Cod Fishery Collapses and North Atlantic GLOBEC

by Michael Sinclair and Fred Page

The collapse of the groundfish fisheries in the northwest Atlantic has focused the attention of both management and science on critical gaps in knowledge. Reported landings of Atlantic cod from 1970 to 1992 are shown in Figure 1. There are differences between the patterns for the northeast and northwest Atlantic stocks. Off North America there was some rebuilding of the cod resource following extension of jurisdiction to 200 miles in 1977, and the removal of foreign fishing. However, cod abundance has declined sharply since the late 80s in most areas of its distributional range. For some cod management units, spawning stock biomasses are sufficiently low that fisheries have been closed since 1993.

The socio-economic impacts of the closures of these groundfish fisheries (particularly in Newfoundland, eastern Nova Scotia, and northern New Brunswick) are yet to be fully understood. There is no doubt, however, that the collapse of the groundfish fisheries is a crisis of historic proportions for Atlantic Canada. Perhaps comparison to the Highland Clearances of the late 1700s to early 1800s in Scotland, when crofters were displaced from their rented land to facilitate sheep farming by the property owners, is not inappropriate. Many of the displaced crofters, under great hardship, emigrated to the colonies (including Atlantic Canada). The displaced fishermen and processing plant workers of today, however, have no new frontier to move to. They are experiencing, at a very personal and emotional level, the tragedy associated with inadequate management of renewable resources.

Governments are now faced with the task of how to improve the management of these resources. As part of this, a number of fundamental questions involving population regulation of marine fish must be addressed. In this article, we pose several of these management questions, summarize the relevant conceptual framework in the ecological literature, and discuss how GLOBEC research in the North Atlantic may generate explanatory power and answers. For this latter part, research results from our modelling team are used to illustrate issues.

Four Management Questions

What criteria should be used to define and open a species/area management unit to fishing? An ecological issue relevant to this question is an understanding of the patterns of geographically distinct self-sustaining populations and their respective "minimum spawning stock" levels. Population patterns are defined by the spatial location and scale of birth site fidelity and we need to better understand the time scales of recovery of spawning components that have been fished close to commercial, and perhaps biological, extinction.

Does the management strategy need to include a multi-species ecosystem approach in order to meet the objectives of fisheries management? Those who argue in the affirmative infer that the abundance of groundfish population is regulated by food chain interactions, and that the impacts of fishing on one species affects predator/prey relationships of co-occurring commercially important species.

Does the trawling activity of the groundfish fishery diminish the benthic food supply to commercially important species? Those who infer that there is an impact of trawling on the sustainability of the groundfish fishery assume that groundfish populations are regulated by intra-specific competition for food, in this case at the juvenile and adult stages of the life cycle.

What management objectives are practical given the degree of environmental variability in the North Atlantic? The ecological issue includes the role of physical oceanographic processes on marine fish population regulation, and the degree to which ecosystem regime shifts occur. The practical issue concerns the degree to which we are able to monitor and regulate fishing in light of these population processes and shifts.

Conceptual Framework Underlying Regulation of Marine Populations

Responses to these management questions require scientific understanding of population regulation. This component of ecology was a hot area of debate in the 1950s and 1960s. Unfortunately, the debate produced little explanatory power concerning what factors regulate abundance, and consensus on the relative importance of density-dependent and density-independent factors is still lacking. Eventually ecologists moved on to other issues.

The International GLOBEC program provides an opportunity for renewed focus on population regulation of marine species, with a particular emphasis on cod for the North Atlantic GLOBEC/ICES component. For marine fish there is a rich conceptual literature, including a number of competing hypotheses on various aspects of population biology (Table 1). It is helpful to consider these hypotheses in the context of three aspects of population regulation: pattern, abundance, and variability (Sinclair 1988; Sinclair and Iles 1989). The hypotheses and their associated physical oceanographic processes are listed in Table 2.

