Prior to the 1950s, the mighty Missouri River would annually flood. The slow march of waters across the valley distributed fertile sediment for agriculture, but also fueled a mutualistic relationship with the river valley’s vast stands of cottonwood trees. These trees support wildlife, provide the river with important nutrients to fuel aquatic life, and support a wide array of ecosystem services to the surrounding towns and cities. However in the mid-1950s, the Army Corps of Engineers completed a multi-decade flood control project. The Corps dammed the upper Missouri River from its headwaters in the Rockies to Yankton, SD, a town located near the tri-state border of South Dakota, Nebraska, and Iowa. How these dams change the relationship between the river and its valley is a question that geography Ph.D. student Samantha Greene and the late professor James Knox explored.
Using hydrologic data from the Corps of Engineers and U.S. Geological Survey, remotely sensed imagery from the Missouri River Institute, and their own field data, Greene and Knox showed that dam-mediated changes in river hydrology is leading to the invasion of red cedar in cottonwood stands immediately downstream of the final Missouri River dam. While red cedars are typical of the uplands in the region, they are not trees associated with river valley habitat. The invasion of red cedar can change the successional trajectory of these cottonwood forests and alter or erase many ecosystem services that the cottonwood forests provide.
Such invasions along regulated rivers are occurring throughout the United States and the world. Owing to this, the researchers decided to take their study a step further and see if they can use the variety of fluvial surfaces and landforms a river created prior to regulation to predict where invasion is likely to occur. Using their field data and remotely sensed information, they created Bayesian models that accurately predicted locations of red cedar invasion.
Greene is optimistic that land managers and other scientists could apply similar methods to quickly predict and map locations susceptible to invasion. With a thorough understanding of river hydrology-ecology relationships, remotely sensed data that provide vegetation and fluvial surface characteristics, and a rapid field assessment to ground truth the remotely sensed data, managers can use a Bayesian model in order to efficiently and economically map invasion and decide which areas to focus management and restoration efforts.
Samantha Green and Jim Knox’s research findings are published in the journal Geomorphology, accessible online here.