1998 Cod Rept. as submitted to COSEWIC

"One of the most bizarre features of any advanced industrial society in our time is that the cardinal choices have to be made by a handful of men: in secret: and, at least in legal form, by men who cannot have a first-hand knowledge of what those choices depend upon or what their results may be."

(Snow, C. P., 1962. Science and government: the Godkin lectures at Harvard University, 1960. New York: New American Library / Mentor by arrangement with Harvard University Press).

Introduction/description go to title toc

This report addresses the status of Gadus morhua primarily within Canadian jurisdictions. Unless otherwise stated, statements refer to Canadian, Canadian transboundary, or Canadian continental shelf stocks or management units. The word "stock" here means those individuals (whether or not they constitute a population) identified with a management area. The existence of unique genetic or other properties pertaining to some populations is not yet known but is of prime importance to the conservation of Cod.

Atlantic Cod, Gadus morhua Linnaeus 1758, has been an economic mainstay for hundreds of years in the fisheries of northern Europe, and was important in motivating the initial colonisation by Europeans of north America, where it has remained important since then.

FIG. 1
near here

Scott and Scott (1988) list two species of Gadus in Atlantic Canadian waters: Gadus morhua (Atlantic Cod) and Gadus ogac (Greenland Cod). Gadus morhua (Figure 1) is distinguished by its generally spotted body (vs. unspotted), a leaden-silvery peritoneum with black dots (vs. uniformly brown-black to black), smaller eye and stouter caudal peduncle in Gadus morhua, and lateral line a paler colour than the adjacent areas of the body in Gadus morhua than in Gadus ogac. Gadus morhua is a large species that has been recorded at over 200 pounds (>90 Kg), although Leim and Scott (1966) report the average size as 5 pounds (2.2 Kg). Cod eggs are pelagic, larvae hatch at 3.3-5.7 mm length, the first portion of the life history is pelagic, and they settle to the bottom in the size range of 25-20 mm TL (Scott & Scott 1988). The status of Greenland Cod has not been assessed, but insofar as it overlaps the range of Atlantic Cod and is similarly vulnerable to the fishery, it could be in a similar position.

Inshore catches in Newfoundland and Labrador rose from about 100,000t in 1800 to about 250,000t in 1920 to 1930, thereafter declining back to 100,000t in the early 1960s (Lear & Parsons 1993). Unsuccessful attempts were made in Newfoundland and in New England to raise Cod for release to bolster natural populations. Management and research on the great fisheries of northern Europe, substantially this species, also laid the foundations of systematic fishery science. Catches in the northwest Atlantic reached nearly a million tons annually. There have been, however, fluctuations in apparent biomass both historically and recently, which have motivated attempts to restrict the amounts harvested. The present population status is unprecedentedly poor in most management areas: DFO (DFO 1994; DFO 1996i) describes the situation using phrases like "thought to be low" (2GH), "low" (3Ps), "very low" or "very depressed" (3Pn4Rs, 4Vn) and "historically lowest" or "in jeopardy" or "lowest observed" (2J3KL, 4T, 4VsW, 4X, 5Zjm); see Table 4 . This cannot reasonably be ascribed to natural fluctuation (discussed under Limiting Factors and elsewhere).

Fishery technology has progressed (Keats et al 1986; Lear & Parsons 1993) from hook-and-line operated from schooners and inshore craft, to gillnets, longlines and cod traps used along the coast, to trawl nets towed by increasingly powerful, increasingly efficient motorised vessels equipped with increasingly sophisticated electronics for position-finding as well as fish-finding, and with increasingly efficient communication among vessels to report concentrations of fish. Inshore catches after 1950 declined with respect to offshore catches (Steele et al 1992) in 2J3KL, but in some areas, e.g. 4X (Southwest Nova Scotia/Bay of Fundy) boats under 65 feet still land over 90% of the catch (Doubleday, 1996, pers. comm.). While catches from inshore Newfoundland and Labrador (NAFO [North Atlantic Fisheries Organization] divisions 2J3KL) averaged about 200,000-300,000 tonnes annually 1880 - 1920 (Lear & Parsons 1993) and total catches up to the late 1950s for 2J3KL were still around 200,000t (Steele et al 1992), in the 1960s to 1980s catches soared dramatically to about 800,000 tonnes and thereafter collapsed. The catches from all areas Both papers agree that inshore catches fell precipitously but they differ on details: declining in the 1920s (Lear & Parsons 1993), or around 1950 (Steele et al 1992). There had been optimism (page v: "Canada's 200 mile limit has provided a new measure of security...") surrounding the extension of management over a 200 mile EEZ: that (page 179) the biomass of 2J3KL Cod would triple by 1985 compared to 1979 levels and that TACs of 400,000 tonnes would be reasonable (NORDCO 1981).

Some significant areas (Flemish Cap, Nose and Tail of the Grand Banks) of the American north Atlantic continental shelf extend beyond jurisdiction of any single government, leaving them more difficult to regulate, and stocks more difficult to conserve if they are apt to cross into waters subject to more open multinational access and (consequently) more intense fishing pressure (see Protection). The fishing pressure itself is a complex result of declining world stocks, increasing human populations, increasing prices commanded by fish, and the recent oversupply of fishing equipment due to closures of or more stringent regulation of fisheries in other parts of the world. The surplus of fishing capacity finds the locations of least stringent or least enforceable regulation: so-called international waters; in the case of transboundary stocks (at least some populations within 2J3KL) their exposure to fishing outside the 200 nm EEZ may affect their prognosis.

Table 1
near here

The recent severe decline of the populations of Cod in the northwest Atlantic has severe economic and social implications. The barren sea in Michael Cook's play "Head, guts and sound-bone dance" (Cook 1973) was disturbingly prescient, as were his presentations of social demoralisation that followed. The moratorium has now continued for five years, far exceeding the projection by DFO that the stocks would rebuild so that fisheries could reopen two years afterward, i.e. 1994.

The recovery is impeded by collapsed age structures (Myers et al 1995b; Trippel et al 1997) and also by the poor condition implied by the low weights at age (DFO 1994; DFO 1996i). A moratorium on the fishery for this species was instituted for the 2J3KL area in 1992 (Steele et al 1992; Lear & Parsons 1993) and was since extended to most Canadian areas (Table 1). The Cod moratoria have been expensive for the government and demoralising for the local people. 2GH catches, described as "negligible since 1990" (DFO 1996i), have not exceeded 500t since 1985, even though for a half-dozen years up 94,000t (in 1966) were caught (Lear & Parsons 1993). It is not certain when, if, or to what level populations might recover.

