Sources, Transport, and Nutrient Environment of Toxic "Red Tide" Populations in the Western Gulf of Maine

D.M. Anderson, B. Butman, D.A. Fong, P.J.S. Franks, B.A. Keafer, W.R. Geyer, T.C. Loder III, and R.P. Signell

Article from Summer 1995 "Gulf of Maine News," the Regional Association for Research on the Gulf of Maine newsletter

Franks and Anderson (1992) found that the coastal current in the western Gulf of Maine has a strong influence on the distribution and fate of a toxic dinoflagellate, Alexandrium tamarense. They posed the "plume advection hypothesis", which states that a source population of A. tamarense near the mouth of the Kennebec-Androscoggen Rivers is advected long distances along the coast by the freshwater plume of the rivers flowing into the western Gulf of Maine. In order to test this hypothesis and thus to better understand the coupling between dinoflagellate distributions and physical transport processes, a team of investigators from Woods Hole Oceanographic Institution, Scripps Institute of Oceanography, University of New Hampshire, and the U.S. Geological Survey performed a combined field and modeling study of the physics and Alexandrium distribution in the western Gulf of Maine Coastal Current. This study has been supported by the Regional Marine Research Program since 1993.

Measurements of the physics and biology were conducted during two intensive sampling periods in the spring of 1993 and 1994. The measurements included a total of 14 shipboard surveys on The R/V ARGO Maine, the R/V Anderson, and the R/V Jerry Chase, as well as deployment of current meter moorings and surface drifters, and satellite observations of sea-surface temperature.

The pattern of river discharge was similar between 1993 and 1994, and the wind conditions were generally similar ( Figure 1 and Figure 2 .) However there were several strong, downwelling-favorable (northerly to northeasterly) wind events during the period of peak river discharge in 1993, while in 1994 the winds were upwelling-favorable (southerly to southeasterly) during the critical period of peak river discharge. This seemingly minor difference in wind-forcing produced markedly different plume conditions and correspondingly large differences in the number and distribution of Alexandrium cells. In 1993, the plume generally hugged the coast, and virtually all of the fresh water coming out of the western Gulf of Maine rivers could be observed within the southward flowing coastal current, which extended continuously along the coast into Massachusetts Bay. In 1994, there were episodes of southward transport, but they were interrupted by several upwelling events that caused the plume to spread offshore and actually to head northeastward for several days to a week. Much of the fresh water was transported offshore, and the salinity of the coastal waters were considerably higher than in the spring of 1993, even though the freshwater input was the same.

These differences in hydrographic conditions were reflected in marked differences in the Alexandrium sp. distributions. During 1993, the plume carried Alexandrium cells southward, consistent with the plume advection hypothesis, and a number of coastal stations reported toxicity as the cells contaminated the coastal monitoring stations. During 1994, the upwelling conditions resulted in offshore transport of the early background concentrations of Alexandrium sp., preventing the development of a sufficient population to cause nearshore shellfish toxicity. Later, several shellfish sites near Casco Bay also became toxic, but these outbreaks did not persist due to the strong upwelling conditions. In 1994, no toxicity was detected in Massachusetts Bay, suggesting that the persistent upwelling prevented the low-level populations to the north of Cape Ann from seeding the southern waters via advection.

One implication of these observations is that not only the seasonally averaged forcing conditions, but also the timing of these forcing variables, is critical in determining the transport processes and their influence on populations. The subtle connections between run-off, wind-forcing and toxic plankton distribution emphasizes the need for time-dependent models that can actually account for the covariance between the critical forcing variables and ultimately predict the occurrence of red tides.

Several "levels" of modeling are being pursued in this project to understand and ultimately to predict the variations of the physical regime and the associated red tide population. A simple, two-dimensional model was developed to examine the time-variations of offshore extent of the plume (Fong and Geyer, 1994; Fong et al., 1995). This analysis indicates that wind forcing is indeed the principal agent responsible for variations in the width of the plume, but that the basic Ekman dynamics are significantly modified by time-dependence and friction between the plume and the underlying water.

A more computationally-intensive, 3-dimensional model is also being implemented. This is the same model that was used by Signell and co-workers (e.g., Signell et al., 1993) to simulate the flow in Massachusetts Bay. The western Gulf of Maine domain extends from Penobscot Bay to Cape Cod and out to the 100 m isobath. One of the difficulties of implementing a "limited domain" model such as this is the specification of the open boundary condition. We are using the output of a larger scale model of the entire Gulf (Namie et al., 1994) to "drive" the open boundary. Another critical issue is the parameterization of vertical mixing, which is accomplished by a second-order turbulence closure scheme (Mellor and Yamada, 1982). The idealized, two-dimensional model indicated that friction was important, implying turbulence in the stratified shear layer at the base of the plume. The comparison between the 3-D model and observations show discrepancies that may be explained by an underestimate of the friction term. This study of the turbulence parameterization of the model is important both for the predictive skill of this model and for the general problem of mixing in river plumes.

In addition to these physics-oriented model studies, we are also modeling the coupled physics-biology problem ( Figure 3). So far, our simulations have addressed the advection and growth of cells, with a simple, temperature-dependent growth function. These runs have shown that nature is more complicated than the model. One of the difficulties is adequately parameterizing the source function, which we believe to originate from a "seed population" of cysts in the bottom sediments near the mouth of the Kennebec River. Other questions involve the role of physical aggregation at fronts (Franks, 1992), the influence of nutrients on growth, and predation. Ultimately, this coupled modeling effort will lead to a better understanding of the dynamics of the Alexandrium population as well as a potential predictive tool for forecasting "red tide" outbreaks.


Fong, D.A., Geyer, W.R, and R.P. Signell, 1995. Ekman dynamics of the western Gulf of Maine plume. Journal of Marine Systems, submitted.

Fong, D.A., and W.R. Geyer, 1994. Wind-forcing of river plumes. American Geophysical Union Fall Meeting, San Francisco, CA. Abstract: EOS Supplement, 75: 052-7.

Franks, P.J.S., 1992. Sink or swim: accumulation of biomass at fronts. Marine Ecology Progress Series, 82:1-12.

Franks, P.J.S. and D.M. Anderson, 1992. Alongshore transport of a toxic phytoplankton bloom in a buoyancy current: Alexandrium tamarense in the Gulf of Maine. Marine Biology, 112:153-164.

Mellor, G.L. and T. Yamada, 1982. Development of a Turbulence Closure Model for Geophysical Fluid Problems, Review of Geophysical Space Physics, 20: 851-875.

Naimie, C.E., Loder, J.W., and D.R. Lynch, 1994. Seasonal-Variation of the 3-Dimensional Residual Circulation on Georges Bank, J. of Geophysical Research-Oceans, 99(8): 15967-15989.

Signell, R.P., Jenter, H.L., and A.F. Blumberg, 1994. Modeling the Seasonal Circulation in Massachusetts Bay, Proceedings of the 3rd International Conference on Estuarine and Coastal Modeling III, Publ by ASCE, NewYork, NY, USA., p 578-590.