Processes controlling plankton abundance on Georges Bank:

Proc= esses controlling abundance of dominant copepod species on Georges Bank:

Local dynamics and large-scale forcing

PIs: Cabell Davis (WHOI), Robert Beardsley (WHOI), Changsheng Chen (UMassD),

Rubao Ji (WHOI), Edward Durbin (URI), David Townsend (UMaine),

Jeffrey Runge (UNH), Charles Flagg (SUNY), Richa= rd Limeburner (WHOI)

Proj= ect Summary

A fundamental goal of Biological Oceanography is to understand how underlying biological-physical interactions determine abundance of marine organisms. For animal populations, it is well= known that factors controlling survival during early life stages (i.e., recruitme= nt) are strong determinants of adult population size, but understanding these processes has been difficult due to model and data limitations. Recent advances in numerical modeling, together with new 3D data sets, provide a unique opportunity to s= tudy in detail biological-physical processes controlling zooplankton population size. We propose to use an ex= isting state-of-the-art biological/physical numerical model (FVCOM) together with = the recently-processed large 3D data set from the Georges Bank GLOBEC program to conduct idealized and realistic numerical experiments that explore the deta= iled mechanisms controlling seasonal evolution of spatial patterns in dominant zooplankton species on G= eorges Bank. We will examine a series of hypoth= eses that address how dominant copepod species populations are maintained on the bank, including local dynamics and large-scale forcing. Specifically we will determine whe= ther the observed characteristic sea= sonal-spatial pattern of each species (long-term and inter-annual) is predictable from the interaction between its characteris= tic life-history traits and physical transport. The extent to which the copepod populations are controlled by food-availability (bottom-up) or predation (top-down) processes will be examined, including the influence of Warm Slope Water versus Labrador Slope Water (NAO-dependent) on nutrient influx through the Northeast Channel and subsequent upwelling and biological enhancement on the bank. Self-sustainability= of each species population on the bank itself and in the Gulf of Maine will be studied by controlling immigration from specific source regions. Large-scale forcing including NAO and catastrophic global warming (e= .g. complete polar ice melt) will be examined explicitly by forcing the model at the boundaries, using scenarios based on basin-scale data and from concurre= nt basin-scale modeling efforts. Intellectual Merits: The proposed modeling study will pr= ovide new insights into the role of local and large-scale processes controlling zooplankton abundance in the ocean. The dominant copepod species to be studied include small species that are the dominant prey for larval cod and haddock in this region, thus provi= ding critical information for concurrent larval fish modeling studies. This detailed, process-oriented, regional-scale modeling with boundary forcing will lay the groundwork for integration with models of the entire ocean basin. The resulting model will be a lega= cy of the GLOBEC Georges Bank program by providing a powerful new tool for understanding how local and large-scale forcing interact to control plankton production in the sea. Broader Impacts: Results of t= he proposed work will be broadly disseminated to the general oceanographic community, the fishing industry, K-12 institutions, and to the population at large, through web-based servers using existing infrastructure at the proposers’ institutions. Web-based users will be able to ac= cess model results and run the model using chosen parameter settings to obtain predictions of currents, hydrography, and plankton abundance patterns given selected climate forcing scenarios. Collaboration with the WHOI/UMASS COSEE program will foster communication with K12 students and the public both nationally and internationally.

1. Back= ground

The GLOBEC approach — Und= erstanding complex marine ecosystems requires use of simplifying assumptions, which historically has involved trophodynamic analysis of energy or mass flow by = measurement and modeling (Lindeman, 1942; Teal, 1962; Odum, 1957; Steele, 1974). In regions of high diversity such = as the oligotrophic ocean this approach remains the only feasible method of analys= is (e.g., Sarmiento et al., 1993). As an alternative in low diversity regions, it is possible to model the population dynamics of a few dominant species to understand processes controlling their abundance and to obtain information about system level functioning (e.g., <= /span>Davis, 1987a). = The trophodynamic concept has been used to relate primary and second= ary production to fish production (e.g., Clarke, 1946; Cushing, 1975; Sissenwin= e et al., 1984; Cohen and Grosslein, 1987; Nixon, 1988), but this method is of limited utility in predicting abundance of particular fish species. It is well known that adult fish populations are dominated by a few strong year classes that are set by recruitment from early life stages (Hjort, 1914; Grosslein and Hennemuth, 1= 973). The GLOBEC program recognizes recruitment as the dominant factor controlling population abundance in mari= ne animal species (GLOBEC, 1991a). The = regional GLOBEC programs focus on target species and their dominant prey, with empha= sis on individual organisms, population dynamics, and interactions with the physical environment, especially as it relates to global climate change (GL= OBEC, 1991b, 1992, 1996, 1997, 2000).

Georges Bank GLOBEC: dominant zooplankton species — Georges Bank was chosen as the first GLOBEC study site due to its sensitivi= ty to climate change, definable populations, importance as a fishing ground, a= nd significant historical database (GLOBEC, 1992). The goal of this program is to understand the biological and physical processes controlling abundance of c= od and haddock larvae and their dominant prey species. In the program implementation plan= we emphasized the copepods Calanus finmarchicus and Pseudocalanus spp. as target species, since these are dominant prey items for larval cod and haddock. Subsequent studies h= ave found that small copepods (Pseudoca= lanus spp., Oit= hona similis, C= entropages spp., = Temora longicornis) are dominant prey ite= ms for cod and haddock larvae on Georges and Western Banks (Lough and Mountain, 19= 96; Lough et al, 1996; McLaren and Avendano. 1995; = McLaren et al., 1997). This contrasts= sharply with the situation in the eastern Atlantic (North Sea and Norwegian S= ea) where cod larvae feed almost solely on Calanus finmarchicus (Sundby, 1999), and decadal shifts in C. finmarchicus abundance affect larval cod survival (Beaugrand et al., 2003). This difference in larval fish prey preference appears due to prey availability since the larval fish diets ref= lect ambient copepod species composition, and the shallow banks have high concentrations of small copepod species (Davis, 1987; Sherman et al., 1987;= Meise, and O’Reilly, 1996., McLaren et al. 2001; Reiss et al, 2003). In the shallow <= span style=3D'font-size:11.0pt'>Baltic Sea, cod larvae also feed on small copepod species (Hinrichsen et al, 2002).<= span style=3D'mso-spacerun:yes'> In general, small copepods are an important component of marine ecosystems (e.g., Davis, 1987b; Turner, 2004)= . The present proposal focuses on understanding the biological-physical processes controlling the abundance of dominant copepod species in the Georges Bank/<= /span> Gulf= of Maine region, including Calanus finmarchicus and the small= er species, Pseudocalanus spp. (moulton= i and newmani), Oithona similis, T= emora longicornis, Centropages spp. (typicus and hamatus).

1.1.=    The <= /span>Georges Bank Physical Environment =

Local dynamics — Geo= rges Bank (GB), the Gulf of Maine (GOM), and Scotian Shelf (SS) are part of a si= ngle regional coastal current system, driven in large part by upstream mass and buoyancy forcing (Fig. 1). The bank itself is a quasi-flow-through system, = with water from offshore and upstream sources arriving= on the bank, being modified locally by surface forcing and tidal mixing before moving off-bank to the Mid-Atlantic Bight or re-circulated northward along = the Great South Channel. <= /span>

The strongest currents o= ver the bank are of tidal origin, and turbulent mixing associated with the tidal bottom boundary layer is most intense over the shallow cap of the bank, effectively eliminating local vertical temperature and salinity stratificat= ion throughout the year. As seaso= nal stratification increases on the flanks of the bank, the tidal mixing front forms with associated secondary flow. The clockwise around-bank residual fl= ow increases with seasonal stratification, becoming partially closed from June through

Figure 1. Circulation:= 1 flow across GSC into the north flank jet, 2<= /b> tidal-pumping of deep water onto GB, 3 wind-driven near-surface flow, 4<= /b> small-scale cross-frontal proces= ses, 5<= /b> SS cross-over

Figure 2. Schematic of the western North Atlantic shelf-break current system in summer,

= (Fratantoni and Pickart,= 2005)

September until fall sto= rms and surface cooling destroy the local stratification. Surface heating drives the develop= ment of the seasonal thermocline. = Salinity on the bank is controlled by advective and mixing processes along the north= ern and southern flanks. On the southern flank salinity is influenced by on-bank intrusions of saline shelf= -break frontal water and very saline warm-core ring water (Fig. 1). Along the northern flank salinity = is controlled by advection from the western Gulf across the northern Great Sou= th Channel (Fig. 1, 1), tidally-driven near-bottom residual flow (the “ti= dal pump”, Fig. 1, 2), wind-driven near-surface flow (Fig. 1, 3), small-scale cross-frontal processes (Fig. 1, <= span style=3D'color:red'>4), and intermittent cross-over of low salin= ity SS surface water (Fig. 1, 5). The tidal pump in particular provides a strong mechanism for bringing deep water from Georges Basin up onto th= e bank (Chen and Beardsley, 1998; Chen et al., 2003).

Figure 3. Schematic showing the strong westward penetration of LSW and northward position of the Gulf Stream following low winter North Atlantic Oscillati= on (NAO) index (left) and weak penetration of LSW and more southward positio= n of Gulf Stream following high winter NAO index. (Drinkwater et al, 2003)

Large-scale forcing — Water enters the GOM via two primary paths: (1) the flow of relati= vely fresh water above 100m from the SS and (2) warmer, more saline Slope Water (SW) at depths greater than 100m through the Northeast Channel (NEC). The primary source of SS water is the West Greenland/Labrador Current system, with additional input from the St. Lawrence system (Fig. 2). As the Labrador Cur= rent flows around the Grand Banks, the large shoaling in shelf-break depth from ~300m in the north to ~100m to the southwest helps force the deeper Labrador Current water to flow along the upper slope, thus forming Labrador Slope Wa= ter (LSW), which flows west into the Laurentian Cha= nnel and along the Scotian upper slope. The westward extent of LSW depends on its source strength, thought to depend on basin-scale forcing (NAO) (Fig 3), and degree of mixing with ambient Warm Slope Water (WSW) of <= span lang=3DEN-GB style=3D'font-size:11.0pt;mso-ansi-language:EN-GB'>Gulf Strea= m origin. Si= nce WSW is warmer, more saline, and nutrient rich than LSW, the relative mix of the= se two end members entering the NEC, and its transport relative to the inflow = of SS water, strongly influences circulation, water property distributions, and n= utrient content in the GOM/GB system.

