Why Lake Itasca may not be the headwaters of the Mississippi River

ARE THE HEADWATERS OF THE MISSISSIPPI RIVER IN SOUTH DAKOTA?

By Wendell A. Duffield

November 2012

A "no" answer to the question posed by the title of this essay likely seems obvious to anyone who has been exposed to even just a smattering of education in current geography. Check any reputable published map of rivers in Minnesota and you will see that the Mississippi originates at Lake Itasca, in the northern part of that state. From there, a small stream meanders first northward, and then eastward and finally southward, growing in size along the way, until it joins one of the state's other large rivers, the Minnesota, in the twin cities of Minneapolis and St. Paul.

Until I became a professional geologist, via a B.A. from Carleton College followed by MS. and Ph.D. degrees from Stanford University, I never thought to question the validity of the Itasca origins for the Mississippi. Shucks, as a six-year-old kid I even successfully jumped from stone to stone placed across the stream outlet from Itasca at the site called the source of the Mississippi. I, similar to thousands of nimble-footed people before and after me, could therefore congratulate myself for jumping across the mighty Mississippi without even getting my feet wet! But now, nearly seven decades later, I know enough of the geologic story behind the development of Upper Midwest river systems to suggest that the Little Minnesota River originating near the South Dakota town of Veblen is a candidate for the headwaters of the Mississippi. Look out, Itasca!

I'm a son of South Dakota, born in Sisseton, which is not far from Veblen. OK. I'm not truly a South Dakotan. My home town is Browns Valley, Minn., barely across the state boundary ten miles east of Sisseton. Browns Valley had neither doctor nor hospital in May of 1941. And Mother was not interested in home birthing. So Dad drove her up and out of town to the nearest suitable medical facility. The valley in which that town is nestled is a key feature supporting my contention that South Dakota can lay claim to the source of the Mississippi. Here's how the background story goes.

Chapter 1: Circa 14,000 to Circa 12,000 Years Ago

The most recent North American continental glacier covered almost all of Canada (maybe a few mountain peaks poked through out west) and most of the northern tier of our USA states. This huge mass of ice began to melt and retreat slowly northward around 14,000 years ago. With continuing melt and retreat, a certain gigantic lake formed in front of the glacier. At its maximum size, this lake (named Agassiz for a famous nineteenth century Swiss glaciologist who carried out much of his research in the United States) covered much of northwestern Minnesota, eastern North Dakota, a bit of northeastern South Dakota and a whole lot of Canada.

Along its southerly-facing margins, Lake Agassiz's water was held in primarily by long ridges of rock debris (called glacial moraines) that had been pushed and carried into place by the glacier. To the north, the retreating front of the glacier itself completed the shoreline encompassing the lake.

With more and more ice melting and a concomitant rise in lake level, the rocky moraine deposits on the south were overtopped. About 12,000 years ago, water began to flow through a breach outlet at the south tip of the lake and quickly eroded a broad deep channel across the landscape. This river (the forerunner of today's Minnesota River) would much later be named glacial River Warren for a mid-nineteenth-century land surveyor of the region. Small tributaries began to feed in from both sides. One would much later be named the Little Minnesota River and flow smack dab through my hometown.

The outflow through glacial River Warren was of gigantic epic proportions. At its maximum, the volume of water per unit time (often reported for rivers as cubic feet per second) was far greater than that pouring into the Gulf of Mexico at the mouth of today's Mississippi River. This massive flow eroded out an impressive valley — locally as much as 400 feet deep and four miles wide — the pathway for today's Minnesota River. Further downstream, glacial River Warren's water was largely responsible for creating the valley that contains today's Mississippi River southward from the Twin Cities area.

While glacial River Warren was sculpting this landscape, a much smaller tributary was feeding in from the north at what would much later become the site of the Twin Cities. This tributary today is recognized as the furthest upstream reach of the Mississippi. However, the volume of flow and erosive power of glacial River Warren were so much greater than those of this tributary that downward erosion of the bed of the tributary could not keep pace with that of Warren. I'll explain more about the possible significance of this disparity later.

Chapter 2: Circa 12,000 to Circa 9,500 Years Ago

The melting front of the glacier eventually retreated far enough to open outflow channels eastward and northward from Lake Agassiz, with connections to the North Atlantic and Arctic Oceans. For some uncertain period of time, water simultaneously escaped at the east north and south outlets of Agassiz. With continuing outflow, about 9,500 years ago the lake surface dropped and stayed at a level below the elevation at the head of glacial River Warren, the future site of Browns Valley. After that, Warren was fed simply by rainfall and snowmelt, including input from downstream tributaries.

Two shallow finger-like lakes, which would much later be named Traverse and Big Stone, occupied part of Warren's valley near its now-dry connection to Agassiz. Many Native Americans must have watched in awe through this evolution from gigantic river to a couple of small lakes. Imagine the colorful and perhaps mystical tales of mind-challenging oral history passed from generation to generation!

One of these early inhabitants of the region was ceremoniously buried with some of his possessions in a gravel bar of River Warren. It's uncertain what the depth of flow, if any, was at that time; the top of the gravel bar is below the nearby rim elevation of Warren's channel. The surface of Lake Agassiz must have been significantly below its maximum elevation by burial time.

The burial site is within my hometown. The human remains are called the Browns Valley Man, and were discovered in 1933 by William Jensen, a local amateur archaeologist. Carbon-14 dating of the human bones suggests that the Browns Valley Man is one of the oldest documented human inhabitants known for North America.