Distribution (and Subpopulations) go to title toc

FIG. 2
near here

Gadus morhua occurs on the western and eastern sides of the north Atlantic (Leim & Scott 1966), and is known by the names Atlantic Cod, Cod, Codfish, Lutefisk, Torsk, Morue Commune, Cabillaud, Morue Franche. In the western north Atlantic it occurs coastally and on the continental shelves from Cape Hatteras to the Davis Strait (between Baffin and Greenland), and eastward to the Flemish Cap. The distribution continues along the east Greenland (below about 70¡N) and Icelandic coasts, and in the eastern north Atlantic from the Bay of Biscay to Spitsbergen and Iceland (Figure 2). From Cape Breton northward in the western north Atlantic the range of Gadus morhua partially overlaps with that of the only other species of Gadus in Canadian waters, Gadus ogac (Greenland Cod, Ogac, Uvak, Pilot, Rock Cod, Morue de Roche, Poufin). Gadus ogac is less abundant and is less commercially important than Gadus morhua, and is more restricted to coastal waters and inlets but has a range extending north and westward through the Canadian Arctic as far as Alaska.

FIG. 3
near here

Uniqueness (commonly referred to in the genetic sense, although non-genetic uniqueness can be acknowledged for conservation purposes) is a critical issue in a conservation context. Sub-populations have been referred to or evaluated on the basis of growth rates, vertebral numbers, spawning patterns, and movements of tagged individuals (Leim & Scott 1966).

Tagging and other early information led to the idea that there were local stocks, e.g. (Lear & Parsons 1993) "Templeman (1962) postulated that each shelf region, e.g. the Bonavista Shelf, Fogo Shelf and St. Anthony Shelf, projecting seaward with deep water on each side, had a basic stock of its own and that overfishing on these local stocks could cause local depletion (Templeman 1959; Fleming 1959). Tagging studies during 1954-55 and 1962-66 (Templeman 1974, 1979) produced continued evidence of homing of local populations of cod to original inshore tagging locations along the coast from the Avalon Peninsula to northern Labrador". Templeman prophesied: "in the future enough differences will be found to indicate a number of north-south and inshore-offshore sub-stocks of this Labrador-Newfoundland stock which either do not intermingle greatly or separate out at certain seasons" (Templeman 1962).

Individual fish may move considerable distances, even from the North Sea to the Grand Banks; this has only been observed once although movements between Greenland, Iceland, the Faroes and the North Sea are more common (Gulland & Williamson 1962). It seems clear that certain groups which are genetically distinguishable to mix with other groups at some times, yet still sort themselves out (as it were) and show strong fidelity (from tagging and satellite DNA data) to spawning locations (Taggart & Ruzzante 1997b; Taggart & Ruzzante 1997a). This would be consistent with Sinclair's member-vagrant hypothesis (Sinclair 1988), and the finding of the occasional stray or vagrant, or a group of cod of several genetic types in the same place, would not necessarily mean genetic mixing. Whether depleted or vanished stocks can gain members from neighbouring stocks with stronger populations is not a certainty; if movement patterns are 'learned' from older fish by younger fish (as suggested by Rose 1993) then there can be a unique and irreplaceable component to stock even if it is not genetic. This reservation was also expressed by a recent DFO scientific Workshop (DFO/Rice 1997).

Recent genetic (Mork et al 1985; Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b) and otolith microchemical (DFO/Rice 1997) work has shown differences among a number of stock components, although the early genetic work of this decade on Cod (Pepin & Carr 1993) did not detect clear differences at least within NAFO divisions (3)K,L,N and O.

History: this report originally addressed Atlantic Cod in Canada as a single unit. At the time (two to three years ago) it seemed that DNA techniques must hold the ultimate answer (to everything), and because some early attempts to use DNA had failed to show differences where it had been applied it seemed there might not be structure. But the very large range occupied by Cod was -- and perhaps this was not justification in itself -- troublesome in the context of possible uniformity. It was difficult to continue to think that tag returns and all the structure they suggested could be dismissed as reflective of short-term environmental habituation, and at the same time difficult to think that DNA techniques had missed all the differences in the genome (the entire genome is not compared, only selected parts for which there are convenient probes, and not all probes are equally useful for all taxa). This dilemma was resolved when W. Doubleday (DFO/Doubleday 1996 pers. comm.) quite correctly commented:

"[a] serious flaw in the work is that it treats cod in Atlantic Canada as one unit. It appears that ESC/IUCN standards permit the definition of species populations which are geographically distinct. The accumulated body of evidence which indicates limited movement between various cod stocks in Atlantic Canada is cavalierly ignored. We might recognize that there is uncertainty in the origins of larvae which eventually inhabit some stock area, but even here, I think the evidence would not support widespread mixing. This flaw pervades the document" (DFO/Doubleday 1996 pers. comm.) .

The new evidence supported DFO on this point: pre-publication findings by Ruzzante and others had tended to point to differences among groups of fish showing different migrations. So, the report was modified to accommodate the DFO criticism -- a change that proved unpopular with DFO after all (April 1997 COSEWIC meeting, where DFO's representative Jake Rice argued for a one-group treatment, offering no explanation for the change and claiming not to have seen the previous comments).

Popular or otherwise, consideration of status must be done on the best available information, and that information is summarised in the DFO workshop proceedings, alluded to both in the latest comments (DFO/Anonymous 1997 pers. comm.) and by a DFO representative at the April 1997 COSEWIC meeting (puzzlingly, to argue against consideration on a stock-by-stock basis):

p11 "in light of the tabulated information [whether all the presented types of data supported a single- or multi-component stock structure, where the 'homogeneous' Model I(one) scored 36.5 'unlikely' vs. 17 'likely', therefore 'homogeneous' was rejected], the Workshop concurred that a precautionary approach to assessment would have to recognise the complex stock component structure explicitly. More importantly, a precautionary approach to management would have to ensure that any harvesting should be carefully distributed among spawning components, so that none of the components would be consistently over-exploited. These approaches would require both assessments and management approaches which are much more complex than have been implemented in the past." (DFO/Rice 1997).

