WORKING GROUP REPORTS

Special Tools

Technological Needs for Eastern Boundary Current Experiments

Cochairs: J. Jaffe and M. Mullin

Participants: K. Bailey, A. Bucklin, F. Chavez, D. Hedgecock, B. Hickey, V. Holliday, R. Iturriaga, K. Parker, and L. Walstad

A recent GLOBEC report on acoustical and optical technology (Holliday et al. 1991) detailed the needs common to all GLOBEC study sites. In this report we discuss the methodologies specific to an eastern boundary current program, including data processing, storage, retrieval, and sharing by the community, acoustic and optical techniques; and molecular biological techniques.

ACOUSTICS

Two approaches discussed by Holliday et al. (1991) and currently supported by GLOBEC are particularly important to eastern boundary current studies. D. Van Holliday is developing multifrequency acoustics to map size classes at a single location as a function of depth. His instrument, which can be used in either a moored or drifter mode, contains a string of eight dual-frequency sonars at 165 kHz and 1.1 MHz that will be deployed vertically in the water column. This distribution of frequencies allows the system to map the densities of centimeter-sized and millimeter-sized organisms. In addition, a dual beam system capable of judging target strengths will be deployed. Finally, an eight-frequency system will measure biomass and size distribution. The system uses real-time two-way telemetry.

Another system being developed under GLOBEC by J. Jaffe at the Scripps Institution of Oceanography's Marine Physical Laboratory (MPL) is a three-dimensional acoustic imaging system. Other systems are being developed at Woods Hole Oceanographic Institute (T. Stanton) and at the Applied Physics Lab at the University of Washington (E. Belcher). These systems can rapidly estimate biomass in a three-dimensional volume. The MPL system, operating at dual frequencies of 420 kHz and 1 MHz, is being designed to track individual animals in three dimensions and to provide a synoptic view of small-scale phenomena such as patch morphology and evolution.

The group concluded that an acoustic device, towed from a ship, that would permit mapping of zooplankton throughout the water column was especially desirable for the eastern boundary current - an area of changing and evolving oceanographic features. The design of such a device would have to account for the traditional trade-offs between range as a function of frequency (i.e., higher frequencies=shorter ranges) and the kinds and sizes of animals to be studied. The potential also exists for defining a three-dimensional volume at a fixed locality and surveying the volume repeatedly over time to observe the time-varying evolution of structures within the volume.

Group members also noted that because relatively high frequencies were needed to obtain a back-scattered signal from zooplankton, any correlation with mesoscale features would have to be obtained by a towed acoustical device. Maximal ranges for such devices to see the largest animals (euphausiids) would be no greater than 1 km.

OPTICS

The group supported the view of Holliday et al. (1991), who highlighted the synergistic relationship between acoustics and optics. Presently, optics is the only remote sensing technique that provides a unique identification of animal species and size. Unfortunately, optics are useful only for an extremely small volume of water - typically with dimensions of several centimeters. But an in situ optical imaging system which could identify animals that were being ensonified could provide ground truth for the acoustic system.

An intriguing possibility, briefly discussed, concerned the many advantages of airborne platforms - ease of deployment, short notice, and the potential for more control of the sampling area with higher resolution. Methods for resolving three-dimensional structure in the surface layers must be developed for such an approach. Futuristic approaches include stereo viewing from two platforms, and laser ranging, which uses time delay to judge distance.

DATA COLLECTING, PROCESSING, AND ARCHIVING

Clearly, a program encompassing a large variety of devices ranging from moorings to satellites will require diverse data formats. Assimilating these different types of data will present a challenge for the research community.

The need for real-time data was addressed. Both satellite and mooring data are important and could be used to help scientists select survey sites. Dates of a cruise usually become fixed far in advance of departure, but environmental information could be used to determine the exact location for a survey immediately before the cruise. Since the eastern boundary current (EBC) is an area of diverse, ever-changing, oceanographic features, a timely and opportunistic determination of survey locations would be highly advantageous. In addition, satellite information is needed at sea to guide the field studies, since site selection could be partly based on the satellite information. Subsurface events detected from moorings, if available in real time, could also guide fieldwork.

To facilitate the science, it is essential to consider the types of data and the storage techniques that will be used. As a preliminary step, the EBC community should determine whether generally accepted standards exist for storing and processing oceanographic data. Many institutions have already considered this issue. For example, both the University of Miami and the Jet Propulsion Lab have on-line data bases that can obtain data over a network. A recent research initiative funded by the Office of Naval Research deals with new methods for managing data and visualizing oceanographic data. Other research initiatives with similar requirements (such as JGOFS) may develop data-handling methods that could be adopted by the GLOBEC community. The National Center for Atmospheric Research (NCAR) may also have data-handling techniques that could be used. Finally, the 40-year CalCOFI data base should be considered to ensure compatibility.

