Predicting Mysis relicta vertical distribution in lake Ontario

A model predicting the distribution and diel vertical migration (DVM) of Mysis relicta was constructed based on inferred temperature preferences determined from published feeding experiments and light preferences estimated using observed in situ distributions under isothermal conditions. Model predictions were compared to detailed (on a meter scale) observations of mysid vertical distributions using 420 kHz hydroacoustics at two stations during three seasons in Lake Ontario. The model described the spring and summer distributions but failed to predict fall distributions when the thermocline was below 40m depth. Addition of a term representing avoidance of temperature gradients derived independently from the literature improved predictions substantially during fall. We hypothesize that the avoidance of a temperature gradient may be the result of the presence of fish kairomones in the upper surface waters. Overlap between distributions, predicted by the model, including the temperature gradient and observed distributions was over 60 % for most time periods and seasons. This model can be used to calculate likely overlap between mysid and their zooplankton prey and between mysids and their fish predators based on readily measured temperature and light profiles.