Smayda (1989) has compiled extensive evidence in support of the hypothesis that the worldwide increase in nuisance algal blooms is related to increased nutrient availability. For instance, a 2.5-fold increase in nutrient loadings accompanied an 8-fold increase in the annual number of red-tide blooms in a harbor in Hong Kong between 1976 and 1986. Increased nutrient concentrations in the North Sea, the Baltic Sea, and in waters between Denmark and Sweden (the Skagerrak and Kattegat) have co-occurred with increased primary production and increased incidence of blooms in these waters (Smayda 1989). The green-tides which occurred in the Great South Bay of Long Island in the 1950s were also clearly associated with nitrogen loading from duck farms there (Ryther 1954), and the reduction of nutrient loadings and the opening of a channel to increase water exchange between the bay and ocean have greatly reduced these blooms (Ryther 1989). Also, nuisance algal blooms are much more likely to occur in nutrient-rich estuarine waters than in more coastal or shelf waters (Cosper 1991, Prego 1992). The new dinoflagellate discovered by Burkholder et al. (1992b), which produces toxins only in the presence of fish, seems to be stimulated by phosphorus additions.
On the other hand, there is little if any evidence to show a direct connection between either nitrogen or phosphorus concentrations and blooms of most brown-tide or red-tide organisms (Cosper 1991, Wells et al. 1991). Red-tide blooms in Florida are not correlated with concentrations of any measured form of nitrogen or phosphorus (Rounsefell and Dragovich 1966). Similarly, the brown-tide blooms of the mid-1980s along the northeastern coast of the United States did not appear to be correlated with higher levels of nitrogen or phosphorus (Cosper et al. 1989, Cosper 1991). However, it is important to note that the concentration of a nutrient at any given point of time may not be correlated with its availability to phytoplankton (Howarth 1988), and phytoplankton can grow for long periods of time off of internally stored pools of nutrients (Andersen et al. 1991).
Perhaps more importantly, it may not be the availability of nitrogen or phosphorus alone that matters in controlling nuisance algal blooms but rather the relative availability of these nutrients in comparison to silicon (Officer and Ryther 1980, Smayda 1989). When Si:N and Si:P ratios are relatively high, silicon is relatively available, favoring the growth of diatoms, which have a high requirement for silicon. However, as the Si:N and Si:P ratios decrease, competition begins to favor other algae with no silicon requirement, such as the red-tide, green-tide, and brown-tide organisms. Most silicon comes from natural sources, and as inputs of nutrients from sewage increase, the Si:N and Si:P ratios decrease (Officer and Ryther 1980). Eutrophication itself can decrease the abundance of silicon by increasing sedimenta-