Invited ReviewUp in the refrigerator: Geomorphic response to periglacial environments in the Upper Mississippi River Basin, USA

- Permafrost was extensive in the Upper Mississippi basin during the LGM.
- James Knox argued that periglacial conditions strongly affected landscape evolution.
- Hillslope deposits record accelerated erosion under periglacial conditions.
- Streams aggraded in response to accelerated erosion.
- Eolian activity was also significant under periglacial conditions.

James C. Knox was best-known for his work on Holocene and historical changes in fluvial systems, but he also had a long-standing interest in the effects of late Pleistocene periglacial environments on landscape evolution in parts of the Upper Mississippi River basin that were just outside the Laurentide Ice Sheet margin, or as Knox put it, 'up in the refrigerator.' Knox and others in the Quaternary community of the Midwestern U.S. often suggested that hillslope erosion was accelerated under periglacial conditions, so that glacial periods have had a dominant effect on the landscape we see today. This paper reviews the evidence and reasoning supporting that view in a study area of the Upper Mississippi basin bordered on three sides by ice margins of the last glaciation, including the Wisconsin Driftless Area and adjacent landscapes. Sparse but compelling paleoecological data and relict ice- or sand-wedge polygons provide clear evidence for a cold climate and widespread permafrost around the peak of the last glaciation. In highly dissected, relatively high-relief parts of the study area, the loess and soil stratigraphy on ridgetops and the colluvial mantles on steeper slopes are best explained by highly effective hillslope erosion, including solifluction, during and just after the Last Glacial Maximum. Knox used the post-depositional truncation of a loess unit to quantify contrasting late Pleistocene and Holocene sediment yields from a small Driftless Area watershed. While the late Pleistocene yield indicates accelerated erosion, it is still lower than modern sediment yields in many tectonically active or semiarid landscapes, and it may reflect deposition of highly erodible loess as well as effects of periglacial conditions.The views of Knox and other Midwestern geomorphologists on landscape evolution through glacial-interglacial cycles were highly influenced by the work of Robert V. Ruhe. Ruhe proposed that an episode of widespread erosion during and just after the Last Glacial Maximum can explain enigmatic aspects of Quaternary stratigraphy and the soil landscape on the Iowan Erosion Surface, a very low relief landscape of the study area. Ruhe's key evidence is still valid, though it needs to be separated from an implausible model of landscape evolution. Interpretation of the Iowan Erosion Surface and other low-relief landscapes just outside the ice margin also requires recognition of the profound effect of eolian processes on those landscapes under periglacial conditions. Many new insights on landscape evolution in the study area could result from wider application of cosmogenic nuclide-based methods to assess glacial-interglacial changes in basinwide rates of erosion and residence time of soils. Just as important, a need exists for much more field-based characterization of hillslope, fluvial, and eolian sediments for comparison with those of modern permafrost regions and past periglacial environments in Europe.