[This paper was presented at the AFS “Fisheries in a Changing Climate” symposium

[This paper was presented at the AFS "Fisheries in a Changing Climate" symposium

held in Phoenix, AZ during August 2001. To be published in a proceedings volume]

Potential Consequences of Climate Change for the Fish Resources

in the Mid-Atlantic Region

David G. Mountain

Northeast Fisheries Science Center/NOAA/NMFS

Woods Hole, MA 02543

ABSTRACT

Anticipated changes in climate for the mid-Atlantic region of the United States seaboard likely will result in increased water temperatures, more intense development of seasonal stratification and changes in the regional circulation. Each of these factors may affect the fish stocks in the region. Direct effects would include a general northward shift in stock distributions in response to increased water temperatures, reduced reproductive success for some cold water species due to temperatures rising above the optimum for larval growth, and greater frequency of hypoxic conditions in coastal areas due to stronger stratification. Indirect effects may occur through changes in the processes controlling phytoplankton productivity and species composition and, subsequently, the productivity of the zooplankton populations that are the primary prey of larval fish. By influencing larval survival and recruitment, these bottom-up effects could represent the most important pathway for changes in climate to affect fish resources. However, evaluating the consequences of bottom up changes will require an improved ecosystem modeling capability.

INTRODUCTION

The northeast shelf of the United States has diverse finfish and shellfish communities that have supported economically important commercial and recreational fisheries for hundreds of years. The purpose of this paper is to consider the potential consequences of anticipated changes in climate for these fish resources. The region being considered is the mid-Atlantic continental shelf from Cape Hatteras northward to the Gulf of Maine (Figure 1). The consequences of climate change on the estuarine areas in the region are the subject of another paper in this symposium (see Wood et al. 2001) and are not considered here.

A number of previous analyses (e.g., Colton 1972; Frank et al. 1990; Mountain and Murawski 1992; Murawski 1993; Boesch et al. 2000) have considered the influence on the fish stocks of environmental variability or changes in climate in the mid-Atlantic region. A major focus of these efforts has been distributional changes by the stocks in response to changes in water temperature. Identifying the influence of environmental variability on the productivity and abundance of the stocks over recent decades, while desirable, is quite difficult. As pointed out by Frank et al. (1990), since the 1960s the environmental influence on abundance has been largely overshadowed by the effects of heavy fishing pressure. Estimating the consequences of climate change on fish productivity and abundance, therefore, requires projecting the likely effects of environmental changes on fish reproduction, growth and survival, without the benefit of clear analogs in the observational record.

Regional background: Physiographically the mid-Atlantic shelf region is comprised of two very different areas. The Middle Atlantic Bight (MAB), from Cape Hatteras to about 70° W, is a smoothly sloping shelf from the coast to the shelf break that occurs at about 100m depth. The shelf is about 100 km wide and the bottom is largely coarse sand. By contrast the Gulf of Maine is a deeper, semi-enclosed area. Within the gulf are several basins >200 m deep, separated by steep and rocky ridges. Georges Bank is an eastward extension of the MAB shelf and forms the southern boundary for the Gulf of Maine. The only deep connection from the Atlantic Ocean to the gulf is the Northeast Channel, between Georges Bank and Browns Bank. Oceanographically, the waters throughout the region originate from two primary sources. Relatively cold, low salinity water from the Scotian shelf enters the gulf near the surface around Cape Sable (Smith 1983). Relatively warm and saline water from the offshore slope region (Slope Water) enters the gulf at depth through the Northeast Channel (Ramp et al. 1985). These two water masses mix as they flow around the gulf to form the Shelf Water that then flows clockwise around George Bank and southwestward through the MAB. In the MAB the inflow of waters from the Gulf of Maine/Georges Bank represents an advective cooling that partially counteracts local surface heating in summer, and results in a ‘cold-pool’ of water at depth along the outer part of the shelf (Houghton et al. 1982). The north-to-south flow is part of a long, continental scale coastal current system that can be traced northward to Labrador (Chapman and Beardsley 1989). The annual cycle in surface water temperature ranges from about 10 °C over the deep Gulf of Maine to greater than 20 °C on the shallow shelf south of Delaware Bay (Figure 2). The region has experienced interannual variations in water temperatures of about 2 °C over recent decades (Figure 3). Stratification develops in the summer over most of the region due to the combination of increased surface heating and river runoff, except on the shallow (< 60 m deep), central part of Georges Bank where strong tidal currents keep the water column well mixed year round.