Fig. 4. GB salinity anomaly (Mountain, 2005)

Fig. 5. Percent LSW in bottom water (150= -

200 m) in 1998. (Drinkwater et al, 2003)<= span lang=3DDE style=3D'font-size:11.0pt;mso-ansi-language:DE'>

Data from the 1995-1999 = GB GLOBEC field study provide an excellent example of the flow-through nature = of the GB/GOM system and its linkage to larger basin-scale forcing. The salini= ty on GB exhibited two significant freshening events between early 1996 to ear= ly 1997 and between late 1997 through 1998, with a net drop of ~1 psu (Fig. 4). These two events also were observed in surface (0-30m) GOM waters, suggesting significant increases in freshwater influx from the SS (Smith et al., 2001). In the NEC, WSW was replaced by coo= ler, fresher LSW in January 1998 as the leading edge of LSW flow extended west d= ue to an increase in the Labrador Current associated with a low NAO. As LSW entered the GOM during early 1998, it mixed with resident GOM water (Fig 5). By early 1999, WSW was ag= ain flowing into the GOM through the NEC. Since the tidal pump mechanism can carry deep water up on the northe= rn flank of GB, advection of LSW in Georges Basin onto the bank can occur on relative short time scales (>=3D1 month), suggesting that part of the freshening = on GB observed during 1998 was due = to the influx of LSW into the GOM. Clearly, variations in the primary upstream sources (the SS Water, t= he mix of WSW versus LSW) linked to basin-scale forcing strongly control the w= ater properties (including heat, salt and nutrients) through the GOM and onto GB= .

Fig. 5a. Scenario of nutrient input onto GB from the NEC (Townsend et al., 2004).

1.2.=    Conne= ction between NAO and plankton productivity on GB

NAO-dependent intrusions of LSW and WSW through the NEC greatly influence the nutrient (N= , Si) input into the GOM. = It is believed that the tidal pumping mechanism along the northern edge of the ba= nk can quickly transport nutrients from Georges Basin up onto the NE peak of GB, impacting the productivity of the bank very ra= pidly (Fig. 5a). The ratio of N and= Si in the inflow water can influence the duration of = the diatom bloom. In general, the= NEC inflow can affect the total salt and nutrient budget of the GB/GOM, determining the thermohaline circulation and influencing upwelling along the western GOM and subsequent nutrient transport to western GB. The potential impact of this inflo= w will be modeled in the proposed study.

1.3.   Characteristics of GB zooplankton

GB zooplankton is dominated by Calanus finmarchicus, Pseudocalanus spp., Oithona similis, Centropages spp., and Temora longicornis, and Paracalanus parvus (Bigelow, 1926; Davis, 1984, 1987b; Sherman et al, 1987; Durbin = et al., 2003; Durbin and Casas, submitted<= /span>). Each species exhibits a characteristic life cycle and seasonal/spatial pattern in the GB/GOM region. Calanus finmarchicus and Pseudocalanus<= /i> spp. are cold-water species that avoid the warm surfa= ce layer (>10-12^oC) during summer and fall and produce large spr= ing populations. Centropages spp, Temora, and Paracalanus are warm water spec= ies and are most abundant during late summer and fall. Oithona is plentiful throughout the GB/GOM region year round.

Fig. 6. Calanus finmarchicus abundance (log₁₀(#= /10m²) in the GB/GOM, mean 1977-1987, (MARMAP data redrawn from Meise and OReilly, 1998).

Calanus finmarchicus— This species spends the warmer months in a state of diapause as stage CV in cool= er waters (5-7^oC) at depth in the GOM (Fig. 6). During late Decemb= er, it emerges from diapause (mechanism unknown), swims to the surface and molt= s to adult. Subsequent egg product= ion depends on availability of phytoplankton (Durbin et al., 2003), with the first generation born in late December-early January (Durbin et al., 1997). Generation time is ~2 months at th= e cold winter/spring temperatures (~5^oC), so G1 adults appear in March, and there is time for a total of 3 generations by the end of its growing se= ason in July. Overwintering females produce eggs = for a prolonged period, smearing out cohorts (Durbin and Casas, submitted).   This annual cycle in the GB/GOM appears stable, having persisted for many decades (Bigelow, 1926; Clarke et al, 1946; Meise-Munns et al., 1991), but the extent to which th= is population is self-sustaining is unknown.

Fig. 7. Calanus finmarchicus abundance i= n the N Atlantic from CPR data (Spears 2005) and nets/OPC (Hea= th et al. 2004). Note high densit= ies in the Labrador and Norwegian Seas.

Calan= us finmarchicus is an open ocean species, occurring throughout = the northern North Atlantic<= /span> from the eastern US to the Barents Sea, with centers of population abundance in the Norwegian and = Labrador seas (Fig. 7). Immig= ration into the GB/GOM population from upstream sources may contribute to the appa= rent stability of the population. = C. finmarchicus enters the GOM fro= m the SS and possibly from the SW through the NEC.&= nbsp; C. finmarchicus in the SW originate from the Labra= dor Sea and to an unknown = extent from spill-over of the productive shelf populations. Scotian Shelf C. finmarchicus originate from Labrador Sea via the SW (Head et al., 1999) and to some degr= ee from the productive Gulf of S= t. Lawrence population (Zakardjian et al. 2003). It is unlikely tha= t immigration directly determines the large spring abundance peak in C. finmarchicus in the GB/GOM, since the water mass turnover ti= me is long relative to the generation time, but a seeding type of immigration during the “off-season” could be important in determining the s= ize of the startup population of CVs in December (<= span style=3D'font-size:11.0pt;mso-bidi-font-size:12.0pt'>Saumweber and Durbin, submitted= ). = Since the diapausing SW population resides at a depth of 500m (below the NEC sill depth), and high concentrations are not observed in the upper water column until April (Miller et al., 1991), well-after initiation of the GB/GOM population growth, it is unlikely that a SW population entering through the= NEC contributes significantly to the GB/GOM C. finmarchicus population.

It may well be that the resident diapausing GOM population is sufficient to produce the large spring GOM/GB population. At the end of the growing season, downward migrating diapausing CVs cannot reach their normal open ocean dept= hs of 500-2000m, and they become trapped in the GOM basins. A similar effect has been observed= on the SS basins (Sameoto and Herman, 1990) and fo= r C. pacificus in the Santa B= arbara basin (Osgood and Checkley= , 1997). Lagrangian modeling st= udies have found retention of diapausing CVs in the GOM to be high, especially if= the animals stay below 150 m (Johnson et al., submit= ted). Thus it appears possible that the = GB/GOM population could be self-sustaining, but further 3D modeling work is needed, using a concentration-based approach to quantify population dynamics and transport continuously throughout this region during the year. This self-sustainability is in sha= rp contrast with the situation in the North Sea, whi= ch has no deep basins and is dependent on annual input from offshore waters (e.g. Heath et al., 1999).

Fig. 8. GLOBEC monthly mean Calanus abundance

       &nbs= p;   on GB, showing abundance “hole”

Fig. 9. Pseudocalanus abundance in the GB/GOM. MARMAP bi-monthly means 1977-1987 (from McGillicuddy et al., 1998).

Although C. finmarchicus may be able to sustain itself in the GB/GOM system, it does not maintain itself on GB, sin= ce the bulk of the population disappears from the bank during the off-season (= Fig. 6) and even during the growing season its abundance appears to be driven by= GOM concentrations, with a well-defined “hole” in abundance in the center of the bank (Fig. 8, Durbin and Casas, submitted). This hole results= from a combination of advection around the bank of the large GOM population and possibly high predation over the crest.&nb= sp; A huge literature exists for C. finmarchicus egg production, development and growth as a function of fo= od and temperature including several studies done as part of the GLOBEC GB pro= cess work (Campbell and Head, 2000; Campbell et al., 2001a,b; Runge et al. submi= tted). Such data can be incorporated into= the population model for this species and used together with the large field da= ta base to conduct targeted forward modeling to examine biological/physical processes controlling observed patterns.

Pseudocalanus spp.— Like Calanus, the growth season for Pseudocalanus is winter/spring (Fi= g. 9, 10). Its abundance is higher = in shallower areas (<100m) and is highest in the crest region of the bank in June. Pseudocalanus do= es not overwinter in the GOM as does Calanus an= d is not normally present in the central Gulf during winter. This genus is an egg carrier and consequently has lower egg production and egg mortality rates (Corkett and McLaren, 1978; Ohman= et al., 2002). Pseudocalanus in the GB/GOM

Fig. 10. GLOBEC data for Pseudocalanus

= Fig. 11. A) Temora, B) Centropages hamatus C) Centropages typicu= s, D) Oithona similis. Mean monthly GLOBEC data for 1995-1999, A,B,D May; C January; log₁₀= (#/m³+1)

region comprises two species: P. newmani and P. moultoni (Frost, 1989; McLaren= et al., 1989a; Bucklin et al., 1998, 2001; McGillicuddy and Bucklin, 2002).   P. moultoni appears to be a colder= water species and more abundant during winter/spring, while P. newmani is more plentiful during spring/summer (McLaren et a= l., 1989a,b). = P. moultoni is a coastal species a= nd P. newmani an offshore one (Frost, 1989). Thus P. moultoni may be carried onto the bank from western GOM coast= al waters (e.g., Cape Cod Bay), while P. newmani may have a source from SS water, either crossing over the NEC, = or indirectly through the central GOM (McGillicuddy and Bucklin, 2002). Inverse model= ing, representing the biology by a single source/sink term, revealed that the two species are likely to have different source locations on the bank but intermingle by June (McGillicuddy and Bucklin, = 2002). The use of a simple cost function instead of a population model, however, m= eant that the independent species-specific life history information could not be included in the model. Substa= ntial information on development, growth and egg production as a function of food= and temperature is available for Pseudo= calanus (e.g. Corkett and McLaren, 1978; Vidal, 1980; D= avis, 1983, 1984a,b; McLaren et al., 1989b; Ban et al., 2000; Lee et al., 2003; Dzierzbicka-Glowacka, 2004). Use of these laboratory and shipbo= ard derived rates together with field data on distributional patterns of the combined species (plus a subset of species-specific data, Bucklin et al., 2001) can = be used in numerical models to explore the biological/physical processes contr= olling seasonal development of the spatial distribution patterns. Targeted forward modeling experime= nts are required to determine the potential source locations and interactions controlling subsequent spatial patterns of each population, given their distinctive temperature-food dependent growth and reproductive rates.<= /o:p>

Other Dominant Copepod Species &#= 8212; Each dominant copepod species on GB has its own characteristic temporal-spatial patterns and life histories (e.g., <= st1:City>Davis, 1987a). Copepods that lay bottom resting eggs, including Centropages hamatus and Temora longicornis, have well defined populations on the crest of the bank (Fig. 11A, B) (Davis, 1987; Sherman et al., 1987). It has been sugge= sted that these species use resting eggs as a strategy for “gluing” their populations to regions that are favorable for growth (Davis, 1987; Li= ndley and Hunt, 1989; Lindley, 1990; Marcus, 1996; Marcus and Lutz, 1998), but the necessary biological/physical dynamics have yet to be verified via modeling= . Centropages typicus abundance is highest during late summer and fall and decreases markedly dur= ing winter and spring. It is most abundant in the warm surface layer above the thermocline and not restricted= to the crest of GB like C. hamatus. Oithona similis is abundant year round and is not restricted to GB but has a pa= ttern similar to C. finmarchicus, wit= h a large off-bank population.   A substantial literature on life history information is available for these species, including egg production, development and growth rates as functions of food= and temperature (e.g., Davis and Alatalo, 1992; Klein Breteler and Gonzalez, 1986; Klein Breteler et al., 1982, 1996; Sabatini and Kiørboe, 1994; Maps et al. 2005).

