Years ago, many attributed sudden red tide outbreaks to the rapid, in situ growth of the causative algae, yet numerous studies have since shown that much of the variability in the distribution of HABs can be explained by large-scale advective processes such as wind-driven, buoyancy-driven, and tidally-generated motions of water masses. Several prominent examples illustrate the importance of this type of transport, the different mechanisms involved, and the areas where our knowledge is deficient.
The Rias Bajas of northwest Spain are a group of oceanic bays noted for their prolific production of blue mussels. This productivity is due in large part to intermittent enrichment of the rias by nutrient-rich deep water during upwelling, driven by persistent winds from the north [ Otto, 1975; Tenore and Gonzalez, 1975]. Beginning in 1976, these mussels have been affected by outbreaks of PSP which are often quite sudden; toxicity can rise from undetectable to extremely high levels in a few days. Fraga et al., [1988] argue that the PSP toxicity is linked to the relaxation of upwelling following a change in wind direction and speed. Upwelling is driven by north or northeast winds, but winds from the south or west cause downwelling that transports offshore surface waters into the rias, carrying established dinoflagellate populations with it. The resulting changes in the composition of the phytoplankton community are not due to in situ growth of the red tide algae, but instead to the transport or delivery of blooms that originated elsewhere.
Documentation of this linkage between PSP outbreaks in Spain and upwelling relaxation relied heavily on quantification of the degree of upwelling using an index [ Bakun, 1973] representing the onshore or offshore transport of surface water by winds. This ``upwelling index'' is a simple calculation from meteorological pressure fields, but it has tremendous practical importance since it can serve as the basis for a predictive capability for toxic blooms [ Fraga, et al., 1988]. Further research is needed to confirm the validity of this concept in the Spanish rias [ Figueiras and Pazos, 1991] and to better define other necessary conditions. Nevertheless, the concept of linking toxic outbreaks to easily measured meteorological or oceanographic parameters is compelling and worth exploring, not only in Spain, but in other areas of the world where seasonal upwelling is a prominent feature.
In this context certain similarities between the hydrography, meteorology, and patterns of PSP in California and northwest Spain are noteworthy. In both areas: a) the dominant hydrographic feature is coastal upwelling, driven by persistent northerly winds; b) sudden outbreaks of PSP toxicity occur, with toxicity increasing far faster than is possible from localized, in situ growth of the causative dinoflagellates; and c) PSP typically occurs during months when a cessation or relaxation of upwelling is common [ Price et al., 1991]. Studies are now underway to see if certain California PSP events can be correlated with upwelling relaxation. If so, state health authorities can look forward to the time when they can manage affected shellfish resources using remote sensing or moored instrument arrays to detect such events in real time. The first step, however, is to develop a fundamental understanding of the linkage between large-scale physical forcings and the pattern of PSP outbreaks.
Another example of how long-distance bloom transport mechanisms can be identified and characterized is found in the southwestern Gulf of Maine. There, the temporal and spatial pattern of persistent PSP outbreaks have been linked to the behavior of a buoyant plume or coastal current originating in several rivers that empty into the Gulf [ Franks and Anderson, 1992a,b]. Concentrations of the toxic dinoflagellate Alexandrium tamarense are much higher within the lower salinity waters of the plume than without, and toxicity in coastal shellfish rises and falls with the movement of the plume. This is in turn driven by the local wind stress, by rainfall and snowmelt patterns, and by the general circulation of the Gulf. As in Spain, downwelling conditions are conducive to toxicity development, since such conditions trap the plume and its associated cells tightly against the coast and accelerate them to the south. In contrast, upwelling favorable winds push the plume offshore, spreading it out laterally and dramatically decreasing the concentration of A. tamarense in nearshore waters due to the upwelling of deeper, saltier waters that contain no toxic cells.
Studies of this coastal current are ongoing, with the objective of better characterizing the nature of the association between the toxic cells and the low salinity water mass, as well as the physical forcings that determine plume behavior. Circulation models are under development that should prove very useful in a predictive capacity, and additional insights are coming from remote sensing detection of the plume using its warm surface temperature signature [ Keafer and Anderson, 1993]. Here, as in Spain, basic oceanographic studies of large-scale physical transport mechanisms may someday lead to a predictive capability of great practical importance.