In this section we summarize the recommendations of the GLOBEC meeting on Marine Animal Populations and Climate Change in the Southern Ocean, distilled from the Working Group Reports which are presented in their entirety in Section 7. These recommendations are intended to serve as guidelines for specific GLOBEC studies.

For each taxonomic group - zooplankton, benthos and top predators - the recommendations are organized according to Site Selection criteria, with the exception that criteria relating to climate change are first discussed in general as applicable to all populations considered, and aspects of international, inter-program and inter-agency activity are discussed in a separate section following this one (Section 3). An outline of the suggested logistics of the field study is presented in Section 4.

2.1 Relation to Climate Change

The Southern Ocean is in many respects an ideal region in which to study marine animal populations in the context of global climate change. Meteorological dynamics are likely to impact coastal zones, areas covered by sea ice, and may affect the large-scale circulation. Present ocean-atmosphere models of the earth suggest that the Southern Ocean may be the last part of the ocean to experience warming, but other effects of global atmospheric warming could take precedence.

Coastal ocean regions are believed to be the places in which climate change would most influence marine animal populations, primarily through changes in meltwater input and solar radiation. Meltwater input would be induced by melting of the polar icecap, and solar radiation is expected to decrease over the ocean. Either of these processes will affect water column stability, leading to changes in vertical mixing and the primary productivity which fuels higher trophic levels. Animal populations are most concentrated in coastal regions.

Sea ice covers roughly half of the Southern Ocean during winter and approximately 10% during the summer. The annual cycle of accretion and melting contributes significantly to primary production, again by altering water column stability. Furthermore, the ice edge is a region where marine animals congregate in large numbers. Long term effects of global warming are expected to reduce the seasonality of sea ice and could result in the eventual absence of summer ice. These effects are likely to greatly reduce the productivity and habitat for marine animal populations.

Changes in the flow intensity of circumpolar current may also be anticipated, brought on by a reduced temperature contrast between the equator and the poles which would reduce the strength of prevailing circumcontinental westerly winds. It is clear that circumpolar current interacts with ocean bathymetry to yield areas of high primary production, but it is not clear how global climate change would impact overall production in the circumpolar region.

The key recommendations identified with respect to climate change are:

  1. To organize and analyze existing historical data to compensate for the lack of long term observations;
  2. To make consistent and synoptic observations of sea ice and currents in the Southern Ocean, to optimize the ability to detect the effects of a climate change;
  3. To improve the understanding of how meteorological conditions drive variability in sea ice extent; and
  4. To improve observations of coastal circulation, which determines the distribution of marine animal populations.

2.1.1 Approach

The general approach is to undertake studies which will address the role that climate plays in determining local and regional episodic events, mass transport, and total energy of the marine system. Such studies should employ satellites, moored instruments, and drifters. Coupled with a better understanding of how physical mechanisms affect marine populations, this approach will lead to the basis for predicting how climate change will affect population dynamics.

2.2 Zooplankton, Including Krill

2.2.1 Target species

Krill (Euphausia superba) has clear economic and ecological importance, and is suggested as the primary target species. Other species of primary interest include Salpa thompsoni, which can be especially dominant but about which little is known, Euphausia crystallorophias, a coastal and high Antarctic species, and two abundant copepod species, Calanoides acutus and Calanus propinquus. These species together represent the spectrum of different life strategies and the bulk of the zooplankton biomass in the Southern Ocean. Other species of interest would include Themisto gaudichaudi, Metridia gerlachei, Rhincalanus gigas, Thysanoessa macrura, and Sagitta gazellae.

2.2.2 Definable populations

The Bellingshausen Sea is a region where relatively discrete populations of krill and other holozooplankton might be maintained, by virtue of a regional gyral circulation, which may restrict communication with adjacent seas. An area directly west of the Ross Sea, bounded by the continent to the south, 65S, and 140 to 160E, has supported a consistent krill fishery for some time, and may also possess populations definable in space and time. The details of regional circulation in both areas are poorly known, and will require study. However, geostrophic circulation patterns presented in Stein (in press) suggest the presence of two gyres in this region that partly overlay the continental shelf (Figure 1). These features may be persistent Key objectives of a field study would include:
  1. Sampling surveys of sufficient frequency to define the temporal and spatial extent of discrete populations, and developmental cohorts;
  2. The use of biochemical and genetic marker techniques to clearly identify populations.

