface water salinity existed in the Late Pennsylvanian Midcontinent Sea during times of maximum flooding by analysing oxygen isotopes of conodont apatite from the maximum flooding surface in black shales of the Midland, Midcontinent, Illinois and Appalachian basins. In the Midland and Midcontinent basins, average δ¹⁸O values for conodonts from the Oakley, Hushpuckney and Heebner shales are 18.8 ±0.3‰, 19.1 ±0.2‰, and 18.9 ±0.1‰, respectively. Oxygen isotope ratios of conodonts from the Nuyaka Creek shale of the Midcontinent Basin are 19.8 ±0.3‰, and thus significantly higher in comparison to the other core shales. Conodonts from Ohio and Pennsylvania are depleted in ¹⁸O on average by 2.5‰ if compared to conodonts from the Midcontinent Basin. This pattern is also evident for conodonts from the Heebner and Hushpuckney shales, although conodonts from the Appalachian Basin show only 0.4 to 0.7‰ lower δ¹⁸O values compared to conodonts from the Midcontinent Basin. The differences in δ¹⁸O are interpreted as reflecting primarily salinity, assuming that temperature was relatively constant across the tropical Midcontinent Sea.

Lower salinities in the Appalachian Basin and a significant salinity gradient in the Midcontinent Sea were reconstructed for the Oakley shale (Desmoinesian), supporting the idea that freshwater discharge may have resulted in a salinity-stratified water column during everwet paleoclimate times, at least in the eastern part of the epeiric sea. However, the salinity gradient is considerably reduced during the subsequent periodically dry interglacial periods (Hushpuckney, Heebner). This suggests that surface waters in the Midcontinent Sea were not significantly influenced by continental runoff during the Missourian and Virgilian, and calls into question whether a halocline was necessary for the formation of phosphatic core shales.