This review provides an integrated synthesis with timelines and evaluations of ecological responses to eutrophication in Chesapeake Bay, the largest estuary in the USA. Analyses of dated sediment cores reveal initial evidence of organic enrichment in ~200 yr old strata, while signs of increased phytoplankton and decreased water clarity first appeared ~100 yr ago. Severe, recurring deep-water hypoxia and loss of diverse submersed vascular plants were first evident in the 1950s and 1960s, respectively. The degradation of these benthic habitats has contributed to declines in benthic macroinfauna in deep mesohaline regions of the Bay and blue crabs in shallow polyhaline areas. In contrast, copepods, which are heavily consumed in pelagic food chains, are relatively unaffected by nutrient-induced changes in phytoplankton. Intense mortality associated with fisheries and disease have caused a dramatic decline in eastern oyster stocks and associated Bay water filtration, which may have exacerbated eutrophication effects on phytoplankton and water clarity. Extensive tidal marshes, which have served as effective nutrient buffers along the Bay margins, are now being lost with rising sea level. Although the Bay's overall fisheries production has probably not been affected by eutrophication, decreases in the relative contribution of demersal fish and in the efficiency with which primary production is transferred to harvest suggest fundamental shifts in trophic and habitat structures. Bay ecosystem responses to changes in nutrient loading are complicated by non-linear feedback mechanisms, including particle trapping and binding by benthic plants that increase water clarity, and by oxygen effects on benthic nutrient recycling efficiency. Observations in Bay tributaries undergoing recent reductions in nutrient input indicate relatively rapid recovery of some ecosystem functions but lags in the response of others. KEY WORDS: Eutrophication · Nutrients · Chesapeake Bay Resale or republication not permitted without written consent of the publisherChesapeake Bay is a large estuary which has undergone many changes in its ecological properties and processes in response to nutrient enrichment over the last 2 centuries. Susceptibility of the Bay to eutrophication arises in part from the long dendritic shoreline that intimately connects it to its large watershed (covering an area 15 times that of the Bay) which contains expanding human population centers and extensive agricultural activities. (Satellite image from MODIS,

No abstract

It is generally believed that the high productivity of many estuaries is a consequence of both allochthonous nutrient inputs and autochthonous recycling of nutrients among producers and decomposers of organic matter. However, the mechanisms by which nutrients are recycled between sources and sinks are not clear We have documented time-dependent variations in density structure along a transect normal to the main axis of Chesapeake Bay which may be important in this regard. These variations influenced lateral distributions of dissolved inorganic nutrients, oxygen, chlorophyll a, and bacterioplankton and appeared to be responsible for high phytoplankton production over the flanks of the main channel relative to production over the channel. Vertical and seasonal variations in bacterial abundance were correlated with phytoplankton biomass, and increases in bacterial abundance were related to the development of phytoplankton bloon~s with lags of 0 to 6 d. Bacterial biomass and production were high throughout the study period, averaging 20 O/ O to over 50 O/ O of phytoplankton biomass and production, respectively. Absolute levels of bacterial abundance were among the highest reported and suggest that bacteria are responsible for a large function of the carbon flux in Chesapeake Bay. Relations among phytoplankton production, bacterial abundance and sediment trap collections indicate that variations in density structure reflect a transverse circulation which may explain how nutrients regenerated below the pycnocline are rapidly recycled into the euphotic zone above. This could be an important mechanism by which river-borne nutrient inputs during spring are coupled to high phytoplankton production during summer.

