WORKING GROUP REPORTS

Latitudinal Gradients Within the Eastern Boundary Current System

Cochairs: M. Ohman and B. Hickey

Participants: V. Holliday, A. Huyer, P. Kremer, A. MacCall, D. Mackas, K. Parker, W. Pearcy, J. Schumacher, T. Strub, R. Tipper, and D. Ware

INTRODUCTION

The California Current system (CCS) comprises three broadly definable regions: I, Vancouver Island to Cape Blanco; II, Cape Blanco to Point Conception; and III, Point Conception to northern Baja California (Fig. 5). These regions are characterized by differences in wind stress, intensity of coastal upwelling, coastal morphology, freshwater inflow, and influence of long-time-scale advection. The regions also differ in seasonality of planktonic production cycles. For some planktonic species, the regional boundaries represent biogeographic boundaries, although other species range throughout the three regions.

Among the most striking biological contrasts is the paucity of surface spawning by epipelagic fishes in the central region (II; see Parrish et al. 1981). This region is characterized by strong seasonal upwelling near coastal promontories and by a variety of mesoscale features such as coastal jets, eddies, and filaments that tend to transport organisms offshore. The fish species that do reproduce in this region tend to brood eggs or larvae, or to spawn demersally in protected waters. Equally striking is the concentration of spawning activity in Region III: about 90% of the epipelagic fish biomass (hake, sardine, anchovy) in the southern part of the California Current system spawns in the Southern California Bight and waters offshore. Primary and secondary production in Region III is therefore important to a significant fraction of California Current fishes.

A number of critical physical processes are likely to govern much of the variability in marine populations over time in the CCS and other eastern boundary currents:

These processes may be influenced by future changes in the global ocean; some evidence suggests that changes have already occurred in recent decades. The differences in physical-biological linkages in Regions I, II, and III provide a natural basis for comparative study of the changes in marine populations that may accompany different scenarios for climate change.

We strongly recommend conducting comparative studies in all three regions of the CCS. Forcing functions differ spatially, but some marine species depend on processes in all three regions, so a successful GLOBEC program will have to incorporate the significant processes operating on a broad latitudinal scale. One component of this program should include intensive field studies that focus on processes thought to be important in each region. As a second component, we recommend time series sampling in each of the three regions for at least a decade. The time series might include a combination of moored arrays, satellite observations, and ship surveys. Suitable localities within each region might include the west coast of Vancouver Island, or the coasts of Washington or northern Oregon (Region I); Point Arena to Point Sur (Region II); and the Southern California Bight (Region III). In each region there are oceanographic institutions and shipboard resources that could be applied to the task. It is likely that existing field programs in each of the three regions could be expanded.

california current regions

Figure 5. Generalized regional variations in physical and biological processes within the California Current System. The boundaries between Regions I, II, and III are only approximate and vary over time. The generalizations regarding Region III apply primarily to the Southern California Bight.

REGIONS OF THE CALIFORNIA CURRENT SYSTEM

The California Current system spans more than 20 degrees of latitude along the west coast of North America. Its range exceeds the scales of dominant atmospheric pressure systems as well as the scales of regional coastal morphology. Thus the variability of the California Current and associated planktonic populations inherently includes significant and sometimes dramatic latitudinal differences. Many of these differences (e.g., wind forcing, temperature, photoperiod) are roughly a linear function of latitude, whereas others have distinct biogeographical boundaries. Examination of the dominant characteristics suggests that the California Current has three principal regions (Fig. 5). The boundaries between these regions are not exact, particularly between Regions I and II, where the area of seasonally reversing monthly mean winds extends farther south (to about 39°N) than the area of relatively straight coastline (about 42°N). The boundaries are known to change during El Niño/Southern Oscillation and other long-term events, including intrusions of subarctic water from the north. The boundaries might also be expected to shift under some scenarios of global change.

In Region I, coastal wind stress is relatively strong, and wind direction reverses seasonally as well as on shorter time scales. Winter storms are particularly strong and frequent, leading to intense mixing and alongshore northward advection (Huyer et al. 1978; Thomson 1981; Hickey 1989; Thomson et al. 1989). Except for the region near the Strait of Juan de Fuca, the coastline is relatively straight, and the shelf is continuous, though narrow, over large alongshore scales (hundreds of km). Significant freshwater input is provided throughout most of the year by the Strait of Juan de Fuca and by the Columbia River. Large estuaries are relatively common and are thought to provide nurseries for several important species (e.g., Dungeness crab, McConnaughey et al., in press; Pacific herring, Haegele and Schweigert 1985). Primary production rates (Perry et al. 1989) and zooplankton biomass (Mackas 1992) have strong seasonal variations in this region. Some of the major copepod species overwinter at depth, then reappear in the surface layer for relatively short periods of growth in spring and summer. Several species (e.g., Neocalanus plumchrus, N. cristatus, Eucalanus bungii, Calanus pacificus oceanicus) enter Region I from the Subarctic Pacific or the West Wind Drift (Fleminger 1964; Fleminger and Hulseman 1973) and rarely extend south of this region.

