W. W. Shilts

Multiple deglaciations of the Hudson Bay Lowlands, Canada, since deposition of the Missinaibi (Last-integlacial?) formation

Abstract The stratigraphic record in the James and Hudson Bay Lowlands indicates that the sequence of glacial events at the geographical center of the 12.6 × 10 6 km 2 Laurentide Ice Sheet may have been more complex than hitherto imagined. Isoleucine epimerization ratios of in situ and transported shells recovered from till and associated marine and fluvial sediments cluster into at least 4 discrete groups. Two alternative explanations of the data are offered, of which we strongly favor the first. Hypothesis 1: Setting the age of the “last interglacial” marine incursion, the Bell Sea, at 130,000 yr B.P. results in a long-term average diagenetic temperature for the lowlands of +0.6°C. Using this temperature enables us to predict the age of shells intermediate in age between the “last interglaciation” and the incursion of the Tyrrell Sea 8000 yr ago. Between these two interglacial marine inundations, Hudson Bay is predicted to have been free of ice along its southern shore about 35,000, 75,000, and 105,000 yr ago based on amino acid ratios from shells occurring as erratics in several superimposed tills and fluvial sediments. These results suggest (1) that traditional concepts of ice-sheet build-up and decay must be reexamined; (2) that “high” sea levels may have occurred during the Wisconsin Glaciation; and (3) that a critical reappraisal is required of the open ocean δ 18 O record as a simple indicator of global ice volume. An alternative, Hypothesis 2 , is also examined. It is based on the assumption that the 35,000-yr-old deposits calculated on the basis of Hypothesis 1 date from the “last interglaciation”; this, in effect, indicates that the Missinaibi Formation, commonly accepted as sediments of the “last interglaciation,” are about 500,000 yr old and that the effective diagenetic temperature in the lowlands during approximately the last 130,000 yr has been close to −6°C. We argue for rejection of this alternative hypothesis.

Keewatin Ice Sheet—Re-evaluation of the traditional concept of the Laurentide Ice Sheet

Patterns of dispersal of distinctive Proterozoic and Paleozoic erratics across terrain formed on Archean and Aphebian crystalline rocks indicate that (1) ice never flowed from Hudson Bay into Keewatin in the region from the Manitoba border (lat 60°N) northward at least to lat 65°N; (2) westward-southwestward flow out of the bay, probably from a Labradorean dispersal center, interfaced with southward- and southeastward-flowing ice from the Keewatin dispersal center somewhere between Nelson River and Churchill; and (3) at least 300 km of dispersal of distinctive erratics observed from the vicinity of the last position of the Keewatin Ice Divide to the present coast of Hudson Bay required considerably more time than the 1,000 to 3,000 yr that the divide has traditionally been thought to have existed. In fact, the Keewatin Ice Divide and its precursors represent the centers of an independent, land-based ice sheet that probably existed throughout the period of Wisconsin Glaciation.

