ters, where long-term water quality monitoring programs exist. The precise causes of any bloom event are difficult to ascertain, but there is increasing evidence that unusual phytoplankton blooms are related to changes in nitrogen-silicon ratios caused by eutrophication. The food-web consequences of the global epidemic of noxious phytoplankton blooms could be severe in some areas.
Fishing activity also changes the structure of marine ecosystems, although the effects of overfishing are often difficult to resolve from. long-period cycles in organism abundance. Overfishing of the Georges Bank off the northeastern United States has changed the composition of fish species. From the mid-1960s until the early 1970s, herring and haddock declined by about a factor of 10 due to severe fishing pressure. At the same time, squid, dogfish, and sand lance increased, probably because they filled the ecological niches of depleted haddock and herring stocks (Sissenwine, 1986).
Long-term studies of some coastal benthic communities suggest that they have changed significantly owing to eutrophication. A 20-year time series of benthic species abundance data at a station in Puget Sound suggests that eutrophication may be causing shifts in the dominant species as well as increasing the magnitude of population fluctuations (Nichols, 1985). Yet, even with decadal time series, cause and effect are difficult to ascribe unambiguously, in part because anthropogenic effects are difficult to distinguish from natural changes.
The complexity of biological systems and their variability in time and space pose practical problems for designing programs and setting research priorities. Research based on the theme that food-web variability controls variability in fluxes of biologically important elements in the global ocean could take many forms; efforts must then focus on a subset of key questions and approaches.
One possibility is to take a comparative approach and focus studies on regions or times of the year with clearly distinguishable food-web structures, and to examine processes in the euphotic zone and in deeper waters. A second possible strategy is to plan biological studies to resolve seeming inconsistencies or contradictions obtained from geochemical measurements and models. For example, recent interest in vertical fluxes in the North Atlantic (Altabet, 1989) were inspired in part by geochemical studies indicating that conventional views of productivity and particle