Fig. 12. Abundance trends in GB copepod, GLOBEC 1995-1999.

Abundance trends during the GLOBEC years — Abundance of the dominant copepod species, except Calanus, increased significantly during the five-year GLOBEC GB field program (Fig. 12.). Whi= le this trend may reflect a change toward smaller species, indicative of a war= ming trend, no concomitant increase in temperature was observed. These changes were negatively correlated with salinity (Durbin and Casas<= /span>, submitted.). High abundances in 1999 appear to be related to a phytoplankton bloom that took place during winter in the central GOM, leading to high reproductive rates and abundances (Durbin et al 2003).= The elevated abundances may have been advected onto GB. It is possible that low surface sa= linity during winter increased stability of the water column and led to the bloom.= A negative correlation between salinity and chlorophyll on GB also was found = at this time (Durbin and Casas, submitted), the re= ason for which is unknown. The pro= posed modeling work will examine the potential causes of the zooplankton increase and its relationship to large-scale forcing and= climate change.

1.4.=    Avail= able Data Sets

GLOBEC GB — The GLOBEC GB field program was conducted from 1995-1999 and includ= ed a combination of monthly broadscale and process-oriented cruises (GLOBEC, 199= 2; Wiebe et al, 2002).   Broadscale cruises provided 3D maps of the plankton species based on= net tows (1-m² MOCNESS, Wiebe et al., 1985) at a set of 41 standard stations covering the bank and adjacent waters, and plankton pump sampling (Durbin et al 1997, 2000) at ~18 of the standard stations. CTD bottle casts= were made at all stations. MOCNESS= (0.15 mm mesh nets) zooplankton samples were collected from 0-15, 15-40 m, and 40= -100 m or the bottom (if shallower than 100 m) and 100 m to the bottom or 450 m = (if >100 m). Pump (0.035-mm ne= ts) samples were collected over the same depth ranges as the MOCNESS in the upp= er water column but the maximum depth it was deployed to was 70-100 m dependin= g on wind and tides, or to the bottom on shallower parts of the bank. For complete description of collec= tion and processing methods see (Durbin et al. 1997; 2000). Samples were sorted, and processed= data now are available for all life stages of Calanus finmarchicus and Pseudocalanus spp. (N1-N6, C1-Adult). For other copepod species life stages were sorted as adult males and females, copepodids, and nauplii. These data are stored in an Oracle database together with all of the CTD and chlorophyll data (http://globec.gso.uri.edu). The complete GLOBEC GB data set for broadscale zooplankton includes >3500 net samples (41 stations, 3 depths= , 30 cruises), plus >1500 pump samples (18 stations, 3 depths, 30 cruises). In addition, vertically stratified= data from several other cruises have been collected in deeper regions of the GOM= .

Other Data Sets — Data from the bimonthly cruises of= the MARMAP program 1977-87 and its follow-on program, ECOMON (1992-present), include shelf-wide distributions of hydrography and zooplankton (0.333mm bongos). The complete data se= t is available to us via an NMFS Oracle database (D. Mountain, pers. comm.) Although the zooplankt= on data are from integrated hauls, these data cover a broader area than the GL= OBEC GB data, and, together with GLOBEC vertical data from deeper GOM, will allo= w us to approximate 3D spatial patterns of each species throughout the GOM region. Further data from the= GOM CPR transect, transatlantic CPR, UNH Coastal Observing Center, the 1939-41 GB study, and Canadian AZMP (see H= ead letter), allows the 1995-1999 GLOBEC data to be viewed in a larger context.=

1.5.   Previous modeling

Several previous biological/physical models of the GB/GOM region have been developed. <= st1:place>Davis (1984) showed that the interaction between arou= nd-bank advection/diffusion and temperature-dependent development of the copepod Pseudocalanus could explain observ= ed spatial patterns in population structure.&= nbsp; Subsequent studies have shown that this scenario is most likely for = P. moultoni, with an admixture of = P. newmani later in the season (McGillicuddy and Bucklin, 2002). McGillicuddy<= /span> et al. (1998) had used the same inverse approach to locate source/sink areas for the Pseudocalanus population.   Klein (1987) modeled the 2D NPZD-advection-diffusion on the bank and found that phytoplankton production was removed from the bank via detritus rather than grazing. Lewis et al., (1994,= 1991) used 3D modeling to show that strong wind forcing can cause a partial washo= ut of plankton from the bank, increasing the phytoplankton/zooplankton ratio (= NPZ model), and causing spatially-temporally sensitive larval populations (particle-based) to be lost from the bank, affecting recruitment. The Calanus finmarchicus population in the GB/GOM has been modeled using the Dartmouth model to show that the deep basins and SS can c= ontribute to the bank population; these models used 2D transport of concentration-bas= ed (Lynch et al. 1998), individual-based (Miller et al., 1998), and particle trajectories (Hannah et al 1998). The C. finmarchicus population thro= ughout the Gulf of St. Lawrence/SS/G= OM region was modeled to examine local dynamics and exchange between regions within the model interior (Zakardjian et= al., 2003). The latter model was a= 3D concentration-based 13-stage population model with temperature-dependent development and was run for a full-annual cycle. The results indicated that horizon= tal transport was dominant on the SS, but that the population is self-sustainin= g in the Gulf of St. Lawrence. = Food-dependence effects on population dynamics during the active season were not examined, however.

1.6.   Combining existing models and data

Over the past decade, the GLOBEC GB program has acquired and processed an exceptional 5-year 3D data set of plankton and physical variables, while at= the same time developing a high-resolution state-of-the-art prognostic 3D biolo= gical/physical model of this region (FVCOM). Data processing and model development recently have reached the point where they= can be effectively combined. By t= he start of the proposed study, the FVCOM model will have been ported to a massively-parallel supercomputer. We now have the unique opportunity to use= the data and model together to study the detailed mechanisms controlling zoopla= nkton abundance patterns on GB, explicitly including boundary forcing determined = by basin-scale dynamics.

We have assembled a team of PIs who are leading experts on the plankton and physical dynamics of this area. Beardsley and Flagg are physical oceanographers with a long backgrou= nd in this region. Chen is a physical oceanographer and developer of the FVCOM model. Durbin is the scientist in charge of the broadscale data acquisition= and processing. Runge is an exper= t in copepod biology specializing in Cal= anus fertility and population dynamics. Townsend is the lead scientist studying nutrient-plankton production= in the GB/GOM region. Davis is a zooplankton biologist who has done field, laboratory, and biological-physical modeling studies of copepod population dynamics in the GB/GOM region. Davis and Beardsley helped formulate the theoretical underpinnings a= nd approach for the GLOBEC GB program (GLOBEC, 1992).

2.      Prop= osed Research

2.1.   Hypotheses

Working Hypothesis – The seasonal evolution of characteristic mean spatial abunda= nce patterns of each dominant copepod species on GB is predictable from the interaction between its characteris= tic life-history traits and physical transport. These life-history traits include = egg production, development, and growth rates (temperature/food dependent) as w= ell as other traits such as vertical migration and diapause. Both the long-term, multi-year, me= an and year-to-year variations in seasonal-spatial patterns are predictable by the= se interactions. Within this wor= king hypothesis, we will address three specific null hypotheses:

H1₀: Th= e abundance of copepod species on the bank is controlled by food availability (bottom-up control). Here we will examin= e the scenario that GB productivity, and thus food availability for the dominant copepod species, is controlled by nutrient input into the GOM through the N= EC, which is determined by the intrusions of Labrador Slope Water versus Warm Slope Water. Alternative hypotheses include: 1) predatory control= of copepod seasonal cycles (top-down control), 2) a combination of food-limita= tion and predation (time-space dependent), and 3) purely physical control by dir= ect effects of temperature on vital rates or advection. In 3), we will examine causes of t= he observed increase in warm-water copepod species during the GLOBEC years (1995-1999).=

H2₀:   Copepod populations on GB and/or the GOM region are not self-sustain= ing. We will examine the need for immig= ration from different sources to maintain the copepod populations over multiple ye= ars. Key source regions will be ex= amined including the SS and SW. For self-sustainability on GB itself, we will examine potential sources from the coastal regions (e.g., Cape Cod = Bay, Penobscot Bay), the GOM, and from bottom resting eggs.

H3₀:   Catastrophic global warming (e.g., total polar ice melt), parameteri= zed as a lack of Labrador Sea water at the NEC, causes a regime shift on GB fro= m cold-water copepod species to warm-water ones.

2.2.   Objectives

The overall goal of the proposed study is to understand the underlying biological-physical mechanisms controlling the seasonal development of spat= ial patterns of dominant copepod species on GB. Our specific objectives are: 1) to examine how local-dynamics and external forcing control the abundance of th= ese species on GB, 2) to determine the degree to which top-down versus bottom up proces= ses control the dominant copepod species on GB, and 3) to use existing state-of-the-art 3D physical/biological numerical models together with exis= ting high-quality 3D data sets from the GLOBEC GB field program (and other historical data sets), to conduct targeted numerical experiments that explo= re the likelihood of the hypotheses listed above.

2.3.   Methods

2.3.1.      The Integrated Model System

The UMASSD-WHOI re= search team has developed an integrated model system for the GOM/GB region (Fig. 1= 3). The major components of this system include: (1) the modified fifth-generat= ion community mesoscale atmospheric model (MM5), (2) the unstructured grid Finite-Volume Coastal Ocean circulation Model (FVCOM), (3) a generalized lower trophic level food web model, and (4) multi-stage zooplankton models. The 4-D optimal interpolation (OI) and nudg= ing data assimilation methods are used with FVCOM to incorporate satellite-deri= ved sea-surface height (altimeter), insolation, SST= , and remote sensing reflectance (RSR) as well as in-situ oceanographic data (moo= red and shipboard hydrographic and current data) in model simulations conducted with realistic forcing and boundary conditions and observed ocean response = for specific periods of time. The integrated model system can be run with ideal= ized forcing and boundary conditions to investigate specific processes. A brief description of the integra= ted model system is given below.

