A dynamic model to predict responses of millets (Echinochloa sp.) to different hydrologic conditions for the Illinois floodplain-River

Millets grow on floodplain mud flats exposed when seasonal floods recede, and the seeds of this plant are an important food source for waterfowl during their spring and autumn migrations in the Mississippi Flyway. Productivity of millets along the Illinois River has declined because of unnaturally frequent floods that inundate the mud flats and drown the plants during the summer growing season. These small floods are caused by operation of the navigation dams on the main channel and by alterations of the floodplain and tributary watersheds and channels. Predictive models are needed to evaluate the most cost‐effective combination of approaches for restoring plant productivity. We developed a moist‐soil plant model that simulates millet growth on 1 m2 in response to daily water levels during the summer growing season. The model responds to daily water depth, flood timing (within the growing season), and flood duration, and was qualitatively verified using historical (1938–1959) water levels and plant coverage for three areas along the Illinois River. In the absence of untimely floods, the model predicts net above‐ground primary productivity of ∼500 g m−2 yr−1 and plant heights of up to 130 cm by the end of the growing season. As expected, growth declines with decreasing land elevation or with more frequent flooding (or a shorter duration of the dry period) at the same elevation. A dry period of >85 days is required to achieve at least 50% of maximum production during the growing season, which is somewhat longer than the 70‐day recommendation based on reported field observations. The model predictions of plant success or failure agree with historical observations, indicating that water regime is a major factor limiting plant success. Copyright © 2004 John Wiley & Sons, Ltd.