The group discussed the advantages and disadvantages of central and distributed data storage. With the advent of high-speed computing networks, it is possible to store and retrieve relatively large amounts of data across networks. This would warrant a distributed network in which individual scientists are responsible for their own data, but allow access to other members of the community. This system has the advantage of allocating more funding to individual investigators but requires greater coordination. The creation of a central data facility would reduce the funds available for research but could have the advantage of a small staff to assist users, thus ensuring that the data would be available to both naive and sophisticated users, and that necessary preprocessing would be responsibly handled. Other schemes that combine both of these options are also possible.

WISH LIST

The following instrumentation needs, not met by existing or newly planned instrumentation, could be considered in a proposal request.

DEVELOPMENT AND USE OF BIOTECHNOLOGICAL TOOLS

Biotechnological tools offer a wide variety of techniques for analyzing specific classes of organic molecules, particularly those making up or closely controlled by an organism 's genetic identity. Many questions of central importance to GLOBEC can be addressed only with biotechnological tools, or can be answered more efficiently with such tools than by more classical methods, or can be answered on spatial or temporal scales at which classical methods are inadequate.

Such techniques can be used in at least two basic ways: (1) To identify species and subpopulations. GLOBEC is concerned with changes in abundance and distribution of marine populations; thus correctly and efficiently identifying such units is critical. A related use is in identifying body parts or other remains in guts of predators (to determine diets) or in sediments. (2) For proxy measures of physiological or reproductive states (e.g., age, sexual maturation) or of individual metabolic rates (e.g., ingestion, respiration) or demographic rates (e.g., natality). Knowledge of states can reveal responses to sublethal stresses, or facilitate the assignment of critical events, such as first reproduction, to a specific age. Metabolic rates are used to calculate material balances (income minus outgo of organic matter), which can lead to changes in populations, and demographic rates permit the direct calculation of such changes.

The above measurements are needed in all GLOBEC regions of study, and are not specific to eastern boundary currents. But identifying organic remains in anoxic sediments is more critical in the California Current system than in regions without a usable sedimentary record. Particularly important are techniques robust and sensitive enough to be applied to small amounts of material (micrograms) and to historical collections preserved for other purposes (e.g., in Formalin or alcohol), because past environmental events could be analyzed, and present collections of material simplified.

A wide range of biotechnical tools exists, and many more tools are being developed (though generally for nonoceanographic purposes); the field is evolving rapidly. In choosing techniques, researchers must recognize that the number of samples to be analyzed for any question on the GLOBEC scale will range from at least hundreds to many thousands, so the cost in time and money per analysis must be fairly low. Some present techniques may meet all the requirements, most notably analysis of mitochondrial and nuclear DNA amplified by the polymerase chain reaction to determine genetic identity. More tools must be developed, and existing ones need to be further calibrated, for different organisms, so that interpretations are more exact. but several techniques are ready to be applied to GLOBEC studies.

GLOBEC issued a request for proposals (RFP) in biotechnology and funded two projects for assessing metabolic health. The RFP, in the opinion of the committee, correctly and fully identified GLOBEC's needs. The group recommends that, rather than designing new criteria, GLOBEC should rerelease the same basic RFP, backed up with funds to support additional biotechnological projects.

POLICY ISSUES

The group discussed several policy issues that affect the applications of technology in GLOBEC. The way proposals for sea-going research are prepared and reviewed in NSF has a potentially stifling influence on the development and wide acquisition of technological tools. The cost of ship time is not a line item in the budget of a research proposal, but the cost of a new tool usually is. Therefore some solutions that would be cost-effective in substituting tools for ship time, or in making use of ship time already funded elsewhere, may not be proposed because of the way they affect the budget of an individual proposal. This situation has been discussed many times, within and outside of NSF, and mechanisms exist for developing and purchasing expensive equipment, but even if the problem is only one of perception, it must he considered.

The issue of using ships of opportunity (or other platforms, such as drilling rigs) within GLOBEC was not discussed specifically, although one of the major time series in biological oceanography - the continuous plankton recorder survey of the eastern North Atlantic - was constructed for such use. Proposals exist to greatly expand this approach to monitor the biological marine environment.

A related question is whether the diverse (and diffuse) needs of biologists, their technological innocence, or their relatively modest per capita funding has prevented the formation of a group large enough to support an expensive communal instrument. Some technologies (acoustic Doppler current profilers and bottom swath-mapping seabeams) have become institutionally purchased equipment, supplied with the ship an investigator uses. Some of the more elaborate tools recommended in this report may have to be managed and financed in this way.