The Workshop also recognised that recolonisations could not be relied upon, and gave their reason for doubt:

p. 10 (under hypothesis/scenario/model II: underlying nature of stock structure is "small inshore components as well as large offshore stock components") "In this scenario, strong growth and increases in abundance of inshore stock components may contribute little to rebuilding of the offshore stock components. Only if the inshore stock components began to migrate to offshore banks when densities were high, would offshore stock components gain recruits. This would be a new and complex set of behaviours for the traditional inshore stock components, and Rose (1993[migration highway/Nature], 1994[et al Changes in distribution...]) suggests several reasons why such migrations are not assured, even when inshore abundances are high" (DFO/Rice 1997)

The number of separate, biologically meaningful, discrete populations is not yet known. It is crucial that work on delineation continue so that the risk of loss of individual components can be addressed. 2J3KL is genetically resolvable at least into both inshore and offshore components as well as north and south components, and multiple units have been distinguished in the Gulf of St. Lawrence (Ruzzante et al 1995; Ruzzante et al 1996; DFO/Rice 1997; Ruzzante et al 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b).Protection

Fisheries are usually common resources, and are susceptible to the problems of all commons, namely that harvesters tend to be opportunistic and follow short-term strategies of maximised capture rather than long-term strategies of conservation. The mobile nature of fish adds a dimension of uncertainty to both regulation and data gathering where the fish cross boundaries between differing management and harvesting regimes.

Protection comprises the regulation of fishing gears, seasons, effort and catch, and the formulation or development of such regulations at both the political and scientific levels. The science, assessment, and political decisions have come into question in connection with the 'what happened' questions, so some discussion of these processes is needed.

Management structures, entire range of Gadus morhua go to title toc

Gadus morhua is a coastal and shelf species, and has a wide distribution which is predominantly within various national jurisdictions of North America and northern Europe. Where stocks occur outside national jurisdictions (i.e. shoals or banks outside 200-mile EEZs) they are managed by either NAFO (Northwest Atlantic Fisheries Organization; approximately, west of 40¡W longitude) or ICES (International Council for the Exploration of the Sea; east of 40¡W) geographic units. The EU (European Union) sets TACs for cod in most ICES areas (ICES 1994), exceptions being those areas managed by Iceland, Norway and the Russian Federation.

Europe

For most important species, fisheries in waters pertaining to coastal waters of members of the EU are managed centrally. The EU as a member of NAFO also negotiates allocations for its member states in the NAFO Regulatory Area. European member states of the EU generally have access to all EU-managed waters. Fishermen from different EU states sometimes take a poor view of what they see as ruthless fishing practises, or object to the new presence of vessels (of new EU member states) in what previously were national waters. This has generated considerable friction and at times capture of foreign vessels by fishermen and cutting of nets (Bowcott & Gooch 1994; Bowcott & Jury 1994).

Under-reporting and mis-reporting is cited as the reason that, for some areas, assessments are suspect and/or could not be completed (ICES 1994). The recent dispute surrounding Turbot (Reinhardtius hippoglossoides (Walbaum 1792), or Greenland Turbot, Greenland Halibut) involved similar charges by Canada: of undersized mesh, undersized fish, false records and hidden compartments on an EU (Spanish) vessel, the ESTAI; the claims and the evidence presented tested the relationships among the NAFO members.

NAFO

The Northwest Atlantic Fisheries Organization (NAFO) has authority for stocks outside Canada's 200-mile EEZ; the NAFO Fisheries Commission sets quotas, by stock, and allocates portions of such quotas to the Contracting Parties. NAFO has fifteen Contracting Parties based on parties fishing the area at the time of its inception: Bulgaria, Canada, Cuba, Denmark/Faroes/Greenland, Estonia, European Union (includes Spain and Portugal), Iceland, Japan, Republic of Korea, Latvia, Lithuania, Norway, Poland, Romania, and the Russian Federation. Of these, only Canada and Greenland are coastal states in respect of the waters fished by member states. Membership does not require reciprocal fishing permission.

The politics of NAFO have gained wide attention following the recent Turbot conflict. NAFO rules contain an 'objection' clause, which permits a member state to disregard (e.g. Rowe 1993) the consensus scientific advice and TAC, and to set unilateral quotas its fleet. Policing appears to have been less rigorous than within Canadian jurisdiction, and unilateral quotas and catches have often exceeded the NAFO allocations (see table below). Gear infractions also occur and are reputed to be widespread; for example, for the illegal use of a liner of small mesh, the Spanish freezer-trawler ESTAI was arrested in March 1995 on the Nose of the Grand Bank. The ESTAI was also said to have parallel but different logs found on board, and about twice as much Turbot on board as declared by captain (according to statement by the Honourable Brian Tobin, Minister of Fisheries, aired on CBC [Canadian Broadcasting Corporation], Mar. 11, 1995).

The EU, possibly under internal political stresses (see Europe, above), has invoked this procedure often in the last decade. The following quote usefully contrasts the EU catch in NAFO areas with the NAFO allocations:

"From 1985 to 1990, total NAFO allocations and catches (000 t) of groundfish outside 200 miles by the EU and reported to NAFO were as follows (these include transboundary stocks of northern Cod, southern Grand Bank Cod, American Plaice, Yellowtail Flounder, Witch Flounder and Redfish as well as stocks completely outside 200 miles on Flemish Cap (Redfish, Cod and American Plaice):

Year 1986 1987 1988 1989 1990 1991 1992 1993

Allocation 26 23 19 13 15 20 19 16

EU Catch 172 141 85 93 46 57 30 15

It must be pointed out that the EU used the 'Objection Procedure' during these years, not accepting the allocations decided by the Fisheries Commission, but setting their own quotas instead. They have the legal right to do this under the NAFO Convention.

There is ample evidence in the Reports of the Scientific Council of significant unreported catches during most of this period." (courtesy D.B. Atkinson, DFO St. John's)

Only recently has agreement been reached with the EU and other NAFO Contracting Parties on monitoring of gear, catches, minimum size limit conformance and fishing areas. The agreement has yet to be fully implemented.