Nutrient fluxes and oxygen consumption (SOD) across the sediment-water interface were measured in situ along with vertical profiles of dissolved and particulate-phase nutrients in sediments and overlying water at 8 locations along the salinity gradient of Chesapeake Bay during spring and summer. Strong spatial and temporal patterns were evident. Highest rates of sediment NH: regeneration and SOD occurred in summer at mid-salinity (12 to 17 %a) stations. Sediment fluxes of dissolved inorganic phosphorous (DIP) were always low, possibly due to relatively oxidized conditions in surficial sediments. Nitrate (NO;) fluxes generally were directed into the sedirnents in spring and from the sediments during summer, and in both seasons fluxes were proportional to NO3 concentrations in overlying waters. Seasonal shifts in sediment 0:N flux ratios suggest that denitrification may have been important in spring but not summer. Significant relations were inferred between C:N:P composition of suspended materials and surficial sediments and the magnitude and direction of sediment-water fluxes. Although accretion of particulate nitrogen in sediments was less than 6 % of NH: regeneration in the lower bay, it was similar to rates of NH: flux in the low salinity region, suggesting that burial represents a significant sink for N in some estuarine zones. SOD constituted an important term in water column O2 budgets at all stations (1G50 % of total respiration), and sediment regeneration of NH: was capable of supplying 13 to 40 % of calculated phytoplankton N requirement, being more important during the summer period of higher productivity.

Seasonal O2 budgets were developed for the mesohaline region of Chesapeake Bay (USA), which experiences bottom water O2 depletion in summer. Rates of O2 production and consumption by the planktonic community and O2 consumption by the benthos were measured at 1 to 4 wk intervals from March to October at 2 stations. Under summer anoxic conditions, rates of sulfide diffusion from sediments were also measured directly with in situ chambers. Weekly observations of water column temperature, salinity and 02, combined with wind data and regression models, allowed calculation of air-sea gas exchange. Using these rates in mass-balance analyses for the upper and lower water column layers, we were able to compute net physical O2 transport across the pycnocline and longitudinally through the bottom layer. Mean monthly estimates of these net physical O2 transports were highly correlated to their respective O2 gradients. Slopes of these correlations provided estimates of the average spring-summer vertical dispersion coefficient (0.2 cm2 S-') and net gravitational water velocity (5 cm S-'), both of which correspond to previous reports. Vertically integrated planktonic respiration rates in the lower water column layer were compared to benthic O2 consumption from April to August. In general, planktonic processes dominated 0 2 consun~ption, comprising almost two-thirds of the total. Oxygen consumption associated with benthic processes, however, exceeded planktonic rates in early spring prior to vernal warming and in late August when large S2-fluxes resulted from release of accumulated pore water pools. By combining our respiration data with values from other coastal environments and plotting rates versus water-column depth, we find a general relation in which planktonic respiration exceeds benthic respiration for systems deeper than 5 m. Hence, for stratifled estuaries with bottom layers thicker than 5 m, seasonal O2 depletion is dnven primarily by planktonic respiration rather than benthic consumption of accumulated organic pools. A comparison of mean monthly rates for bottom respiration (plankton plus benthos) and net physical O2 replenishment here revealed that the 2 processes were highly correlated between March and October; both rates increased through July and declined thereafter. This strong correlation underscores a fundamental interdependence of biological O2 consumption and net physical transport, w h~c h is based on the O2 gradient by which the 2 processes are coupled. Consequently, relatively large reductions In respiratory O2 consumption (e.g. with decreased organic inputs) would lead to substantially smaller decreases in the extent of bottom water O2 depletion because of an inherent adjustment between the coupled biological and physical processes.

Eight experimental ponds containing submersed vascular plants (predominantly Potamogeton perfoliatus and Ruppja maritirna) were subjected in duplicate to 4 levels (including controls) of fertilization from June to August 1981. Seston and phytoplankton chlorophyll a increased with fertilization, and pronounced algal blooms were evident under high dosage. Of the total seston. phytoplankton exerted the greatest influence on attenuation of photosynthetically active radiation (PAR), such that there was insufficient light for submersed vascular plant growth at the sediment surface during bloonls. An extensive epiphytic community developed on plants in all nutrient-treated ponds at densities similar to those observed in nature on senescent plants. At high nutrient treatments the accumulation of epiphytic material resulted in > 80 % attenuation of the incident radiation at the leaf surface. Biomass of submersed macrophytes decreased significantly under high and medium nutrient treatments compared to control and low treatments within 60 d following initial fertilization. Apparent production of vascular plants (based on oxygen production and I4C-bicarbonate uptake) was reduced at the higher nutrient treatments for both R. maritirna and P. perfoliatus. Most of this reduction in macrophyte photosynthesis could be explained by attenuation of PAR associated with epiphytic material. However, without PAR attenuance in the overlying water, observed levels of epiphytic growth would be insufficient to reduce light below compensation levels needed to sustain vascular plant growth. At the high fertilization rates, integrated primary production of pond communities was significantly reduced with the loss of the vascular plants, even though phytoplankton and epiphytic growth were enhanced.