Bedforms of the Keewatin Ice Sheet, Canada

Abstract By compiling glacial bedforms on a map that covers most of one sector of the Laurentide Ice Sheet, it is possible to make some suggestions about their genesis based largely on spatial relationships. It can be concluded that drumlins and ribbed moraine form at the base of actively flowing ice under similar dynamic conditions. For either landform to exist, however, there must have been enough sediment available in the base of the glacier to leave or form a feature large enough to be recognizable. The presence or absence of sufficient load is related to the geology of the glacier bed and has little to do with regionally changing dynamics of the ice-water system. Likewise, given sufficient load, it is evident that whether drumlins formed or whether ribbed moraine formed in a certain area is a function of the physical nature of the load which is, again, related to geology of the source outcrops. Whether the physical characteristics come into play after the sediment has been released from the ice and is being reshaped by basal drag, streamlining, etc., or whether the nature of the load while entrained changes the behaviour of the basal part of the ice is unclear. Physical characteristics of the basal sediment load have apparently promoted internal thrusting of coherent slabs of entrained debris and ice to form ribbed moraine on melting, whereas drumlins may reflect moulding of plastic subglacial debris or erosional streamlining of both the unconsolidated glacial substrate and bedrock. The observation that many eskers cross drumlin fields at nearly right angles to their orientation suggests that conditions producing streamlining and those pertaining to subglacial drainage are separated in time and circumstance. The general occurrence of drumlins and eskers throughout the sediment-rich portions of the Keewatin Ice Sheet, from Zone 1 to its edge, is difficult to reconcile with the restriction and intimate association of these forms with ribbed moraine almost exclusively in Zone 2. Because such a zonal relationship exists to some extent around other ice divides, at least in Labrador/Nouveau Quebec and Newfoundland, it seems that some condition changed or existed in this zone throughout or at some specific time during the existence of the Keewatin Ice Sheet. Possibly around the last ice divides, reactivation of the ice sheet by alimentation associated with climatic deterioration may have promoted flow of thin, brittle ice. This may have “shattered” the glacier, forming stacked thrust plates of ice and debris. With subsequent stagnation, the thrust plates may have melted out without substantial deformation.

Quaternary Stratigraphy and Events in Southeastern Quebec

Quaternary stratigraphic units have been mapped in the Appalachian region of southeastern Quebec, and formal stratigraphic names for these units are proposed. Evidence exists for four separate glacial phases, the last three of which are represented by tills. The three tills, from oldest to youngest, have been named Johnville, Chaudiere, and Lennoxville, respectively. Stratified sediments interbedded with the tills record significant nonglacial intervals between each of the glacial phases. It is suggested that the last three glacial phases are of Wisconsin age and that the Lennoxville Till represents the entire late Wisconsin. Ice-flow directions were determined using dispersal shadows (indicator trains), till fabrics, and striations. Johnville ice flowed from the northwest; Chaudiere ice flowed initially from the northeast, but later from the northwest; Lennoxville ice flowed from the northwest. Late Wisconsin glaciers did not flow northward or northwestward into Quebec from New England. Pre-Johnville stratified sediments probably record pre-Wisconsin deposition. The Massawippi Formation, recording the nonglacial interval between the Johnville and Chaudiere glacial phases, may correlate with the St. Pierre peat beds of the St. Lawrence Lowlands. The Gayhurst Formation, recording the nonglacial interval between the Chaudiere and Lennoxville glacial phases, may correlate with some of the Port Talbot interstadial sediments of southern Ontario. The Quebec Appalachians were finally deglaciated by about 12,500 C 14− yrs B.P.

Till geochemistry in Finland and Canada

Abstract Glacial geology has been much more thoroughly integrated into mineral exploration programs in Finland than in Canada. In Finland, the compositional properties of glacial sediments are utilized to find buried ore, whereas in Canada glacial deposits have been regarded more often than not as a hindrance to exploration. The application to mineral exploration of the latest theories on ice-sheet dynamics, sedimentation from ice, and till composition is presently regarded as somewhat more important in Canada than in Finland. Three specific geochemical aspects of till have been investigated recently by the author and should have considerable relevance to techniques of exploration: 1. (1) Mineral and chemical partitioning in till is marked because of the tendency of minerals to crush to certain specific sizes during glacial comminution. For most metals investigated, highest concentrations are in the clay-sized fraction, presumably held within the structure of phyllosilicates or scavenged by secondary oxides, both of which phases occur preferentially among particles finer than about 10 μm. 2. (2) The more labile ore minerals, such as sulphides, are destroyed by weathering to depths several metres below the postglacial solum. This destruction is often accompanied by a concomitant increase in metal concentration in the clay-sized ( μ m) fraction. 3. (3) A till sample is a composite of compositions inherited from over-lapping dispersal trains at several different scales. The sizes of trains range from continental scale, covering tens of thousands square kilometres, to local scale which can be mapped for distances of no more than a few hundred metres. It is essential in mineral exploration to be able to differentiate between the rare, but sometimes distinctive or easily concentrated clasts from far away and those that might have been generated by a small, but economic, local source.