The Role of Models in the Study of Eastern Boundary Current Systems

Chair: L. Walstad

Participants: B. Hickey, B. Jones, P. Smith, and D. Ware

We recommend that a suite of ecosystem models for the California Current system be developed and that studies of these model systems begin. A number of considerations affecting the form and function of these models are outlined here, and a subset of needed studies is discussed.

The long-term goal is to understand and predict the effect of climate change upon eastern boundary current marine ecosystems. To reach this goal, the initial emphasis should be on understanding the dynamics of model ecosystems rather than on re-creating oceanic ecosystems. Models must include ecosystems as well as specific biological and physical interactions. Coordination between model developers should be required to ensure that components are interchangeable where appropriate. This should lead to more robust models and substantially increase the probability of attaining our long-term goal. As our understanding improves, the models can be modified to more closely reproduce the marine ecosystem.

ECOSYSTEM EQUATIONS: CHOICE OF VARIABLES

Although an ecosystem comprises individual organisms, description of a full system at the level of the individual is not expected to be feasible in the near future. Rather, an aggregate measure of the population will be required. Historically, models have used nitrogen or carbon as the quantitative measure of subpopulations within the ecosystem. Promising alternatives include the use of more than one quantitative measure and the use of different measures at separate trophic levels. It should be emphasized that the variables in such ecosystems are amalgamations of similar species and that the particular quantitative measures are simply a means of counting the organisms. Multiple measures may be useful in situations where the C/N ratio varies or where the behavior or response of organisms cannot be related to the biomass alone. Fundamental to the strategy is the premise that the importance of a unit of carbon will depend upon where it is located in the ecosystem. Within trophic levels, subdivisions by body size or functional group may be necessary or desirable.

Table 2 lists potential biological variables for a model of upper-water-column interactions within the California Current system. The phytoplankton trophic level forms the base of the food chain and, through interaction with the water column's physical structure and nutrients sets the overall input of organic matter. A primary focus of GLOBEC research will be on secondary production and zooplankton population dynamics. To understand the variations of important individual zooplankton taxa it is necessary to include them explicitly. We suggest at least four groups representing major but distinctive components of the California Current zooplankton biomass.

Fish populations will also be studied in GLOBEC. The "fish" subdivisions in Table 2 were chosen for their societal importance and because each exemplifies a different migratory or reproductive strategy, or behavior, or feeding pattern. In particular, hake, anchovies, and sardines make up a substantial fraction of the fish biomass in the California Current system and are representative of closely related species that are important in all major eastern boundary current systems. Hake also feed preferentially upon euphausiids and consume much of the productivity of the euphausiid community. Recruitment success in hake and sardine is affected by the availability of suitable spawning habitat and small zooplankton, both of which are affected by upwelling. In addition, the survival rate of some fish larvae may depend upon the availability of specific phytoplankton taxa, so subdivisions of the phytoplankton trophic level may be needed.

Even this relatively complicated trophic and taxonomic subdivision of the population may not be sufficient; in some cases specific age classes or identifiable genetic pools within fish and zooplankton species may be needed. Also, the zooplankton community may have to be further divided according to body size or reproductive pattern. Meroplanktonic larvae and their benthic adult stages may also have to be included for some GLOBEC research objectives.

Rates of the fundamental life processes are also needed to complete the model. For each variable, uptake (as fixation or consumption of organic matter), excretion, recruitment, death rates, and (in some cases) migratory and reproductive strategies must be provided.

Table 2. An Example of Trophic Subdivisions

	Fish*	   	   Zooplankton		Phytoplankton	 Nutrients
	Hake (8)	   Euphausiids		Dinoflagellates	 Nitrate
	Anchovies (3)	   Copepods		Diatoms		 Carbon
	Sardines (1-2)	   Salps
	Mackerel (1-2)	   Microzooplankton

*Numbers in parentheses indicate the potential number of year classes.

REQUIREMENTS FOR NEW OR ADDITIONAL MODEL INPUT

Extensive research has already been carried out for most of the suggested biomass pools and some of the required rates (e.g., Francis 1983; Livingston 1983; Tanasichuk et al. 1991), as have model studies of pieces of this or similar ecosystems (Walsh 1975; Wroblewski 1977, 1980, 1982). Similar field, laboratory, and modeling studies will be needed to identify the remaining rates and the dependence of these rates on the physical and biological environment. Identifying how key rates and strategies depend on physical variables (e.g., stratification, temperature, turbulence) that may vary with climate change is particularly important for GLOBEC.