Canada go to title toc

Legal Regulatory Framework (Canada)

Within Canadian jurisdiction, regulation of the fishery is under the authority of the federal Minister responsible for the Department of Fisheries and Oceans (DFO) via a complex web of linkages with other departments and agencies (Appendix: figure 1). DFO conducts research, and proposes regulations for gear, season, size and catch limits on many species; these regulations in theory arise from (or at least are supported by) scientific study or opinion. Because of the diverse interests involved, it is almost inevitable that there will be conflicts among scientific judgement, public opinion and interest groups in the making of policy. For example, the mesh size in Japanese Cod Traps is considered by some fishermen (at 1995 FRCC public meeting, Holiday Inn, St. John's August 31) to be too small and to increase catch and mortality (cod die when air-sacs distend during hauling of traps) of small cod, but is said to have been reduced in response to pressure from other fishermen. CAFSAC (Canadian Atlantic Fisheries Scientific Advisory Council) provided DFO's scientific advice on catch levels for Atlantic fisheries until 1992, when this function was assumed by a new organization, the FRCC (Fisheries Resource Conservation Council). The FRCC consists of "not more than 14 members with an appropriate balance between 'science' and 'industry' ... 'Science' members are drawn from government departments, universities or international posts, and are of an appropriate mix of disciplines, including fisheries management and economics ... 'Industry' members are knowledgeable of fishing ... and understand the operational and economic impacts of conservation decisions ... All members of the Council are appointed by the Minister ... Members have to disclose any interest ... so as to avoid ... conflict of interest" (FRCC 1996). Interestingly (page A45, section 5.9), the delegates and ex-officio people are not asked to officially endorse recommendations, although they can participate fully in generating them: it would seem there can be input without accountability. Canada (through DFO, FRCC) regulates the gear, for example, mesh, hook sizes, etc. (DFO 1996a) used to catch the fish, but not technical aids to fish detection, navigation and communication within the fleets (fish-finders, Loran/GPS, radio traffic), which undoubtedly improve fishing efficiency. In short: DFO scientific advice on commercial stock status is formulated from research surveys and commercial catch data, analyses are discussed at assessment meetings where consensus is formed and later expressed in a report that is sent for approval and editing (Hutchings et al 1997) by DFO's senior executives, approved reports are then forwarded to FRCC which develops recommendations that are sent to Minister of Fisheries & Oceans who sets the TAC. Thus the ultimate authority is political.

Enforcement within Canadian jurisdictions is by a combination of monitoring of landings (usually reported by the fishing industry), gear, and fishing monitored by on-board observers on larger craft. Within Canadian jurisdiction there are rumours of data falsification (double log books) by fishing vessels, reports of nets set with camouflaged markers (floats covered with garbage bags in prohibited areas; only sophisticated navigation technology permits these to be found again, and thus permits such practises), loose pieces of net (which can later be claimed to have been ghost nets or flotsam caught during the trawl) thrown into codends to increase retention of smaller fish. It is widely accepted that various abuses are commonplace (Napier 1996). They are acknowledged by DFO as well, e.g. in reference to 3Pn, 4R and4S: "It is suspected that many landings were not reported during the 1980s, and that many small fish were discarded during this period; consequently, the data on catch at age may not be entirely accurate" (DFO 1994). The nature of such occurrences is that they are more likely to occur where there is a low probability of detection and attribution of responsibility to a particular person or vessel, low probability of successful prosecution, and where penalties are small enough to be tolerable as routine costs of doing business.

Effective regulation must comprise adequate estimation of biomass and replenishment rate, must set biologically reasonable (conservative) TACs, and must be able to enforce these. Enforcement appears to require either more resources allocated to monitoring, or penalties which are drastically more severe for infringements that are more difficult to detect. If cheating is to be controlled, the average cost of cheating must be set so as to clearly exceed its average profit.

The UN convention on the law of the sea (UNCLOS) currently prohibits any state from extending jurisdiction beyond 200 miles. This has created problems for management of Cod and other groundfish on the Grand Banks because some individuals of the stocks (2J3KL and 3LN) move outside a 200-mile limit and become vulnerable to fishing practises not under the jurisdiction of the coastal state. To deal with this kind of situation groups of nations making some claim to legitimacy of participation have formed cooperative management structures whereby, in theory, the conservation need can be fairly balanced against the demands of each participant. However, NAFO has had only limited success dealing with overfishing. That groups such as ICES partners and NAFO partners have effectively reserved the North Atlantic and take action against non-member nations fishing in NAFO areas (NAFO web page) means that, effectively, jurisdictions have been extended beyond 200 miles but have been masked by the multinational character of the cooperative regulatory agreement. This soft precedent makes it more likely that if these structures fail in their purpose unilateral actions may be taken by individual states to avoid the alternative: obliteration of fish stocks. Just as Iceland's declared EEZ was illegal according to international law at the time but set a new standard, a number of current problems not adequately dealt with by existing structures may force a revision of UNCLOS.

Scientific Input to Management Decisions (Canada) go to title toc

In an ideal world all fish would be visible and an exact census would always be available. It would then be possible to manage the fishery in terms of a base population and a surplus, the base chosen so as to maximise the sustained production of surplus. Because this is logistically not possible, considerable theory is required both for deciding what the level of biomass should be and for estimating what the population is at any given time. The application of this theory involves assumptions about basic biology and ecology, and occasionally these need to be revised as knowledge improves.

Scientific input to management is primarily through a complex stock assessment conducted by DFO. Assessment and management are based principally on investigations on numbers, ages and sizes of adult fish or fish of commercially catchable size in each managed stock. The perception of stock is evolving (see discussion of populations, genetics), so it is increasingly recognised that movements of stocks cause them to overlap with each other either in space and not time, or in both. This can reduce the effectiveness of surveys and assessments:

"However carefully the boundaries of the major stocks were determined when they were established, distribution of major stocks appear [sic] to have changed in recent years, so geographical structures are not captured well in current assessment and management practices" (DFO/Rice 1997).

