Benn and Evans argue that the basis for the interpretation of the Livingstone Lake drumlins was the form and the presence of sorted sediment in the drumlins. This is a simplification: the interpretation was based on the form of drumlins, the style, not the presence, of sediment, clast lithology and rounding, landform association and landform sequence.

They also write that the meltwater hypothesis does not predict any systematic difference in the morphology of cavity fill and erosional drumlins. In fact, Shaw (1996) gives the two types of drum-lins two different names because they are distinctly different in form. Shaw (1983) had previously drawn attention to the remarkable difference in form between the Livingstone Lake drumlins and classic drumlins. It was this difference that prompted the meltwater hypothesis.

As far as the Livingstone lake drumlins are concerned, Boulton's (1987) paper is contradicted by the field evidence; where he shows attenuation of fold limbs, the sediment is unde-formed (Shaw et al., 1989). Tulaczyk et al. (2001) specifically state that their ploughing mechanism does not explain drumlins. How could ploughing explain crescentic scours around the upstream ends of drumlins? As well, Tulaczyk et al. (2001) point out that the deep, pervasive deformation required by Boulton to explain drumlins is unlikely in light of observations beneath modern glaciers and experiments on deformation. Clarke et al. (2003) write on landforms without any observations on their internal structure. They deal exclusively with form and ignore our field observations on structure.

Benn and Evans state incorrectly that the megaflood hypothesis relies very heavily, if not exclusively, on morphological similarity with other forms. In addition to morphology we studied landform pattern, and detailed sedimentology including sedimentary architecture, landform associations, clast lithology and roundness (Shaw et al., 2000), computational fluid dynamics (Pollard et al., 1996), stone lags (Rains et al., 1993; Munro & Shaw, 1997; Beaney, 2002), hydraulic modelling (Beaney & Hicks, 2000), and valley profiles (Rains et al., 2002; Beaney, 2002).

As an example, they state that Munro & Shaw (1997) interpret Rogen and hummocky moraine as flood landforms because hummocks are associated with other landforms interpreted as flood forms and therefore assumptions become based upon assumptions. This is misleading because Benn and Evans are forcing the reader to believe that Munro & Shaw (1997) presented form analogy as their only evidence. They do, in fact, miss the point and ignore the main evidence presented: hummocks are truncated at their surfaces regardless of internal structure and without deformation of the immediate underlying sediments (see Munro-Stasiuk and Sjogren, this volume, Chapter 5, for images), and boulder lags heavily pitted with percussion marks sit on the surface of many hummocks. Although form was the starting point in formulating the megaflood hypothesis and remains an essential element, our hypothesis testing has clearly become more sophisticated.

Benn and Evans have used aerial photograph mosaics to contradict the morphological work of Munro & Shaw (1997) in south-central Alberta. It should be noted that mosaics provide less information on form than stereo aerial photographs or DEMs because they do not give a three-dimensional view of the landscape. It is also surprising then that Benn and Evans criticize us for using form analogy because they also use analogy. When they compare landforms around McGregor Lake Reservoir in south-central Alberta with forms in the forefield of Breidamerkurjokull, they give no indication of internal sediment, or sediment-landform relationships. Instead, they ignore the sedimentary evidence presented by Munro & Shaw (1997) (see previous paragraph). Here we present an aerial photograph with an interpretive map (Fig. 8.3) that lies within the mapped region presented by Benn and Evans. Clearly there are a number of different 'transverse' landforms on this photograph, yet Benn and Evans show all ridges as push moraines. They map undifferentiated mounds, ridged mounds (some with central depressions), linear ridges and some eskers as push moraines. The eskers show a flow direction towards the southwest, towards McGregor Lake Reservoir (transverse to the main regional flow direction). Of all the features on Fig. 8.3, few resemble the features on the forefield of Breidamerkurjokull (compare Fig. 8.3 with fig. 7 in Evans et al., 1999). It is possible that the linear transverse ridges are moraines, as they are quite similar to Icelandic features. There are other features like these in southern Alberta only 20 miles from the McGregor Lake

Reservoir near the town of High River that strongly resemble many of the moraines around Vatnajokull in Iceland. However, all other features are dissimilar and require an alternative interpretation. We have tried to provide an alternative explanation based on the similarity of these forms with large-scale ripple marks, and based on sedimentology (Munro & Shaw, 1997). It appears that Benn and Evans are confusing landforms. There are some broad arcuate ridges that have a very restricted distribution in Alberta. Although they may represent the ice margins they cannot be used as evidence for ubiquitous ice recession on the Canadian Prairies because they are not present everywhere. On the other hand the transverse features described by Munro & Shaw (1997) as hummocky terrain and the giant ripples described by Beaney & Shaw (2000) are distinctly different. The giant ripples are sinusoidal, and in places show rhomboidal patterns. The forms described by Beaney & Shaw (2000) are eroded from unde-formed bedrock so how could they be push moraines?

Benn and Evans refer to the use of Reynolds numbers in our hypothesis as spurious. Considering erosional marks in bedrock (Kor et al., 1991), we find rounded, lag boulders, over 1m in diameter and with distinctive percussion marks. These boulders rest on erosional marks that are identical in form to those inter

0 0

Post a comment