Studies of a wind mechanism influencing the recruitment of blue crabs in the Middle Atlantic Bight

This report forms part of an on-going effort to understand the large yearly variations in blue crab harvest of Chesapeake Bay. Recent sampling programs have indicated that the larvae are transported out of the bay immediately after being spawned, and spend their first month offshore at the sea surface. Although it is well established that a mid and outer shelf southward flow occurs during all seasons in the Middle Atlantic Bight, very little is known of the nearshore currents. This study constitutes an effort to determine if the characteristically light, but northward, wind stress during the critical summer months is sufficient to drive northward counter flow at the surface and, hence, to reduce the chances that the larvae are being advected south and lost from the area of Chesapeake Bay.We investigate a local model of wind-driven currents on the continental shelf with vertical decoupling at the pycnocline. Additional driving forces include an alongshore sea surface slope and horizontal pressure gradients. With characteristic forcing values, it is found that the wind stress is indeed sufficient to drive a light northward flow within 25 to 50 km of the shoreline. We expect, then, that blue crab larval recruitment back to Chesapeake Bay may be partially dependent on summer wind stress. A comparison between a wind index time series and harvest several years later is strongly suggestive of such a dependency.     

Occurrence of Irgarol 1051 and its major metabolite in Maryland waters of Chesapeake Bay

Irgarol and its major metabolite (GS26575) were measured in Maryland waters of Chesapeake Bay: (1) in and near 10 marinas, a mainstem Bay site and two Severn River locations during a general survey in July and December of 2002; (2) at various sites in the Port Annapolis Marina and the Severn River area during March of 2002 before the boating season began; and (3) during July (peak boating season) in the same Port Annapolis Marina and Severn River sites area during both an ebb and flood tide. Irgarol concentrations ranged from 1.82 ng/l at the mid-Bay site to 585 ng/l in Port Annapolis marina during the July and December general survey. An Irgarol 90th centile of 239 ng/l was reported for the 10 marina sites, two Severn River sites and one mainstem site sampled during the general survey conducted in July and December. Temporal analysis of all pooled data showed that 90th centiles were over seven times higher in July when compared to December. A comparison of Irgarol concentrations at 12 sites in the Port Annapolis marina and Severn River area during both an ebb and flood tide in July showed no consistent trend with tidal cycle by site although significant reductions in concentrations were reported with distance from the three Port Annapolis marina sites. Ecological risk from Irgarol exposure was judged to be low for most Chesapeake Bay sites sampled. Possible exceptions were Port Annapolis marina, Severn River sites in close proximity to this marina and Chesapeake Harbor marina where Irgarol concentrations exceeded a conservative effects threshold during the peak boating season in July. Ecological risk from GS26575 exposure was low for all sites.     

Resonance and sea level variability in Chesapeake Bay

A numerical model is used to determine the resonant period and quality factor Q of Chesapeake Bay and explore physical mechanisms controlling the resonance response in semi-enclosed seas. At the resonant period of 2 days, the mouth-to-head amplitude gain is 1.42 and Q is 0.9, indicating that Chesapeake Bay is a highly dissipative system. The modest amplitude gain results from strong frictional dissipation in shallow water. It is found that the spatial distribution of energy dissipation varies with forcing frequency. While energy at tidal frequencies is dissipated around topographic hotspots distributed throughout the Bay, energy dissipation at subtidal frequencies is mainly concentrated in the shallow-water lower Bay. An analytic calculation shows that the bottom friction parameter is much larger in Chesapeake Bay than in other coastal systems with strong resonance response. The model-predicted amplitude gains and phase changes agree well with the observations at semidiurnal and diurnal tidal frequencies. However, the predicted amplitude gain in the resonant frequency band (34–54 h period) falls below that inferred from band-passed sea level observations. This discrepancy can be attributed to the local wind forcing which amplifies the sea level response in the upper Bay. The model is also used to show that rising sea levels associated with global warming will shift the resonance period of Chesapeake Bay closer to the diurnal tides and thus exacerbate flooding problems by causing an increase in tidal ranges.     

