Nutrient biogeochemical cycles in the Gulf of Riga: scaling up field studies with a mathematical model

A box model has been implemented to understand the large-scale biogeochemical cycles of nitrogen, phosphorus, and silicon in the Gulf of Riga. The large data sets collected within the international Gulf of Riga Project in 1993/1995 were used to validate the model. The comparison to data was useful in scaling up to the gulf-wide level and scrutinizing the conclusions based on short-term field surveys and experimental studies. The simulations indicate that the limiting role was passing from silicon to phosphorus to nitrogen over the seasons of organic production. However, on an annual scale, nutrient limitation was close to the “Redfield equilibrium”. Mass balance considerations, based on modeled coupled fluxes, disagree with the conclusions on low sediment denitrification and high phosphorus retention in the pelagic system, which were derived from isolated measurements.Nutrient budgets constructed with the model revealed the high buffer capacity of the Gulf of Riga. The nutrient residence times span a range from 6 years for N to 70 years for Si. The buffering arises from intensive internal recycling in the water body and by the bottom sediments. The budgets indicate that the Gulf retains about two-thirds of external nitrogen and silicon inputs, while phosphorus retention is only 10%.A slow response to external perturbations is demonstrated with numerical experiments run for 15 years under 50% reductions of terrestrial nutrient inputs. These experiments imply that the most effective is the N+P reduction scenario, which resulted in a 20% decrease of primary production after 12 years. A reduction of P resulted in only a 6% decrease of primary production; however, it yielded an 80% drop in the amount of nitrogen fixation.     

Application of a validated primary production model (BLOOM) as a screening tool for marine, coastal and transitional waters

In order to manage aquatic systems, it is necessary to apply methods relating the environmental variables and system-state parameters with external factors that affect the system. External factors can be natural (i.e. the movement of water) or partly-anthropogenic (i.e. nutrient loads). In addition to the national authorities, who have been implementing environmental policies for several decades, the EU is presently implementing the Water Framework Directive (WFD) aimed at establishing a new set of standards for the ecological and water quality of water systems. Among these are the phytoplankton biomass and composition. Phytoplankton affects turbidity, oxygen depletion, total productivity of the system and the occurrence of (harmful) algal blooms. A range of methods is available to relate phytoplankton to the controlling environmental conditions. Among these are statistical relations for instance of the Vollenweider type as well as deterministic simulation models. At the end of the 1970s, a generic deterministic phytoplankton module called BLOOM was developed, which has since been applied to a wide range of fresh water and marine systems. Here we test the applicability of this model as a screening tool for coastal waters. We conclude that the model is able to reproduce observed chlorophyll levels adequately under a wide range of conditions. Subsequently the model is applied to demonstrate the potential impacts of reductions in nitrogen, phosphorus or both nutrients simultaneously. Depending on which factors are initially controlling, the impacts of these reductions vary considerably both between locations and during the season. While this type of application lacks explicit relations between nutrient concentrations and external loadings, it does consider a number of relevant conditions in a consistent way and requires remarkably little data and effort. It is therefore a valuable screening tool.     

Modelling nutrient fluxes from sub-arctic basins: Comparison of pristine vs. dammed rivers