The species composition and distributions of the fish community in the region are well documented. The National Marine Fisheries Service has conducted standardized trawl surveys throughout the region each spring since 1968 and each fall since 1963 (Azarovitz 1981). Grosslein and Azarovitz (1982) describe the region as a transitional faunal province with significant overlap of cold temperate and warm temperate fish species and attribute dramatic seasonal shifts in the species composition to the large seasonal temperature changes indicated in Figure 2. They identify 180 fish species caught on the NMFS trawl surveys, with about 60% being warm water species, 35% being cold water species and only 10 species that are year-round residents on the Mid-Atlantic region from Cape Hatteras to Georges Bank. Many of the southern species migrate into the region from the south only during the summer. The region represents the northern limit in distribution for many of the warm water species and the southern limit for the cold water species.

ANTICIPATED CHANGES IN CLIMATE

A summary of likely changes in climate along the mid-Atlantic coastal region is given by Boesch et al. (2000), as part of a nation-wide analysis of the effects of climate change in coastal areas. The climate projections are based on two general circulation models – the Canadian Global Coupled Model and the Hadley Centre Coupled Model. Anticipated changes include warmer surface temperatures, with increases of 2-3 °C in winter and 3-4 °C in summer for the mid-Atlantic region. The report points out that because of the variety of physical processes influencing water temperature distribution, sea surface temperatures are not well projected in the global models. Both models suggest enhanced low pressure systems along the east coast in winter. This could increase the southward water transport along the shelf (Boesch et al. 2000). In summer, a stronger anticyclonic tendency could have the reverse effect. The models suggest opposite changes in the fresh water input to the coastal system. The Canadian model projects a decrease in runoff, while the Hadley model projects an increase. Both models project a sea-level rise of about 50 cm over the next century. Wright et al. (1986) considered the implications of climate change for oceanic conditions along coastal Canada, which also could have important implications for the U.S. mid-Atlantic region. They anticipate increased precipitation and runoff along eastern Canada that would lead to an increase in the transport of the Labrador Current. They also expect a decrease in the strength of the winds over the north Atlantic, which could weaken the Gulf Stream and decrease the frequency of warm-core Gulf Stream rings that impinge on the outer edge of the mid-Atlantic continental shelf.

The consequences for the fish resources of three aspects of these anticipated changes in climate will be considered: an increase in temperature, an increase in stratification and changes in the regional circulation. Temperature is a primary variable influencing biological processes. With increased temperature (and assumably increased surface heat flux) and possibly increased fresh water inflow to the coastal system, an increase in the seasonal stratification would be expected (Wright et al 1986; Boesch et al 2000). Stratification can influence a number of biological processes, particularly those related to primary production. An increase in the transport of the Labrador Current and of the large-scale coastal current system both could significantly influence the oceanographic conditions in the Gulf of Maine and MAB. The anticipated rise in sea level could have important implications for the estuarine and near coastal areas, but not likely for the open shelf region considered here.

CONSEQUENCES OF CLIMATE CHANGES

Temperature: As indicated above, many of the fish species in the region undergo seasonal migrations in response to seasonal changes in temperature. Therefore, climate-scale changes in temperature would be expected to cause changes in the spatial distributions and migration patterns of many species found in the region. The anticipated change (2-4 °C) is relatively small compared to the annual range (Figure 2). The climate-induced increase in temperature, therefore, would not result in conditions beyond those currently experienced, except for the maximum values in the late summer. Changes in the timing of migration and a general northward shift in distribution might be expected for many species, but not a significant change in the overall species composition in the region as a whole.