Late Winconsinan glacial dynamics, deglaciation, and marine invasion in southern Quebec

Deglaciation patterns of the Laurentide ice sheet in southern Québec were related to climatic and nonclimatic factors. Thinning of the ice sheet and thermolatitudinal ice retreat are directly linked to the global warming at the end of late Wisconsinan time, between 17 ka and 11 ka. However, correlations between regional deglacial events and global climatic oscillations during that period have yet to be established, except for the St. Narcisse Moraine event, which has been assigned to the Dryas III, and perhaps for the reactivation of Laurentide ice in the middle Chaudiere Valley area during an older cold event. Nonclimatic factors also played a major role on the deglaciation of the region. After the last glacial maximum, the Laurentide ice sheet began to decrease, and the St. Lawrence corridor channelized a major ice stream, the St. Lawrence ice stream, which became a major feature of the southeast sector of the ice sheet. The St. Lawrence ice stream is a flow convergence zone caused by a combination of ice dynamics and topographic factors and rapid ablation at its terminus. The head of the flow convergence migrated deeply into the Laurentide ice sheet and caused thinning of adjacent ice masses. As a consequence of this accelerated ablation, an Appalachian sector became differentiated from the main ice sheet. Regionally, the terminus of the ice stream was a calving bay that retreated along the Laurentian channel to the mouth of the Saguenay fjord. The ice stream and the deglaciated estuary generated the well-known flow reversal along much of the northern margin of the Appalachian sector. In addition to these generalized deglaciation processes, local and regional topographic features influenced the ice dynamics and the final deglaciation patterns. Occhietti, S., Parent, M., Shilts, W.W., Dionne, J.-C., Govare, É., and Harmand, D., 2001, Late Wisconsinan glacial dynamics, deglaciation, and marine invasion in southern Québec, in Weddle, T.K., and Retelle, M.J., eds., Deglacial History and Relative Sea-Level Changes, Northern New England and Adjacent Canada: Boulder, Colorado, Geological Society of America Special Paper 351, p. 243–270. E-mails: Occhietti, ochietti.serge@uqam.ca; Parent, parent@gsc.nrcan.gc.ca; Shilts, shilts@geoserv.isgs.uiuc.edu; Dionne, ggr@ggr.ulaval.ca; Govare, e.govare@sympatico.ca; Harmand, harmand@clsh.univ-nancy2.fr on February 28, 2015 specialpapers.gsapubs.org Downloaded from

IRSL dating of Middle Pleistocene interglacial sediments from southern Quebec (Canada) using multiple and single grain aliquots

The IRSL dating of Middle Pleistocene interglacial #uvial sediments from Southern Quebec, correlated with oxygen isotopic stage 7, yields optical dates much younger than the expected geological age. Single grain IRSL measurements on alkali}feldspars, following the fadia protocol developed by Lamothe and Auclair (Earth and Planetary Science Letters, 171, 319 }323, 1999), suggest that anomalous fading is the most probable cause for this severe age underestimation. The IRSL dates corrected for this anomalous fading are in better agreement with the expected ages. ( 2000 Elsevier Science Ltd. All rights reserved.

Surficial geochemistry, south-central Canadian Shield: implications for environmental assessment

