Modelling of Dreissenid Mussel Impacts on Lake Michigan

Abstract

Invasive dreissenid mussel appear to have profoundly altered Great Lakes food webs and nutrient cycles during the past several decades. Recent declines of phytoplankton were supposed to be highly related with the increase of mussel population. These phytoplankton declines were further found to be coincident with declines in the abundance of planktivorous fish. In addition, the resurgence of Cladophora in Great lakes was estimated to be associated with the high density colonization of mussels. More light is available at lake bottom due to the mussels’ graze effect. The mussels further promote Cladophora growth by fertilizing it with nutrient-rich excrement. And their shells provide hard rocky surface on which the algae can grow. Numerical models were applied in this thesis to investigate the ecosystem of Lake Michigan, specifically to explore how invasive mussels affect the nutrient dynamics and distribution of phytoplankton. A 1D biophysical model was successfully developed for the mid-depth zone and it can be easily expanded to any other area with stable horizontal homogeneity. Coriolis force, momentum sink as well as wave effects were fully considered in order to well resolve the physical mixing. The phytoplankton simulation results of 2013 with the 1D model revealed significant reduction of the phytoplankton in the water column with the mussels incorporated at the bottom, 24% and 8% reduction for spring and summer respectively. The mussels’ filtering effect over lake wide was investigated with the 3D biophysical model as well. The physical parameters were solved with FVCOM while the biological part was simulated based on the NPZD model. For the simulation in 2012, the total amount of phytoplankton for the whole lake was estimated to decreased by 6% in spring and 2% in summer with mussel model included. A high concentration layer of nutrient and low concentration layer of phytoplankton were observed with the existence of mussels which was in agreement of the recent field observations. Meanwhile, the vertical distribution of phytoplankton with and without mussels along the Lake Ferry Express transect was discussed for the nearshore and offshore exchange. The mussels in our simulation results proved an important role in keeping a low concentration of phytoplankton in the offshore area where no mussel colonization was found. Finally, The Cladophora particle tracking simulation was conducted based on the physical FVCOM output for the Sleeping Bear Dune national park area. The model results suggested the sloughed dead Cladophora particles were very likely to deposit at the nearby but deeper area comparing with their original positions.