Jan A. Piotrowski, Institute of Geophysics and Geology, University of Leipzig, Talstr. 35, 04103 Leipzig, Germany, jpiotrow@rz.uni-leipzig.de; David M. Mickelson, Department of Geology and Geophysics, Weeks Hall, University of Wisconsin, 1215 West Dayton St., Madison, WI 53711, USA, davem@geology.wisc.edu; Slawek Tulaczyk, Department of Geological Sciences, 101 Slone Bldg., 0053, University of Kentucky, Lexington, KY 40506-0053, USA, smtula0@pop.uky.edu; Dariusz Krzyszkowski, Institute of Geography, WSP Slupsk, Poland, dk2@wro.top.pl; Frank W. Junge, Institute of Geophysics and Geology, University of Leipzig, Talstr. 35, 04103 Leipzig, Germany.

In recent years many papers have suggested that past ice sheets rested on wet deforming beds that contributed significantly to ice movement and sediment transport. Also, modern analogues have been postulated to substantiate inferrences about ancient ice sheets. The presence or absence of subglacial soft-sediment deformation is fundamental to controlling the dynamics of soft-bedded glaciers, and it may significantly influence the course of glaciation. Here we express our scepticism about uncritical application of the deforming bed hypothesis to vast areas of Europe and North America glaciated during the Pleistocene. Although we accept bed deformation as a geological process, we believe that its occurrence, thickness and intensity was much more restricted than postulated by many recent papers. The common treatment of bed deformation as a viscous process has a weak observational basis despite a decade-long effort to confirm viscous till behaviour in laboratory and field experiments. We base our argument on the following observations and logic:

Undeformed, apparently intact bodies of sorted sediments and other heterogeneities, including large glacial "rafts", often occur in basal tills. This precludes pervasive deformation, which should lead to a mixed, homogenized till matrix.

Occurrence of intact fragile materials (including shells and heavily weathered boulders) in subglacial tills is difficult to reconcile with the concept of till deformation.

Bed deformation should be accompanied by rotation and re-orientation of stones and till matrix. However, as indicated by till fabrics, striations on boulder pavements and by other structural elements, many subglacial tills remained relatively stable after deposition from ice.

Beneath some basal tills there are evidently undeformed tills with intact palaeosols on top. It is difficult to explain why the underlying till would remain stable under the allegedly deforming till with very similar mechanical properties above.

In places where tills are underlain by undeformed sediments, as in vast areas of Europe and North America, the maximum possible depth of ductile, wet-bed deformation is constrained by the till thickness, which is often much smaller than the predicted deformation depth. Also, thin tills cannot be considered efficient conveyer belts for glacial debris.

Contacts between homogeneous tills and undeformed sediments beneath are often sharp, which is in contrast with the deforming bed concept predicting a transition zone with diverse styles of glaciotectonic deformations.

Deforming bed hypothesis predicts piles of glacial material at ice stationary positions. However, tills in many regions seem to decline in thickness and gradualy wedge-out in the proximity of past ice margins. Also, in many instances no end moraines or outwash accumulations occur at ice limits.

Subglacial meltwater conduits associated with deforming beds ("canals") are believed to be wide and shallow. This is in contrast with some observations from Germany, where many such conduits are rather narrow and steep, suggesting largely stable conditions.

Deforming-bed tills are typically associated with fast ice flow, which implies that average transport distances of debris should be of the order of 100s of km. This is not consistent with observations of relatively short transport distances for the bulk of the glacial material in tills deposited by fast flowing glaciers in Europe and North America.

Ice rafted debris in the North Atlantic (e.g. Heinrich events) indicates englacial rather than subglacial material transport. If the material was transported mainly in a wet deforming bed, then it would have been deposited at the grounding line of the ice sheet and not carried further in icebergs.

There is lack of unequivocal modern analogues. (1) A deforming bed beneath the West Antarctic Ice Streams has been interpreted largely from geophysical data, which indicates dilated basal tills and high pore water pressure, but does not prove pervasive bed deformation. One direct borehole experiment yielded data inconsistent with the deforming bed model. Furthermore, diatoms in the sub-ice-stream tills are distributed in very inhomogeneous manner, which is also inconsistent with this model, that should produce well-mixed material. (2) Famous experiments under Breidamerkurjkull on Iceland were conducted under the outermost part of the glacier, under few-m-thick ice only. It is difficult to consider these experiments a good analogue for the past ice sheets. (3) High sediment flux rates attributed to a deforming bed are based on a thick deforming layer. Observations under modern glaciers do not document a thick deforming layer.

We suggest that basal sliding, rather than bed deformation, was the major movement mechanism of past ice sheets and bulk of the debris was transported englacially rather than within the alleged deforming bed. Passive, localized deformation of till caused by ploughing and dragging of till material by clasts and ice bumps provides a sufficient explanation for the existing sedimentological evidence of subglacial shear. There is no need to invoke continuous, active deformation of viscously behaving till layers.
 
 This is a contribution of the Quaternary Geology group at the University of Wisconsin--Madison.  For complete citation return to publications list.