The dominant physical characteristics of Region II, which extends approximately from Cape Blanco to Point Conception, are the coastal promontories. Recent research suggests that energetic coastal jets, filaments, and meanders are associated with these promontories (Davis 1985; Kosro & Huyer 1986; Huyer & Kosro 1987; Strub et al. 1991). Current jets commonly extend 200-300 km offshore and may lead to relatively short residence times for plankton in the coastal zone. The strongest equatorward wind stress and, hence, coastal upwelling also occur in Region II (Nelson 1977; Huyer 1983; Strub et al. 1987). Although wind stress varies seasonally, the seasonal mean is always directed toward the equator. The strong coastal upwelling in this region supplies "new" nutrients into the euphotic zone, leading to elevated primary production rates and high standing stocks of phytoplankton (Dugdale & Wilkerson 1989). Satellite imagery suggests that many of the high-chlorophyll features found in Region II are associated with jets, eddies, and other mesoscale features (Flament et al. 1985). Zooplankton biomass varies seasonally (Roesler & Chelton 1987). Zooplankton species composition can shift relatively abruptly at the frontal boundaries associated with mesoscale jets and eddies. Among the most dramatic biological characteristics of Region II is the latitudinal minimum in spawning of pelagic fishes. Whereas epipelagic fishes spawn extensively in Region III and to some extent in Region I, those in Region II appear mainly to brood their eggs or larvae (e.g., rockfishes) or to use nearshore embayments as spawning and nursery grounds (e.g., Pacific herring) that appear to reduce the probability of offshore transport of pelagic larvae (Parrish et al. 1981).

The dominant physical characteristic of Region III is that, because of the coastline bend at Point Conception, local wind stress is relatively weak on the scales of seasons and events (Nelson 1977; Halliwell and Allen 1987). Thus local upwelling is weak in spring and summer, and wind- and wave-induced mixing is relatively weak year-round. Winter storms occur only occasionally. Freshwater input is insignificant. Interleaving of differing water masses occurs in Region III, making it particularly sensitive to large-scale, long-time-scale environmental perturbations such as ENSO (Hickey 1979; Lynn & Simpson 1987,1990; Tsuchiya 1980). Seasonal cycles in zooplankton biomass are relatively weak (Roesler & Chelton 1987). Deep overwintering of calanoid copepods occurs (Alldredge et al. 1984), but it may involve only part of a population while another part grows and reproduces year-round (Mullin & Brooks 1967). The boundary between Region II and III is a biogeographic boundary for some species of nearshore benthic marine invertebrates and pelagic fishes. Region III is the preferred spawning site for over 90% of the epipelagic fish biomass (hake, sardine, anchovy) in the southern part of the CCS.

In addition to these latitudinal patterns, strong cross-shore variations occur in the CCS. For example, the wind field has strong cross-shore gradients at most latitudes, with maximum winds occurring seaward of the continental shelf (Nelson 1977). Vertically integrated primary production rates tend to decrease in the cross-shore direction (P.E. Smith, pers. comm.; F. Chavez pers. comm; Perry et al. 1989). A zone of maximum variability in dynamic height begins approximately 200-300 km offshore (Lynn & Simpson 1987); this zone has been called an eddy alley. The long-term maximum in macrozooplankton biomass occurs offshore in some areas (Roesler & Chelton 1987). This maximum is sometimes dominated by gelatinous zooplankton such as salps and doliolids (Berner 1967).

CRITICAL EASTERN BOUNDARY CURRENT PROCESSES AND BIOLOGICAL LINKAGES

We identify key physical-biological linkages that should be addressed in the GLOBEC eastern boundary current program. Each linkage focuses on physical processes thought to significantly affect population growth rates of metazoans having a planktonic stage. The relative importance of these processes typically differs in the three regions of the CCS. The regional differences may be viewed as a "natural experiment," permitting the relative impact of different processes to be quantified and perhaps used as a basis for projecting responses of marine populations to different scenarios of global change. We also pose a few preliminary hypotheses; we expect these to be modified and others to be put forward in future discussions.

Mesoscale Eddies, Jets, and Meanders
Rationale: Physical features such as eddies, jets, and meanders of coastal currents can be highly energetic and variable over time. Planktonic organisms entrained within these features often experience increased offshore (or onshore) transport, as well as different regimes of food and predators. Behavioral adaptations to such features (e.g., diel vertical migration interacting with current shear) can increase the residence time of organisms in the coastal zone.