MM5R= 12; The current version of the meteorological model utilizes the fifth-generation mesoscale regional weather model (MM5) develo= ped by NCAR/Penn State (Dudhia et al., 2003; Grell et al., 1994) for community use. MM5 uses NCAR/= NCEP or ETA weather model fields as initial and boundary conditions with two-way nesting capability, and can provide continuous hindcasts and three-day forecasts. We have used MM5 to construct the surface weather hindcast and forecast system for fishery studies in the GOM/GB (Chen et al., 2004). This model (called GOM-MM5) is configured with a regional domain (covering the <= /span>Northeast U.S.) and a local domain (covering the SS, GOM, and = New England Shelf) with horizontal grid spacing of 30 and 10 km, respectively, = and 31 sigma levels in the vertical with finer resolution in the PBL. To improve the model-based surface wind field, wind stress and heat flux estimates over the ocean, GOM-MM5 uses the COARE 2.6 bulk algorithm (= Fairall et al, 2003) for the air-sea fluxes, satellite-based i= nsolation, cloud cover, and SST data (International Satellite Cloud Climatology Projec= t) for the radiative fluxes, and all coastal NDBC = and C-MAN surface weather data available in the local domain are incorporated through 4D data assimilation.

GOM-MM5 is presently in operational use, with 3-day forecasts of the surface conditions, (including wind stress, heat flux, P-E) over the GOM/GB region posted on the SMAST website (http= ://www.smast.umassd.edu/research_projects/GB/mm5/mm5_eta/) for research, education, and public use. By this summer, we will complete the hindcast of the surface forcing fields with mesoscale (10-km) resolution co= vering the FVCOM domain for the 1995-1999 GLOBEC field period<= /span>. This is the first calibrated mesoscale meteorological database built in the GLOBEC GB Phase 4 program.

FVCOM— FVCOM is a prognostic, unstructured grid, finite-volume, free-surface, 3D primiti= ve equation coastal ocean circulation model (Chen et al., 2003; Chen et al. 2004a). In common with other coastal models, FVCOM uses the modified Mellor= and Yamada level 2.5 (MY-2.5) and Smagorinsky turbu= lent closure schemes for vertical and horizontal mixing, respectively (Mellor and Yamada, 1982; Galperin et al., 1988; Smagorinsky, 1963), and a sigma coordinate to follow = bottom topography. The General Ocean Turbulent Model (GOTM) developed by Burchard’s research group in Germany (Burchard, 2002) ha= s been added to FVCOM to provide optional vertical turbulent closure schemes. Unli= ke existing coastal finite-difference and finite-element models, FVCOM is solv= ed numerically by flux calculation in the integral form of the governing equat= ions over an unstructured triangular grid. This approach combines the best featu= res of finite-element methods (grid flexibility) and finite-difference methods (numerical efficiency and code simplicity) and provides a better numerical representation of momentum, mass, salt, heat, and tracer conservation. The ability of FVCOM to accurately= solve scalar conservation equations in addition to the geometrical flexibility provided by unstructured meshes (Fig. 14) and the simplicity of the coding structure makes FVCOM ideally suited for interdisciplinary applications (Ch= en et al., 2005b-c; Ji 2003; Ji et al., 2005). Examples can be viewed on Chen’s research group website at http://codfish.smast.umassd.edu).

FVCOM has been validated through direct comparison with analytical solutions for idealized cases (Chen et al., 2005b; Huang et al., 2005a-c) and other models for application in estuaries (Chen et al., 2005d-e; Huang et al., 2005d), inter-bays (Zhao et al., 2005) and the GOM/GB region (Chen et al, 2003b; Ch= en et al., 2005c). These studies sh= ow that different physical processes controlling currents and stratification in the coastal ocean have inherent time and space scales that must be carefully considered when determining model grid resolution for accurate simulation. = This is particularly important with freshwater discharge, buoyancy-driven coastal plumes and currents, tidally-forced flows and upwelling, and fronts in the = GOM/GB region (Chen et al 2005a). For example, model dye experiments made to simulate Houghton’s May/June 1= 999 dye dispersion observations on GB suggest that the horizontal resolution ne= eded to resolve the diffusive flux is 500 m (Chen et al., 2005c). This resolution is also required to= have a convergence solution of the tidal-induced residual current and buoyancy-induced current at the shelfbreak on GB (Chen et al., 2005b). These requirements will be used to refine the final FVCOM grid used in the propos= ed work. FVCOM has been used to hindca= st currents and hydrography in the GOM/GB region for 1995 and 1999 using GOM-MM5 surface forcing, 4D data assimilation of 5-day averaged satellite SST and available moored current data, and open boundary conditions (Chen et al., 2003b). The model tidal currents compare v= ery well with available surface elevation and current data, with overall uncertainties for the dominant M2 component of less than 3 cm in amplitude, 5° in phase, and 3 cm/s in the tidal current major axis (Chen et al., 2005c). The model subtidal currents and stratification also compare wel= l with existing in-situ measurements, capturing the seasonal cycle in vertical stratification and increased around-bank circulation during June-September. These two hin= dcasts clearly illustrate significant short-term (daily to weekly) and long-term (seasonal and interannual) variability in the <= span class=3DSpellE>subtidal currents on GB. For example, surface winds in Marc= h 1999 were stronger and more variable than in 1995, resulting in stronger monthly-mean offbank near-surface flow in 1999 = than in 1995 (Fig 15). =

FVCOM Biological Module—= Various ecosystem models have been implemented in FVCO= M, including NPZ, NPZD, NPZDB, and water quality models. To make FVCOM more flexible for ecosystem studies, we have built a generalized biological modu= le into FVCOM to allow users to select either a pre-built biological model (su= ch as NPZ, NPZD, etc) or construct their own biological model using the pre-defined pool of biological variables and parameterization functions, including zooplankton life-stage models.&n= bsp; This module acts like a platform that allows us to examine the relat= ive importance of different physical and biological processes under well-calibr= ated physical fields.

Upstream Boundary Conditions<= span style=3D'font-size:11.0pt'>—   For this study, we plan to mo= ve the “upstream” boundary of the GOM/GB FVCOM domain eastward to cut across the SS and upper slope through Banquereau Bank. This choice simplifies the cross-shelf bathymetry and separation of along-s= helf flow into inner-shelf and shelfbreak components= (Han et al, 1997), and was used in the Hannah et al (2001) model simulations of = the seasonal circulation on the western and central SS. They used the Dartmouth circulation model (QUODDY4) to produce dynamically-consistent seasonal-mean solutions based on historical hydrographic, wind stress, and M2 tidal elevation data. The model solutions show generally= good agreement with moored current data and capture the strong seasonality of the major currents, indicating that the seasonal-mean baro= tropic and baroclinic pressure gradients specified alo= ng the upstream boundary were realistic. We plan to use the Hannah et al (2001) seasonal-mean boundary conditions, a statistical method to compute the along-shelf wind-driven time-dependent transport over the inner-shelf (Schwing= , 1992), recent hydrographic and moored data, and the new shelfbreak current system climatology (Fratantoni and Pickart, 2005a,b) to construct the best set of upstre= am boundary conditions for both process and long-term hindcast simulations. In addition to the extensive hydrographic data set collected in GLOBEC GB, additional data have been collected in the GB/GOM by the ongoing NMFS ECOMON program and on the SS by= the Canadian Atlantic Zone Monitoring Program (AZMP) (see letter from Dr. Erica Head). Starting in 1997, a comprehensive suite of hydrographic, nutrient, and biological data is being collected 3-4 times a year along transects through Browns Bank, off Halifax, through Banquereau Bank (the “Louisbourg= Line”), across Cabot Str= ait, and others further upstream. = The overlap of AZMP with GLOBEC GB field effort begins in 1997, the collocation= of our upstream boundary with the AZMP Louisbourg= Line will allow us to compute a climatology of physic= al and biological properties (1997-2005) which will help determine realistic bound= ary conditions and their interannual variability. In formulating boundary conditions= we will collaborate with Haidvogel (Rutgers) who is modeling basin-scale dynamics and with McGillicuddy et al. who have proposed a concurrent basin-scale adjoint modeling study of= Calanus.[1]

We will not include the potential influence of warm core rings (eg., Flierl and Wroblewski, 1985) and other eddy features originating in the Gulf Stream (Fig. 1= ), because the larger regional and basin-scale models do not yet produce these features accurately enough in this region for us to use them to construct t= he boundary conditions along the open ocean part of FVCOM. The influence of rings may be minor (e.g., Churchill et al., 2003), h= owever, and we will be able to infer their potential importance by their omission.<= span style=3D'mso-spacerun:yes'> As the larger-scale models mature = in this respect, follow-on studies of these processes can be developed.=

FVCOM Computational Aspects— Wit= h GLOBEC GB Phase 4 support, FVCOM has been converted into a FORTRAN 95/2K paralleli= zed program to take advantage of multi-processor computing (Cowles et al., 2003). This implementation us= es a SPMD (Single Program Multiple Data) approach with a message-passing model to perform the necessary inter-processor communication and synchronization. The physical domain is decomposed into sub-domains using the METIS graph partitioning libraries. Each sub-domain is assigned to a processor for integration of the model equation= s. The exchange subroutines utilize non-blocking sends and receive from the MPI (Message Passing Interface) 2.0 library. The efficiency of the code can be measured in terms of its speedup and/or scalability on a multiprocessor computer. Chen’s modeling lab will install a new high-performance 256-processor super-cluster computer this summer. With this computer, 1-yr model run with data assimilation and the existing GOM/GB FVCOM grid should = take less than 1-day clocktime. We plan to use this computer with attached mass storage for the proposed Phase 4B model experim= ents (1995-1999 hindcasts and process studies), model/data comparisons, model re= sult analysis and visualization, and archiving model output and results.

2.3.2= .      =       Biolo= gical Models

We will incorporate a copepod population model and a simple food web model (NP= Z) into the plug-in modules of FVCOM. We will draw from our previous modeling work involving Calanus, Pseudocalanus<= /i>, and other copepod species (Davis, 1984a,c, 1987;= Lynch et al., 1998; McGillicuddy et al., 1998; Zakardjian et al., 2003), as well as our food= web modeling (Davis, 1987b; Flierl and Davis, 1993; Lewis et al., 1994; Davis a= nd Steele, 1994; Zeldis et al., 1995; Dadou et al. 1996; Ji, 2003, submitted a,b,c).