The stock assessments can use data from multiple sources (research surveys, commercial trawler catches, inshore catches; acoustic data are of recent appearance and have not been fully integrated into this process) to generate and tune estimates. Integrating the various kinds of data requires a judgement on the quality, appropriate weighting, and relevance of each source of data. Not all data sources equally estimate some other quantity of interest. Philosophies on this question have varied over time, and the weighting of each data source has therefore varied also. The question of weighting of data for 2J3KL, including data sources said to have been given zero weight has been claimed (Harris 1993) to have contributed to overestimation of Cod stocks in the years leading up to 1991, leading to continually high TACs and the collapse of the stock necessitating a moratorium on Cod fishing.

Research surveys and Sequential Population Analysis (SPA), which includes a number of estimation routines (Rivard 1982) , one of which is Virtual Population Analysis or VPA (Ricker 1975), are conducted by DFO to evaluate stock biomass and mortality rates. The research surveys are randomised within strata (depth, bottom type, etc.) in order to reduce bias and allow extrapolation from area swept by research trawls to the area of the stratum. Stock assessments are routinely conducted using SPA, which calculates the size of the exploited population from removals (principally offshore commercial fishery data). Commercial catch data in any given year does not generally provide good data on total population size because the areas are sampled non-randomly, seeking the highest concentrations of fish; however, the sum of catches for a given cohort (here, cohort = all fish hatched in a given year, hence all the same age in subsequent years) over years (i.e. {Na,y + Na+1,y+1 + Na+2,y+2 ... Na+n,y+n}, or until there are no more fish of age a+n in year y+n) is a conservative estimate the original size of that cohort, and by definition the size of the Virtual Population which gives the method its name. That this is the principle on which the cohort analysis method is based is succinctly stated in Pope's comment "when we have caught them all [i.e. all of a cohort] we will know how many there were". The catch data can be standardised by the units of effort they represent, giving catch-per-unit-effort (CPUE) which can be an indicator of changes in availability of fish and hence an indicator of changes in biomass; typically, where the method used to harvest remains constant the CPUE declines as a stock becomes fished out. "There was a reliance on commercial catch per unit effort to estimate the population numbers in the last year of the VPA; however the catch per unit effort estimates of abundance were positively biased" (Myers et al 1995b). The effect of changing technology is to distort CPUE: more efficient technology returns fish more easily, and therefore tends to obscure depletion of the stock (Myers et al 1995b). Similar comments are made by CAFSAC (1989):

"Information from research surveys has, for example, indicated a smaller stock, subject to higher fishing mortality, than has the information from the offshore commercial fishery. In most years the differences in stock size implied by these two sources of data have been considerable, but both data sets have indicated steady growth of the stock. Information from the inshore fishery has been of little value to the assessment of stock status because of the lack of data on and amount of effort deployed. The relatively poor inshore catches in 1984-87 had been interpreted in some quarters as evidence of declining stock size, although the scientific assessments, using either research or offshore commercial catch data, concluded that the low availability of cod inshore could be due to a large number of possible factors, many associated with environmental conditions.

"The Scientific Council of NAFO and, in 1987, CAFSAC handled the different indications of fishing mortality, as estimated separately from research data and from offshore commercial data, by choosing a point midway between the two values, there being no justification for assigning greater confidence to one set of data than the other. As noted above, the research survey series has given more pessimistic indications of stock size than the commercial series and CAFSAC has reported (Adv. Doc. 86/25) that retrospective analysis has shown that, in hindsight, the annual assessments have been optimistic. The 1986 survey gave a much higher index value for the biomass than previous surveys. Consequently, the difference between the estimates of fishing mortality obtained from the use of the two data series was much less than in the most recent assessment ... the research survey data for 1987 and 1988 throw considerable doubt on the 1986 value which had led to the more optimistic evaluation in 1987. Furthermore, the new commercial catch data (1987 and 1988) also support a less optimistic conclusion ... it is the opinion of CAFSAC ... that the indications of stock status provided over the years by the research surveys have more accurately reflected historical trends in the cod population. Surveys do not, however, provide a reliable estimate of the level of fishing mortality in the last fishing year. The conclusion is therefore that the almost constant increase in offshore commercial fishing success (catch per hour) between 1977 and 1985 was due to factors in addition to stock growth and that the extent of stock recovery was much less than the improvement in fishing success would indicate."

The two main sets of data available for stock evaluation reflect radically different activities. Research surveys are randomised and designed to harvest information, while commercial fishing has as its objective the maximising of the rate of return on investment. The research surveys are, by design and intent, methodologically conservative so that different years can be interpreted without adjustments for technique. Conversely, the commercial fishery is methodologically progressive, and has benefited from the introduction of superior equipment including electronic fish-detection and location-finding devices (e.g. compass, log and sextant to Loran to GPS) and improved communication among vessels to share information. The result of these differences is that the two sets of data are very different, and difficult to integrate, as acknowledged by CAFSAC (1989).

FIG. 4
near here

Table 2
near here

The difficulty of integrating data from two sources can be illustrated by using data from Myers and Cadigan (1995), to calculate a ratio R/C (for each year and by age group, the ratio of research catch per tow to the commercial catch). Fall research surveys are used in this example. Ideally, R/C would be constant over time (years), at least within each age group. However, R/C is not constant over time, and not only does it vary, but it varies systematically such that the change in the relationship can be explained by significant linear regressions against time (Figure 4) for ages 6 and over (e.g. r2 = 0.6 for age 10; Table 2). This suggests that the commercial fishery improved its ability to select and harvest some ages. The lack of such a trend for ages 3-5 may indicate, for example, a lack of improvement of skill of the commercial fishery relative to the research surveys, or better policing of mesh sizes over time, or under-reported discarding or catches of small fish, or avoidance of those ages in some other manner. Whatever the cause, the difference means that they do not respond in the same way to population size; and since the research surveys follow a standardised protocol it is likely that the commercial fishery became more efficient such as to postpone the reduction in catches that would signal a depletion of the stock.