The deterministic Riverstrahler model of river functioning is applied for the first time to sub-arctic catchments. Seasonal nutrient (N, P, Si) deliveries to the coastal zone are simulated, and nutrient annual fluxes are established for the nearly pristine river Kalix (hereafter called Kalixälven) and the heavily dammed river Lule, (hereafter called Luleälven) both located in Northern Sweden and draining into the Bothnian Bay, Baltic Sea.For Kalixälven simulations are performed with a runoff calculated from precipitation, evapo-transpiration and temperature data for the period 1990–1999, using a hydrological model calibrated on observed monthly discharges at the river outlet. The same hydrological parameters are used to calculate specific runoff for the Luleälven basin in absence of dam regulation. Reservoir filling and emptying are simulated using a simplified representation of their management rules. Diffuse sources of nutrient are evaluated according to land cover of the catchment. The simulated seasonal trends are within the range of the observed data, in particular for discharge, dissolved silica, total phosphorus, inorganic nitrogen and total organic carbon. Specific runoff is 50% higher in the Luleälven than in the Kalixälven watershed due to higher altitudes and precipitations. Average silica, nitrate and phosphorus concentrations are much lower in Luleälven than in Kalixälven. Comparison of model results for the Luleälven with and without dams shows a reduction of respectively 25% and 30% in silica and phosphorus fluxes delivered at the outlet, while nitrogen delivery is increased by 10% in the dammed vs. undammed river system. The model allows assessing the respective role of reservoir trapping of nutrient in the reservoir through algal uptake and sedimentation, and of changes in the vegetation induced by flooding the valley formerly covered by forests and wetlands.     

Influence of model domain size,wind directions and Ekman transport on storm surge development inside the Chesapeake Bay: A case study of extratropical cyclone Ernesto, 2006

The Chesapeake Bay is vulnerable to severe flooding caused by hurricanes and strong Northeasters. A 3D storm surge model of the Chesapeake Bay is developed for studying the impact of model domain size, wind directions and Ekman transport on the storm surge in the Chesapeake Bay. The model encompasses the Chesapeake Bay and the US East Coast shelf to reduce the influence of model domain size on surge prediction inside the Chesapeake Bay and to account for both local and remote wind effects. This study used 3D model experiments, with respect to different wind directions, to diagnose the relative influences of the local and remote wind effects and Ekman transport on spatial surge distribution during storm events. The model results confirmed that spatial surge distribution can be well explained by the superposition of two distinct physically driven mechanisms during a storm event: incoming surge wave caused by remote effects and local wind forcings. A large model domain is a necessity for predicting storm surge accurately inside the Chesapeake Bay. The model results suggest that the interactions of the incoming surge propagating into the Bay and the local wind forcing from N and NE directions result in an enhanced setup in the lower to middle portions of the Bay, whereas the superposition of incoming surge and the local wind forcing from S and SE directions enhance the surge in the upper Bay region. A combined northwesterly wind over the middle to upper portions of the Bay and southwesterly wind over the lower Bay can cause a large setdown throughout the entire Bay. The Ekman setup along the coast contributes significantly to the water level variations during storm events. It enhances (reduces) surge inside the Bay under the wind forcings from N and NE (SW, S, and SE) directions.     

Navigating Institutional Constraints in the Process of Computational Environmental Modeling

Abstract

The use of computational models has proliferated in the last few decades due to the practical benefits that they offer in terms of understanding complex systems. However, the process of developing a computational model is not only shaped by scientific and practical concerns. Modelers must also navigate the institutional structures in which the models are constructed and implemented. In the process, they must often draw upon existing or create new social relationships that can, themselves, alter the institutional structures they are attempting to navigate. In this article, I use an ethnographic approach to examine three examples in which the process of building and implementing a computational model was constrained by institutional factors and the strategies the modelers used to navigate them. The research was conducted with computational modelers at the Chesapeake Bay Program who are involved in constructing the Chesapeake Bay modeling system for nutrient management in the watershed. Modelers in this context had to: build relationships with the broader scientific community to reinforce the “believability” of the model, draw upon institutional relationships in order to work around limitations on data-sharing between organizations, and manage differing incentive structures to motivate and coordinate the research needed to complete the model     

