Sensitivity of 1-D hydraulic models of fish passage in culverts to descriptions of fish swimming performance
Nixon, Kyle Marshall.
MetadataShow full item record
One way culverts become barriers to the upstream movement of fish is by creating excessive velocities exceeding a fish's swimming ability. FishXing, a common tool for indirectly assessing fish passage, uses fish swimming ability information with one-dimensional culvert hydraulics to predict barrier status of culverts. However, since fish swimming ability data is scarce for many fish species, predictions of a culvert's barrier status can be inaccurate and overly conservative, possibly leading to misclassification or uneconomical design. Additional fish swimming ability research is necessary to strengthen these models. The primary goal of this study was to determine the effects of different swimming ability algorithms on velocity barrier flow rates predicted by one-dimensional culvert hydraulics models. A one-dimensional culvert hydraulics model was created in Visual Basic. This model was designed to mimic FishXing's fish swimming algorithm, or use more complex fish swimming algorithms. Three diverse test culverts were selected to show how varying culvert properties (length, geometry, flow regime, and embedment) influences which fish swimming ability algorithm most affects the predicted velocity barrier flow rate. A "test fish" was designed based upon fish swimming ability literature. Each culvert was subjected to six tests, each testing the sensitivity of a particular fish swimming algorithm. This study determined that for different types of culverts, different components of fish swimming ability algorithms substantially affect the velocity barrier flow rate. The study needed only three test culverts to show that accurate quantification of the fish species' burst speed, burst duration, the burst speed/duration relationship, prolonged swimming speed, and constant deceleration time from burst to prolonged speed is necessary to model diverse fish passage situations. This study also showed that if a fish has a substantial deceleration time, a constant deceleration is probably sufficient to model it. In the future, if programs like FishXing adapt to include deceleration in fish swimming models, constant deceleration is an adequate addition. With this analysis, fish swimming ability variables substantially affecting fish passage were determined. The study can be used to guide further research so swimming ability studies can gather swimming data that is most crucial to predicting fish passage.