Population model = 212; Copepod dynamics will be modeled using a stage-structured population model containing 15 life stages (egg, N1-6, C1-3, C4M-F, C5M-F, Adult) plus additional diapause stages (e.g. eggs or CV) as needed. A standard formulation for the stage-based model will be used (e.g., Zakardjian et al., 2003). A potential problem with using only= 15 life stages is artificial diffusion of individuals through the life stages, and, to prevent this, age-within-stage models have been developed (<= st1:City>Davis 1984c). If necessary we will use a 150-age-stage class model (e.g., <= st1:City>Davis 1984c) which will still run efficiently given t= he supercomputing power available. Egg production, molting, and growth rate parameters as functions of temperature= and food will be taken from the vital rate measurements made during the GLOBEC process cruises and from the large body of available literature (see citati= ons in Background section). The model will be initialized and validated using observed patterns of copepod abundance according to the numerical experimen= ts described below. The copepod population then will grow based on ambient temperature and food levels. Mortality rates will be inferred by adjusting them to match model and field abundances (Hall, 1964; Davis, 1984a). Mortality is stage-sp= ecific and typically higher for younger stages (e.g., = Davis, 1984a). = Mortality rates as a function of life-stage have been found for Calanus and Pseudocalan= us as part of the GLOBEC GB studies (Ohman et al., 2002). We will use relative stage-dependent mortality curves, from this and other studies, when inferri= ng mortality rates. We also will explore the possibility of conducting targeted inverse modeling to obtain t= he stage-dependent mortality rates (e.g. Li et al, in prep). Vertical migration behavior observ= ed from the GLOBEC data sets during process and broadscale cruises will be inc= orporated into the model using a simple biological advection term. The temperature fields are given b= y the physical model. For the food-= fields we will use two approaches as described in the next section: 1) empirically-forced 3D static phytoplankton fields derived from 3D chlorophyll measurements and ocean col= or data, and 2) a dynamic 3D prey field based on a simple NPZ-type of food web model. In the former case the copepods will be transported through static prey fields, and in the latter = case the copepods and prey field will be transported together.=

Lower trophic level food web model — We will explore the possibility of generating temporally evolving 3D phytoplankton fields from a lower trophic level food= web model (e.g. NPZ or NPZD) to provide food for the copepod population model. = We will use one-way coupling whereb= y the copepods depend on the food field but do not affect it (e.g. Carlotti, 1998; Batchelder et al., 2002). In this case, the= Z will represent microzooplankton grazers used solely as a closure term. We can generate realistic 3D phytoplankton fields using a simple NPZ model (Ji, 2003). It may also be possible to allow t= he copepods to graze the P and Z (e.g., Carlotti, = 1996), but this approach may not be necessary or feasible in 3D. The NPZ model will be adjusted so = that the resulting fields approximate the 3D chlorophyll and satellite data. The concept of using an NPZ model = to generate spatially explicit prey fields has been used with an IBM of larval pollock in the Gulf of A= laska (Hermann et al., 2001), where the Z represented copepod prey. Here we propose= to use the NPZ to generate the P fields for spatially explicit population mode= ls of dominant copepod species. = The resulting copepod distributions will be used in a concurrent modeling study= of Werner et al. who are proposing to develop an IBM of larval cod (see Werner letter).

Plankton food web models are variously complex, ranging from simple NPZ to models wi= th multiple subcomponents. In the GOM/GB a simple NPZ (Klein, 1987; Lewis et a= l., 1994; Franks and Chen, 1996; Franks and Chen, 2001) and a nine-compartment model (Ji, 2003) have been used. The NPZ model coupled with a 3-D nonlinear, primitive equation, finite-difference ocean circulation model (E= COM-si, Blumberg and Mellor, 1987) can successfully appro= ximate 3D phytoplankton fields observed on GB from summer cruises and satellite da= ta. Robust features, such as the subsur= face maximum and mixing-front induced productivity were produced with this model (Franks and Chen, 2001). More recently, a nine-compartment model (Ji, 2003), coupled to ECOM_si and FVCOM, captured the basic= seasonal and spatial patterns of nutrients and phytoplankton on GB. This model includes 3-N (nitrate, ammonia and silicate), 2-P (large and small), 2-Z (large and small), and tw= o detrital pools (N and Si)= . Silicate can limit the spring diatom bloom on GB (Townsend and Thomas, 2001; Townsend and Thomas, 2002, Ji, 2003).

In the proposed study, the simple NPZ model will be used due to its capability= and robustness, as well as the availability of observation data for the initial= and boundary conditions. We also will explore the use of separate pools in the model for Si and N and for P (diatom, non-diato= m). The model will run continuously fo= r 3 years from January 1997 to December 1999, a period when we have both nutrie= nt and phytoplankton data from the GLOBEC GB program. Initial horizontal distributions of nitrogen, phytoplankton and zooplankton during winter will be derived from climatological data (e.g. Petrie and Yeats, 2003), satellite imagery and MA= RMAP data, respectively. An initial homogenous vertical distribution also will be specified (as observed in win= ter). While the NPZ model is being = tested against the observed data, these same data will be used to generate spatial= ly and temporally interpolated 3D static prey fields for the copepod population model. This empirical approac= h is complementary to the model-based approach.

Upstream data – Biological data for the SS and SW will be obtained from our Canadian colleagues, = who have been actively involved in the GB and Canadian GLOBEC programs as well = as in other studies of the SS and Labrador Sea including the AZMP= (see Head letter). To the extent available we will use interpolated data from 1995-1999 in the SS and SW. We also will use idealized boundary scenarios based on observed data from other years and larger scales (eg. CPR data). =

Biological transport = 212; We will use concentration-based (Eulerian) rather than individual-based (Lagrangian) transport for the copepod and food web models= . While we have used both types of models in the past, the use of concentration-bas= ed models allow us to compute fluxes and mass balances directly and more accurately. This approach is necessary for quantifying such factors as sustainability of populations in particular areas. The concentration-based approach lends itself easily to stage-structured popula= tion models. This straightforward = method avoids the necessity of using super-individual particles or spawning and removing particles at each time step.

2.3.3.Numer= ical experiments

We will conduct a series of targeted numerical experiments using prognostic fo= rward model runs (rather than inverse methods) to address each of the above hypotheses. Data and model wi= ll be compared using a maximum-likelihood method (Stock, 2005; Stock et al., submitted). These process studies will help interpret the basic behavior and inter-annual variability shown in the 1995-99 hindcast. Detailed tasks and time table for = this work are given in the required supplement on project management and data exchange.

Working hypothesis— We will initi= alize the model with the mean winter concentrations of each copepod species (in separate model runs). The same model structure will be used for all species, changing only the parameter values (temperature/food/life-stage dependent egg production rate, developm= ent rate, growth rate, and normalized stage-dependent mortality) and behaviors,= and thus expediting the model runs. The inputs of characteristic life history traits of each species together with = its initial abundance patterns should generate its observed characteristic seasonal/spatial patterns. The model will be run for a complete annual cycle and compared with the 5-year monthly mean abundance patterns. To examine inter-annual variation = we will model each copepod species during the complete 5-year GLOBEC period 1995-19= 99, first year by year, then continuously over the 5 years. The result will be a complete 5-ye= ar biological/physical hindcast.

Food versus PredationR= 12; We will examine the degree of food-dependence and mortality with regard to inter-annual variability for each species. First we will confirm that the NPZ= model can be used to approximate the 3D distribution of phytoplankton for the multi-year monthly mean and year-to-year changes. In parallel, we will use fixed 3D phytoplankton fields approximated from monthly averaged satellite and in si= tu data. We will interpolate the phytoplankton fields between monthly values to avoid discontinuities. We will use the NPZ model together= with scenarios of salinity and nutrient (N, Si) input through = the NEC, to determine the extent to which phytoplankton biomass/production on GB is determined by this forcing and the extent to which the copepod species are affected by it. We will exami= ne possible effects of temperature and transport during 1995-1999 on the trend= s in mean abundances of each species.

Self-sustainabilty— We will initi= alize the model with the observed distribution of each species and exclude input from other regions, to determine whether the local population is self-sustaining= . In particular we will examine the importance of Calanus input fro= m SW and/or SS to the GOM/GB population, the input of Pseudocalanus from western GOM to the GB population, the key so= urce regions for Oithona similis, and whether Centro= pages typicus depends on immigration to sustain its population on GB. We will further examine whether re= sting eggs can explain high abundance of Centropages hamatus and Temora longicornis on the crest of GB. We will initialize the population = as resting eggs on the crest and determine whether the resulting plume of nauplii and subsequent copepodid= s matches the observed distributions. We will examine the formation of the Calanus “hole” on GB by initializing with CVs in the GOM during December and determining the extent to which the hole forms as a result of gradient advection versus high crest mortality.

Catastrophic global warming— Finally we wi= ll use boundary forcing that represents a scenario of catastrophic warming (see FV= COM section). We will run the mod= el for single and multiple years to determine the impact on the NPZ fields and on = each dominant copepod species.

2.3.4.Work = schedule— The schedule for the proposed modeling work tog= ether with a description of individual tasks, project management, dissemination, = and timeline are described in the required supplemental documents.

3.      Sign= ificance of Proposed Research– Intellectual Merit

The proposed work will provide new insights into the role of local dynamics and large-scale forcing in controlling population dynamics of marine copepods.<= span style=3D'mso-spacerun:yes'> We believe that the physical/biolo= gical model resulting from the proposed work will be a legacy of the Georges Bank GLOBEC program by providing a valuable tool that subsequent researchers can= use to study the dynamics of this system in hindcast, nowcast, and forecast mod= es. The results of the targeted experiments will provide insights into the degree of bottom-up and top-down control and the degree of sustainability of copepod populations under different conditions of external forcing. The work will provide a new understanding of the impact of basin-scale forcing, including catastrophic change, on local-scale plankton dynamics.&= nbsp; The resulting spatially explicit model of small and large copepod species will provide dynamic prey fields for concurrent and subsequent larv= al fish modeling studies, leading to a better understanding of recruitment in = fish populations.

4.      Broa= der Impacts

The New England Regional Center for Ocean Science Education Excellence (NER-COS= EE) at WHOI will work with us to facilitate development of the education and outreach component of this project. COSEE = is a national program, with regional centers, designed to facilitate K-12, colle= ge, and public education and outreach. <= /span>The proposed work will provide a state-of-the art coupled biological/physical m= odel of an important marine area. = We envision a broad user-base for this final product, which will be served via= the web to our scientific colleagues as well as to fishermen, K-12 educators, a= nd the general public. The main web-page will be located at WHOI with links to the other institutions. Chen’s group at UMD will cre= ate a website specifically for this work, expanding their existing FVCOM site, wh= ich now has user-friendly point-and-click modeling capacity. Users will be able to examine 4D m= odel data from the hindcast for the GLOBEC years as well as from pre-run nowcasts and forecasts. Users will be = able to change model parameters and create new model runs, with the level of sophistication depending on skill level. Further discussion of these broader impacts is given in the required supplementary section on this topic.

References

Ban, S.,= H. Lee, A. Shinada, and T. Toda. 2000. In situ egg production and hatching success of= the marine copepod Pseudocalanus newmani in Funka Bay and adjacent waters o= ff southwestern Hokkaido, Japan: associated to diatom bloom. J. Plankton Res. 22, 907-922.

Batchelder, H.P., C.A. Edwards and T.M. Powell, 20= 02. Individual-based models of copepod populations in coastal upwelling regions: implications of physiologically and environmentally influenced diel vertical migration on demographic success and nearshor= e retention. Prog. Oceanogr., 53, 307-333.

Bi= gelow, H. B. 1926. Plankton of the offshore waters of the Gulf of Maine. Bull. US Bur. Fish., 40, 1–507.

