| Sensitivity analyses are performed on numerical algorithms which attempt to simulate the active vertical motion of nekton. Stochastic Monte Carlo techniques are applied to simulate swimming, swarming and settling. Vertical migration is stimulated by assumed irradiance below the water surface and by a light level that stimulates swimming. Since nekton typically have a planktonic egg stage, sensitivity analysis is conducted on Monte Carlo derived settling velocities. Data concerning Morone saxatilis are used to demonstrate the algorithms using gross estimates of swimming speeds (estimated from video image processing) and using settling velocities of eggs measured in an experimental chamber.;Vertical distributions of active particles (simulated nekton) are constructed using output from the modeled processes. The distributions do not include the effects of physical processes which may be important in specific estuarine compartments. However, the results of the model sensitivity analysis indicate the relative importance of modeled biological and optical variables as conceptualized in the model. Model variables that had the greatest influence upon constructed distributions of the nekton were, in order, the assumed water depth, a vertical swimming dispersion constant, a mean sustainable swimming constant, assumed estuary width, nekton swimming speed, a light attenuation coefficient and the model time step.;The underlying hypothesis for this research is that it is possible to simulate the active vertical motion (swimming and swarming) and the settling of nekton in an idealized estuary. This research represents the first attempt to address this topic. The current data and algorithms are simplified, limiting model validations and practical applications. The modeling and research methods provide new insights into the relative importance of modeled variables that may ultimately affect the vertical distributions of nekton in the estuarine environment. The results may be useful for future data collection and research, and the model can be refined with each new increment of information. |