Alterations in nutrient limitations — Scenarios of a changing Baltic Sea

Previous trend studies have shown increasing nitrogen and phosphorus as well as decreasing silica concentrations in the water mass of the Baltic Sea. This has had an impact on the amount of primary production, but also on the quality and succession of plankton species. Present study examines the spatial and temporal patterns of potential nutrient limitations in the Baltic Sea for the time period 1970–2000. Generally, low concentrations of DSi can limit the diatom blooms and such conditions are found in the Gulf of Riga and Gulf of Finland during spring and summer. Nutrient ratios, DSi:DIN, DSi:DIP and DIN:DIP, are often used to determine which nutrient may limit the primary production. Annual long-term temporal trends of silica to inorganic nitrogen and phosphorus respectively show consistent decreasing patterns. The largest slopes are detected during spring and summer for DSi:DIN and during spring for DSi:DIP ratios. For the DIN:DIP ratio significant slopes are only found in a few locations despite increasing levels for both nutrients, displaying a large variation in trends. In the open Baltic Proper the present trends are positive during winter and negative during spring and autumn. Gulf of Finland and Gulf of Riga are areas where both DSi:DIP and DSi:DIN ratios are found close to the Redfield ratios for diatoms. Together with the evaluated trends these suggest that the Gulfs may become silica limited in a relatively near future. These findings give some implications on the development and impact of changing nutrient concentrations.     

Ecological modeling as a tool for coastal zone management in Dokai Bay, Japan

An ecological numerical model has been developed for coastal zone management in Dokai Bay, Japan. The numerical model experiment reproduces the occurrence of red tides and oxygen-deficiency in Dokai Bay during summer. Moreover, the model experiments forecast that when the phosphorus or nitrogen load from land will be cut by 90% or 95%, respectively, red tides and oxygen-deficiency will not occur in Dokai Bay. If the sewage facility is not sufficient to cut the nutrients load to the necessary degree, it is necessary to consider other countermeasures such as cultivating bivalves in order to decrease the nutrients in the bay.     

Budget calculations of nitrogen, phosphorus and BOD5 passing through the Oder estuary

The Oder River estuary is a large and complex system composed of lagoons, lakes and river branches in which numerous biogeochemical processes lead to modification of loads of dissolved/suspended material brought in with the riverine waters. Budget calculations show that on an annual basis, 71–88% of total nitrogen, 73–89% of total phosphorus and 72–101% of BOD₅ inflowing to the estuary are exported to the Baltic Sea. Among the inorganic nutrient species, nitrates exhibit the highest net transformation rate into organically bound forms (over 60%). The transformation could have been equally high or even higher in the case of ammonia and phosphates but these processes may have been compensated by intensive mineralization. The mechanisms responsible for the nutrient transformation patterns, as well as their net effect on the annual loads delivered into the Baltic Sea, are discussed in the paper. Phosphorus seemed to play a limiting role in phytoplankton production in the estuary in spring, while nitrogen did the same in summer.     

浅水湖泊平面二维水质模型研究

磷氮元素在水体中迁移对湖泊富营养化研究具有重要性。采用有限元伽辽金加权余量法建立了简单的二维水质模型,并计算了在风场作用下的苏州市澄湖丰水期和枯水期的氮磷的分布以及输移。计算中引入了可由实测资料率定的可为正负的综合衰减参数,综合考虑了氮磷元素的沉降速率和释放速率,这样减少了水质多参数难确定的困难。计算的结果表明,流态与实际情况吻合较好,氮磷元素的浓度计算值与实测值相差不大。     

The mid-Atlantic load centre: Baltimore or Hampton Roads?

Competition between closely situated seaports in not new. The ports of Baltimore and Hampton Roads, located at either end of Cheaspeake Bay, have been locked in a fierece battle to gain dominance in the lucrtice mid-Atlantic container traffic for a number of years. The probability that ech will succeed is not great, as shipping lines, concerned with the high costs associated with idle port time, will select one,rther thatn serve both. Indeed, evidence is beginning to mount that, dispite massive expenditures by the Stte or Maryland, the Virginia ports a reemerging as the region' load centre. This paper examines recent trends in container movements through these two major ports and suggests the long-termprospects for each. among measures considered are the location of eachi in relation to other ports both North and South, sizes of the port cities and their immediately adjacent hinterlands, proximity to inland population centres (markets), connectivity between the ports and these inland centres (by both rail and truck), expenditures by port agencies and land crries, labour relations, and availabillity of space for expansion.     