Interannual variations in water temperature over recent decades (e.g., Figure 3) when detailed fish survey data were collected provide examples of the adjustments the fish stocks have made in their distributions in response to temperature changes. As suggested by Murawski (1993), insight into the response to climate change may be gained through analogy to these historic conditions. Colton (1972) compared the distributions of four goundfish species (American plaice (Hippoglossoides platessoides), haddock (Melanogrammus aeglefinus), yellowtail flounder (Limanda ferruginea) and butterfish (Peprilus triacanthus)) during warm conditions in the 1950’s and cool conditions in the 1960’s. With the cooling, the southern limit of plaice was extended and the northern limit of butterfish was reduced. Haddock and yellowtail, however, did not exhibit evident shifts in distribution, which Colton attributed to their location being more influenced by the bottom type appropriate for spawning or feeding than by temperature. Frank et al. (1990) summarized the results from a number of studies, identifying a variety of species whose distributions have shifted in earlier cool and warm periods. They pointed out that in addition to north/south movement, some species changed depth to find desired temperature conditions. For example, Taylor et al. (1957) showed that during the warming in the 1940s silver hake (Merluccius bilinearis), which generally favored the warmer, deeper waters of the Gulf of Maine, were found year round in the shallow waters of Georges Bank.

Mountain and Murawski (1992) analyzed the temperature, depth, and latitude of the catch for 30 fish and squid species from NMFS spring surveys from 1968 to 1989. Comparing the weighted average temperature where the fish were caught with the area averaged temperature in the region showed that many species compensated for the interannual temperature changes. For example, the catch temperature for haddock exhibited a significant relation to the Georges Bank bottom temperature, with a regression slope of 0.34 (Figure 4). For each degree the average bottom temperature changed, the haddock population changed its distribution to experience only about a third of a degree change. Fourteen of the stocks considered compensated for at least half of the interannual changes in temperature. Haddock appeared to accomplish this compensation by changing depth, moving shallower in years Georges Bank was warmer (Figure 5). The movement to shallower waters represents a movement to cooler water because in spring the shallow portions of the Bank are cooler than the deeper areas. A number of species moved north-south along the shelf, changing their latitude to minimize the temperature change experienced, for example Atlantic mackerel (Clupea harengus) (Figure 6).

Murawski (1993) analyzed the same type of temperature, depth, and latitude data from both the spring and fall NMFS trawl surveys in a principle component analysis. Three groups of species were identified from the scores of the first two components: a warm-water, highly migratory group, a deep-water sedentary group and a shallow-water sedentary group. The migratory group changed latitude with change in temperature. The sedentary groups experienced high variability in bottom temperature, but did not exhibit major shifts in location. Murawski pointed out that an important aspect of the temperature induced movements would be changes in the distributional overlap between species that could alter predator-prey interactions. For example, the interaction between a sedentary predator like Atlantic cod (Gadus morhua)and its migratory prey (Atlantic mackerel or Atlantic herring (Scomber scombrus)) might be effected (Murawski 1993). At the same time, the mackerel migration in the spring is timed such that it may pass over Georges Bank when the larvae of cod and haddock are in the water column and the mackerel can be a significant predator on the developing gadid year classes (Michaels 1991). The temperature-induced changes in predator-prey interactions could affect the recruitment success of some fish stocks, as well as the growth and mortality of their adult predators.

Other important life history events, such as spawning, can be sensitive to temperature changes, too. Marak and Livingston (1970) reported the timing of haddock spawning on Georges Bank was about a month earlier during a group of warm years compared to that of other years when the water temperatures were about 2 °C cooler. Changes in spawning time could alter predator-prey interactions between the resulting larvae and their adult fish predators. A more complicated temperature dependent trophic relationship influencing larval survival was suggested by Frank et al. (1990). Warmer Gulf of Maine surface waters in winter can lead to an early development of the ctenophore populations, which feed on the same zooplankton prey as the gadid larvae. Increased competition for food could result in lower larval survival, as may have been the case in 1983 when the winter water temperatures were warm, ctenophore populations were early and abundant and the subsequent haddock recruitment was poor.