Cross-Media Models of the Chesapeake Bay Watershed and Airshed

A continuous, deterministic watershed model of the Chesapeake Bay watershed, linked to an atmospheric deposition model is used to examine nutrient loads to the Chesapeake Bay under different management scenarios. The Hydrologic Simulation Program - Fortran, Version 11 simulation code is used at an hourly time-step for ten years of simulation in the watershed. The Regional Acid Deposition Model simulates management options in reducing atmospheric deposition of nitrogen. Nutrient loads are summed over daily periods and used for loading a simulation of the Chesapeake estuary employing the Chesapeake Bay Estuary Model Package. Averaged over the ten-year simulation, loads are compared for scenarios under 1985 conditions, forecasted conditions in the year 2000, and estimated conditions under a limit of technology scenario. Limit of technology loads are a 50%, 64%, and 42% reduction from the 1985 loads in total nitrogen, total phosphorus, and total suspended solids, respectively. Urban loads, which include point source, on-site wastewater disposal systems, combined sewer overflows, and nonpoint source loads have the highest flux of nutrient loads to the Chesapeake, followed by crop land uses.on assignment from NOAA Air Resources Laboratory     

Upper Eocene ejecta of the New Jersey continental margin reveal dynamics of Chesapeake Bay impact

Evidence for the Chesapeake Bay Crater as the source for New Jersey continental margin ejecta is provided by fine-grained tektites and coarse-grained unmelted ejecta. The Upper Eocene ejecta deposit, now demonstrated to be part of the North American strewn field, occurs on the New Jersey continental margin at Ocean Drilling Program (ODP) Sites 904 and 903. The mineralogy, major oxide composition of the ejecta materials, and biostratigraphic age of the enclosing sediments link the origin of these ejecta to the recently recognized Chesapeake Bay impact crater, located only 330 km away. Sediments associated with the ejecta provide information about the dynamics of impact events. The 35-cm-thick ejecta-bearing layer can be subdivided into three subunits that indicate a sequence of events. Bottom subunit III documents sediment failure and deposition of gravel-sized fragments, middle subunit II records deposition of abundant sand-sized ejecta by gravity settling, and upper subunit I contains a 12-cm-thick sedimentary deposit containing rare silt-sized tektites and evidence of waning currents. These events are interpreted by linking sediment deposition to seismic ground motion and subsequent tsunami waves triggered by both the Chesapeake Bay impact and slope failures.     

A multivariate analysis of planktonic molluscan distribution in the Middle Atlantic Bight

Faunal structure is described for the planktonic molluscs of the Middle Atlantic Bight based on two years of seasonal data from the continental shelf. Collection and taxa groups are constructed using numerical classification and reciprocal averaging ordination. Discriminant analysis is used to relate surface collection groups to physical variables, then taxa group distribution among these collection groups is analyzed by nodal fidelity analysis. The areal distribution of dominant species is presented by season, as is their surface temperature-salinity distribution.Four communities are recognized on the continental shelf. A subarctic community, including Limacina retroversa, Paedoclione doliiformis, and Clione limacina, is advected down the central shelf region from the northeast. A Gulf Stream community of weak vertical migrators, including Limacina trochiformis, Cavolinia longirostris, Creseis conica, Atlanta peroni, and A. gaudichaudi, is introduced onto the shelf in occasional intrusions across the shelf-edge front. A depth-limited warm-water community of strong vertical migrators, including Limacina inflata, L. bulimoides, L. lesueuri, and Cavolinia inflexa is generally confined offshore of the 100-m isobath since the extent of their daily vertical migration is greater than the bottom depths on the continental shelf. A coastal community, including the larvae of Loligo pealei and of Ensis directus is found in coastal water of local origin and is generally confined within a coastal boundary layer.     

Circulation in the Great Barrier Reef Lagoon using numerical tracers and in situ data

Numerical hydrodynamic models of the northeastern Queensland shelf, forced by regional winds and modelled boundary currents in the northern Coral Sea, are used to provide improved estimates of general flow trajectories and water residence times within the Great Barrier Reef (GBR) shelf system. Model performance was checked against a limited set of current metre records obtained at Lark Reef (16°S) and the Ribbon Reefs (15.5°S). Estimates of water parcel trajectories are derived from a series of numerical tracer experiments, with daily releases of neutrally buoyant, un-reactive particles at 320 sites along the coast between Cape York (10.7°S) and Hervey Bay (25°S). Flow trajectories and residence times for tracer particles introduced to the GBR lagoon in the southern—ca. 22°S, central—19°S, and northern reef—14°S are emphasised. For purposes of the analysis, the year was divided into two seasons based on mean alongshore current direction. Most coastal sourced tracers entering the central GBR lagoon between 16° and 20°S during the northward-current season (January–August) primarily encounter the outer-shelf reef matrix after exiting the lagoon at its northern “head” (nominally 16°S), after 50–150 days. Up to 70% of tracer particles entering in the southward-current season (August–December) eventually crossed the lagoon to the outer-shelf reef matrix, with median crossing times between 20 and 330 days. During favourable wind conditions, tracers introduced at the coast may move rapidly across the lagoon into the reef matrix. The tracer experiments indicate that most coastal-sourced tracers entering the GBR lagoon remain near the coast for extended periods of time, moving north and south in a coastal band. Residence times for conservative tracer particles (and implied residence times for water-borne materials) within the GBR shelf system ranged from ca. 1 month to 1 year—time frames that are very long relative to development times of planktonic larvae and cycling times for nutrient materials in the water column, implying they are transformed long before reaching the outer reef matrix.     