A box model approach for a long-term assessment of estuarine eutrophication, Szczecin Lagoon, southern Baltic

We develop a layered “box model” to evaluate the major effects of estuarine eutrophication of the Szczecin lagoon which can be compared with integrating measures (chlorophyll a (Chl a), sediment burial, sediment oxygen consumption (SOC), input and output of total nutrient loads) and use it to hindcast the period 1950–1996 (the years when major increase in nutrient discharges by the Oder River took place). The following state variables are used to describe the cycling of the limiting nutrients (nitrogen and phosphorus): phytoplankton (Phy), labile and refractory detritus (D_N, D_Nref, D_P, D_Pref), dissolved inorganic nitrogen (DIN), dissolved inorganic phosphorus (DIP), and oxygen (O₂). The three layers of the model include two water layers and one sediment layer. Decrease of the carrying capacity with respect to the increased supply of organic matter of the system with advancing eutrophication over the period studied is parameterized by an exponential decrease of the sediment nitrogen fluxes with increasing burial, simulating changing properties from moderate to high accumulating sediments. The seasonal variation as well as the order of magnitude of nutrient concentrations and phytoplankton stocks in the water column remains in agreement with recent observations. Calculated annual mean values of nutrient burial of 193 mmol N m⁻² a⁻¹ and 23 mmol P m⁻² a⁻¹ are supported by observed values from geological sediment records. Estimated DIN remineralization in the sediments between 100 and 550 mmol N m⁻² a⁻¹ corresponds to SOC measurements. Simulated DIP release up to 60 mmol P m⁻² a⁻¹ corresponds to recent measurements. The conceptual framework presented here can be used for a sequential box model approach connecting small estuaries like the Szczecin lagoon and the open sea, and might also be connected with river box models.     

Influence of local and external processes on the annual nitrogen cycle and primary productivity on Georges Bank: A 3-D biological–physical modeling study

Georges Bank is one of the world's most highly productive marine areas, but the mechanisms of nutrient supply to support such high productivity remain poorly understood. Intrusions of nutrient-poor Labrador Slope Water (LSW) into the Gulf of Maine (NAO-dependent) potentially can reduce nutrient delivery to the bank, but this mechanism has not been quantitatively examined. In this paper, we present the first whole-year continuous model simulation results using a biological–physical model developed for the Gulf of Maine/Georges Bank region. This high-resolution three-dimensional coupled model consists of the Finite Volume Coastal Ocean Model (FVCOM) and a Nitrogen–Phytoplankton–Zooplankton–Detritus (NPZD) model, and was used to examine the influences of local and external processes on nitrogen and phytoplankton dynamics on Georges Bank. The model captured the general pattern of spatial-temporal distributions of nitrogen and phytoplankton and provided a diagnostic analysis of different processes that control nitrogen fluxes on Georges Bank. Specifically, numerical experiments were conducted to examine seasonal variation in nitrogen transport into the central bank (new nitrogen supply) versus nitrogen regenerated internally in this region. Compared with previous observation-based studies, the model provided a quantitative estimate of nitrogen flux by integrating the transport over a longer time period and a complete spatial domain. The results suggest that, during summer months, internal nitrogen regeneration is the major nitrogen source for primary production on the central bank, while nitrogen supply through physical transport (e.g. tidal pumping) contributes about 1/5 of the total nitrogen demand, with an estimated on-bank nitrogen transport at least 50% less than previous estimates. By comparing the model runs using different nitrogen concentrations in deep Slope Water, the potential influence of NAO-dependent intrusions of LSW was examined. The results suggest that the change of nitrogen concentration in the deep Slope Water may not have a significant impact on nitrogen and phytoplankton dynamics on the well-mixed central bank, largely due to limited nutrient exchange across the tidal mixing front and enhanced near-frontal nutrient uptake. However, relatively more significant impact was observed in the model simulations if both well-mixed and seasonally-stratified areas (inside 100 m isobath of the bank) were considered in flux calculations.     