2.2.3 Population dynamics

The primary goal of population dynamics studies on krill and other zooplankton is to better define demographic parameters, particularly in the context of regional circulation. Such studies will require year-round sampling, with particular emphasis on the role of sea ice in structuring the community. The following particular studies are indicated:
  1. Much more data are required on populations in the winter, especially on the role of demographic parameters of populations in determining the size of populations during the productive summer season;
  2. Identification and study of those demographic parameters which may be especially sensitive to climate change, and to temperature increases in particular.

2.2.4 Focus on process and mechanisms

Process studies would be carried out on cruises designed specifically for that purpose, as well as at the numerous shore-based laboratories in the Antarctic Peninsula region. Particular attention should be paid to measuring rates of metabolism, egg production, feeding, growth and development, as well as investigating the diapause phenomenon. Research might include the following studies:
  1. Determination of the environmental triggers for metabolic and behavioral events, noting that very small changes (i.e., 0.5C) may trigger change;
  2. Comparison of metabolic responses between extremes in environment (e.g. summer vs. winter);
  3. Measurement of physiological responses to conditions outside the normal environmental range;
  4. Determination of the relative sensitivity of various developmental stages to environmental variables, to understand which stages are most vulnerable.

2.2.5 Historical database

Relatively little information exists on plankton distributions in the Bellingshausen Sea, although the nearby waters of the Antarctic Peninsula region are perhaps the best studied in all the Southern Ocean. This is particularly important because waters from the Bellingshausen Sea provide some of the flow through the northern reaches of the Antarctic Peninsula coastal region, and thus the fauna of the Bellingshausen are already reasonably well known. The BIOMASS data base, centered at the British Antarctic Survey in Cambridge UK, may prove a valuable resource.

2.2.6 Modeling

Specific modeling studies recommended include:
  1. Design of models to investigate life cycles of zooplankton, with particular emphasis on determining the results of different life-history strategies (e.g. seasonally migrating vs. non-migrating species);
  2. Development of coupled biological-physical numerical models for krill and other zooplankton populations at the study site, with particular emphasis on interactions with regional scale circulation, and with finer-scale resolution, especially in the vertical;
  3. Development of models regarding the formation, maintenance and dissolution of patches, with particular emphasis on krill.

2.2.7 Technology

Certain developments in technology will be applicable to all taxonomic categories of interest, particularly in the case of field sampling instruments. Those of special interest to zooplankton and krill studies would include:
  1. Improvement of instrumentation needed to sample the upper 10 rn of the water column as well as under sea ice, where current instrumentation is inadequate;
  2. Improvement of large volume sampling techniques to determine the abundance, biomass and distribution of salps with minimal disturbance to aggregates;
  3. Development of non-invasive techniques to observe distributions of krill and other zooplankton in both ice-covered and ice-free areas.

2.2.8 References

Stein, M. 1991. Variability of local upwelling off the Antarctic Peninsula, 19861990. Archiv far Fischwiss. (In Press)

2.3 Benthos

2.3.1 Target species

Five characteristics were considered as criteria for the selection of benthie species, primarily that the species: (1) have measurable growth parameters, (2) be abundant, (3) have either a wide or restricted distribution, (4) have a known life history, and (5) be amenable to reproductive studies. Given these constraints, the following species are particularly recommended (p and b denote pelagic and benthie larval forms, respectively):
  1. Bivalves: Adamussium (p), Laternula (p), Mysella (b), Gamardia (b)
  2. Echinoderms: Odontaster (p), Sterechinus (p), Ophionotus (p), D iplasteria (b)
  3. Crustaceans: Notocrangon (p), Chorismus (p), Glyptonotus (p)

2.3.2 Definable populations

Populations that would be definable in time and space would be likely to occur in the following areas, distributed from high to low Antarctic:

High Antarctic:

  1. Ross Sea/McMurdo Sound
  2. Southeast Weddell Sea Davis Sea
Low Antarctic:
  1. South Orkney/South Shetland Islands
  2. Antarctic Peninsula/Bellingshausen Sea
In particular, genetic studies would be desirable for distinguishing between populations.