Bucklin, A., A. M. Bentley and S. P. Franzen. 1998. Distribution and relative abundance of the copepods Pseudocalanus moultoni and P. new= mani on Georges Bank based on molecular identification of sibling sp= ecies. Mar. Biol., 132, 97–106.

Bu= cklin, A., M. Guarnieri, D. J. Mc= Gillicuddy and R. S. Hill. 2001. Spring-summer evolution of Ps= eudocalanusspp. abundance of Georges Bank b= ased onmolecular discrimination of P. moultoni and P. newmani. Deep-Sea Res. II., <= i>48, 589–608.

Burchard, H., 2002. Applied turbulence modeling in marine waters.= Springer:Berlin-Heidelberg-New= York-Bar= celona-Hong Kong-London-Milan Paris-Tokyo, = 215pp.

Ca= mpbell R W, Head E J H 2000 Egg produc= tion rates of Calanus finmarchicus i= n the western North Atlantic: effect of gonad maturity, female size, chlorop= hyll concentration, and temperature. Can= J Fish Aquat Sci 57:518–529

Campbell<= span class=3DGramE>, R.G., M.W. Wagner, G.J. Teega= rden, C.A. Boudreau and E.G. Durbin. 2001b Growth and development rates of the copep= od Calanus finmarchicus reared in the laboratory. Mar. Ecol. Prog. Ser., 221, 161–183.

Campbell, R.G., Runge, J. A., Durbin, E.G., 2001a. Evidence for food limitation of Calanus finmarchicus = production rates on the southern bank of = Georges Bank during April 1997. Deep-Sea Research II 48, 531-549.

Carlotti, F. and G. Radach, = 1996. Seasonal dynamics of phytop= lankton and Calanus finmarchicus in = the North Sea as revealed by a coupled one dimensional model. Limnol= . Oceanogr.,= 41, 522-539.

Carlotti, F. and K.U. Wolf, 1998. A Lagrangian ensemble model of Calanus finmarchicus coupled with a 1-D ecosystem model.= Fish. Oceanogr., 7, 191= -204.

Chen, C, G. Cowles and R. C. Beardsley, 2004a. An unstructured grid, finite-volume coastal ocean model: FVCOM User Manual. SMAST/UMASSD Technical Report-04-0= 601, pp183.

Chen, C. and R. C. Beardsley. 1998. Tidal mixing over finite-amplitude banks:= a model study with application to Georges Bank. Journal of Marine Research, 56 (6), 1163-1203.

Chen, C., H. Liu, and R. Beardsley, 2003a. An unstructured grid, finite-volume, three-dimensional, primitive equations ocean model: Application to coastal ocean and estuaries. Journal of Atm= ospheric and Ocean Technology, 20 (= 1), 159–186.

Ch= en, C., H. Liu, R. C. Beardsley, G. Cowles, J. Pringle, R. Schlitz, and B. Rothschild, 2003b. Application of FVCOM to the Gulf of Maine/Georges Bank: Simulated and assimilated modeling studies of stratification and subtidal circulation. EOS Trans, AGU= , 84(52), Ocean Science Meeting Suppl., Abstract OS51G-08.

Ch= en, C., J. Qi, H. Liu, C. Li, H. Lin, R. Walker and= K. Gates, 2005d. Tidal Flushing Dynamics in the Satilla River, Georgia: A Comparison between FVCOM and ECOM-si= . Journal of Geophysical Research= , to be submitted.

Chen, C., Q. Xu, R. = C. Beardsley, and P. J. S. Franks., 2003. Modeling Studies of the Cross-Frontal Water Exchange on Georges Bank: A 3D Lagrangian Experiment, Journal of Geophys= ical Research, 108(C5), 3142, 10.1029/2000JC000390 .<= o:p>

Chen, C., R. C. Beardsley, H. Huang, H. Liu, and= Q. Xu, 2005b. A finite-volume numerical approach for coastal = ocean studies: Comparisons with the finite-difference models. Journal of Geophysical Research, in revision.=

Ch= en, C., R. C. Beardsley, Q. Xu, G. Cowles, and R. Limeburner, 2005c. Tidal Dynamics in the <= span style=3D'font-size:11.0pt'>Gulf of Maine and New England Shelf: An Application of FVCOM. Deep Sea= Researc= h II-GLOBEC/GB special issue, submitted.

Ch= en, C., R. Houghton, R. C. Beardsley, Q. Xu, and H.= Liu, 2003c. Preliminary results of model dye experiments on Georges Bank. EOS Transactions, AGU, 84(52), Ocean Science Meeting Suppl., Abstract OS32C-13. <= /p>
Chen, C., Wang, T, L. Wang, J. Blanton, C. Li, H. Hua= ng and H. Lin, 2005e. Tidal-induced fl= ushing process over the estuarine-tidal creek-salt marsh complex of the Okatee/Collection River in South Carolina<= span style=3D'font-size:11.0pt'>: An application of FVCOM. Journal of Geophysical Research. To be submitted.

Chen, C., Z. Wu, R. C. Beardsley, S. Shu, and C. Xu, 2005a. Using MM5 to hindcast the o= cean surface forcing fields over the Gulf of Maine and Georges Bank<= /st1:place> region. Journal= of Atmospheric and Oceanic Technol= ogy, 22(2): 131-145.

Churchill, J. H., J. P. Manning, and R. C. Beardsley, 2003. S= lope water intrusions onto Georges Bank. Journal of Geophysical Research, (GLOBEC special section), 108(C11), 8012, doi:10.1029/2002JC001400.

Clarke, = G.L., 1946. Dynamics of production in a marine area. Ecol. Monogr. 16, 321–337.

Cohen, E. B., and M. D. Grosslein. 1987. Production o= n Georges Bank Compared with other shelf ecosystems. Pp. 383–391, In R.H. Backus and D.W. Bourne (eds.) Georges Bank. MIT Press. Cambridge, MA. 593 p= p.

Corkett C J, and I. A. McLaren. 1978. The biology of Pseudocalanus. Adv. Mar Biol., 15, 1-231.=

Cu= shing. D. H. and J. J. Walsh. 1976. The Ecology of the Seas. London: Blackwell. 467 pp.<= o:p>

Dadou, I. V. Garcon, V. Anderson, G. Flierl, and C. Davis. 1996. Impact of the north equatorial current meandering on a pelagic ecosystem: A modeling approach. J. Mar. Res.,<= span style=3D'mso-spacerun:yes'> 54, 311–342.

Da= vis, C. S. 1983. Laboratory rearing of marine calanoid copepods. J. Exp. Mar.= Biol. Ecol., 71, 119-133.

Da= vis, C. S. 1984a. Predatory control of copepod seasonal cycl= es on Georges Bank<= /st1:place>.   Mar. Bio= l., 82, 31-40.

Da= vis, C. S. 1984b. Food concentrations on Georges Bank: Non-limiting effect on growth and survival of laboratory reared Pseudocalanus sp.&nb= sp; and Paracalanus parvus. &nb= sp; Mar.&nbs= p; Biol., 82, 42-46.

Da= vis, C. S. 1984c. Interaction of a copepod population with t= he mean circulation on Georges Bank.   J. Mar. Res., 42, 573-590.

Da= vis, C. S. 1987a. Components of the zooplankton production cycle in the temperate ocean.   J. Mar. Res., 45, 947-983.

Da= vis, C. S. 1987b. Zooplankton Life Cycles. pp. 256-267. In: Geor= ges Bank. = R. H. Backus (ed.), MIT Press.

Davis, C. S. and J. H. Steele. 1994. Modeling upper ocean biological-physical processes. URIP Workshop Report, WHOI TECH RPT 94-32, 65 pp.

Davis, C.S. and P. Alatalo. 1= 992. Effects of constant and intermittent food supply on life history parameters= in a marine copepod.   Limnol= . Oceanogr., 37, 1618–1639.

Drinkwater, K. F., D. B. Mountain, and A. Herman= . 2= 003. Variability in the Slope Water Properties off Eastern North America and their Effects on the Adjacent Shelves. J. Geophys. Res. (submitted)

Dudhia, J., D. Gill, K. Manning, W. Wang, C. Bruyere, J. Wilson, and S. Kelly, 2003: PSU/NCAR mesoscale modeling system tutorial cla= ss notes and user’s guide, MM5 modeling system version 3, Mesoscale and = Microscale Meteorology Division, National Center for Atmospheric Research.

Du= rbin, E. G., J. A. Runge, R. G. Campbell, P. R. Garrahan, M. C. Cascas and S. Plourd= e. 1997. Late fall-early winter recruitment of Calanus finmarchicus on Georges Bank., Mar. Ecol. Prog. Ser., 151, 103-14.

Durbin, E.G., Garrahan, P.R., Casas, M.C. 2000. Abundance and distribution = of Calanus finmarchicus on Georges Bank during 1995 and 1996. ICES Journal of Marine Science 57, 1664-1685. <= o:p>

Durbin, E.G., M.C. Casas. Abundance, spatial distribution, and interannual variability of copepods on Georges Bank during the winter spring period. Deep-Sea Research (Submitted)

Durbin, E.G., R. Campbell, M. Casas, B. Niehoff, J. Runge, M. Wagner.<= span style=3D'font-size:11.0pt'> 2003. Interannual V= ariation in phytoplankton blooms and zooplankton productivity and abundance in the <= /span>Gulf of Maine during winter. Mar Ecol Progr Ser. 254:81-100

Dzierzbicka-Glowacka, = L. 2004. Growth and development of copepodite stages of Pseudocalanus spp., J. Plankton Res., 26, 49-60.

Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson= , and G. S. Young, 1996. Bulk parameterization of air-sea fluxes for tropic ocean global atmosphere coupled-ocean atmosphere response experiment. Journal of Geophysical Research, 101 (C2), 3747-3764.<= o:p>

Fairall, C. W., E. F. Bradley, J.E. Hare, A.A. Grachev, = and J. B. Edson, 2003. Bulk parameterization of air-sea fluxes: updates and verificat= ion for the COARE algorithm. J. Climate, 16, 571-590.

Flierl, G. R. and C. S. Davis. 1= 993. Biological effects of Gu= lf Stream meandering. J. Mar. Res.,<= span style=3D'mso-spacerun:yes'> 51, 529–560.

Flierl, = G.R. and Wroblewski, J.S. 1985.= The possible influence of warm core Gulf Stream rings upon shelf water larval fish distribution. Fish. = Bull., 83: 313-330=

Franks, P.J.S. and C. Chen, 1996.<= span style=3D'font-size:11.0pt'> P= lankton production in tidal fronts: A model of Georges Bank in summer. Journal of Marine Research 54, 631-651.

Franks, P.J.S. and C. Chen, 2001.<= span style=3D'font-size:11.0pt'> <= span class=3DGramE>A 3-D prognostic numerical model study of the <= st1:place>Georges Bank<= /st1:place> ecosystem. Part II: biological-physical model. Dee= p Sea= Rese= arch II 48, 457-482.