Seasonal variation of riverine nutrient inputs in the northern Bay of Biscay (France), and patterns of marine phytoplankton response

Seasonal variations in nutrient inputs are described for the main rivers (Loire and Vilaine) flowing into the northern Bay of Biscay. The river plumes are high in N/P ratio in late winter and spring, but not in the inner plume during the summer. Conservative behavior results in most nutrients entering the estuary and eventually reaching the coastal zone. Temporal and spatial aspects of phytoplankton growth and nutrient uptake in the northern Bay of Biscay distinguish the central area of salinity 34 from the plume area. The first diatom bloom appears offshore in late winter, at the edge of the river plumes, taking advantage of haline stratification and anticyclonic “weather windows.” In spring, when the central area of the northern shelf is phosphorus-limited, small cells predominate in the phytoplankton community and compete with bacteria for both mineral and organic phosphorus. At that period, river plumes are less extensive than in winter, but local nutrient enrichment at the river mouth allows diatom growth. In summer, phytoplankton become nitrogen-limited in the river plumes; the central area of the shelf is occupied by small forms of phytoplankton, which are located on the thermocline and use predominantly regenerated nutrients.     

Bathymetric influences on the lower Chesapeake Bay hydrography

Observations of salinity and density in the lower Chesapeake Bay are used to describe the bathymetric influence on the transverse hydrographic structure in the area. Current velocity observations of high spatial resolution are also used to relate the flow structure to the hydrographic structure. Tidal flow characteristics in the lower bay are affected by the combination of bathymetry and hydrography. Increased stratification over channels relative to shoals may increase M₂ ellipticity with depth over channels but not over shoals. It is found that three consistent hydrographie features can be related to the transverse structure of the longitudinal flow: (1) persistent stratification over channels due to differential tidal advection of density gradients, (2) development of bottom front separating net inflows from net outflows at the region south of Chesapeake Channel, and (3) outflow of low salinity water at the northern end of a lower bay section. Based on these hydrographie features, two basic hydrographic regimes are proposed to exist throughout the year in the lower Chesapeake Bay: (1) a low buoyancy-high mixing energy regime of stratification restricted to channels, a northward monotonical increase in salinity, and a weak bottom front, and (2) a high buoyancy-weak mixing energy regime of stratified conditions everywhere, a large region of northward salinity decrease at the northern half of the section, and a robust bottom front. The dynamics in the transverse direction for the former regime is ageostrophic, and in the latter regime the contribution by geostrophy is approximately 50% as bathymetric influences become less evident.     

Simulated responses of the Gulf of Mexico hypoxia to variations in climate and anthropogenic nutrient loading

A mathematical model was used to simulate monthly responses of the Gulf of Mexico hypoxia to variations in climate and anthropogenic nutrient loading over a 45-year period. We examined six hypothetical future scenarios that are based on observed and projected changes in the Mississippi River discharge, Mississippi River nitrate concentrations, and ambient water temperatures. In particular, we investigated the implications of a 30% decrease in the Mississippi River nitrogen flux, which was recently proposed by the Mississippi River Watershed/Gulf of Mexico Hypoxia Task Force as a measure to reduce the size of the hypoxic zone. Model simulations suggest that the frequency of hypoxia in the northern Gulf of Mexico is highly sensitive to variations in riverine nitrate flux, but also to variations in freshwater discharge and ambient water temperatures. A 30% decrease in the Mississippi River nitrate flux, for example, would reduce the frequency of hypoxia by 37%. Nevertheless, a 20% increase the Mississippi River discharge, which may occur under some climate change scenarios, would produce an increase in the frequency of hypoxia of the same magnitude. Thus, if the potential climatic variations are taken into account, a 30% decrease in the nitrogen flux of the Mississippi River may not be sufficient to accomplish the proposed hypoxia management goal.     