Horizontal and vertical migration at the higher trophic levels pose an especially difficult modeling task. Fish, which migrate a significant horizontal distance and appear to select spawning sites, must be represented by models that include migratory and spawning parameterizations. These parameterizations will contribute to the horizontal redistribution of fish through specification of swimming rate and direction. Identification of spawning regions is also important because these regions represent sources for larval stages and thereby determine the distribution of juvenile fish.

Physical models are needed to provide three-dimensional fields of temperature, salinity, velocity, and turbulence. Fish eggs and larvae and the lower trophic levels will be advected through these fields, and uptake, excretion, and survival rates will vary, partly as a function of local physical conditions. Larger fish tend to be more independent of these fields, except for spawning, which is closely coupled to upper-ocean temperature distribution. But larger fish may be indirectly coupled through their food supply. For example, the spatial and trophic association of hake with the euphausiid stocks that are abundant along the upwelling front suggests that hake populations may be influenced by the physical dynamics of the upwelling front. A goal of the physical ocean models should be to reproduce the gross features of the upwelling front, including seasonal and alongshore distribution of upwelling, the local upwelling rate, and the persistence of the front.

Daily cycles of atmospheric forcing, and the ocean's response to this forcing may be important. These cycles include intensification of alongshore winds off northern California, and set-up of sea breeze. Alongshore winds are critical for correctly reproducing the upwelling. The set-up of the sea breeze and resulting increase in turbulent forcing of the upper ocean may significantly affect the plankton community by increasing the turnover rate within the mixed layer or deepening the mixed layer at the time of day when photo-inhibition would affect phytoplankton confined to the upper few meters. Although physical models are improving in their capacity to reproduce mixed-layer cycling, depth, and turbulent intensity, a continuing effort will be required.

Sensitivity studies and scaling of proposed model systems should begin as soon as possible. Sensitivity studies may be used to help rank the importance of components of the observational plan, but only if sufficient lead time is provided. Studies should address the form of uptake rates at each trophic level and should characterize the system's sensitivity to changes in total biomass or increased interannual variability. Because this is a nonlinear system, chaotic behavior should be investigated with the predictability of the system in mind. It is likely that some aspect of the system will be sensitive to initial or boundary conditions or to parameter choices; however, some quantities may prove to be well predicted. Quantities of interest - zooplankton biomass, for example - should be determined and examined for predictability. Scaling of the system of equations may reveal fundamental balances that can be used to define the system's basic dynamics. The physical oceanographic analogues of this process include the derivation of quasi-geostrophy, Ekman balance, and wind-driven and thermohaline circulation theories. These early studies will necessarily be simple, but the experience gained will lead to better understanding in the future and may help improve the field program. In several areas (listed below) we recommend, based on present information, focused sampling and special-purpose submodels of particular ecosystem features or components.

Biological patches - both their biological consequences and the appropriate methods for including them in ecosystem models - are not adequately understood. Theoretically, the importance of spatial heterogeneity is attributed to nonlinear coupling of ecosystem components. Field sampling on fine horizontal and vertical scales will be important to understanding patch phenomena. Also, model studies of small regions on fine horizontal scales may help determine how to include the effects of patches in models with coarser spatial resolution.

Because the ecosystem comprises individuals, the most direct means for reproducing ecosystem behavior would seem to be through modeling of individuals. Current laboratory studies should yield improved quantitative descriptions of individual behavior (e.g., predator-prey interaction). Studies of small ocean regions with models that can track the position of 'individuals" (Hofmann et al. 1991) might improve parameterizations in biomass models focused on larger scales. Improvement and application of these individual models should be encouraged, especially when the emphasis is on improving biomass models.

Various kinds of historical data provide a basis for validating model response to a wide range of climatic forcing. The paleo-oceanographic record in the California Current system is the subject of a separate working group. But we note here that individuals of long lived-species (e.g., Sebastes spp.) can provide up to a 100-year record of their growth pattern. Many of these species show considerable territorial fidelity and thereby provide a proxy record of local environmental conditions.

SUMMARY

An eastern boundary current model suite is needed to further our understanding of the marine ecosystem. The long-term goal is a model that reproduces the behavior of the ecosystem, including the response to climate change. An appropriate set of equations has not yet been put forth, but considerable field and laboratory and model-subsystem studies have been conducted. A preliminary suite of models should be developed and adopted, and sensitivity experiments should be conducted. This is an ambitious goal, but the process of developing ecosystems models is likely to identify needed studies as well as to focus some aspects of the field program. Because of the system's complexity, such a model is expected to comprise multiple trophic levels with subdivisions in each level. The initial objective should be to understand the dynamics of this model system rather than to reproduce the marine ecosystem. As field and laboratory studies develop, our understanding of the ecosystem will be corrected at the process level. All components - including the physical model, patch dynamics, physiological rates, migration pattems, and spawning patterns - will need considerable development.