Fratan= toni, = P. S. and R. S. Pickart, 2005a. T= he western North Atlantic <= span class=3DSpellE>shelfbreak current system in summer. J. Phys. Oceanography, submitted.

Fratantoni, = P. S. and R. S. Pickart, 2005b. S= tructure and alongstream evolution of the shelfbreak jet in the western North Atlantic= .<= /span> J. Phys. Ocean= ography, submitted.

Fr= ost B W. 1989. A taxonomy of the marine calanoic copepod genus = Pseudocalanus. Can J Zool 67: 525±551

Galperin, B., L. H. Kantha, = S. Hassid, and A. Rosati, 1988. A quasi-equilibrium turbule= nt energy model for geophysical flows. Journal of the Atmospheric Sciences, 4= 5, 55–62.

GLOBEC, 1991a. U. S. GLOBEC: Initial Science Plan, Report Number 1.

GLOBEC, 1991b. U. S. GLOBEC: N= orthwest Atlantic Program, GLOB= EC Canada/U.S. Meeting on N.W. Atlantic Fisheries and Climate, Report Number 2= .

GLOBEC, 1992. U. S. GLOBEC: N= orthwest Atlantic Implementation Plan, Report No. 6,

GLOBEC, 1996. U. S. GLOBEC: Northeast Pacific Implementation Plan, Report No. 17.

GLOBEC, 1997. U. S. GLOBEC: Workshop on Modeling the Southern Ocean Ecosystem, Report No. 18.

GLOBEC, 2000. GLOBEC in the G= ulf of Mexico: Large Rivers a= nd Marine Populations, Report No. 19.

Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the Fifth-Generation Penn State/NCAR Mesoscale Model (MM5). = NCAR Technical Note, NCAR/TN 398+STR, 117pp.

Grosslein, M. S., and R. C. Hennemut= h, 1973. Spawning stock and other factors r= elated to recruitment of haddock on Georges Bank. Rapp. P.-v. Réun. Co= ns. Int. Explor. Mer, 163, 77–88.

Hall, D. J. 1964. An experimental approa= ch to the dynamics of a natural population of Daphnia gal= eata mendotae . Ecology 45 (1), 94-1= 12.

Han, G., C. E. Hannah,= J. W. Loder, and P. C. Smith, 1997. Seasonal varia= tion of the three-dimensional mean circulation over the Scotian Shelf. J. Geophysical Research, 102(C1), 1011-1025.

Hannah, C. G., J. A. Shore, J. W. Loder, and C. E. Naimie, 2001. Sea= sonal circulation on the western and central Scotian Shelf. J. Phys. Oceanography, 31(2), 591-615.

Ha= nnah, C.G., Naimie, C.E., Loder<= /span>, J.W., Werner, F.E. 1998 Upper-ocean transport mechanisms from the Gulf of Maine to Georges Bank, with implication= s for Calanus supply. Cont. Shelf Res., 7, 18871911.

He= ad, E.J.H., Harris, L.R., Petrie, B., 1999. Distribution of Calanus spp. on and around the Nova Scotia Shelf in April: ev= idence for an offshore source of Calanus finmarchicus to the central and western regions. Canadian Journal of Aquatic Sciences 5= 6, 2463-2476.

He= ath, M.R., Backhaus, J.O., Richardson, K., McKenzie, E., Sl= agstad, D., Beare, D., Dunn, J., Fraser, J.G., Gallego, A., Hainbucher, = D., Hay, S., Jónasdóttir, S., Madden, H., = Mardaljevic, J., Schacht, A., 1999. Climate fluctuations and the spring invasion of the North Sea by Calanus finmarchicus<= /i>. Fisheries Oceanography 8 (= Suppl. 1), 163-176.

Hermann, A. J., S. Hinckley, B. A. Megrey and J. M. Napp. 2001 Applied and theoretical considerations for constructing spatially explicit individual-b= ased models of marine larval fish that include multiple trophic levels, ICES J. Mar. Sci. 58:1030-1041

Hinrichsen, H.H., C. Möllmann, R. Voss, F.W. Köster, and G. Kornilovs. 2002. Biophysical mode= ling of larval Baltic cod (Gadus morhua) growth and survival. Can. J. Fish. Aquat<= /span>. Sci., 59, 1858–1873.

Hjort, J . 1914. Fluctuations in the great fisheries of Europe viewed in the light of biological research. Rapp . P.-v . Reun . Cons. int. Explor. Mer, 20, 1-228.

Huang, H., C. Chen, G. Cowles, R. C. Beardsley, = and K. Hedstrom, 2005a. Sensitivity of the numerical solution to unstructured triangular grids: A validation experiment of FVCOM for the Rossby= equatorial soliton. Dee= p Sea= Rese= arch II-GLOBEC/GB special issue, to be submitted.

Huang, H., C. Chen, J. O. Blanton, and F. A. And= rade, 2005d. Tidal current asymmet= ry and implication to variability of the dissolved oxygen over shallow tidal creek= s: an application of FVCOM to the Okatee River, So= uth Carolina. Journal of Geophysical Reseqarch= , to be submitted.

Huang, H., C. Chen, R. C. Beardsley, and K. Hedstrom, 2005c. Validation experiments of FVCOM for the wind-driven flow in an elongated rotating basin: A comparison with analytical solution. = Deep Sea Research II-GLOBEC/GB spe= cial issue, to be submitted.

Huang, H., C. Chen, R. C. Beardsley, and K. Hedstrom. 2005b. Capability and Accuracy of the unstructured grid model to simulate t= he coastal hydraulic jump: A validation experiment of FVCOM via horizontal advection schemes. Deep Sea= Rese= arch II-GLOBEC/GB special issue, to be submitted.

Ji, R., 2003. Biological and physical processes controlling the spring phytoplankton bloom dynamics= on Georges Bank. = Ph.D. Dissertation, University of Georgia= , = Athens<= /span>, Georgia= .<= span style=3D'font-size:11.0pt;color:black'>

Ji, R., C. Chen, P. J. S. Franks, D.W. Townsend,= E.G. Durbin, R. C. Beardsley, R.G. Lough, and R.W. Houghton, 2004.= Spring bloom and associated lower trophic level= food web dynamics on Georges = Bank: from observation to model. Submitted to Deep-S= ea Research II

Ji, R., C. Chen, P. J. S. Franks, D.W. Townsend,= E.G. Durbin, R. C. Beardsley, R.G. Lough, and R.W. Houghton, 2004.= Effects of topography and tidal mixing front on= the spring phytoplankton bloom on Georges Bank: 2-D model experi= ments. Submitted to Deep-Sea Research II.

Ji, R., C. Chen, P. J. S. Franks, D.W. Townsend,= E.G. Durbin, R. C. Beardsley, R.G. Lough, and R.W. Houghton, 2004.= The impact of Scotian Shelf Water “cross-over” on the plankton dynamics on Georges Bank: A 3-D experiment for the 1999 spring bloom. Submitted to Deep-Sea Research II

Johnson, C., J. Pringle, and C. Chen. <= span class=3DGramE>Transport and retention of dormant copepods in the Gulf of Maine. (submitted to DSRII)= .

Klein Breteler WCM, = Fransz HG, Gonzalez SR (1982) Growth and development = of four calanoid copepod species under experimenta= l and natural conditions. <= span style=3D'font-size:11.0pt;mso-bidi-font-weight:bold'>Neth J Sea = Res 16: 195–207

Klein Breteler WCM, Gonzalez SR (1986) Culture and development of Temor= a longicornis (Copepoda<= /span>, Calanoida) at different conditions of temperature and= food. Syllogeus 58:71–84

Klein Breteler WCM, Gonzalez SR, Schogt N (1995) Development of Pseudocalanus elongatu= s (Copepoda, Calanoida) cultured at different temperature and food conditions. Mar Ecol Prog Ser 119:99–110

Kl= ein, P., 1987. A simulation of some physical and biological interactions. Georges Bank (eds. by R.H.Backus= and D.W.Bourne), pp. 395-402. MIT Press, Cambridge, = Massachusetts.

Lee= , H., S. Ban, T. Ikeda and T. Matsuishi. 2003. Effect of temperature on development, growth and reproduction in the marine copepod Pseudocalanus newmani at satiating food condition. J. Plankton = Res. 25, 261-271.

Lewis, C. V. W., C. Chen, and C. S. Davis. 2001. Variability in wind forcing and its effec= t on circulation and plankton transport over Georges Bank. = Deep-Sea Research II, 48, 137–158.

Lewis, C. V., C. S. Davis, and G. Gawarkiewcz. 1994. Wind-forced biological-physical dynamics = on an isolated off-shore bank. Deep-Sea Res. 41, 51–73.

Li, X., D. J. McGillicuddy<= /span>, E. G. Durbin, P. H. Wiebe. Biological control of the vernal population increase of Calanus finmarchicus on Georges Bank. In prep for Deep Sea Resea= rch II.

Lindeman, R. L. 1942. The trophic-dynamic aspect of ecology. Ecology, 23, 399-418.

Li= ndley J A, and = H G Hunt. 1989. The distribution of Labidocera wollastoni and Centropages hamatus in the North Atlantic Ocean and the North Sea in relation to the role of resting eggs in the sediment. In: Ryland JS, Tyler PA (eds) Reproduction, genetics and distributions of marine organisms. Olsen and Ols= en, Fredensborg, pp 407–413

Li= ndley JA 1990 Distribution of overwintering calanoid copepods eggs in sea bed sediments around so= uthern Brit= ain. Mar Biol 104:209&#= 8211;217

Lough, R. G., and D. G. Mountain.<= span style=3D'font-size:11.0pt'> 1996. Effect of small-scale turbulence on feedi= ng rates of larval cod and haddock in stratified water on Georges Bank. = Deep Sea= Res. II, 43, 1745-1772.

Lo= ugh, R.G., E. M. Caldarone, L. J. Buckley, E. A. Broughton, M. E. Kiladis, and B. R. Burns, 1996. Vertical distribution of cod and haddock eggs and larvae feeding and condit= ion in stratified and mixed waters on southern Georges Bank, May 1992. Deep Sea Res.= , 43, 18751904.

Ly= nch, D. R., Gentleman, W. C., McGillicuddy, D. J. an= d C. S. Davis, 1998. Biological/ physical simulations of Calanus finmarchicus population dy= namics in the Gulf of Maine. M= arine Ecology Progress Series, 169, 189–210.

Maps, F., J.A. Runge, B. Za= kardjian and P. Joly. 2005. Egg production and hatching success of = Temora longicornis (= Copepoda, Calanoida) in the southern Gulf of St. Lawrence. Mar.-Ecol. Prog. Ser.285:117-128=

Marcus, N. 1996. Ecological and evolutionary significance of resting eggs in marine copepods: past, present, and future studies. Hydrobiologia, 320: 141–152.