The migration triangle hypothesis states that the geographic patterns of marine fish populations are established and maintained by residual currents linking spawning locations to juvenile nursery areas. In contrast, the member-vagrant hypothesis states that the patterns are maintained by areas that limit the dispersal and advection of eggs and larvae during the early part of the early life history (i.e., areas of retention of eggs and young larvae). The competing hypotheses identify different physical/biological coupling processes as being critical to the definition of spawning populations and as such are mutually exclusive.

The match-mismatch hypothesis states that mean population abundance is regulated in a density-dependent manner by the plankton food available along the drift route. At high population levels the larvae become relatively more food limited, and vice versa. The member-vagrant hypothesis states that mean abundance differences between populations of the same species are defined by the size of the physical oceanographic features that restrict dispersal of eggs and early stage larvae. Furthermore, it is argued that density-dependent vagrancy (i.e., an increase in loss rate at higher spawning stock levels, and vice versa) can regulate abundance without density dependent trophic processes. Again, the competing hypotheses identify different oceanographic processes.

Three of the four hypotheses which address temporal variability in the abundance of year-classes focus on food availability during the larval stage. The match-mismatch hypothesis states that the variable timing of the seasonal phytoplankton bloom in relation to a fixed period of spawning generates interannual differences in the match between the zooplankton production cycle and the period of fish larval feeding. The key physical oceanographic process is the seasonal development of vertical stratification in the water column that permits phytoplankton blooms to develop. The stable ocean hypothesis states that vertically stratified (i.e., low mixing) conditions are needed to generate high local concentrations of food at the pycnoclines for favourable larval survival rates. These concentrations, however, are broken down during strong wind events. Thus, the physical process of importance is the frequency and intensity of wind mixing during the larval feeding stage, with low winds being considered favourable for larval survival. The encounter rate hypothesis is almost the opposite of the stable ocean hypothesis; increased turbulence enhances the encounter rate between fish larvae and their prey. Thus, years of increased wind mixing and areas of strong tides should improve larval feeding success and generate relatively higher survival rates.

In the member-vagrant hypothesis both food chain and spatial displacement processes contribute to variable loss rates from the appropriate geographical area for the population. The two categories of processes can act in a density-dependent or a density-independent manner. If vagrancy is itself density dependent for a particular population, then there is no necessity for density dependent trophic limitation of abundance.

In sum, the five hypotheses identified above involve differences in oceanographic processes, physical/biological coupling mechanisms, and their characteristic time scales. Several of the hypotheses assume that larval feeding is the key process. Much of the fisheries research during the past two decades has been on the year-class variability aspect of population regulation. Prediction of the impacts of climate change are dependent upon which hypothesis or hypotheses best capture the realistic dynamics for a given population and time.

Georges Bank Modelling Study

In the North Atlantic component of U.S. GLOBEC there is an opportunity to generate consensus on which processes are critical to the three aspects of population regulation of cod, and thus to enhance our explanatory power and predictive capability. We will briefly describe the approach that our modelling team is taking and some of the results to date. Our approach is to develop a suite of physical and biological models using realistic geography, forcings, and boundary conditions for the Georges Bank cod and haddock populations that capture the processes implied under the various hypotheses. In this way, we hope to evaluate which of the hypotheses are more appropriate for explaining population regulation processes for this geographic area. The results to date are described in Lough et al. (1994), Lynch and Naimie (1993), Naimie et al. (1994), Lynch et al. (1992), Ridderinkhof and Loder (1994), Tremblay et al. (1994), Werner et al. (1993), and Werner et al. (1994). Here we discuss several of the biological results of significance to the conceptual literature on marine population regulation.

Werner et al. 1993 investigated the relative importance of circulation and behaviour on the distribution at age of eggs and larvae of cod and haddock that spawn on the northeast peak of Georges Bank. A 3-d flow field comprising the dominant M2 tidal current and the seasonal-mean circulation associated with tidal rectification, winter-spring wind stress, and Scotian Shelf inflow was used. Eggs released in the surface layer are rapidly advected off the bank, whereas a large proportion of eggs released at mid-depths persist on the bank for a couple of months. Losses of deeper releases are along shelf into the mid-Atlantic Bight, whereas surface releases are lost in the cross bank direction into slope water. The position of larvae on the southern flank of Georges Bank, both with respect to depth in the water column and cross-bank horizontal position, influences their subsequent fate (Figure 2). With the inter-annual differences in circulation on the bank, one would expect variable distributions on the southern flank, and thus variability in loss rate from the bank.