Coastal and shelf circulation in the vicinity of Camamu Bay (14°S), Eastern Brazilian Shelf

The Camamu Bay (CMB) is located on the narrowest shelf along the South American coastline and close to the formation of two major Western Boundary Currents (WBC), the Brazil/North Brazil Current (BC/NBC). These WBC flow close to the shelf break/slope region and are expected to interact with the shelf currents due to the narrowness of the shelf. The shelf circulation is investigated in terms of current variability based on an original data set covering the 2002-2003 austral summer and the 2003 austral autumn. The Results show that the currents at the shelf are mainly wind driven, experiencing a complete reversal between seasons due to a similar change in the wind field. Currents at the inner-shelf have a polarized nature, with the alongshore velocity mostly driven by forcings at the sub-inertial frequency band and the cross-shore velocity mainly supra-inertially forced, with the tidal currents playing an important role at this direction. The contribution of the forcing mechanisms at the mid-shelf changes between seasons. During the summer, forcings in the two frequency bands are important to drive the currents with a similar contribution of the tidal currents. On the other hand, during the autumn season, the alongshore velocity is mostly driven by sub-inertial forcings and tidally driven currents still remain important in both directions. Moreover, during the autumn when the stratification is weaker, the response of the shelf currents to the wind forcing presents a barotropic signature. The meso-scale processes related to the WBC flowing at the shelf/slope region also affect the circulation within the shelf, which contribute to cause significant current reversals during the autumn season. Currents at the shelf-estuary connection are clearly supra-inertially forced with the tidal currents playing a key role in the generation of the along-channel velocities. The sub-inertial forcings at this location act mainly to drive the weak ebb currents which were highly correlated with both local and remote wind forcing during the summer season.     

Long-term trends in the macrobenthos and water quality of the lower Chesapeake Bay (1985–1991)

Long-term trends in macrobenthic communities of the lower Chesapeake Bay, USA, were examined using data collected quarterly (March, June, September and December) from 1985 to 1991 at 16 stations along a salinity gradient from tidal freshwater regions of the major tributaries (James, York and Rappahannock rivers) to the polyhaline region of the main-stem of Chesapeake Bay. A non-parametric trend analysis procedure was applied to five parameters characterizing macrobenthic community structure: community biomass, species richness, abundance of individuals, proportion of biomass composed of opportunistic species (opportunistic biomass composition) and proportion of biomass composed of equilibrium species (equilibrium biomass composition). For the parameters tested 36 trends were detected. For community biomass, five trends were significant; all had positive slopes and occurred in the James and York rivers. For species richness, six trends were significant; all had positive slopes with three trends in the James River, two trends in the York River and one trend in the main-stem of Chesapeake Bay. For abundance of individuals, 17 trends were detected; all abundance trends were seasonally dependent, had positive slopes and occurred at 12 of the 16 stations. For opportunistic biomass composition, four trends were significant; all had positive slopes with one trend in the lower Rappahannock River and three trends in the main-stem of Chesapeake Bay. For equilibrium biomass composition four trends were significant; two trends had positive slopes (one in the James River and one in the York River) and two trends had negative slopes (one in the Rappahannock River and one in the main-stem of Chesapeake Bay). Trends in the James and York rivers were considered to indicate improving conditions for the benthos, while trends in the lower Rappahannock River and the main-stem of the Chesapeake Bay were considered to indicate deteriorating conditions. Deteriorating conditions for the benthos were associated with regions exposed to summer, low dissolved oxygen events. The trends in the indicators of benthic biological community health were inferentially related to trends observed in water quality conditions in the tributaries and main-stem of Chesapeake Bay. All major water quality and biotic trends appeared to correspond in an ecologically meaningful manner.     