Temperature affects the metabolism of organisms. Brander (1995) showed that cod growth in different stocks around the North Atlantic basin increased significantly with temperature. Boesch et al. (2000) pointed out that this increased growth could influence the reproductive output of the population since fecundity often is a function of size. Increased metabolism also means an increased food requirement. For the adult cod populations studied by Brander (1995), the increase in growth occurred throughout the temperature range encountered, implying that the available food was sufficient to meet the increased metabolic demand. Buckley et al. (2001), however, show that for larval cod and haddock on George Bank, growth rate reaches a maximum at about 7 °C and decreases at higher temperatures. In laboratory studies with excess food, the larval growth rate increases up to temperatures above 14 °C. The authors concluded that in the field the gadid larval metabolism and the available food are in optimal balance at about 7 °C, and the larvae become food limited at higher temperatures. Temperature-induced changes in the growth and survival of the early life stages could influence the reproductive success and sustainable population level of some fish stocks.

Increased water temperature could alter the nutrient dynamics supporting the primary production in the region. Townsend and Thomas (2001) show that the winter-spring bloom of diatoms on Georges Bank in 1999 ended by May due to silicate limitation. However, diatom abundance again increased in June, evidently on regenerated silicate after rising water temperatures increased the dissolution rate of diatom frustules from the earlier bloom. Higher water temperature would enhance the recycling of silicate and, perhaps, contribute to continuous diatom production through out the season.

Stratification: Wright et al. (1986) predict an increase in stratification with a changing climate due to increased surface heating, increased coastal runoff and a decrease in the mean winds. The degree of change in stratification in the coastal regions would depend on the interaction of a variety of processes controlling the air-sea exchange and vertical mixing within the water column. The potential change in stratification could be quite large, with important biological implications. The non-linearity of the equation of state for seawater results in 40% more buoyancy from a unit of heat at 6.5 °C than at 4.5 °C (i.e., the expansion coefficient for sea water increases with temperature). If a water column comes out of the winter warmer, when spring heating begins, each unit of heat will generate more surface layer buoyancy and more stratification. The development of stratification has a very strong feedback mechanism – a small degree of stratification tends to trap subsequent heating in the surface layer, increasing the stratification and increasing the trapping of further heating. A subtle difference in the dynamics can be amplified to result in a large difference in the ultimate stratification. Using a simple one-dimensional water column model (James 1977) with surface heat flux and wind forcing appropriate for the Gulf of Maine (Mountain et al. 1996), the development of seasonal stratification was determined for water columns initially at 4.5 °C and at 6.5 °C (figure 7). Through the spring, the density difference over the upper 50 m of the water column for the warm case is about 30-50% greater than for the cold case and at the summer maximum, about 0.5 sigma-t units greater. This difference is due solely to the temperature dependence of the expansion coefficient of seawater. With an increased surface heat flux, increased freshwater runoff and decreased wind stress, as projected by Wright et al. (1986), the stratification under the warmer climate scenario could be increased substantially more.

An increase in coastal stratification would favor the growth of smaller dinoflagellates rather than larger diatoms, and likely increase the steps in the food chain, with an associated loss of secondary production (Mann and Lazier 1996). Changes in phytoplankton species composition and size spectrum could affect the productivity and composition of the zooplankton community, in which many species select food more on the basis of size rather than taxonomy (Viliela 1995). These changes, in turn, could affect the growth and survival of fish larvae by changing the size and composition of their prey field. Greater stratification also could inhibit the transfer of organic material to the bottom and adversely affect the benthic productivity (Frank et al. 1990), particularly if combined with higher water temperatures which would increase the metabolic requirements of the benthos. Reduced benthic production could be particularly important because of the large demersal fish community that relies upon benthic prey.

Stronger stratification could result in more frequent occurrence of hypoxic events in the MAB by inhibiting the vertical exchange of oxygenated surface waters with the deeper layers. In 1976 the unusually early and intense development of stratification led to a major hypoxic event over a large portion of the northern bight, leading to major mortalities of the shellfish populations (Swanson and Sinderman 1979). Less dramatic events are not uncommon in the coastal waters of New Jersey and Delaware (Reid et al. 1987). Even when not lethal, low oxygen levels can reduce the growth rates of demersal fish species (Bejda et al. 1992).