The trophic state of the Baltic Sea a century ago: a model simulation study

We apply a 3-D circulation model with a biogeochemical module (ERGOM) for the simulation of trophic conditions in the Baltic Sea a century ago. One aim is to provide reference or background data for nitrogen, phosphorus and chlorophyll, which is required for the implementation of the European Water Framework Directive (WFD). We assume that the situation a century ago serves this purpose well. Model input for this long-term simulation study are the regionally differentiated riverine and atmospheric nutrient loads to the Baltic Sea, which were compiled and calculated for a situation 100 years ago on the basis of various literature sources. For the mixed surface layer of the open Baltic Sea, we suggest maximum winter concentrations for dissolved phosphorus (dissolved inorganic nitrogen) of 0.23–0.35 (2.7–3.7) mmol/m³. Maximum chlorophyll-a concentrations are between 1.8 and 2.4 mg/m³. The concentrations of all parameters for different coastal waters vary in a wide range, depending on exposure to nutrient sources. Our nutrient concentrations for the situation a century ago are close to early measured data (1950–1960) and suggest that this data is suitable as reference data, as well.     

Differences in emission of nitrogen and phosphorus into the Vistula and Oder basins in 1995-2008 — Natural and anthropogenic causes (MONERIS model)

The aim of the modeling studies (MONERIS) was to estimate annual source apportioned nitrogen (N) and phosphorus (P) emissions into the Vistula and Oder basins in 1995-2008, thus, during the transition period in Poland, characterized by changes in both agricultural sector and handling of point source pollution. N and P emissions into both basins showed declining tendencies. Between the sub-periods 1995-2002 and 2003-2008, the overall N emission into the Vistula and Oder basins decreased by 16-17% (i.e. by ca. 26,900 tons in the Vistula and by ca. 18,000 tons in the Oder basin); P emission declined by 23% in the Vistula and by 32% in the Oder basins (i.e. by ca. 3400 tons in the Vistula and by ca. 2200 tons in the Oder basin). The temporal patterns of N and P emission into the Vistula and Oder basins, as well as the percentage contribution of N and P pathways (particularly: overland flow, tile drainage, groundwater, waste water treatment plants) showed great differences between the basins. Natural (type of bedrock, soil type, lake area) and anthropogenic (regionally and temporarily different type and intensity of agricultural activity, spatially different structural changes in agriculture during the transition period, regionally and temporarily different investment in waste water treatment plans) factors were found to be responsible for the differences, and the relationships are extensively discussed in the paper. In 1995-2008, 70% of N emission into both river basins was via groundwater and tile drainage, with the former playing more important role in the Vistula basin, and the latter playing more important role in the Oder basin; contribution of N emission from point sources was comparable in both rivers and it reached 11-12%. In 1995-2008, point sources, erosion, overland flow, and urban systems were found the most important P pathways in both basins, with a higher percentage contribution of point sources in the Oder basin.     

Comparison of Beach Nourishment along the U.S. Atlantic,Great Lakes,Gulf of Mexico,and New England Shorelines