Modelling the potential transport of tropical fish larvae in the Leeuwin Current

Application of an individual-based particle tracking model to the migration of tropical fish larvae along the continental shelf between the Houtman Abrolhos Islands and Rottnest Island (Western Australia) has shown that there is potential for the southwards advection of passive particles/larvae in the Leeuwin Current system throughout the year. However, seasonal variations in the prevailing wind field result in corresponding seasonal changes in the surface current flow (both alongshore and cross-shelf) on the continental shelf, leading to a pulse of modelled particles arriving at Rottnest Island during the autumn months. This matches, within a month, the observed April/May peak in annual recruitment of 2 species of damselfish (Abudefduf sexfasciatus and A. vaigiensis), at the time when the Leeuwin Current is strengthening. It is assumed that the larvae are in the uppermost 20 m of the water column and that there are no vertical diurnal movements.The model has a 10 km grid spacing, and so can resolve some of the current gradients across the continental shelf. Comparison of the modelled near-surface currents with ADCP measurements at 2 current mooring sites as well as with a broader range of historical current measurements off south-western Australia indicates that the alongshore net transport is reasonably well reproduced by the model, whereas agreement with the cross-shelf flow is not as good (this may be partly attributed to the paucity of high-quality near-surface current measurements in the area of study). Because of limitations in our knowledge of the swimming ability, choice of swimming direction and habitat selection of larval fish, as well as the inability of the model to reproduce the small-scale circulation around Rottnest Island, the swimming capacity of the late-stage larval fish is not specifically included; they are considered as potential settlers once they have reached within 20 km of the Island.The observed inter-annual variability in recruitment is not as well matched by the model as is the seasonal pattern, but this is almost certainly due to uncertainties in biological factors such as spawning strength, food and predation en route, which are not known.The modelled results are also applied to a more general discussion of the transport of eggs and larvae of commercial fish and invertebrate species on the Western Australian continental shelf, and it is shown that the seasonality and position on the shelf of the spawning may play a large role in the movement (and hence survival and ultimate recruitment) of different species.     

Assessment of Impact of Storm on Point Source Pollutant Transport in Estuary by Dissolved Tracer Modeling

A continuously discharged dissolved conservative tracer was simulated with the Chesapeake Bay Estuary Model Package to study pollutant transport in the estuary in response to point source loads and the impact of the November, 1985 storm. A visualization technology is applied to show 3-dimensional concentration variations in a continuous daily time sequence. The differential responses of daily net transport during storms versus inter-storm periods can be observed from an MPEG movie. It may take 2–3 months for a tracer to travel from the fall-line to the mouth of a river during relatively dry seasons, only 2 weeks in some medium storms, and less than 5 days in a big storm. Plots of daily concentrations from eleven selected locations in the estuary provide quantitative information on the response of tracer concentration to flows. The magnitude and time of tracer peaks related with different weather events in these locations reflect the combined effects of flows from various directions to these locations. The lower tributaries (which are closer to the Bay mouth) are affected more than the upper tributaries by a source discharged at a mid-tributary. A storm can transport materials more effectively to the Bay and affect adjacent tributaries more severely.     

Shelf sediment dispersal during the dry season,Princess Charlotte Bay,Great Barrier Reef,Australia

Princess Charlotte Bay, located on the northern Great Barrier Reef, is an environment of terrigenous and carbonate deposition. The dynamics on this shelf are controlled by the Great Barrier Reef at the edge of the shelf, and the mid-shelf, shore-normal reefs. This study examines the dynamics during the dry season, with six time-series records from instrumented tripod deployments and numerous hydrographic stations.The shallow nearshore waters and the estuaries prove to be the sites where most active sediment resuspension and transport takes place. Sediment resuspension is effected primarily by waves in the nearshore, and channeling of tidal currents in the estuaries. Bedload transport did not occur during this study, mainly because current velocities were too low. Suspended particulate matter (SPM) transport in the bay is governed by tides and winds. Strong tidal flow imparts a strong offshore component to the transport, and strong southeast winds impart an alongshore component that transports SPM out of the bay to the northwest. Rattlesnake Channel, east of Princess Charlotte Bay, is another route by which SPM leaves the bay. Flow through this channel is predominantly tidal, with ebb waters (leaving Princess Charlotte Bay) carrying higher SPM concentrations than flood waters.SPM flux in the nearshore was an order of magnitude higher than at offshore stations, with highest fluxes generally occurring at times of sustained southeast winds. Transect data show that SPM drops to average bay values in water 11 m deep, indicating most SPM is transported in nearshore waters.     

