CGRG Bibliography of Canadian Geomorphology
Search Results
Author : Brown, P.A.; Kennett, J.P.; and Teller, J.T.
Date : 1999.
Title : Megaflood erosion and meltwater plumbing changes during last North American deglaciation recorded in Gulf of Mexico sediments: reply.
Publication : Geology
Issue : 27(5):
Page(s) : 479-480.
Abstract
The Brown and Kennett (1998) paper attempted to link the continental record of North American meltwater runoff during the last deglaciation with changes in marine proxies (siliciclastic grain size, clay minerals, oxygen isotopes, and reworked microfossils) observed in well-dated Gulf of Mexico (Orca Basin) sediments. Our study employed recent syntheses of meltwater runoff variability based on terrestrial studies of proglacial lakes and Midwest fluvial sequences, such as summarized by Teller (1987, 1990, 1995), Kehew and Lord (1987), and Knox (1996), as a chronological and conceptual framework to guide our interpretations of megaflood erosion and meltwater plumbing changes. Karrow's Comment focuses on apparent discrepancies in the details of our correlation with deglacial events documented in the Great Lakes basin. He stresses the complexity of deglacial drainage history and associated difficulties in correlating with the complicated Great Lakes chronology. It is not surprising that disagreements exist for the major events that we are trying to resolve and correlate at century scale, given the uncertainties about the timing of continental events and the vast distances and different types of records involved. Indeed, the chronological ambiguities in the continental record of the last deglaciation underscore the complementary value of marine sediments in providing well-dated, uncomplicated proxy sequences sensitive to changes in large-scale continental runoff processes. Although dates cited by Karrow differ from those used in our study, we maintain that the general sequence of continental drainage events known from terrestrial data is consistent with the Orca basin record. The latest terrestrial-based synthesis of deglacial drainage history (Licciardi et al., 1999) generally supports the marine evidence for at least three erosive megaflood episodes that flowed south to the Gulf of Mexico at about 15, 13.5, and 12.6-12.0 ka. The two earlier episodes correspond to the beginning and end of the Port Bruce glacial readvance (15.2-13.5 ka), while the later, much larger episode occurred during the Port Huron readvance (13.0-12.3 ka), both spanning intervals when meltwater was redirected southward to the gulf from its eastward route to the North Atlantic (Licciardi et al., 1999). Further support for our megaflood model comes from terrestrial evidence for deep erosion of clay-rich midcontinental terraces interpreted by Knox (1996) to record the passage of several massive flood bursts down upper Mississippi River tributaries before 12 ka. This interval most likely correlates with the pulse of fine-grained sediment, high sedimentation rates, most depleted d 18 O, and reworked nannofossils recorded in the Orca basin at 12.6-12.0 ka. We believe that abrupt rerouting from large drainage basins more likely explains meltwater runoff changes inferred from the marine record than catastrophic failure of ice dams. Abrupt switchovers from eastward drainage to gulf routing, when discharge increased from near zero values to 0.02-0.03 Sv (10 6 m 3 /s) (Licciardi et al., 1999), were sufficiently large to account for inferred high-magnitude flows through the southern outlets of proglacial lakes. A substantial body of continental data suggests that following the climax of meltwater input into the gulf at ca. 12 ka, Lake Agassiz drainage continued to flow southward until it was rerouted eastward at 11-10.7 ka (Clayton and Moran, 1982; Fenton et al., 1983). This switchover was abrupt and resulted in reduced meltwater flow to the gulf, as indicated by oxygen isotopic and grain size proxies in the Orca basin. Although Karrow correctly states that the Great Lakes record does not seem to indicate a strong burst of outflow from Lake Agassiz at 11-10 ka, many studies do identify a new source of outflow waters at that time, which researchers ascribe to Agassiz (e.g., Rodrigues and Vilks, 1994; Lewis et al., 1994; Colman et al., 1994). The record in the Agassiz basin strongly indicates an abrupt decline in lake levels there during this interval, and it has been concluded that at ca. 11-10 ka, outflow from the extensive Agassiz basin (2 000 000 km 2 ) was rerouted away from the Gulf of Mexico into the Great Lakes (e.g., Clayton and Moran, 1982; Fenton et al., 1983; Teller, 1987). We have considered the possibility that clay mineral changes in deglacial-age Orca basin sediments may represent regional North American provenance shifts, but we find no evidence in the available clay data to support such a conclusion. The only interval associated with significant shifts in clay mineral assemblages occurred at ca. 15 ka, when higher abundances of illite were replaced by a dominance of smectite, probably reflecting an initial flushing of glacial sediments from the ice margin near the beginning of significant meltwater flow to the south. This general pattern of abundant smectite and lesser amounts of illite and other clays continued through the late-glacial period into postglacial times. Since the clay mineral assemblage remained relatively constant even during inferred megaflood episodes, there is no need to invoke major changes in regional sediment sources at these times. Any such provenance changes, if they occurred, are unresolved by the marine record. Rather, we observe a significant decrease in clay-size quartz and an associated increase in total clay abundance during the main meltwater interval from 14.0 to 11.0 ka. This we interpret to represent increased sediment contributions from clay-rich terraces along upper Mississippi River tributaries diluting quartzose sediment from the main channel (Knox, 1996). As expected, total clay abundance varies with siliciclastic sediment grain size changes such that a major clay abundance peak accompanies the ultrafine megaflood sediments deposited in the Orca basin at 12.6-12.0 ka. In summary, we believe that the terrestrial chronological framework of North American deglaciation is consistent with our interpretations of variations in Gulf of Mexico marine proxies and that the Orca basin record remains a valuable link in reconstructing meltwater drainage history.
Bibliography of Canadian Geomorphology