ABSTRACT-Nutrient enrichment and consequent alteration of nutrient biogeochernical cycles is a serious problem in both freshwater and marine systems. The response of aquatic systems to additions of N and P is generally to increase algal biomass. The partitioning of these nutrients into different functional groups of autotrophic organisms is dependent upon both intrinsic and extrinsic factors. A common response to nutrient loading in northern temperate aquatic ecosystems is an increase in diatom biomass. Because nutrient enrichment generally leads to increases in water column concentrations of total N and total P (and not Si) such nutrient loading can lead to transient nutrient limitation of diatom biomass due to lack of dissolved silicate (DSi). Increased production of diatom biomass can lead to an increased accumulation of biogenic silica in sediments, ultimately resulting in a decline in the water column reservoir of DSi. Such biogeochemical changes in the silica cycle induced by eutrophication were first reported for the North American Laurentian Great Lakes. However, these changes are not a regional problem confined to the Great Lakes, but occur in many freshwater and marine systems throughout the world. Here we summarize the effects of anthropogenic modification of silica biogeochemical cycles for the North American Laurentian Great Lakes, describe some of the biogeochemical changes occurring in other systems, and discuss some of the ecological implications of a reduction in water column DSi concentrations, including changes in species composition, as DSi concentrations become limiting to diatom growth and biomass, changes in food web dynamics, and altered nutnent-recycling processes.

Seasonal variations in the distribution of dissolved inorganic nitrogen, sihcon, and phosphorus along the salinity gradient of Chesapeake Bay from 1984 to 1988 suggest that dissolved silicate (DSi) controls the magnitude of diatom production during the spring bloom, causes the collapse of the spring bloom, and leads to changes in floristic composition. High sedimentation rates of chlorophyll biomass observed during this per~od could be due to Si-deficiency, suggesting that the supply of DSi may also control the flux of phytoplankton biomass to the benthos, an important parameter of seasonal oxygen depletion in the Bay.

Experimental studies of intact cores from the Baltic Sea were conducted to determine the response of sediment nutrient recycling processes to varied inputs of organic matter. A 2 mo enrichment experiment was carried out in the laboratory on sediment cores held at 4°C using a flow-through system where overlying waters were continuously replaced at a rate of 1 d.' The experiments were designed to simulate the deposition of organic matter that occurs during a typical spring diatom bloom ( l x ) and under enriched conditions with eutrophicatlon at approximately 3 times (3x) a normal spring bloom utilizing added organic matter from a natural phytoplankton assemblage collected in a eutrophic coastal fjord during the spring diatom bloom. Low and constant sediment-water fluxes were observed throughout the duration of the experiment in control cores with no added organic matter. In all cases an immediate response was noted when a single pulsed addition of algal material was added to the sediment surface. Sediment-water fluxes of ammonium (NHdt), and dissolved inorganic phosphate (DIP) increased significantly (ANOVA, p < 0.01). For nitrite + nitrate (NO; + NO3-) and dissolved silicate (DSi) sediment-water fluxes, differences were initially observed; however, only the NO2-+ NO3-fluxes were significantly different over time (ANOVA, p < 0.01). Fluxes of NOz-+ NO3-were into the sediment for 3 to 10 d after addition of organic material, followed by small fluxes out of the sediment. The addition of algal material proportionate to a normal spring bloom ( l x ) had only a minor effect on porewater nutrient concentrations, whereas the 3x treatment substantially modified both the short-and long-term response of sediments. A greater proportion of anaerobic decomposition products, e.g. NH,' and DIP, were observed with an expansion of more reducing conditions resulting from the addition of organic matter. The percentage of Si remineralized decreased as the flux of n~aterial to the sediment increased. Deposition rates similar to a typical spring bloom did not have long-term effects on the nutrient recycling processes; however, increases in the present level of deposition (as simulated in this study), which are forecasted with further eutrophication in the Baltic Sea, may have a significant impact on nutrient biogeochernical cycles.

The winterkpring bloom of 1990 in Chesapeake Bay. USA, was prolonged and well developed, relative to other recent years, along the axis of the Bay. However, the bloom did not occur uniformly along the axis of the Bay, but rather developed and dissipated at different times in different regions of the Bay. The peak of the bloom progressed northward and was observed in late March in South Bay, early April in Mid Bay, and not until mid May in North Bay. We measured biomass and nutrient concentrations and the rates of carbon, nitrogen, phosphorus, and silicon utilization during the development and dissipation of the bloom, and compared ratios of these rates to the elemental ratios of the incoming nutrients and the resulting particulate material. In North Bay, bloom development was probably delayed due to light limitation of carbon uptake. Nitrogen was delivered and utilized in excess of stoichiometric proportions in the northern part of the Bay, eventually leading to phosphorus and/or silicon limitation. In the mid portion of the Bay, the mean stoichiometric proportions of the particulate nutrients were similar to Redfield proportions, but ratios of uptake of nitrogen and phosphorus exceeded Redfield proportions by more than 20-fold, reflecting both the high uptake rates of nitrogen and low uptake rates of phosphorus in that region. However, only at the peak of the bloom in mid April did transient phosphorus limitation of growth occur at Mid Bay. In contrast, ratios of nitrogen to phosphorus uptake rates in South Bay were considerably below Redfield proportions, primarily due to the low availability and low uptake rates of nitrogen. Concentrations of Si(OH), in South Bay were also extremely low through the bloom period, and thus Si(OH), and nitrogen, as well as pod3-, limited growth there. In addition, temperature appeared to play a key role in the collapse of the diatom assemblage in mid May. During the early stages of the bloom in South Bay, No3-+ NOzcontributed >60% of the total nitrogen utilized, but by the end of the spring bloom period in May, over 50% of the nitrogen uthzed was urea alone. These data underscore the need to understand how freshwater flow, ambient nutrient concentrations, temperature, and light dlffer along the axis of the Bay to understand the differential timing and magnitude of bloom development in different regions of the Bay.

No abstract