Effect of the mesoscale hydrographic features on larval fish distribution across the shelf break of East China Sea

Surveys of fish larvae and oceanographic conditions were conducted along transects across the shelf break region in the East China Sea (ECS) in May 2001. The objective of this study was to investigate the distribution and assemblages of fish larvae across the shelf break region and their relationships to mesoscale hydrographic features. There was a warm surface streamer from the Kuroshio which extended toward the shelf region of the ECS; concurrently, we observed the intrusion of the less saline shelf water into the subsurface layer towards the offshore. In all 66 taxa (65 families and 1 order) of larvae were collected by oblique net tows using a bongo net, the larval fish density and number of families sampled were lower in the low-temperature area. Based on cluster analysis and environmental factor, two larval fish assemblages were identified: off-shelf/Kuroshio and shelf break. In shelf break assemblage, Auxis spp. and Diaphus spp. were abundant in the warm surface streamer. In contrast, Maurolicus japonicus, Synagrops spp., Bregmaceros sp. and Champsodon spp. were found in the off-shelf region where the offshore intrusion of less-saline water occurred. This different pattern of the horizontal distribution would reflect vertical distribution of the larvae in the water column. Moreover, copepod nauplii density was higher in the upper layer of the shelf region. Thus, dynamic interactions between the Kuroshio and shelf waters at the frontal boundary are concluded to potentially affect larval fish transport and prey availability.     

Regularities of hydrological processes in the Chesapeake Bay (USA): Case study of a classical estuary

The main regularities of hydrological and hydrological-environmental processes occurring within the complex estuary, the Chesapeake Bay and the mouths of its tributaries, are discussed. The peculiarities of the estuary morphological structure, including the structures of tidal and net currents, salinity and water turbidity fields and their variability, the environmental conditions, and their human-induced changes. Using the Chesapeake Bay as an example, it became possible to reveal the basic features of classical estuaries subject to a considerable impact of river runoff and featuring mixing of river and sea water and moderate stratification of the water mass. It is shown that the regularities of hydrological processes in the Chesapeake Bay are typical of many mouth water bodies of estuarine type (inlets, drowned river valleys, lagoons, and tidal estuaries proper).     

The influence of wind forcing on across-shelf transport in the Florida Keys

Acoustic Doppler current profiles and current meter data are combined with wind observations to describe the transport of water leaving Florida Bay and moving onto the inner shelf on the Atlantic side of the Florida Keys. A 275-day study in the Long Key Channel reveals strong tidal exchanges, but the average ebb tide volume leaving Florida Bay is 19% greater than the average flood tide volume entering the bay. The long-term net outflow averages 472 m³ s⁻¹. Two studies in shelf waters describe the response to wind forcing during spring and summer months in 2004 and during fall and winter months in 2004–2005. During the spring–summer study, southeasterly winds have a distinct shoreward component, and a two-layer pattern appears. Surface layers move shoreward while near-bottom layers move seaward. During the winter study, the resultant wind direction is parallel to the Keys and to the local isobaths. The entire water column moves in a nearly downwind direction, and across-shelf transport is relatively small. During the summer wet season, Florida Bay water should be warmer, fresher, and thus less dense than Atlantic shelf waters. Ebbing bay water should move onto the shelf as a buoyant plume and be held close to the Keys by southeasterly winds. During the winter dry season, colder and saltier Florida Bay water should leave the tidal channels with relatively high density and be concentrated in the near-bottom layers. But little across-shelf flow occurs with northeasterly winds. The study suggests that seasonally changing wind forcing and hydrographic conditions serve to insulate the reef tract from the impact of low-quality bay water.     

Overlap between Atlantic bluefin tuna spawning grounds and observed Deepwater Horizon surface oil in the northern Gulf of Mexico

The 2010 Deepwater Horizon oil spill impacted the northern Gulf of Mexico (GOM) during the spring spawning season of Atlantic bluefin tuna (BFT). Overlap between BFT spawning habitat and surface oil in the northern GOM was examined using satellite-derived estimates of oil coverage, and spawning habitat models. Results suggested that although eggs and larvae were likely impacted by oil-contaminated waters in the eastern GOM, high abundances of larvae were located elsewhere, especially in the western GOM. Overall, less than 10% of BFT spawning habitat was predicted to have been covered by surface oil, and less than 12% of larval BFT were predicted to have been located within contaminated waters in the northern GOM, on a weekly basis. Our results provide preliminary but important initial estimates of the effects of the spill on larval BFT mortality, as concern continues over the appropriate management responses to impacts of the spill.     