Circulation Changes: Wright et al. (1986) conclude that one consequence of climate change would be an increase in the transport of the Labrador Current. Increased Labrador Current transport has been linked to westward penetration of cold, Labrador Slope Water south of the Scotian Shelf and the mid-Atlantic shelf (Petrie and Drinkwater 1993; Rossby and Benway 2000). In the 1880’s the occurrence of cold, slope water along the edge of the mid-Atlantic Shelf is believed to have been the cause of a large mortality of tilefish (Marsh et al. 1999). In the mid-1960’s Labrador Slope Water extended westward to Northeast Channel and flooded the bottom layers of the Gulf of Maine, reducing the water temperatures by about 2 °C (Petrie and Drinkwater 1993). Increased Labrador Current transport and westward penetration of Labrador Slope Water in the 1880’s and 1960’s are believed related to large-scale atmospheric forcing associated with low values in the North Atlantic Oscillation (NAO) index (Marsh 2000 and Loder et al. 2001). A model analysis indicates that the westward transport along the southern edge of Geoges Bank into the MAB also increased during the cold, low NAO 1960’s period (Loder et al. 2001).

As in observed in the 1880’s and 1960’s, an increase in the transport of Labrador Slope Water likely would cause a cooling in the mid-Atlantic region. This advective cooling would be opposite in sign to the atmospherically driven warming discussed above and, in general, have opposite implications for the fish stocks in the region.

DISCUSSION

The consequences of climate change for the fish populations in the mid-Atlantic region will be determined by the combined effects of the different aspects of climate change identified above. When combined, these effects may tend to reinforce or to counteract each other in their influence on the productivity and abundance of the fish stocks.

As noted above, the climate models suggest increases in surface water temperature of 2-3 °C in winter and 3-4 °C in summer. These changes would not necessarily extend uniformly to the bottom, where many of the important fish species spend most of their lives. During winter when the water column generally is well mixed, the anticipated increase in temperature likely would be uniform through the water column. In summer, however, stratification inhibits the downward penetration of surface heating, and the temperature increase at the surface likely would be greater than that at the bottom – particularly if the stratification itself increases, as suggested above. Still, with increased winter temperatures and increased surface heating throughout the year, the bottom temperatures during the summer would be expected to increase at least as much as the bottom temperatures in winter.

An increase in water temperature could result in the temperature always being above the current optimum for growth for some cold water species. For example, a 3 °C increase would cause the winter minimum temperature on Georges Bank to be at or above 7 °C. This would insure sub-optimal conditions for the growth and survival of the larval cod and haddock populations on the bank, despite any shift in spawning time. The consequence likely would be a reduced reproductive potential for these stocks. Comparable information on the temperature dependence of larval growth in the field does not exist for other species, but all cold water species in the region would be vulnerable to this effect. However, the optimum temperature for growth represents a balance between the organism’s metabolism and its available food. A climate-induced change in lower trophic level productivity or species composition could change the optimal temperature for growth. Evaluating the net effect of a changing climate would need to consider both the temperature effects on metabolism coupled with potential changes in the prey field.

The changes in the regional circulation could tend to counteract the anticipated warming associated with atmospheric heating. If Labrador Slope Water does extend westward and fill the deeper levels of the Gulf of Maine as in the mid-1960’s, the bottom temperatures might become cooler (Fgure 3) while the surface temperatures become warmer. A cooling could influence demersal species in the gulf, such as the shrimp population that prefers temperatures less than about 7 °C (Clark et al. 2000). During the Labrador Slope Water period in the mid-1960’s the shrimp population exhibited a wider spatial distribution and an apparent higher level of production than during periods of warmer bottom conditions (Clark et al. 2000). A greater water transport through the Gulf of Maine system would increase the advective cooling in the MAB that leads to the seasonal development of the ‘cold-pool’ (Houghton et al. 1982) and could counteract a local increase in surface heating. An increase in stratification also could more efficiently trap surface heating in the upper portion of the water column and reduce the effect of increased surface heating on bottom temperature conditions. The net result might be either a warming or cooling of the bottom waters along the outer MAB shelf, with different implications for the local fishery resources. For example, the surf clam and scallop populations avoid areas where temperatures rise above about 15 °C (Cargnelli et al. 1999 and Packer et al. 1999), and their distributions in the MAB are constrained between warm oceanic slope water at the shelf edge and seasonally warming waters on the inner shelf. A net cooling of the bottom waters would be expected to contribute to an increase in the distribution and productivity of these stocks, and a net warming to a decrease.