H1: Nearshore eddies, jets, and current meanders are significant dispersal mechanisms for coastal populations.

H2: Frontal zones associated with these mesoscale features are sites of enhanced production and concentration of planktonic prey.

H3: Offshore mesoscale eddies are retention sites that reduce spatial losses and enhance population growth rates of some planktonic populations.

Timing, Duration, and Intensity of Coastal Upwelling
Rationale: Coastal upwelling influences planktonic organisms through a number of mechanisms. These include: (1) food-web effects (mediated by the input of "new" nutrients and elevated primary production); (2) seeding of epipelagic populations from dormant stages in deeper water; and (3) offshore transport of organisms entrained in upwelling filaments and jets. We emphasize the first two mechanisms here.

H1: Seeding from dormant stages is more important to population growth in Region II, where offshore transport is more frequent, than in Region I, where transport is more frequently alongshore.

H2: A significant fraction of the primary and secondary production in Region II is advected off the shelf and is unavailable to pelagic fishes that normally inhabit the continental shelf and slope (e.g., Pacific hake and northern anchovy).

Vertical Mixing Events
Rationale: (1) The shape and depth of the pycnocline affects the supply of new nutrients into the euphotic zone. This alters the magnitude of primary production as well as food-web structure (i.e., the size and species composition of phytoplankton and microzooplankton). Vertical mixing can change the pycnocline topography and thus food-web structure. (2) Lasker's "stable ocean hypothesis" suggests that "turbulent" mixing can erode the microscale aggregations of microplankton that are essential for the first-feeding success of some larval fishes. The intensity and frequency of storms can alter the availability of prey to a variety of different zooplankters. (3) Turbulence can affect the encounter rates of prey and predators. Recent studies suggest that encounter rates increase with turbulent kinetic energy dissipation (Yamazaki et al. 1991).

H1: There is an optimal wind speed that maximizes primary and hence secondary production. The optimal speed is about 7-8 m s-l.

H2: Survival is greatest for fish larvae that hatch during calm conditions.

Alongshore Advection: Long Time Scales
Rationale: Biogeographic boundaries exist within the CCS, e.g., at Point Conception (the boundary of Regions II and III) and sometimes at the boundary between Regions I and II (Brinton 1962). With alterations in the large-scale circulation of the CCS, these biogeographic boundaries may shift, thereby affecting the survival patterns and spawning success of indigenous populations. Circulation changes may also introduce faunal elements from different biogeographic provinces (e.g., the subarctic Pacific or tropical waters; Pearcy and Schoener 1987). Benthic populations dependent upon particular substrates or sediment characteristics may change abruptly and nonlinearly if the required substrates are discontinuous along the coast and animals are displaced considerable distances alongshore.

H1: Large-scale advection from the north alters the species composition and secondary production of CCS zooplankton assemblages.

H2: Large-scale advection from the south during El Niño/Southern Oscillation alters the species composition and secondary production of CCS zooplankton assemblages.

Alongshore Advection: Shorter Time Scales
Rationale: "Spatial losses" of organisms from a desirable habitat or spawning grounds can result in marked depression of recruitment success (e.g., Bailey et al. 1982). Conversely, alongshore advection can also serve as an essential mechanism by allowing planktonic organisms to return upstream to favorable spawning grounds or habitats. Some aspects of alongshore advection are relatively predictable (e.g., seasonal reversal of winds and currents in Region I; the seasonal in Region I but not in Region III.

H1: Some pelagic species need the poleward-flowing California Undercurrent and Davidson Current to complete their life history.

Each of these processes may fluctuate on a variety of time scales. For example, the frequency or intensity of upwelling-favorable winds may change within a single season, as well as over decades (Bakun 1990).

RESEARCH STRATEGIES

We recommend that two research strategies be pursued in parallel. The first is to study a long-lived, dominant species throughout its life history as it migrates through the different regions of the CCS. This might be characterized as a Lagrangian mode of study. The study site will shift with the ontogeny of the organism to investigate the processes leading to spawning success, larval and juvenile survival, latitudinal migrations, and adult feeding success. A particularly good candidate for this approach is the Pacific hake. Hake spawn in a geographically restricted area in the west of Region III, yet eventually migrate northward to Region I, where they are found as adults (Fig. 6). They may be seen as "integrators" of different processes in different regions of the CCS and therefore particularly sensitive to changes in ocean circulation and food-web structure. Preliminary evidence suggests that their spawning grounds moved northward in the late 1970s (inset, Fig. 6). The timing of this shift corresponds to a warming period in the ocean (A. MacCall, pers. comm.). Thus Pacific hake may be particularly sensitive to global change.

hake migration

Figure 6. Migration of the Pacific hake, Merluccius productus, from Bailey et al. 1982; interpretation of timing modified by D. M. Ware. Inset illustrates the geographic shift in spawning area from 12975 to 1978 (P. E. Smith, pers. comm.). The northward displacement of spawing occurred at a time of warming in the California Current system (A. D. MacCall, pers. comm.).