Simulations using the observed vertical distribution of eggs, and realistic vertical migration behaviour of larvae, result in considerable losses of eggs and larvae from the bank, and are not consistent with empirical observations on larval distributions older than about 60 days. An important behavioral characteristic that influences on-bank retention is the depth at which fish spawn (i.e., egg release depth). Vertical migration of the larvae does not influence loss rates. However, for the model results to reflect the on-bank displacement of older larvae that is observed from field studies, some horizontal swimming behaviour is required. With realistic swimming speeds, the distribution of larvae that swim in the on-bank direction is similar to field observations of two-month-old larvae.

Summary points for the first cod study are as follows:

We are presently examining the MARMAP data on cod and haddock egg and larval distributions to better define the location and time of spawning, and how spawning features change as a function of abundance. In addition, we are summarizing the composite distributions of eggs and larvae at age for the entire eleven years of data. The aim is to evaluate whether spawning is associated with periods and sites of minimal dispersal and to investigate the degree to which density-dependent vagrancy may regulate abundance of these spawning populations.

The composite distribution of stage 1 eggs shows interesting differences between the location of spawning of cod and haddock. Cod spawn more broadly along the northern flank of the bank, whereas haddock spawning is concentrated on the northeastern peak (Figure 3). From an analysis of the composite centers of mass of eggs and larvae at, respectively, 2.5, 8, 15, 24, 37, 51, and 60 days for cod and haddock, the horizontal scales of displacement with time can be summarized. There are differences in the composite distributions at age between cod and haddock, but key similarities are the limited horizontal displacement over the three-month time period and the on-bank movement of the older larvae (Figure 4).

Using the bimonthly circulation results of Naimie et al. (1994), in which seasonal baroclinic circulation is included with the flow components in Werner et al. (1993), the whole bank was seeded with eggs and seasonal loss rates from different parts of the bank estimated (Figure 5). If persistence of eggs and larvae on the bank is a "good thing," model results infer that the observed time and location of spawning are about optimal. Eggs released on the northern flank of Georges Bank at intermediate depths during the late winter/early spring have a high probability of being retained on the bank for a couple of months. We are presently modelling the loss rate of eggs and larvae from the bank as a function of spawning stock biomass, assuming that there is an expansion and contraction in time and space in spawning as a function of abundance.

Lough et al. (1994) compare year-class strengths with egg and larval distributions from the MARMAP database. They show that some of the years with good recruitment were associated with low losses of eggs and larvae from the bank due to favourable wind conditions, and vice versa. For example, loss of eggs and larvae in 1982 was high, there was a strong and unfavourable northeastward wind stress in April, and the year-class was weak. In contrast, in 1985 the losses were relatively low, the winds more favourable, and the year-class strong. Modelling work evaluated the degree to which contrasting winds of 1982 and 1985, as well as variable inflow from the Scotian Shelf, generated differences in egg and larval distributions. There was a considerably higher loss of cod eggs and larvae from Georges Bank in 1982 than 1985. The results indicate that between year differences in circulation and mixing can substantially impact retention of early life-history stages on Georges Bank, and that such processes contribute to variable recruitment.

The final cod/haddock biological study completed by the team establishes a modelling framework to evaluate the importance of circulation and mixing on larval feeding success (Werner et al. 1994). This study was summarized in U.S. GLOBEC News No. 7. The principal conclusions are:

These modelling results give a flavour of one aspect of the Georges Bank work. Overall GLOBEC activities on the bank include a broad range of field studies on processes underlying population regulation of zooplankton, cod, and haddock, with an emphasis on early life-history stages. In addition, circulation is being monitored at key locations. This mixture of field observations on processes, along with the long-term monitoring of distributions of fish, plankton, and the oceanographic environment, allows the modelling activities to focus on central problems.