Historic analogy does provide important insight into distributional adaptations by the adult fish stocks, as illustrated by Murawski (1993). With an increase in temperature due to climate change, a general northward shift in many of the fish populations would be expected. The species composition observed at a particular location likely could change. In addition, the highest temperature anomalies observed since the routine fish surveys began in the 1960’s rarely approached the values anticipated by the current climate change models, and then only for short periods of time (Figure 3). Therefore the accumulated effect of a shift in distributions, such as through altered predator-prey interactions, has not been observed. The same is true for the accumulated effects of changes on lower trophic level productivity and species composition – through stratification or temperature-dependent nutrient regeneration rates. These changes could directly influence recruitment processes and the reproductive success of fish stocks, and may represent the most important path for climate change to affect the fisheries in the region.

Our ability to evaluate the consequences of climate change on the ecosystem through a bottom-up pathway is very limited. Lacking an observed, historic analog, evaluation can be accomplished only through modeling. Until recently our understanding of ecosystem dynamics has been inadequate to support a meaningful, integrated modeling exercise. However, large interdisciplinary research programs in the region, such as the GLOBEC Georges Bank program (GLOBEC, 1992) and the ECOHAB program (Townsend et al. 2001), are beginning to provide process level understanding of the system dynamics and coupled physical-biological models incorporating that understanding. In the future these models can be the basis for a more informed evaluation of the bottom-up consequences of climate change.

Given an appropriate ecosystem model, better information would be needed in at least two areas to evaluate the consequences of a changing climate. The first is a better understanding of the changes that would occur in climate forcing, particularly for precipitation. The Canadian and Hadley climate models referenced by Boesch et al. (2000) provide opposite predictions concerning precipitation, the former predicting a decrease and the latter an increase. Changes in freshwater input would contribute to changes in seasonal stratification and in the buoyancy driven coastal circulation patterns, including the Labrador Current. Second, a better understanding of the changes in regional circulation is needed, particularly changes in the westward penetration of the Labrador Slope Water and in the strength of the coastal circulation along the Scotian Shelf, through the Gulf of Maine and into the MAB.

CONCLUSIONS

With the currently anticipated changes in climate, the primary consequences for the fish populations in the mid-Atlantic region are likely to be:

1) A northward shift in the distribution pattern for many species in response to a general increase in water temperatures;

2) An increase in seasonal stratification, that may lead to more frequent and more intense episodes of hypoxia in the coastal regions of New Jersey and Delaware, and cause changes in the phytoplankton species composition and productivity;

3) A reduction in the reproductive potential of some cold water species due to temperatures rising above the optimum for larval growth and survival.

The largest effects on the fisheries may come through changes at the lower trophic levels affecting the recruitment success for some stocks. Evaluation of these potential effects will require an improved ecosystem modeling capability.

References

Azarovitz, T.R. 1981. A brief historical review of the Woods Hole Laboratory trawl survey time series. Pages 62-67 in W.G. Doubleday and R. Rivard, editors. Bottom Trawl Surveys. Can. Spec. Publ. Fish. Aquat. Sci. 58.

Bejda, A.J., B.A. Phelan, and A.L. Studholme. 1992. The effect of dissolved oxygen on the growth of young-of-the-year winter flounder, Pleuronectes americanus. Env. Biol. Fishes 34:321-327.

Boesch, D.F., J.C. Field and D. Scavia (Eds.) 2000. The Potential Consequences of Climate Variability and Change on Coastal Areas and Marine Resources Sector Team, U.S. National Assessment of the Potential Consequences of Cliamte Variability and Change, U.S. Global Change Research Program. NOAA Coastal Ocean Program Decision Analysis Series No. #21. NOAA Coastal Ocean Program, Silver Spring, MD.

Brander, K. 1995. The effect of temperature on growth of Atlantic cod (Gadus morhua). ICES Journal of Marine Science 52:1-10.