Identification and characterization of lipolytic and proteolytic bacteria isolated from marine sediments

Lipolytic and proteolytic bacteria were isolated from sediments at two sampling stations in the New York Bight Apex and one sampling station each in Sandy Hook Bay and Great Bay, New Jersey. The stations in the Bight Apex and Sandy Hook Bay have received industrial wastes and sewage for several decades, while Great Bay has received little of such materials. Proteolytic counts were 2–4 times higher and lipolytic counts generally 4 times higher in the polluted areas. Of the isolates taken from casein and lipid plates, 76% were Gram-negative rods; 80% of the latter were identified as Vibrio and Pseudomonas. The vibrios comprised more than 60% of the isolates from Station 4 (Great Bay) and Station 1 (Bight Apex), and were tested for their ability to break down casein, lipid, starch, and chitin. From Station 1, 75% of the Vibrio were active in degrading one or more substrates in addition to the substrate of the initial isolation medium; from Station 4, 52% of the Vibrio were active.     

美国Chesapeake湾撞击坑研究进展及其启示

位于美国弗吉尼亚东海岸直径85 km的Chesapeake湾撞击坑,是十几年前发现的由一颗陨星撞击形成的一个复杂撞击坑. 该坑的研究经历四个阶段：地下水调查、撞击坑的发现、美国多学科多部门的综合研究和即将进行的国际钻探取心项目. 钻井岩心中的角砾成份和微体化石,提示撞击坑的存在,并确定撞击发生在35 Ma前,即始新始晚期. 地震反射剖面资料帮助寻找到撞击坑的具体位置,确定撞击坑的结构和形态特征. Chesapeake湾撞击坑埋藏在新生界沉积层之下,是全球已知最大的、保存最好的撞击坑之一. Chesapeake湾撞击坑主要形态像一顶倒置的宽边大草帽,包括外缘、环状洼地、峰环（内缘）、内盆和中央峰. 撞击坑的形成破坏了原来的含水层,撞击坑当时即被富含咸水的抛射角砾岩和海啸角砾岩充填,再被后来的沉积层覆盖. Chesapeake湾撞击坑导致地面沉降、河流变向、海岸含水层的中断、内陆咸水楔的出现、地震,决定Chesapeake湾本身的位置,至今仍然影响当地居民的生活. 了解Chesapeake湾撞击坑对我国撞击坑研究具有借鉴作用.     

Cold anticyclonic eddies formed from cold pool water in the southern Middle Atlantic Bight

AVHRR satellite imagery of the southern Mid-Atlantic Bight during May 1993 revealed a large area of cold water over the shelf break and slope that appeared to spin up into a series of southward propagating anticyclonic eddies. The eddies had diameters of 35–45 km at the surface and moved southward at about 20 cm/sec. A radial TOYO CTD (to 50m) and ADCP velocity (to 400m) transect was conducted across the southern-most of these eddies. The upper 50 meters had minimum temperatures of less than 7°C and salinities of about 33 pss, characteristics similar to cold pool waters usually found over the continental shelf. ADCP velocity data from one of the eddies revealed anticyclonic flow extending to a depth of about 250m. The transport of cold pool water by the eddies was estimated to be 0.1 to 0.2 Sv which is of the same order as the annual mean alongshore transport of shelf water in this region. The origin of the deeper water within the eddy is unlikely to be the continental shelf because the shelf break is less than 100 m. The depth and velocity profiles along the TOYO transect were consistent with the constant potential vorticity eddy model of Flierl (1979) although the source of the eddy kinetic energy is uncertain. The cause for the exodus of cold pool water from the shelf, which extended northward to at least 38°N, is unclear but must involve the establishment of an alongshore baroclinic pressure gradient against the usual southwestward shelf flow. It is possible that the intrusion of Gulf Stream waters onto the shelf near Cape Hatteras was a precursor of this off shelf transport. The southern-most eddy was marked by high biological productivity and very high oxygen supersaturation. The phytoplankton bloom detected within the exported cold pool water, located over the continental slope, suggests a mechanism whereby production fueled by nutrients derived from the shelf can be locally exported into deep water.