Retrospective pattern: it has, apparently, been known since 1949 (DFO/Anonymous 1997 pers. comm.) that, as shown in table 2 of Keats et al (1986), the VPA process tends to initially (relative to subsequent revisions by the same process) overestimate biomass. This means that the initially derived F0.1 catch is larger (1.5-3 times) than that derived retrospectively (Keats et al 1986). This unpleasant behaviour of VPA, if it truly can be attributed to VPA and not to surrounding assumptions, may be a greater problem when exploitation rates are increasing. It has been pointed out (Keats et al 1986; Steele et al 1992) that differences between inshore and offshore catch data are not incorporated into the assessment process and the setting of TAC. Those authors argued that in 2J3KL data before the 1950s the inshore and offshore trends were similar, but that after the 1950s there was an inverse correlation between inshore and offshore catch trends. The long history of catches from 1800 to the 1990s, from various sources, is shown in their (Steele et al 1992) figure 1; after the 1950s the behaviour of the trends changes markedly, with the offshore catches peaking briefly at around 800,000 tonnes while the inshore catches fall. It would be desirable to have an objective way of determining where the change occurred before applying an objective statistical test to hypotheses about similarity of the inshore and offshore trends, or about difference in behaviour of the data before and after the change, which Steele et al date at 1948. The graph suggests that the post-1948 increase in offshore catches was at the expense of inshore catches. It has recently been acknowledged (DFO/Rice 1997) that a precautionary approach requires a management that explicitly recognises stock components, to separate inshore and offshore at least.

The management target (F0.1) was adopted by Canada in 1977 (CAFSAC 1989; Steele et al 1992) to limit fishing mortality. F0.1 is the level of fishing mortality at which the slope of the yield-per-recruit curve is 1/10 of its slope at the origin (ICES 1992); the curve is a graph of Yield-per-Recruit (Y/R) against F (Rivard 1982). In 1979 a level of F was chosen slightly lower than F0.1 (Lear & Parsons 1993), and in 1984 F0.1 was resumed. Despite its adoption as a management target, "actual values of F for northern Cod have invariably exceeded the target F0.1 with predictable catastrophic consequences" (Steele et al 1992). This is partly due to a problem known as 'retrospective pattern' in VPA, which is that for some stocks the calculation of F0.1 gives a lower value in hindsight, that is, later re-evaluations with more data show the original F0.1 to have been optimistic. It is disputed (D.B. Atkinson, DFO, pers. comm.) that harvests in excess of the F0.1 catch have "predictable catastrophic consequences". W. Doubleday (1996, pers. comm.) has also caviled at this point:

"we do not contest the F0.1 target was not attained. However, this was a rebuilding target, not a status quo target. Missing a rebuilding target does NOT necessarily have the same 'disastrous consequences' [quote referring to Steele et al 1992]. A conservation target was not consistently achieved. The author should differentiate a management target from a conservation limit. ...".

Obviously whatever meaningful targets existed were not achieved, because stock declines did follow, and moratoria were eventually required.

The SPA process is directed towards identifying a level of removals which will equate to a management target, in this case F0.1. If, however, a population is subject to, as has been claimed in the media by the Assistant Deputy Minister, L.S. Parsons (Unland 1996), high natural fluctuation, then any target F is inappropriate: the target then should be in terms of Z (Z=M+F, i.e. the sum of instantaneous natural mortality and fishing mortality). In a system where for recruited size classes M fluctuates greatly (this is the only way that a population can fluctuate greatly over a term shorter than a generation), the standard practise of using a fixed value (M = 0.2) in the estimation of F is then also inappropriate.

Age determinations are an essential component of VPA. Without proper assignment of fish to age classes, mortality calculations are meaningless and the VPA can be compromised. Also liable to be compromised by age errors and biases are indices such as age-at-maturity, and weight-at-age. Generally, for most organisms, age can be crudely estimated from size, but there are year-to-year and place-to-place variations in these indices. Size is therefore a poor indicator of age (e.g. see Figure 3), and better ways of determining age have been sought. Otoliths generally present the best opportunities for estimating age (e.g. Stevenson & Campana 1992), but processing of otoliths (extraction, grinding, examination) is expensive and prone to difficulties arising from evaluator subjectivity (Neilson 1992). It is important that otolith readers be isolated from other data about the specimens being evaluated. A useful and reasonable compromise often used is, within a large sample or within a geographic area and a time frame, to use otolith-derived age estimates in an age-length key which is subsequently applied to lengths (to correct for changes due to temporal and spatial variation), resulting in estimates better than those based on length alone, but not as good as those directly obtained from otoliths. The meaning of the otolith features (annual checks, spawning checks) needs to be assumed in order to interpret otoliths, and the meaning attributed to features should be verified or derived from independent studies of the fish concerned, and applied uniformly during interpretation. Otolith and age-length key methods are quite unlike length-frequency-analysis, which assumes that successive modes in a local population are successive major units of age. While length-frequencies may provide interesting information for comparison of the success of either technique, and may shed light on the nature of certain otolith features, they should not be permitted to influence routine otolith readings. Recommended protocols require that otolith reading be done blind (without reference to specimen size, locality, or other information) and quality-control of otolith reading should include blind re-evaluation of previously-read otoliths to allow identification of operator differences. My request for information on how cod is aged by DFO was refused (DFO/Atkinson 1996 pers. comm.; DFO/Doubleday 1996 pers. comm.), but it nonetheless would appear from early papers (Wells 1980; Wells 1981; Wells 1983; Wells 1986) and from the refusal to supply information that ages are not read blind, and therefore cannot be considered immune to intrusion of subjective impressions against which age data might unconsciously be evaluated during processing; therefore, any data depending on age must be considered suspect. Scientific results require the support of documented methods; it cannot be understated that to refuse to divulge methods is utterly inconsistent with proper science. Here, the doubt therefore extends to age groupings in SPA and weight-at-age. An objective evaluation of the process should be conducted as soon as possible, using existing archived otoliths. Most importantly, all otoliths that are collected should be archived for two reasons: for future checking of age determinations, and to make it possible to look at long-term changes in growth rates and other features.

The extent to which bycatches and illegal catches affect the Cod population is impossible to evaluate without data, but in scientific reports (DFO 1994) blank lines often indicate no data for bycatches or illegal catches, suggesting that reporting or policing have been insufficient to generate estimates. This is a documented problem in the ICES management areas as well (ICES 1994). The effect of the less strict policing of fishing on the Nose and Tail of the Banks is, whether or not there is more overfishing there, to contribute uncertainty to the interpretation and reconciling of commercial and research survey data.

Political-Level Input to Management Decisions (Canada) go to title toc

Scientific consensus is currently not the final arbiter of quotas, and whatever shortcomings exist(ed) in the science have been amplified by political considerations, domestic and international, explicit or otherwise. Political decisions have had a strong influence on this fishery (Steele et al 1992) in recent decades, whether we consider the approval or amendment of biological advice within or beyond Canada. Politics also permit and to some extent drive the unilateral quotas set by the EU, since 1986, using the objection procedure of NAFO.