Buckley, L.J., E.M. Caldarone and R.G. Lough. 2001. Optimum temperature and food-limited growth of marine fish larvae: implications to climate change. Fisheries Oceanography (submitted).

Cargnelli, L.M., S.J. Griesbach, D.B. Packer, E. Weissberger. 1999. Essential Fish Habitat Source Document: Ocean Quahog, Arctica islandica, life history and habitat characteristics. NOAA Tech. Mem. NMFS-NE-148.

Chapman, D.C. and R.C. Beardsley. 1989. On the origin of shelf water in the Middle Atlantic Bight. J. Phys. Oceanogr. 19:384-391.

Clark, S.H., S.X. Cadrin, D.F. Schick, P.J. Diodati, M.P. Armstrong and D.McCarron. 2000. The Gulf of Maine northern shrimp (Pandalus borealis) fishery: a review of the record. J. Northw. Atl. Fish. Sci. 27: 193-226.

Colton, Jr., J.B. 1972. Temperature trends and the distribution of groundfish in continental shelf waters, Nova Scotia to Cape Hatteras as determined from research vessel survey data. Fisheries Bulletin 75:1-21.

Frank, K.T., R.I. Perry and K.F. Drinkwater. 1990. Predicted response of Northwest Atlantic invertebrate and fish stocks to CO2-induced climate change. Transactions of the American Fisheries Society 119:353-365.

GLOBEC. 1992. U.S. GLOBEC Northwest Atlantic Implementation Plan. U.S. GLOBEC Report No. 6, 69 pp.

Grosslein, M.D. and T.R.Azarovitz. 1982. Fish Distribution. MESA New York Bight Atlas monograph 15. New York Sea Grant Institute. 182 pp.

James, I.D. 1977. A model of the annual cycle of temperature in a frontal region of the Celtic Sea. Estuarine and Coastal Marine Science 5:339-353.

Houghton, R.W., R. Schlitz, R.C. Beardsley, B. Butman, and J.L. Chamberlin. 1982. Middle Atlantic cold pool: evolution of the temperature structure during summer, 1979. J. Phys. Oceanogr. 12:1019-1029.

Loder, J.W., J.A. Shore, C.G. Hannah and B.D. Petrie. 2001. Decadal-scale hydrographic and circulation variability on the Scotia-Maine region. Deep-Sea Res. II 48: 3-35.

Mann, K.H. and J.R.N. Lazier. 1996. Dynamics of Marine Ecosystems. 2^nd Ed. Blackwell Science, London.

Marak, R.R. and R. Livingston. 1970. Spawning dates of Georges Bank haddock. International Commission for the Northwest Atlantic Fisheries Research Bulletin 7:56-58.

Marsh, R., B. Petrie, C.R. Weidman, R.R. Dickson, J.W. Loder, C.G. Hannah, K. Frank, and K. Drinkwater. 1999. The 1882 tilefish kill – a cold event in shelf waters of the northeastern United States? Fisheries Oceanography 8:39-49.

Michaels, W.L. 1991. The impact of mackerel predation on the survival of pelagic age-zero sand lance, cod and haddock on Georges Bank during the spring of 1986. Master’s thesis, Southeastern Massachusetts University.

Mountain, D.G. and T.J. Holzwarth. 1989. Surface and bottom temperature distribution for the northeast continental shelf. NOAA Tech. Memo. NMFS-F/NEC-73. 32 pp.

Mountain, D.G. and T.J. Holzwarth. 1990. Surface and bottom temperature and temperature anomaly time series from the Northeast Fisheries Center spring and fall bottom trawl survey program, 1963-1989. NAFO SRC Doc. 90/13.

Mountain, D.G. and S.A Murawski. 1992. Variation in the distributions of fish stocks on the northeast continental shelf in relation to their environment. ICES Marine Science Symposia 195:424-432.

Mountain, D.G., G.A. Strout and R.C. Beardsley. 1996. Surface heat flux in the Gulf of Maine. Deep-Sea Res. II 43:1533-1546.

Murawski, S.A. 1993. Climate change and marine fish distributions: forecasting from historical analogy. Transactions of the American Fisheries Society 122:647-658.