The second research strategy involves selecting species (or sibling species) that occur broadly throughout the CCS. It is hypothesized that the same species are governed by different processes in different regions. This strategy might be viewed as an Eulerian mode of study. In the different regions occupied by "metapopulations," or subpopulations, of the species, the differing effects of processes such as offshore transport, food limitation, vertical mixing, or large-scale advection can be quantified.

A simplified description of the population growth rate for a metapopulation within each of the three regions of the CCS can be expressed as follows:

dNR / dt = birthR - deathR - horizontal migrationR - advectionR + diffusionR

where the subscript R designates the region of interest. The terms in the equation are each rather complex and nonlinear functions of other processes and will vary within (as well as between) regions. Nevertheless, in this Eulerian approach, strong regional contrasts in the importance of these terms should make it possible to identify the most significant population control mechanisms. For example, for cyprid larvae of barnacles, the advection term may predominate in Region II, and the death rate term may predominate in Region I.

TARGET SPECIES

Fish

Approximately 90% of the epipelagic fish biomass in the southern part of the California Current spawns in the southern California Bight and associated offshore waters (D. Ware, pers. comm.). The three primary species are Pacific hake, northern anchovy, and sardine (Ware and McFarlane 1989). Because hake constitute 50-60% of all epipelagic fish biomass, they are a particularly suitable target for this study, in addition to the reasons identified above. Recent evidence for recovery of the sardine (P. Smith, pers. comm.), combined with extensive historical information on the sardine and northern anchovy, suggests that they are also excellent candidates for study. Pacific hake and anchovy have spawning subpopulations in Region I as well as Region III and thus are good candidates for the Eulerian approach.

Although the spawning regions of the Pacific hake and northern anchovy overlap broadly, the two species show markedly different interannual variations in recruitment (Fig. 7). Hake have occasional very strong year classes, while northern anchovy tend to have runs of weaker or stronger year classes. This difference is a likely topic for GLOBEC studies.

recruitment

Figure 7. Comparative recruitment time series of Pacific hake (Merluccius productus; from Hollowed and Bailey 1989) and northern anchovy (Engraulis mordax; from Jacobson and Lo 1989).

Holozooplankton

Target species of zooplankton include representative copepods, euphausiids, and a salp or doliolid. Some species are found in different regions of the CCS, exhibiting different life-history traits such as overwintering strategies and vertical migration in different parts of their ranges. Examples include the assemblage Calanus pacificus oceanicus, C. pacificus californicus, and Calanus marshallae. A copepod genus whose life history differs markedly from that of the genus Calanus, yet is also widely distributed in the CCS is the cyclopoid copepod Oithona. Oithona might be considered a "steady-state" genus in contrast to the "opportunistic" characteristics of Calanus. Eucalanus bungii and E. californicus are also likely to offer interesting contrasts.

With respect to euphausiids, Euphausia pacifica is distributed throughout the CCS, from the Gulf of Alaska to Baja California (Brinton 1962). It is the dominant species at many localities and makes an excellent candidate for contrasting studies in different regions of the CCS. The northern metapopulations of E. pacifica have shorter growing seasons and greater age and size at maturity than the southern metapopulations. Other potential candidates include Thysanoessa spinifera and congeners. Since euphausiids and copepods are the dominant prey for the target fish species identified above, we expect to advance understanding of the coupling between physical processes, zooplankton production, and fish recruitment.

Salps, and to a lesser extent doliolids, have extraordinarily rapid growth rates and colonizing abilities. Historical evidence suggests that they may predominate in some regions of the CCS (Berner 1967). It is also known that major ENSO events affect the total thaliacean (salp, doliolid, pyrosome) biomass much more than the copepod and euphausiid biomass (Smith 1985). Both observations suggest that a salp or doliolid species should also be a focus of study. Experiments should be designed expressly to understand the contrasting processes that select for either thaliacean or crustacean dominance, and the resulting consequences for the pelagic ecosystem.

Benthos

Among inhabitants of the hard-bottom benthos, the Dungeness crab and barnacles of the genera Balanus or Cthamalus have the best historical data bases and are likely candidates for further study. Evidence from some studies suggests that adult-adult interactions on the bottom may generally be more important for controlling community structure in Region I (e.g., Paine 1974), while cross-shore transport of pelagic larvae and migration of upwelling fronts may more significantly govern population growth in Region II (Roughgarden et al. 1988).