Fish communities and other ecological variables were sampled for 6 mo (May to October) in successive years (1979, 1980) at vegetated and non-vegetated areas in 2 distinctively different littoral zones (an open bay and a protected cove) of mid-salinity Chesapeake Bay, USA. Fish abundance, biomass and species richness were h~gher in vegetated areas at both sites, and were significantly correlated with macrophyte biomass Diel patterns of fish abundance varied, but highest catches generally occurred at dusk or at night. At one sampling site fish assemblages were dominated by smaller individuals in the vegetated area, suggesting an attraction of juveniles to macrophyte beds for food or refuge from predation. Larger piscivorous fish, which were also caught in greater numbers in vegetated areas, may have been attracted there by higher densities of forage fish. At the cove site the biomass of Paleomonetes sp. was comparable to that of the fish community towards the end of the plant growing season. Benthic infauna were also more abundant in vegetated areas at both sites, and stomach analyses indicated these organisms to be the dominant food resources for common fishes. Diets were generally non-selective in non-vegetated areas while highly selective for epiphytic fauna in macrophyte beds. Fish stomachs were also significantly fuller in vegetated areas, indicating generally greater feeding success. Fish production varied among major species but was higher overall at vegetated areas, following the seasonal patterns of primary production. Most of the differences in fish production between areas were attributable to higher instantaneous growth rates rather than higher biomass. It appears that the greater abundance and species richness of fish assemblages in vegetated areas of this region of Chesapeake Bay resulted from the attractiveness of these habitats as rich sources of preferred foods.

Bottom water and sediment characteristics and net sediment-water fluxes of oxygen, nitrogen, phosphorus and hydrogen sulphide were measured under oxic a n d experimentally induced anoxic conditions along a depth gradient (47 to 130 m) during summer in the Baltic proper. Temperature, salinity and dissolved nutrient concentrations (particularly phosphorus and nitrate) in bottom waters increased with depth while oxygen concentrations decreased sharply. Sediment organic content was much higher in sediments located beneath the permanent pycnocline (ca 65 m). Sediments at all stations were somewhat depleted in total N relative to total C (C.N -10). Sediments at the shallow station (47 m ) were highly enriched with total P relative to C or N (C:N P = 18:2:1) but were somewhat depleted in P at the deeper stations (C:N:P = 120:12:1). Under oxic conditions oxygen fluxes ranged from 214 to 777 pm01 0 m Q-' and decreased with depth. Phosphorus and nitrite fluxes were always very small and were directed either into or out of the sediments. Ammonium fluxes were small (1 to 30 pm01 N m-' h-') at all stations and did not exhibit a clear pattern with depth; 0 : N -N H , flux ratios were close to expected 'Redfield ratios' at the deep station (16:l) but were much higher (>55:1) at the shallower sites indicating that processes other than only an~monificatlon were taking place in sediments. Nitrate was always consumed by sediments (ca 1 to 16 ktn~ol N m-' h-') and fluxes were proportional to nitrate concentrations in the overlying water. Under anomc conditions there was a dramatic increase in P-PO4 fluxes (2 to 40 pm01 P m-' h-') and a smaller lncrease In N-NH, fluxes (14 to 35 pm01 N m-2 h ' ) . Large hydrogen sulphide fluxes (>40 pm01 S m-' h-') were observed in sediments from the deepest station only. Under oxic conditions sediment recycling of N and particularly of P were small compared to estimated rates of burial of P and burial plus denitrification of N However, under anoxic conditions, sediment recycling of both N and P were similar to or much greater than sedlment loss terms of burial and denitrification. Sediment regeneration under oxic conditions could supply 1 to 8 % and 0 to 2 % of estimated phytoplankton demand of N and P, respectively; under anoxic conditions 12 % of N and u p to 200 O/o of P demand could be met via sediment recycling.