2.3.3 Population dynamics

Measurements relative to the population dynamics of benthic species (e.g. recruitment, life history strategies) should be done in conjunction with measurements on physical processes. Particular studies recommended include:
  1. Colonization processes in areas exposed by recent calving of major portions of ice shelf;
  2. Species succession in areas with high iceberg grounding frequency;
  3. Studies that emphasize observations during winter, a period for which little is known.

2.3.4 Focus on process and mechanisms

Studies should be conducted to understand how fundamental parameters of population dynamics, such as growth, reproduction, larval dispersal, behavior, settlement and survival vary directly and indirectly as a function of physical and biological forcing. Particular processes or parameters which should be studied with reference to potential global change include:
  1. Processes delivering carbon to the benthos via vertical flux of particulate matter;
  2. Ice conditions in the overlying water,
  3. Flow of local currents;
  4. Temperature and salinity;
  5. Light regimes; and
  6. Redox profiles in sediments.
Measurements of the response of individuals and populations should be assessed with particular regard to:
  1. Energy flow;
  2. Physiological response, which would provide information on rates and processes;
  3. Population dynamics;
  4. Community structure, which would assess the effects of environmental change on species composition, abundance and biomass.

2.3.5 Historical database

A large body of data exists on benthic communities near a number of Antarctic field stations. Some effort should be made to gather these data and make them available at an accessible central location.

2.3.6 Modeling

Modeling studies are encouraged which
  1. Evaluate the processes of aggregation, dispersal and settlement of meroplanktonic larvae;
  2. Assess the role of large-scale climatic changes on physiology and population dynamics of discrete populations.

2.3.7 Technology

Developments in technology are particularly required in the following areas:
  1. Quantitative assessment of distribution and abundance using video and camera technology;
  2. Methods for determining the age of individuals.

2.4 Top Predators

2.4.1 Target species

Target species are recommended among fishes, birds and mammals. For fish, these include
  1. Commercially harvested species
  2. Non-harvested holopelagic species
  3. Non-harvested nearshore species
Species in the first group occur primarily in the Atlantic sector and are already included in CCAMLR monitoring studies. The second group are species abundant in food webs of the high Antarctic and represent contrasting ecological and life history patterns. The third group contains species which are conveniently collected from shore stations. Among the birds, key species of interest include Adelie, chinstrap, macaroni and gentoo penguins, cape and Antarctic petrels, Black-broWed albatross, grey headed albatross and South Polar skua. It is noted that roughly 2/3 of the Southern Ocean bird biomass is comprised of Adelie penguins; other species are recommended for a variety of specialized reasons.

Target species recommended among the mammals are the crabeater seal and the Antarctic fur seal. Both are largely dependent on krill as a food resource, and occupy habitats analogous to those of Ade1ie and chinstrap penguins.

2.4.2 Definable populations

In the Atlantic sector of the Southern Ocean, which is generally recommended as a primary study region, it is generally felt that CCAMLR subareas represent reasonable approximations of the distribution of fish populations. Distributions of bird and seal populations may represent distinct populations, but studies are required to verify this assumption. Specific studies should include: (1) Better assessment of species distributions in time and space; and (2) Molecular techniques (e.g. mitochondrial DNA) applied to distinguish popuations.

2.4.3 Population dynamics

Some of the important demographic parameters for target populations can be acquired directly from the CCAMLR monitoring program. These would include data on spawning stock biomass, growth and reproduction of commercially taken fishes, and the growth rate, breeding success and cohort survival of birds and seals. Studies particttlarly encouraged under GLOBEC would include:
  1. Assessment of growth and developmental rates of larval fishes as related to biotic and physical environments;
  2. Foraging dynamics of birds and seals, with special emphasis on winter; and
  3. Marking and tracking studies on birds and seals which assist in identifying populations and observing behavior.