John Crosbie was the Canadian Minister of Fisheries leading up to the first moratoria on Cod fisheries. A recent article (Benson 1995), entitled "No Apologies", reported an interview with John Crosbie. The fishery collapse was addressed:

'... Probably one of the biggest problems he [John Crosbie] ever faced was what had to be done with the collapse of the northern cod stocks. It was next to impossible to manage the stocks because fisheries scientists [quote:] "didn't know what they thought they knew ... It was no good of them coming to you after the fact and saying we have been giving you the wrong information ... no one can manage a system like that" '.

Decisions on fisheries management must be based on science. Any assessment based on sampling less than an entire population is subject to sampling error. In some systems the uncertainty is inherently large. Complex systems and biological systems generate large variability. (Two successive estimates are rarely identical, and this is such a well known feature of complex systems that estimates should and do occasion suspicion if they are too close.) Conclusions based on sampling should, properly, be reported in a way that quantifies the uncertainty or imprecision that is expected in sampling. Commonly, this is a confidence interval, which is not an invitation to choose a value from the most convenient end of the range presented. Revision of an estimate happens when assumptions are corrected, or new data arrives, and it does not equate to "didn't know what they thought they knew ...". Statistical uncertainty must not be a reason to disregard scientifically derived estimates when making management decisions:

"Perhaps one lesson which can be distilled from the experience of northern cod management is the necessity for a cautious, conservation-oriented approach whereby the stock is allowed to grow and the harvest is controlled at a low fishing mortality (e.g., F0.1) level, in conjunction with other targets such as minimum spawning biomass constraints. The consistent pattern of natural stock variability in Canada and worldwide should serve as a grim reminder that we cannot will stability into a naturally chaotic system by imposing arbitrary regulatory mechanisms. If this resource is to serve the best interests of those who depend on it for a livelihood, there has to be a shift in our perception of how we utilize natural resources" (Lear & Parsons 1993).

In the early 1970s the F0.1 level was considered and soon adopted as a management tool and goal to prevent over-fishing. In 1979 it was decided to reduce fishing mortality further (i.e., below the F0.1 level) in order to allow the 2J3KL stock to rebuild; this goal (Steele et al 1992) of managing towards a target spawning biomass of approximately 1.5 million tonnes was thought to have been almost achieved in 1984, and management policy returned to F0.1. The F0.1 levels were often substantially exceeded in following years. Catches are only tenuously related to F0.1. Management recommendations begin with the F0.1 tool, and are transformed into Advised TACs for consideration by the Minister, and then into TACs at the Ministerial level. TACs (Total Allowable Catches) set at the Ministerial level have often exceeded biological advice or advised TACs. This has been rejected as "rumour and speculation" (W. Doubleday pers. comm. 1996); so numbers are presented from DFO data:

Between 1987 and 1994 and over all management areas, the Canadian TACs frequently exceeded the F0.1 management target F, and reported catches over the period frequently exceeded the TACs, with the result that the net excess of catch over F0.1 levels amounted to (depending on calculation method, pair-wise or row-wise) 619,000 or 423,400 tonnes, of which 262,000 or 317,000 tonnes was in 2J3KL (calculated from data in DFO 1994). For 1989, 1990 and 1991, the 2J3KL TACs (235,000, 199,000, 190,000 t) all exceeded the F0.1 level (125,000, 121,000, 100,000 t) and the Advised TAC (125,000, 174,000, 100,000 t) (DFO 1994). For Atlantic Cod for all years 1987-1994 and areas covered by the report (DFO 1994), the sums of differences indicate that, from F0.1 (the target) to advised TAC, to actual TAC, and finally to Reported Catch, at most steps there is a net increase in tonnage:

Quantity All Areas 2J3KL

F0.1 1,992,000 t 885,000 t

..difference: 337,990 t 73,000 t

Advised TAC 2,329,990 t 958,000 t

..difference: 586,210 t 308,000 t

TAC 2,916,200 t 1,266,000 t

..difference: -304,700 t -64,000 t

Reported CATCH 2,611,500 t 1,202,000 t

Rep.Catch - F0.1 619,500 t 317,000 t

Notes: Differences calculated from sums, i.e. rows with incomplete pairs still contribute to numbers. Shortfalls are included and reduce the sums accordingly. Because some areas lack different information, like F0.1, in the DFO report, the grand totals, i.e. Catch-F0.1, don't necessarily match the total of the intermediate steps, and slightly greater or smaller differences can result depending on whether differences are calculated for each row (year and NAFO division) or from sums of columns containing F0.1 level, Advised TAC, TAC implemented, and Catch. These figures also do not allow for under-reporting of catches or discards.

The calculated excess of TAC and catch over F0.1 and Advised TAC underestimates the true excess because it neglects discards and unreported catches, which show as blank lines in report (DFO 1994) tables despite the comment, for example, "it is suspected that many landings were not reported... and that many small fish were discarded..." (page 72 for cod in 3Pn4Rs). The need to address discards and other "data infractions" would have to be recognised at the Ministerial level in order to be granted the funding necessary. A similar increase of TAC beyond scientific recommendation is referred to for Haddock in 3Ps: "Although CAFSAC continually advised that catches should be restricted to by-catch only through the 1980s, TAC ranging between 150 to 3,200 t were put in place from 1987-1992" (DFO 1994).

The F0.1 tool (among others) is a guideline to limit fishing to a range consistent with conservation, and the danger in treating it as a minimum to be negotiated upward should now be obvious. Once fisheries policy has been developed and published through democratic mechanisms, Fisheries Ministers should prevent increases in TACs beyond biological advice developed to meet those policies.