Marcus, N. H. and R. V. Lutz. 1998. Longevity of subitan= eous and diapause eggs of Centropages hamatus (Copepoda: Calanoida) from= the northern Gulf of Mexico<= /span>, Marine= Biology, 131, 249-257.

McGillicuddy, D. J. Jr. and A. Bucklin. 2002. Intermingling of two Pseudocalanus species on Georges B= ank. J. Mar. Res. 60: 583–604.

McGillicuddy, D. J., D. R. Lynch, A. M. Moore, W. C. Gentleman, C. S. Davis, and C. J. Meise. 1998. An adjoint data assimilation approach to= the diagnosis of physical and biological controls on Pseudocalanus spp. populations in the Gulf of <= /span>Maine - Georges Bank Region. Fisheries Oce= anogr., 7, 205–218.

Mc= Laren IA, Laberge E, Corkett CJ, Sevigny J-M 1989a Life cycles of four species of Pseudocalanus in Nova Scotia. Can J Zool 67: 552-558

McLaren <= span style=3D'font-size:11.0pt'>IA, Sevig= ny J-M, Corkett CJ 1989b Temperature-dependent development in Pseudocalanus species. Can J Zool 67: 559-= 564

McLaren, I. A., and P. Avendano= . 1995. Prey field and diet of larval cod on West= ern Bank, Scotian Shelf. Can. J. Fish. Aquat<= /span>. Sci., 52, 448–463.

McLaren, I. A., E. Head, and D. Sameot= o. 2001. Life cycles and seasonal distributions of= Calanus finmarchicus on the central Scotian Shelf. Can. J. Fish. Aquat<= /span>. Sci., 58, 659–670.

McLaren, I. A., P. Avendaño, C. T. Taggart, and S. E. Lochmann, 1998. Feeding by larval cod in di= fferent water-masses on Western Bank, Scotian Shelf. Fish. Oceanogr., 6, 250-265.

Meise, C.J. and J. E. O'Reilly, 1996. Spatial and seasonal patterns in abundance and age-composition of Calanus finmarchicus in the Gulf of Maine and on Georges Bank, 19771987, Deep-Sea Res. II, = 7, 14731501

Meise-Munns, C., J. Green, M. Ingham, and D. Mountain, 1990= . <= span class=3DSpellE>Interannual variability in the copepod populations of Georges Bank<= /st1:place> and the western Gulf of Maine, Mar. Ecol. Progr. Ser., 65, 225-232,

Mellor, G. L. and T. Yamada, 1982.= Development of a turbulence= closure model for geophysical fluid problem. Reviews of Geophysics and Space. Physics, 20, 851–875.

Mi= ller, C. B., T. J. Cowles, P. H. Wiebe, N. J. Copley,= and H. Grigg, 1991. Phenology<= /span> in Calanus finmarchicus: Hypotheses about control mechanisms, Mar. Ecol. Progr. Ser., 72, 79-91,.

Mi= ller, C.B., Lynch, D.R., Carlotti, F., Gentleman,W., Lewis,V.W., 19= 98. Coupling of an individual-based population dynamic model of <= i>Calanus finmarchicus to a circulation model for the Georges Bank<= /st1:place> region. Fisheries Oceanography, 7, 3/4, 219-234.

Ni= xon, S. W. 1988. Physical energy inputs and the comparative ecology of lake and marine ecosystems. Limnol. Oceanogr., 33, 1005-1025

Odum, H. T. 1957. Trophic structure and productivity of <= /span>Silver Springs, = Florida. Ecologi= cal Monographs, 27, 55-112= .

Ohman, M. D., J. A. Runge, E. G. Durbin, D. B. Field = and B. Niehoff. 2002. On birth and death in the sea. Hydrobiol., 480, 55-68.

Osgood, K. E. and D. M. Che= ckley. 1997. Seasonal variations in a deep aggregation= of Calanus pacificus in the Santa Barbara Basin. Mar. Ecol. Prog= . Ser., 148<= /span>, 5969.

Pe= trie, B. and P. Yeats, 2000. Annual and interannua= l variability of nutrients and their estimated fluxes in the Scotian Shelf-Gu= lf of Maine region. Can. J. Fish. Aquat<= /span>. Sci<= /span>. 57, 2536–2546.

Plourde= , S., and J. A. Runge. 1993. Reproduction of the planktonic copepod, <= i>Calanus finmarchicus, in the Lower St. Lawrence Estuary: Relation to the cycle = of phytoplankton production and evidence for a Calanus pump, Mar.\ Ecol. Progr. Ser., 102, 217-227.

Reiss, C. S., I. A. McLaren, and P. Avendaño. 2003. Horizontal and vertical distribution patt= erns, retention rates, and population dynamics of zooplankton on Western Bank, Scotian Shelf. Canadian Journal of Fisheries and Aquatic Sciences, 60,12291244.

Ru= nge, J.A., S. Plourde, P. Joly, E. Durb= in and B. Niehoff.&nb= sp; Characteristics of egg production of the planktonic copepod, Calanus finmarchicus, on Georges Bank: 1994-1999. Submitted.to Deep-Sea Res. II= .

Sabatini, M. and Kiørboe,T. 1994. Egg production, growth and development of the cyclopoid copepod Oithona similis. J. Plankton Res., 16, 1329–1351.

Sameoto= , D. D., and A. W. Herman. 1990. Life cycle and distribution of Calanus finmarchicus in deep basins on the Nova Scotia shelf and seasonal changes in Calanus spp., Mar. Ecol. Progr= . Ser., 66, 225-237,.

Sarmiento, J. L., R. Slater, M. J. R. Fasham, H. W. Ducklow, J. R. Toggeweiler and G. T. Evans, 1993. A seasonal three-dimensional ecosystem = model of nitrogen cycling in the North Atlantic euphot= ic zone. Global Biogeochemical Cycles, 7, 4= 17-450.

Saumweber, W. J. and E. G. Durbin. T= he implications of energetic limitation for diapausing Calanus finmarchicus: Towards a Gulf of Maine Calanus budget. <= span class=3DGramE>submitted to D= eep Sea Res. II

Sissenwine, M. P., E. B. Cohen, and M. D. Grosslein, 1984. St= ructure of the Georges Bank ecosystem. Rapp. P.-v. Reun. Cons. int. Explor. Mer, 183, 243-254

Smagorinsky, J., 1963. General circulation experim= ents with the primitive equations, I. The basic experi-ment. Monthly Weather Review, 91, 99–164.

Smith, P.C., R.W. Houghton, R.G. Fairbanks, and = D.G. Mountain. 2001. Interannual variability of boundary fluxes and water mass properties in the Gulf of Maine and on Georges Bank: 1993-1997. Dee= p Sea= Res.= II, 48, 37-70.

St= eele, J. H., 1974. The Structure of Marine Ecosystems. Harvard Univ. Press, Cambridge, = MA.

Stock, C. A., 2005. Testing hypotheses concerni= ng the Initiation and development of blooms of the toxic dino= flagellate Alexandrium fund= yense in the western = Gulf of Maine using a coupled physical-bi= ological model, PhD Thesis, WHOI.

Stock, C.A., McGillicuddy,= D.J., Solow, A.R. and Anderson, D.M. Evaluating hypotheses for initiatio= n and development of Alexandrium fundyense blooms in the western Gulf of Maine using coupled physical-biological model. Submit= ted, Deep Sea Research II

Sundby, S. 2000. Recruitment of Atlantic cod stocks in relation to temperature and advection of copepod populations. Sarsia= , 85, 277-298.

Teal, J. M. 1962. Energy Flow in the Salt Marsh Ecosystem of Georgia. Ecology= , 43, 614-624.
Townsend, D.W. and A.C. Thomas, 2001. Winter-spring transition of phytoplankton chlorophyll and inorganic nutrients on Georges Bank<= /st1:place>. Deep Sea= Research II 48, 199-214.

Townsend, D.W. and M. Thomas, 2002. <= span class=3DGramE>Springtime nutrient and phytoplankton dynamics on Georges Bank<= /st1:place>. Marine = Ecology Progress Series 228, 57-74= .

Townsend, D.W. and N.R. Pett= igrew. 1997. Nitrogen limitation of secondary production on = Georges Bank. J. Plankton Res. 19: 221-23= 5.

Townsend, D.W.,A.C. T= homas, L.M. Mayer, M. Thomas and J. Quinlan. 2004. Oceanography of the Northwest Atlantic Continental Shelf. pp. 119-168. In: Robinson, A.R. and K.H. Brink (eds). The Sea, Volume 14, = Harvard University Press.

Tu= rner, J. T. 2004. The importance of small planktonic copepods<= /span> and their
roles in pelagic marine food webs. Zoological Studies, 43, 255-266.

Wiebe, P. H., R. Beardsley, D. Mountain, and A. Buckl= in. 2002. U.S. GLOBEC Northwest Atlantic/Georges Ba= nk Program, Oceanography, 15, 12-29.

Wiebe, P.H., A.W. Morton, A.M. Bradley, R.H. Backus, J.E. Craddock, T.J. Cowles, <= span class=3DSpellE>V.A.Barber, and G.R. Flierl. 1985. New developments i= n the MOCNESS, an apparatus for sampling zooplankton and mic= ronekton. Mar. Biol. 87: 313-323.

Zakardjian, Bruno. A., J. Sheng, J. A. Runge, I. McLaren, S. Plourde, K. R. Thompson, and Y. Gratton . 2003. Effects = of temperature and circulation on the population dynamics of Calanus finmarchicus in the Gulf of St. Lawrence and Scotian Shelf: St= udy with a coupled, three-dimensional hydrodynamic, stage-based life history mo= del, J. Geophys. Res., 108(C11), 8016,= doi:10.1029/2002JC001410.

Zeldis, J. R., C. S. Davis, M. R. James, S. L. Ballara,= W. E. Booth, and F. H. Chang. 1995. Salp grazing in a larval fish habitat: effects on phytoplankton abundance, vertical distribution, and species composition.&nbs= p; Mar. Ecol. Prog. Ser., 126, 267–283.

Zhao, L., C. Chen, and B. Rothschild, 2005. Tidal dynamics and eddy shedding in the = Mount Hope Bay and Narragansett Bay: An application of FVCOM. Journal of Geophysical Research, to be submitted.


Figure 1. Circulation:= 1 flow across GSC into the north flank jet, 2<= /b> tidal-pumping of deep water onto GB, 3 wind-driven near-surface flow, 4<= /b> small-scale cross-frontal proces= ses, 5<= /b> SS cross-over


Figure 3. Schematic showing the strong westward penetration of LSW and northward position of the Gulf Stream following low winter North Atlantic Oscillati= on (NAO) index (left) and weak penetration of LSW and more southward positio= n of Gulf Stream following high winter NAO index. (Drinkwater et al, 2003)


Fig. 7. Calanus finmarchicus abundance i= n the N Atlantic from CPR data (Spears 2005) and nets/OPC (Hea= th et al. 2004). Note high densit= ies in the Labrador and Norwegian Seas.