2.4.4 Focus on process and mechanisms

Certain key processes are expected to reveal the response of top predator populations to global change. This calls for studies focused on:
  1. Effects of temperature on growth and development of different ontogenetic stages of fishes;
  2. Overwintering studies of top predators to determine critical mortality periods;
  3. Potential effects of UV radiation on near-surface fish eggs and larvae;
  4. Effects of physical circulation on dispersal of early life stages of fishes;
  5. The importance of food availability on physiological condition and reproductive behavior, and
  6. Foraging dynamics of top predators in relation to prey abundance and aggregation behavior.

2.4.5 Historical database

Relatively sound historical databases have been collected through the CCAMLR and BIOMASS programs, as well as through various national programs in the US, UK, Germany, France, Australia, New Zealand and South Africa. Access to these should be made readily to principal investigators conducting studies in the GLOBEC framework.

2.4.6 Modeling

Specific studies recommended for modeling exercises include:
  1. Effects of the physical environment and fluid dynamics on food supply, growth and development rates, survivorship and dispersal of the early life history stages of fish;
  2. Development of a standard population dynamics model for seabirds, integrating physiological data with environmental variables;
  3. Models of movement and dispersal of foraging predators to determine how seabirds and seals locate food patches;
  4. Effects of climate and fishing pressure on harvested species;
  5. Trophodynamic models of multispecies interactions between fish, higher predators, and their prey.

2.4.7 Technology

Special technological advances which could greatly aid studies of top predators in the Southern Ocean include:
  1. Improved acoustical hardware and Software for locating, identifying and quantifying the abundance of fish;
  2. Underwater visual systems for assessing the distributions of prey items (e.g. krill and pelagic fishes);
  3. Improved satellite tracking and time-depth recording devices for predators;
  4. Improved techniques for remote sensing of sea ice, either by satellite or aircraft, which yield higher resolution and which better differentiate sea-ice conditions;
  5. Biochemical methods for evaluating fish condition factors;
  6. Genetic markers for determining stock identity; and
  7. Increased use of Lagrangian drifters to observe current transport and advective processes.

2.5 General Modeling Issues

Many of the issues relevant to modeling marine populations in the Antarctic have already been considered as part of the discussions for each taxonomic group. However, there are additional issues that are relevant to a GLOBEC program in the Southern Ocean.

First, it is now apparent that many of the components of the Antarctic food web are dependent on sea ice during some or all of their life history. The time scales of this dependence range from days (for phytoplankton) to years (for seals and marine birds) and extend over space scales of a few meters to 100s of kilometers. Thermodynamic models of sea ice, that describe the annual growth and melting of an uniform ice field, are reasonably well developed. Schemes for incorporating thermodynamic sea ice models into general circulation models exist. However, the existing sea-ice models are simple and do not include processes such as rafting of sea-ice, and the existing coupled ocean-ice models do not consider flow underneath the ice, which can be important for biological populations. Assuming that climate change effects in the Antarctic will be reflected in the variability and extent of sea-ice cover, then the development of realistic sea-ice models that can be interfaced with circulation and biological models is a critical area of research. Also models that incorporate the feedbacks between sea-ice cover and higher predators, that are decoupled from the flow field (e.g. penguins) need development.

Second, any GLOBEC initiative planned for the Southern Ocean will likely have a regional focus, i.e., Bellingshausen Sea. However, the circulation models developed for regional studies will need to include the effects of the larger scale circulation of the Southern Ocean. Thus, techniques for combining the results of large-scale circulation models (e.g. FRAM) with results from regional circulation models need development. A related problem is that the space and time scales resolved in physical models are often inappropriate for biological processes. In particular, the vertical resolution of circulation models is frequently not consistent with that needed to adequately model biological processes. Thus, methods for scaling between circulation and biological models need development.

Third, there is a need for the development of models that can simulate the aggregation behavior of animals such as krill. Considerable theory on modeling animal aggregation and swarming behavior has been developed for terrestrial systems. Efforts need to be made to transfer and adapt this theory for marine populations.