LATE QUATERNARY UPPER MISSISSIPPI RIVER ALLUVIAL EPISODES AND THEIR SIGNIFICANCE TO THE LOWER MISSISSIPPI RIVER SYSTEM

The period in the Upper Mississippi Valley (UMV) from about 25 000 yea rs B.P. until the time of strong human influence on the landscape begi nning about 150-200 years ago can be characterized by three distinctly different alluvial episodes. The first episode is dominated by the di rect and indirect effects of Late Wisconsin glacial ice in the basin h eadwaters. This period, which lasted until about 14 000 years B.P., wa s generally a time of progressive valley aggradation by a braided rive r system transporting large quantities of bedload sediment. An island braided system evolved during the second episode, which extended from about 14 000 to 9000 years B.P. The second episode is associated with major environmental changes of deglaciation when occurrences of major hoods and sustained flows of low sediment concentration from drainage of proglacial lakes produced major downcutting. By the time of the beg inning of the third episode about 9000 years B.P., most vegetation com munities had established their approximate average Holocene locations. The change of climate and establishment of good vegetation cover caus ed upland landscapes of the UMV to become relatively stable during the Holocene in comparison to their relative instability during the Late Wisconsin. However, Holocene remobilization of Late Wisconsin age sedi ment stored in tributary valleys resulted in a return to long-term upp er Mississippi River aggradation. The dominance of Holocene deposition over transportation reflects the abundance of sandy bedload sediment introduced from tributaries and the situation that energy conditions f or floods and the hydraulic gradient of the upper Mississippi River ar e much less for the Holocene than they were for the Late Wisconsin and deglaciation periods. Outburst floods from glacial lakes appear to ha ve been common in the UMV during the Late Wisconsin and especially dur ing deglaciation. Magnitudes for the Late Wisconsin floods are general ly poorly understood, but an estimate of 10 000-15 000 m(3) s(-1) was determined for one of the largest events in the northern UMV based on heights of paleo-foreset beds in a flood unit deposited in the Savanna Terrace. For comparison, the great hood of 1993 on the upper Mississi ppi River was about 12 000 m(3) s(-1) at Keokuk, Iowa, near the Des Mo ines River confluence where it represented the 500-year event in relat ion to modern flood series. Exceptionally large outburst floods derive d from the rapid drainage of pro-glacial Lake Michigan and adjacent sm aller proglacial lakes between about 16 000 and 15 500 years B.P. are a likely cause of the final diversion of the Mississippi River through the Bell City-Oran Gap at the upstream end of the Lower Mississippi V alley (LMV). The largest outburst hood from northern extremities of th e UMV appears to have occurred between about 11 700 and 10 800 years B .P. when the southern outlet of Lake Agassiz was incised. Based on the probable maximum capacity of the Agassiz flood channel 600 km downstr eam near the junction of the Wisconsin and Mississippi Rivers, the Aga ssiz flood discharge apparently did not exceed 30 000 m(3) s(-1). Howe ver, if the Agassiz flood channel here is expanded to include an incis ed component, then the flood discharge maximum could have been as larg e as 100,000 to 125 000 m(3) s(-1). The larger flood is presently view ed as unlikely, however, because held evidence suggests that the incis ed component of the cross-section probably developed after the main Ag assiz flood event. Nevertheless, the large Agassiz flood between about 11 700 and 10 800 years B.P. produced major erosional downcutting and removal of Late Wisconsin sediment in the UMV. This flood also appear s to be mainly responsible for the final diversion of the Mississippi River through Thebes Gap in extreme southwestern Illinois and the form ation of the Charleston alluvial fan at the head of the LMV. After abo ut 9000 years B.P. prairie-forest ecotones with associated steep seaso nal climatic boundaries were established across the northern and south ern regions of the UMV. The general presence of these steep climatical ly sensitive boundaries throughout the Holocene, in concert with the n atural tendency for grasslands to be especially sensitive to climatic change, may partially explain why widespread synchroneity of Holocene alluvial episodes is recognized across the upper Mississippi River and Missouri River drainage systems. Comparison of estimated beginning ag es of Holocene flood episodes and alluvial chronologies for upper Miss issippi River and Missouri River systems with beginning ages for LMV m eander belts and delta lobes shows a relatively strong correlation. At present, dating controls are not sufficiently adequate and confidence intervals associated with the identified ages representing system cha nges are too large to establish firm causal connections. Although the limitations of the existing data are numerous, the implicit causal con nections suggested from existing information suggest that further expl oration would be beneficial to improving the understanding of how uppe r valley hydrological and geomorphic events are influencing hydrologic al and geomorphic activity in the LMV. Since nearly 80% of the Mississ ippi River drainage system lies upstream of the confluence of the Miss issippi and Ohio Rivers, there is a strong basis for supporting the id ea that UMV fluvial activity should be having a strong influence on LM V fluvial activity. If this assertion is correct, then the traditional assignment of strong to dominant control by eustatic sea level variat ions for explaining channel avulsions, delta lobes, and meander belts in the LMV needs re-examination. A stronger role for upper valley fluv ial activity as a factor influencing lower valley fluvial activity doe s not disregard the role of eustatic sea level, tectonic processes or other factors. Rather, upper valley fluvial episodes or specific event s such as extreme floods may commonly serve as a ''triggering mechanis m'' that causes a threshold of instability to be exceeded in a system that was poised for change due to sea level rise, tectonic uplift, or other environmental factors. In other situations, the upper valley flu vial activity may exert a more dominant control over many LMV fluvial processes and landforms as frequently was the case during times of gla cial climatic conditions.