Links Between GLOBEC and Management Questions

How are the above results, and the North Atlantic GLOBEC program on cod, relevant to the four management questions posed above? In Tables 1 and 2, the theoretical framework underlying population regulation of marine species is outlined. There are a number of competing hypotheses for the three characteristics of populations. Reaching consensus on which of the hypotheses are more appropriate for particular areas and populations is a prerequisite to providing answers that are broadly accepted by the scientific community. The circulation and mixing models are becoming sufficiently realistic that they can be used to help generate consensus.

To answer the first question requires a better understanding of the oceanographic processes that sustain spawning components. If the member-vagrant interpretation of pattern is more realistic than the migration triangle interpretation, then a different approach is needed to define geographic management units and protect individual spawning components. The drift route concept involving large scale residual currents infers that birth site fidelity is defined at the scale of the currents linking spawning areas to juvenile nursery areas. In contrast, the retention concept infers that fidelity is defined at smaller scales, such as re-circulation features on banks, and within bays and inlets.

The approach to defining minimum spawning stock biomass and its geographical distribution varies according to our understanding of the control of population pattern. Thus, the modelling results indicating why cod spawn in particular areas on Georges Bank are of importance to the broader conceptual issue of maintenance of geographic patterns of spawning. Practical measures such as the location of spawning closures, the minimum spawning biomass needed for each component prior to the opening of fisheries, and the time scale of recovery of extinct components, depend on understanding how population patterns are regulated. Perhaps some cod fisheries collapses have been due to gradual elimination of spawning components within a regulatory approach that does not consider this level of complexity in the biology and physics.

Under GLOBEC, a comparative modelling approach throughout the distributional range of cod in the North Atlantic is envisioned. The increased understanding of the physical/biological coupling processes that establish and maintain geographic patterns in cod spawning will aid managers in dealing with the first question and to implement measures to protect the reproductive potential of the species.

The second and third questions (need for ecosystem management and impacts of dragging) require an improved understanding on the regulation of abundance. The management implications of the two competing hypotheses for regulating abundance (Tables 1 and 2) are quite different. The match-mismatch hypothesis infers a tightly coupled community with density dependence operating through food chain interactions. Thus, one would infer linkages between commercial species, and multispecies or ecosystem management would be needed. In contrast, the member-vagrant hypothesis infers relatively uncoupled marine food chains with each population abundance being primarily controlled by physical oceanographic processes in a density-dependent manner. The populations, in essence, are interpreted to be responding independently to the physics; thus, an ecosystem management approach is not warranted. A caveat, however, should be added for species with more collapsed life histories such as marine mammals. Under the member-vagrant hypothesis, abundance of whales, seals, and porpoises should be more influenced by food chain interactions than is the case for the commercially important finfish.

All of the hypotheses in Table 1 assume that the regulation of population features occurs in the pelagic domain, predominantly during the early life-history stages. If correct, one would not expect dramatic impacts of dragging, which might impact food resources for juveniles, on the abundance of groundfish. GLOBEC research may clarify whether most management objectives can be achieved without taking an ecosystem approach, and the degree to which benthos impacts of dragging can be ignored.

There is good evidence that large scale inter-decadal climate change has an impact on ocean productivity (see Beamish 1994 for a North Pacific synthesis) and on the relative abundance of commercially important species (see Baumgartner et al. 1992 for an analysis of sardine/anchovy fluctuations off California). In the North Atlantic, however, there is considerable evidence of resilience in the biological distributions at both the community and population level. For example, since systematic monitoring of fisheries began off northern Europe in the late 19th century, some populations of cod have been spawning at the same time and locations. However, there is also evidence for dramatic changes in dispersal and migration (Greenland, Iceland cod, Scotia Shelf haddock, northern cod), and cod availability (eastern Scotian Shelf cod).

The GLOBEC studies on cod in the North Atlantic should provide an understanding of which oceanographic processes contribute to the establishment and maintenance of population pattern, as well as the relative abundance of the different populations. From knowledge of key physical features, modelling studies will allow some predictability of the degree to which environmental variability limits the attainment of fisheries management objectives. To accomplish this will also require accurate estimates of abundance and distribution of commercially exploited resources.