Packer, D.B., L.M. Cargnelli, S.J. Griesbach and S.E. Shumway. 1999. Essential Fish Habitat Source Document: Sea scallop, Placopecten magellanicus, life history and habitat characteristics. NOAA Tech. Mem. NMFS-NE-134.

Petrie, B. and K. Drinkwater. 1993. Temperature and salinity variability on the Scotian shelf and in the Gulf of Maine 1945-1990. J. Geophy. Res. 98:20,079-20,089.

Ramp, S.R., R.J. Schlitz and R.W. Wright. 1985. The deep flow thorugh the Northeaast Channel Gulf of Maine. J. Phys. Oceanogr. 15: 1790-1808.

Reid, R.N., JM.C. Ingham, and J.B. Pearce (Eds.) 1987. NOAA’s Northeast Monitoring Program (NEMP): a report of progress of the first five years (1979-1984) and a plan for the future. NOAA Tech. Memo. NMFS-F/NEC-44.

Rossby, T. and R.L.Benway. 2000. Slow variations in the mean path of the Gulf Stream east of Cape Hatteras. Geophys. Res. Letters 27:117-120.

Smith, P.C. 1983. The mean and seasonal circulation off southwest Nova Scotia. J. Phys. Oceanogr. 13:1034-1054.

Swanson, R.L. and C.J. Sindermann (Eds). 1979. Oxygen depletion and associated benthic mortalities in New York Bight, 1976. NOAA Professional. Paper 11.

Taylor, C.C, H.B. Bigelow and H.W. Graham. 1957. Climate trends and the distribution of marine animals in New England. U.S. Fish and Wildlife Service Fishery Bulletin 37:293-345.

Townsend, D.W., N.R. Pettigrew and A.C. Thomas. 2001. Offshore blooms of the red tide dinoflagellate, Alexandrium sp., in the Gulf of Maine. Continental Shelf Research 21: 347-369.

Townsend, D.W. and M.Thomas. 2001. Springtime nutrient and phytoplankton dynamics on George Bank. Marine Ecology Progress Series, in press.

Viliela, I. 1995. Marine Ecological Processes. 2^nd Ed. Springer, New York.

Weeks, ?? (et al). 2001. Potential Consequences of Climate Change on Ecosystems and Fisheries of the Chesapeake Bay and Mid-Atlantic Coastal Waters. This symposium.

Wright, D.G., R.M. Hendry, J.W. Loder and F.W. Dobson. 1986. Oceanic changes associated with global increases in atmospheric carbon dioxide: a preliminary report for the Atlantic coast of Canada. Canadian Technical Report of Fisheries and Aquatic Sciences 1426.

FIGURES

Figure 1. Mid-Atlantic region. Depth contours are in meters.

Figure 2. Annual range in surface water temperature (°C) (derived from Mountain and Holzwarth 1989).

Figure 3. Interannual temperature anomaly (°C) for at the surface (open bars) and bottom (closed bars) in the mid-Atlantic region (figure 1), derived from data collected on NMFS spring trawl surveys (see Mountain and Holzwarth 1990)

Figure 4. Average temperature (°C) at which haddock were caught on spring NMFS trawl surveys (1968-1989) vs the average spring Georges Bank bottom temperature (modified from Mountain and Murawski 1992). The solid line represents the regression of the catch temperature onto the Georges Bank temperature and has a slope of 0.34. For comparison the dashed line has a slope of 1.

Figure 5. Average depth at which haddock were caught on spring NMFS trawl surveys (1968-1999) vs Georges Bank average bottom temperature (°C) (modified from Mountain Murawski 1992). The solid line represents the regression of the catch depth onto the Georges Bank temperature.

Figure 6. Average latitude at which mackerel on spring NMFS trawl surveys (1968-1999) were caught vs average MAB surface temperature (°C) (modified from Mountain and Murawski 1992). The line represents the regression of catch latitude onto the MAB temperature.

Figure 7. Seasonal development of stratification (density difference between 50 m and the surface), starting with water columns at 4.5 °C (solid line) and 6.5 °C (dashed line) on calendar day 70. The same surface forcing is used in both cases.