DEGLACIAL HISTORY OF REMNANT LAKES OF GLACIAL LAKE ROSCOMMON

Glacial Lake Roscommon (GLR) was proglacial to the Mackinac Lobe of the Laurentide Ice Sheet, in north-central Lower Michigan. Several remnant lakes of GLR are the focus of this talk. GLR shifted in extent as the ice front advanced into it from ≈40 to 19 ka. These ages were determined by OSL dating of chiefly ice-contact glacio-deltaic sediment. Here we report on nuances of the postglacial history (< 19 ka). as indicated by 76 ¹⁴C ages of plant carbon sampled from eight sediment cores of five remnant lakes.

Seismic profiles revealed detailed on-lap/off-lap relationships in the stratified record of Higgins Lake – the deepest of these remnant lakes. Cores from our other target lakes (Backus, Marl, Saint Helen, and Pup) were obtained from deep pools, using a motor-assisted Livingston piston corer, typically from an ice platform. The longest core (13.8 m), from Higgins Lake, was obtained in water 12.5 m deep. Basal bedded lithofacies A (9.5 m) for Higgins Lake is composed of fossil-barren, reddish-brown silty clay, sand, and gravelly sand. The fractured, compact upper 0.1 m of this unit has weak evidence of desiccation and incipient pedogenesis. Lithofacies B is about 0.45 m thick and composed of fossiliferous silt, including 4-cm beds of bryophytes, with rare spruce needles and bulrush seeds. Unit C is soft, calcareous and fossiliferous gyttja as much as 3.8 m thick.

Core data suggest that plants began to colonize the region at 15.1, 14.2, 14.0, and 13.4 cal ka at Higgins, Backus, Pup, and Mud lakes, respectively. Sediment structures (e.g., brecciated sediment), tundra plant fossils (notably Dryas), cryophyllic ostracodes (e.g., Limnocythere friabilis), and staggered onset ages collectively point to permafrost and its melting on sediment and fossil records. Except for the terminal late glacial record at Backus Lake, conformable post-glacial sedimentation began between 9.5 and 8.6 cal ka.

Sediment accumulation lacuna for most basins temporally coincide with the Chippewa/Stanley Phases of the Great Lakes, which begs the question: were lake levels in north-central Michigan controlled by shifts in effective moisture, or by groundwater levels perturbed by more than 110 m of base-level drop in Lakes Michigan and Huron? Neither analogs of ostracode nor pollen spectra indicate profound changes in effective moisture - only of temperature.