The U.S. national beach nourishment experience is summarized for the East Coast barrier islands, the Gulf of Mexico, New England, and the Great Lakes. A total of 1,305 nourishment episodes on 382 beaches are recorded at a total estimated cost of approximately $1.4 billion ($2.5 billion in 1996 dollars). In terms of both volume and costs, nourishment has been the most extensive by far on the East Coast barrier islands. Depending on the region, between 65% and 85% of all nourishment projects have a federal funding component. Annual expenditures and sand volumes for beach nourishment are increasing, especially on East Coast barriers. At present, total annual national beach nourishment costs (excluding the Pacific Coast) are on the order of $100 million per year. The cost per cubic yard of nourishment sand as expressed in 1996 dollars has remained more or less constant over time. Additionally, the volumes of sand needed for subsequent nourishment episodes on individual beaches do not decrease, despite contrary assumptions in the shoreface-profile-ofequilibrium concept that subsequent nourishment volumes should diminish. In light of the historical experience of beach nourishment identified in this study, individual state and local coastal communities should reevaluate their plans for future beach nourishment programs. The complete listing of all the data on nourished beaches from this survey is available at www.geo.duke.edu/Research/psds/psds.htm     

Effects of small-scale turbulence on the growth of two diatoms of different size in a phosphorus-limited medium

The effect of turbulence on the nutrient flux towards osmotrophic cells is predicted to be size dependent. This should translate into growth. We experimentally followed and modelled the growth of two marine diatoms of different size (Thalassiosira pseudonana, 6 μm in diameter and Coscinodiscus sp., ca. 109 μm in diameter) under still water and turbulent conditions, using a shaker table. Experiments were done with phosphorus-limited cultures and lasted for ca. 5 days. Turbulence enhanced the growth of Coscinodiscus sp. in agreement with theory but not the growth of T. pseudonana, which was actually slightly lower under turbulence. At the end of the experiments there were about 1.7 times as many Coscinodiscus sp. cells in the turbulent treatment than in the still treatment, while for T. pseudonana almost the same cell concentration was found in both conditions. In addition, the Coscinodiscus sp. cells growing under still conditions presented a higher specific alkaline phosphatase activity than those growing in turbulence which indicates a higher need for phosphorus in the still cultures. A simple dynamic model, based on Michaelis–Menten nutrient uptake kinetics, needed nearly no optimisation other than using observed initial conditions of phosphate and cell concentrations. The model showed how an increased nutrient flux towards the cells translates non-linearly into cell growth, most likely by affecting the half-saturation constant (K_M). However, since Coscinodiscus sp. experienced significant mortality and cells partially settled to the bottom of the containers, unequivocal support for the size-dependent effect of turbulence on nutrient uptake will require further experiments and more sophisticated modelling. The mechanisms to connect an increased nutrient flux towards cells with population growth and whether this process is size dependent are important in parameterizing the effects of turbulence on marine plankton in coupled physical–biological models.     

A spatially explicit ecosystem model of seston depletion in dense mussel culture

A fully-coupled biological–physical–chemical model of a coastal ecosystem was constructed to examine the impact of suspended mussel culture on phytoplankton biomass in Tracadie Bay, Prince Edward Island, Canada. Due to the extent of mussel culture there, we hypothesised that shellfish filtration would control the concentration and distribution of phytoplankton and other suspended particles in the bay. Circulation was delineated with a tidally-driven 2D numerical model and used to drive an ecosystem model with a focus on pelagic components including phytoplankton production, nutrients, detritus, and mussels. The benthos were treated as a sink. Nutrients and seston were forced by tidal exchange and river input, with phytoplankton additionally forced by light. Boundary conditions of seston and nutrients were derived from field studies with an emphasis on the contrast between spring (high river nutrients, low temperature) and summer (low river inputs and high temperatures). Model output was used to map phytoplankton carbon over the bay for each season and in the presence of mussels and river nutrient input. Results indicate severe depletion effects of mussel culture on overall phytoplankton biomass, but no spatial pattern that can be attributed to grazing alone. Primary production generated by nutrient-rich river water created a mid-bay spike in phytoplankton that dominated the spatial pattern of chlorophyll-based carbon. Model results were validated with surveys from a towed sensor array (Acrobat) that confirmed the river influence and indicated bay-wide depletion of 29% between high and low water. Our model results indicate that the farm-scale depletion emphasised in previous studies cannot simply be extrapolated to seston limitation at the ecosystem level.