No abstract

We conducted a field study including a series of cruises over an isolated offshore zebra mussel reef (7-11 m deep) in Western Lake Erie to examine the effect of zebra mussels (Dreissena spp.) on the water column. The horizontal currents over the reef were found to be primarily due to the hydraulic flow and surface gravitational seiches. The turbulence generated by these currents was found to be too weak to fully mix the water column. Although seasonal stratification was not observed, solar heating during the day and intrusions of cold central basin water caused stable stratification of the water column 60% of the time. Results from the seston analysis taken at five depths showed a statistically significant mussel-feeding signature in chlorophyll a and organic seston concentrations measured within 2 m above the reef. Estimates of clearance rates based on field data were consistent with rates measured in a flow chamber using water from the site, which indicated that zebra mussels could remove up to 40% of the total seston. The detection of a zebra mussel-induced concentration boundary layer is due to: (1) reduced vertical mixing as a result of semidiurnal periodic stratification, (2) refiltration of bottom water in zebra mussel populations, and (3) in situ clearance rates that are lower than those observed in the laboratory. Thus, offshore zebra mussel colonies may have less of an effect on the water column than had been previously estimated by simple stirred reactor models, and the role of zebra mussels in the clarification of Lake Erie should be investigated further.

A one-dimensional thermodynamic model of a two-component ice and snow cover is added to an existing one-dimensional lake mixing model. Emphasis is placed on the thermodynamic coupling between the ice and mixing models-absent in previous models. The two-dimensional effects of partial ice cover are incorporated into this one-dimensional framework by using a minimum ice thickness. The model is applied to Lake Laberge, Yukon Territory, and to Babine Lake, British Columbia, for periods covering the formation and demise of full ice cover. The results of the model are compared to snow and ice measurements in the first case and to water column data during the spring period in the second. The comparisons are good, implying that the ice and snow model is performing satisfactorily and emphasizing the importance of the coupling between the ice and the underlying water. The successful simulation of the observed mixed layer under the ice, driven by convective stirring caused by short-wave heating below the temperature of maximum density, is an example of the ability of the model to provide physical insight into processes occurring in lakes.

Field data from two lakes of widely differing geometry and size are analyzed in terms of four nondimensional numbers which allow the principal mixing processes in each lake to be identified. The numbers are based on basin geometry, density stratification, wind stress, and rates of inflow and outflow. The procedure highlights the differences in the dynamics of the two lakes and allows assessment of the validity of the assumption of one-dimensionality. The result is that both lakes were dominated by one-dimensional, but different processes. The dynamics of the epilimnion of the smaller lake were dominated by stirring from surface wind and cooling, whereas shear at the pycnocline was also significant in the larger lake. In neither case did the effects of the earth's rotation, inflow, or outflow generate significant horizontal gradients. A one-dimensional numerical model (DYRESM) was used to simulate the vertical temperature and salinity structures of both lakes over lengthy periods, with good results. The model is based on the parameterization of the important physical processes in a framework of horizontal layers of variable thickness and was applied in both lakes without alteration. The interpretative power of the model is demonstrated by examination of the formation and erosion of a thermal inversion in the larger lake.

Zebra mussels (Dreissena polymorpha) are an invasive species that have been implicated in the reduction of algae stocks in the near-shore environment of western Lake Erie. To determine their basin-wide effects, we applied a twodimensional hydrodynamic and water-quality model for 1994. The model accurately reproduced lake-wide hydrodynamics and water quality. When modeled as true benthic organisms (resting on the bottom), the dreissenids grazed 53% of the western basin May through September net algal growth. This grazing resulted in a ,0.1-mg L 21 reduction in the pelagic algae concentration relative to the case without dreissenids. In comparison, dreissenids grazed 77% western basin net algal growth when the lake was modeled as a fully mixed water column. We found that the biomass grazed was governed by a balance between the timescales of vertical wind-induced mixing and benthic grazing. During calm conditions, weak diurnal stratification (,1uC between surface and bottom waters) was sufficient to suppress vertical mixing, when the mean daily wind speed 4 m above the lake surface (U 4 ) was ,6 m s 21 . These conditions allowed a concentration boundary layer ,1 m thick to form, accounting for the reduced grazing effect relative to the fully mixed case. Entrainment of the concentration boundary layer occurred for U 4 .6 m s 21 (associated with the lake's characteristic 10-d storm cycle) facilitating algae supply to the benthos. We formulated the mean daily biomass grazed in terms of the dreissenid areal pumping rate (a) and U 4 and found that because typically U 4 is ,6 m s 21 , the western basin is weakly stratified thermally and a concentration boundary layer forms when U 4 ,3a or a .2 m 3 m 22 d 21 . The dynamics of both wind-mixing and thermal stratification must, therefore, be considered in mixing models applied to shallow weakly stratified lake basins.