An ongoing problem in evaluating the effects of acid rain is distinguishing natural (spatial) variations of the geochemical environment from anthropogenic (temporal) variations. In other words, it is important to determine whether a lake is naturally acid because of its geological setting or whether it has been acidified as a result of dry and wet acid deposition. To provide a model that might acid in answering these questions, the Geological Survey of Canada has carried out a series of sampling programs on the south-central part of the geologically complex Canadian Shield. With these programs we have tried to link the easily observable lithological variations of bedrock with the chemical compositions of overlying glacial deposits with the geochemical compositions of lake sediments and lake waters. Presumably, once these linkages are understood, variations in life systems that inhabit terrestrial and aquatic environments in the areas sampled with be easier to evaluate. The ultimate goal is to provide a base against which observed variations in life systems can be judged as natural or anthropogenic. In addition, sound geological data can be used to refine estimates of SOx and NOx target loadings and to determine areas where acid deposition may mobilize potentially noxious trace metals into hydrologic cycle. The importance of establishing geological and geochemical linkages is illustrated by the fact that significant parts of the Canadian Shield, while formed of granitoid rock, are overlain by calcareous unconsolidated sediments that have been glacially transported long distances from limestone sources. For instance, in areas north of Lake Superior and Eastern Ontario, high concentrations of carbonate minerals in glacial sediments overlying carbonate-poor bedrock result from glacial transport of carbonate minerals tens to hundreds of kilometres from limestones on their up-ice sides. In these same areas, high concentrations of As, Hg and other trace elements in sediment and water can be related to glacial dispersal from bedrock known to host mineralized occurrences or high background concentrations of these elements. Awareness of bedrock composition and of glacial distortion of the chemical signature of bedrock is essential for insuring that a complete and credible database is provided to the legislators and regulatory agencies charged with establishing realistic target loading for acid deposition.

Geochemistry of aquatic and terrestrial sediments, precambrian shield of Southeastern Ontario

Lake water and sediment samples from approximately 2200 lakes and glacial sediment (sub-solum) samples from about 1800 sites were collected throughout a 38000 km 2 rectangular area extending from Georgian Bay east to the Ottawa and St. Lawrence Rivers, Ontario, Canada. Lake water alkalinity and pH patterns are similar to the distribution of carbonate components in glacial drift. Carbonate-rich drift derived from the Paleozoic limestone terrain on the northeast flank of the Precambrian Frontenac Arch has been dispersed in a southwestward direction across a variety of non-calcareous metasedimentary and igneous rocks of the Canadian Shield, providing a buffering capacity to lakes situated in granitic terrain. The distribution patterns of mobile trace and minor elements are influenced by geochemical processes associated with subaerial weathering, ground and surface water transport, and the geochemical environment within the lakes themselves. Although composition of the drift is generally reflected by lake geochemistry, these post depositional processes can cause significant variations between patterns derived from the two sample types. Anions and cations such as SO~ , CI-, Na +, and Fexhibit concentration patterns thought to reflect both anthropegenic inputs and natural variations due to differences in the geology. All regional geochemical patterns may show evidence of local enhancement caused by high concentrations of chemically distinctive minerals in drift or nearby bedrock.

Reply to A. S. Dyke's discussion

Abstract In our opinion the amino acid data are consistent with the till/nonglacial stratigraphy. We reject Dykes proposal that the plotting of data in 0.04 increments is appropriate as an unwarranted interpretation that errors are cumulative. We also see no grounds for accepting his alternative interpretation that the groups of amino acid ratios reflect various transport (read temperature) histories of a single population of Bell Sea shells. It is our opinion that the relative sequence of marine incursions in Hudson Bay is reliable and we repeat that the evidence favors one or more deglacial events. We stress that the ages of the units between the Tyrrell Sea and Bell sea end membrers are interpolated and that the chronology of events is currently based on the assumption that the Bell Sea represents marine incursion at the onset of marine isotope stage 5. Dyke has raised a number of points which have concerned us since we started our joint research on the aminostratigraphy of the Hudson Bay Lowlands. The answer to many questions will come, not from the amino acid results per se, but from detailed litho-, and biostratigraphic logging of the thousands of kilometers exposed along the large rivers that drain into James Bay and southwestern Hudson Bay. This work is presently going on. Let us say in conclusion that analysis of a further 63 shells and shell fragments resulted in a virtually identical frequency distribution to that discussed in our paper. We are currently evaluating the stratigraphic integrity of these results. Field expeditions by the Geological Survey of Canada in 1982 and 1983 into the Hudson Bay Lowlands were specifically designed to log new sections and make additional shell collections. We hope to report on these new data in due course.