Updating an invasive fish and native fish passage model for locks and dams

This project will create an updated version of Computational Fluid Dynamics Agent-Based (CFD-AB) fish passage model using new field data that can better help stop invasive carp while allowing native fish to pass through Mississippi River locks and dams. 

The new field data is being generated by an ongoing field study of fish behavior and passage at Lock and Dam 2. Parameters on fish behavior will then be updated in the CFD-AB fish passage model developed earlier by [Zielinski et al., 2018] to improve it. We will use this updated CFD-AB model to predict fish passage for invasive carp (silver carp, common carp) and two native fishes (channel catfish, lake sturgeon) at two model lock and dams (2 and 8). The updated CFD-AB model will allow to determine optimum spillway gate positions to stop invasive carps at these sites. We will share these new data with the U.S. Army Corps of Engineers and the Minnesota DNR.


As of January 2019, the code development and validation of the Computational Fluid Dynamics - Agent Based model is complete. The accuracy of the fish swimming calculation was improved. Numerous simulations with common carp, which were trying to pass through Lock and Dam 2, have been performed and provided excellent comparisons between the model and the experimental field data. A modification that considers fish swimming both up and downstream has been finished as well, and includes areas that fish may be attracted to, such as resting, migration, and feeding zones.

Next, this model will be used to create predictions of how both invasive carp and native species pass through locks and dams and suggestions for new gate settings that could stop carp and let native fish pass will be made.


  1. The computational code CFD-AB directed to enhance the simulation of swimming fish trying to pass through the navigation dams was updated/developed. The analysis of different fish passage index (FPI) showed that the values of FPI for the modified algorithm for a model channel were greater than the FPI of the original algorithm at about 16%. At this moment, no essential differences in fish passage index for the original and modified model at LD2 and LD8 have been found. This effect can be explained by the special gate adjustments, which generate a rather high fluid flow prevented fish to pass through the dams. In other words, in case of blocking invasive species, the modified algorithm does not change the final results of FPI at LD2 and LD8. But the modified algorithm could play a positive role to help native fish to pass through the navigation dams in the case of changing gate adjustments leading to decrease flow velocity.
  2. The modified algorithms now account for more realistic fish behavior, including placement of “attraction points,” such as resting zones characterized by low recirculating fluid flow. These parameters have been informed by the literature and unpublished field data collected on other projects.
  3. Based on investigations of (Larson, et al., 2017, Kokotovich et al, 2017) it was reported that the “Invasion Front" is currently positioned in southern Iowa between Pool 14 and Pool 16. Therefore, the strategy of blocking bigheaded carp at Lock and Dams of Minnesota should be reconsidered. It is well documented that the navigational dams have significantly altered the movement, spawning, feeding and other activities of native fish (Wilcox et al. 2004). Hence, managers should consider alternative strategies whereby navigation dams are adjusted to help native fish pass, instead of blocking invasive fish. This strategy could help with ecosystem restoration efforts and potentially improve natural resistance to invasion by bigheaded carps. To evaluate this strategy, simulations of walleye passing through LD2 have been executed. It has been shown that by changing gate adjustments, FPI=4% is for the original algorithm and FPI=12% for the modified algorithm. We have to note, that for current gate adjustments from USACE the FPI=0% for original and modified CFD-AB models. By utilizing active monitoring data of bigheaded carp managers could instantly change gate adjustments at LD2-LD8 by using our CFD-AB approach if the invasion front threatens Minnesota.

Project manager: Anvar Gilmanov   

Funded by: Environment and Natural Resources Trust Fund as recommended by the Legislative-Citizen Commission on Minnesota Resources

Project start date: 2018

Project end date: 2019