In sum, the focus of the North Atlantic component of GLOBEC on population regulation of cod should generate answers to some of the key questions that fisheries managers are asking. There is a need for increased consensus by the scientific community on which of the hypotheses best capture the critical mechanisms. The mix of historical data analysis, field experiments, and modelling studies in GLOBEC provides us with a unique opportunity to make progress in a component of ecology that has remained contentious for several decades. Ecology now needs a culling of competing interpretations, with a concomitant increase in explanatory power. (Mike Sinclair is the director of the Biological Sciences Branch at the Bedford Institute of Oceanography in Halifax, Nova Scotia, and Fred Page is a research scientist at the Department of Fisheries and Oceans at the St Andrews Biological Station in New Brunswick)


Baumgartner, T.R., A. Santon, and V. Ferreira-Bartrina. 1992. Reconstruction of the history of Pacific sardine and Northern anchovy populations over the past two millenia from sediments of the Santa Barbara Basin, California. CalCOFI Rep., 33, 24-40.

Beamish, R.A. 1994. Climate change and northern fish populations. Victoria, October 1992.

Cushing, D.H. 1973. The natural regulation of fish populations. In Sea Fisheries Research, p. 389-412. ed. F.R. Harden Jones, Elek Science, London.

Harden-Jones, F.R. 1968. Fish Migration. Edward Arnold Ltd., London, 375 pp.

Lasker, R. 1975. Field criteria for survival of anchovy larvae: The relation between inshore chlorophyll maximum layers and successful first feeding. Fish. Bull., 73, 453-462.

Lough, R.G., W.G. Smith, F.E. Werner, J.W. Loder, F.H. Page. C.G. Hannah, C.E. Naimie, R.I. Perry, M.M. Sinclair, and D.R. Lynch. 1994. The influence of wind-driven advection on the interannual variability in cod egg and larval distributions on Georges Bank: 1982 vs. 1985. Int. Counc. Explor. Sea. Mar. Sci. Symp., 198, 356-378.

Lynch, D.R., and C.E. Naimie. 1993. The M2 tide and its residual on the outer banks of the Gulf of Maine. J. Phys. Oceanogr., 23, 2222-2253.

Lynch, D.R., F.E. Werner, D.A. Greenberg, and J.W. Loder. 1992. Diagnostic model for baroclinic, wind-driven and tidal circulation in shallow seas. Continental Shelf Res., 12, 37-64.

Naimie, C.E., J.W. Loder, and D.R. Lynch. 1994. Seasonal variation of the three-dimensional residual circulation on Georges Bank. J. Geophys. Res., 99, 15967-15989.

Ridderinkhof, H., and J.W. Loder. 1994. Lagrangian characterization of circulation over submarine banks with application to the outer Gulf of Maine. J. Phys. Oceanogr., 24, 1184-1200.

Rothschild, B.J. and T.R. Osborn. 1988. Small-scale turbulence and plankton contact rates. J. Plankton Res., 10, 465-474.

Sinclair, M. 1988. Marine Populations: an Essay on Population Regulation and Speciation. University of Washington Press, Seattle, 252 pp.

Sinclair, M. and T.D. Iles. 1989. Population regulation and speciation in the oceans. J. Cons. int. Explor. Mer, 45, 165-175.

Tremblay, J.M., J.W. Loder, F.E. Werner, C.E. Naimie, F.H. Page and M.M. Sinclair. 1994. Drift of sea scallop larvae Placopecten magellanicus on Georges Bank: A model study of the roles of mean advection, larval behavior and larval origin. Deep Sea Res., Part II, 41, 7-49.

Werner, F.E., F.H. Page, D.R. Lynch, J.W. Loder, R.G. Lough, R.I. Perry, D.A. Greenberg, and M.M. Sinclair. 1993. Influences of mean advection and simple behaviour on the distribution of cod and haddock early life stages on Georges Bank. Fish. Oceanogr., 2, 43-64.

Werner, F.E., R.I. Perry, R.G. Lough, C.E. Naimie. 1994. Trophodynamic and advective influences on Georges Bank larval cod and haddock. Deep Sea Research.

homepage contents next article