Several earlier generations of Native Americans certainly viewed the river when it was at full strength, many times the size of the present-day downstream Mississippi counterpart. Relatives and other friends on opposite river banks would have had a difficult time getting together for social occasions. There would have been no wading across, even with a gravel bar in mid-channel!

Chapter 3: Circa 9,500 Years Ago to the Twenty First Century

Perhaps not much less than 9,000 years ago, the glacier had melted northward and Agassiz had drained enough to expose the system of lakes and rivers across the Upper Midwest and adjacent Canada in essentially the pattern of surface waters that decorate the landscape today. Glacial River Warren was no longer tapping into Lake Agassiz water. The area that would become Minnesota had its 10,000 lakes. Lake Traverse and Big Stone Lake were long shallow ponds, remnants of a once mighty through-flowing glacial River Warren.

As one result of land reconfiguration caused by glacial scouring, the site of Browns Valley emerged as a north/south drainage divide. The Little Minnesota River flows in there from the west (South Dakota) on its path to Big Stone Lake. It has deposited a low delta on which, together with the burial gravel bar, most of the town is built. The river commonly floods much of the town during spring snowmelt.

At drier times of the year, from some spot near the house where I grew up, half of a theoretical raindrop flows northward while the other half heads south. Via Traverse and the Red River, the north half will arrive at Lake Winnipeg, the largest remnant of Agassiz, and continue on to Hudson Bay. Via Big Stone Lake, the other half will get to the Gulf of Mexico by way of the Minnesota and Mississippi Rivers.

But wait! Could the trip of that semi-droplet be via the true Mississippi River all the way southward? Some evidence contained in the landscapes of the region argues for a yes answer.

Chapter 4: Veblen Versus Lake Itasca

Consider the amount of water that the Minnesota and Mississippi Rivers bring to their confluence at the Twin Cities. The amount varies with seasonal and geographic differences for each of the two watersheds. Averaged over years, the discharge of the Mississippi is somewhat larger. I view this difference as analogous to the relative sizes of a couple of sibling children, rather than that between a parent and child.

As my Grandpa Weeks used to say about comparisons, "Look carefully at the facts and think about what you see; then you shouldn't mistake a branch for the trunk of a tree." If he were still with us, I'm pretty sure Grandpa would decide that there is no substantive bigger-is-better argument to justify using the name Mississippi for the river path from Itasca to the Twin Cities. He'd more likely suggest that Itasca and Minnesota Rivers are quasi-equal tributaries joining forces at the Twin Cities to create the Mississippi. I vote for that idea. With their greater age and experience, Grandpas are commonly wiser than the younger generations!

Now compare the size of the river valleys upstream of their confluence. As mentioned earlier, parts of the Minnesota are up to four miles wide and 400 feet deep. Thank you, glacial River Warren! Meanwhile, the valley of the Mississippi between Itasca and the Twin Cities is like a surficial scratch on an otherwise nearly planar surface.

Next ponder the fact that the Mississippi River within Minneapolis includes the 200-foot-high St. Anthony waterfall (modified in the mid-20th century by a concrete retaining walls plus locks to permit through-going boat traffic). The falls originated, near the present location of Fort Snelling, where the Minnesota and Mississippi Rivers join, in part because the pace of erosional lowering of the river bed by water coming down from Itasca was far less than that of the bed of River Warren. Once Warren lost water input from Lake Agassiz and river level dropped, the bed of the Mississippi at the confluence was left hanging as a waterfall. Grandpa's thoughts about branches and trunks may fit this situation. During the time since that waterfall first became exposed (around 9,000 years ago), it has migrated upstream stepwise, by erosional undercutting at the base that triggers repeated collapse at the top, to its current location in Minneapolis.

Finally, entertain the fact that a few miles downstream from its source at Lake Itasca, the Mississippi flows into the west side of a large lake, Bemidji, from which water finds a river outlet on the east side and eventually heads south to the Twin Cities. Similarly, water of the Little Minnesota River flows into the north end of Big Stone Lake and finds a river outlet at the south end where it is called the Minnesota River. Should we call the stretch of the Mississippi between Itasca and Lake Bemidji the Little Mississippi? Should the Little Minnesota be renamed Minnesota? Hmmmm.

So readers, now you have some new(?) bits of information to chew on. After I masticated and digest the facts, I wanted to rename the stretch of the Mississippi upstream of the Twin Cities as the Itasca River. On my new map, the name Mississippi replaces Minnesota and Little Minnesota.

There are other possibilities. Maybe the Little Minnesota and Minnesota Rivers should retain their present identities and join the Itasca River at the Twin Cities to become the Mississippi? I invite you to get creative with friends and discuss possible name changes at your local chat-and-chew, over a cup of coffee and a slice of homemade apple pie.

However any of us may feel about river names, I understand that an official change for labels of any part of the current Mississippi River is unlikely. Serial approval of local, then state and finally federal agencies would be needed. Politics would surely intervene, especially because the interests of two states are involved.

Change can happen, though. For example, the name for the part of today's Colorado River that is in fact within the state of Colorado was changed from Grand to Colorado in 1921, at the request of the state.

Whatever! I hope South Dakotans, and the citizens of Veblen and Browns Valley in particular, will appreciate and spread the word about the fascinating history of lakes and rivers that have shaped much of their part of the Upper Midwest during the past few thousand years. If he could speak, I think the Browns Valley Man would have a lot to say on this topic. Ditto for my Grandpa Weeks. He was a wise and patient no-nonsense dirt farmer near Peever, S.D., who lived to love and understand his land and how it came to be.