Recent structure and FRCC

The post-1992 model of fisheries management in Canada has DFO science generating data, presenting analyses in assessment meetings, generating reports for Ottawa's approval where content is subject to editing. The edited and approved versions are released, FRCC uses these with whatever it gets from public consultations (at which university and other non-DFO scientists' input is limited by the unavailability of anything but edited summary reports, and even those may not be available far in advance) and reprocesses the information as recommendations in a form alleged to be digestible by the Minister. The process may isolate the Minister from the science component, and elements of doubt and caution in the original work may be lacking and the recommendations presented with undue confidence. Other than by attending the assessment meetings, there is no open peer review of the science, and data is mostly unavailable to outside scientists for scrutiny, and details on the methods are sometimes kept hidden as well. The process is faulty because it is insufficiently open, does not avail of the many sources of outside input and advice, and tends to entrain a single mode of thought and approach. The DFO Official Spokesperson Policy (OSP) or 'gag rule' limits the scope for DFO scientists to release dissenting opinions. Inevitably, diversity of thought is lost. As the late Grant Notley commented across the floor of the House to the last PC government "when everybody thinks alike, nobody thinks very much".

Added to the problem of the scrutiny and integrity of units of intellectual effort in the chain leading from fish to data to the Minister, FRCC has problems with credibility.

It is rumoured that FRCC's decision to open 3Ps was taken before the latest results were available. As of late 1997 there were still no results available to indicate the impact of the re-opened fishery, the state of the stock, or its prognosis.

Despite that the 1996 FRCC document (FRCC 1996) was offered as the basis for Minister Mifflin's 'limited re-opening' of cod fisheries on the east coast, it is rife with internal contradiction. Firstly, there is no possible way that the previous suggested criteria of FRCC (1995) could be met, and those criteria have (therefore) been dropped with minimal comment. The notion of a "Precautionary Principle" is piously introduced, only to be circuitously re-defined or neutralised (page 17) with observations like "reference points based on environmental and fishing conditions which prevailed a decade or more ago may be inappropriate to tomorrow's circumstances" to mean, it would seem, let's just go fishing. For the three areas for which it recommended re-opening, these recommendations are explicitly contradicted by the Appendix 4 in the same document: although 3Ps, 3Pn4RS, 4TVn(N-A) (p.31, 78, 81) are blandly given TACs in amounts of unknown origin, Appendix 4 says, for each of the three, "the fishery should be kept closed" (p. A44-A47). This is despite that Appendix 4 is referred to on page 17 as containing the basis for the recommendations. These contradictions are spectacular and call into question FRCC's constitution and effectiveness. They give grounds to speculate about schizophrenic management, or political machinations leading up to the national election; that the book was written as 'don't re-open', but that at the last minute a decision was made to change it to 'open', and whoever was given that task changed the front bit of the book but in the rush forgot about the back bit.

Population Size and Trend go to title toc

Table 3
near here

In all NAFO management areas, populations of Gadus morhua have declined (DFO 1994; DFO 1995; DFO 1996i), most (Table 3) extremely seriously (DFO 1994; DFO 1995; op. cit. and Myers et al 1995b; DFO 1996i). Declines are serious for many ICES areas as well (true for many ICES areas as well, ICES 1994). The Canadian area showing the least decline is 4X (with a 1994 biomass around 57,000t little changed from 62,000t in 1987), but recruitment is poor in virtually all areas. Divisions 4T and 4VsW are following trajectories worse than would correspond to "critically endangered", and 5Zjm biomass since 1991 has dropped from 49,000t to 17,000t (DFO 1994), with an indicated increase more recently (DFO 1997c) but note that a current year's estimate is often revised the following year (as for 5Zjm) and that, to relax a restriction or increase a harvest, a precautionary approach would require better evidence of 'good news' than a single year's estimate under normal conditions. The change in 5Zjm qualifies it for "critically endangered" (ESC) or "Endangered" (COSEWIC): even though the change 1987 to 1991 was small, the decline from 1991 to 1994 was more rapid than for 4T and 4VsW. The 2J3KL divisions ('northern cod') are the most seriously depleted, with a 1992 biomass estimated at 1 to 2 % (i.e. over 98% collapsed) of its maximum documented level since 1962 (Myers et al 1995b). Recent news (1997) is restricted to very brief summaries (DFO 1997d) and news reports (Whiffen 1997b; Whiffen 1997c), in one of which the outlook was summarised as "grim", quoting the DFO ADM, L.S. Parsons.

Collapses have been sufficiently large and that complete moratoria on Cod fishing were declared. Of the 6 stocks examined by Myers et al. (1995b), at the time moratoria were imposed, all existed at less than 25% of former maxima; three at under 10%; one at under 2% (Table 3). The maxima, it should be noted, are only those substantiated by methodically collected data; populations may well have been larger before population sizes were estimated.

Fish over 8 years old in 2J3KL were virtually absent in 1993 (Myers & Cadigan 1995); compare this with the "not uncommon" occurrence of Cod up to 16 years of age (Leim & Scott 1966). The collapsed ages structures imply, according to recent studies, a reduced potential for stock re-building (Trippel & Morgan 1994; Trippel 1996 ms; Trippel et al 1997). Contemporaneous changes in the populations of other groundfish species suggest that a common contributing factor (see the section "Limiting factors" and subsections).

Inshore stocks have been recognised (Lear & Parsons 1993) on the basis of meristics, physiology, migration patterns, parasite infestation, and biochemistry (numerous authors cited by Lear & Parsons 1993). There are indications that inshore stocks have already been lost in Newfoundland (Myers et al 1997) and in New England (Ames 1997). A 1997 workshop (DFO/Rice 1997) concluded that prudence required the recognition, as supported by the preponderance of evidence brought by the presenters, that Cod in many divisions were unlikely to be a single homogeneous group:

(p10) "There was full agreement that, to different degrees in different Regions, there is substantial structural heterogeneity in all the stocks as presently defined" and (p. 11) "strong growth and increases in abundance of inshore stock components may contribute little to rebuilding of the offshore stock components. Only if the inshore stock components began to migrate to offshore banks when densities were high, would offshore stock components gain recruits. This would be a new and complex set of behaviours for the traditional inshore stock components, and Rose (1993, 1994) suggests several reasons why such migrations are not assured, even when inshore abundances are high"

Divisions 2GH for example have failed to recover after 30 years following intense exploitation the annual catches during the late 1960s averaged 68,000 tons (Lear & Parsons 1993), but although the TACs 1987-1992 were 20,000 tonnes, "since 1985 catches have been less than 500 t and there has been no reported catch since 1990" (DFO 1994). Despite that (DFO 1994) "assessments [of 2GH] have not been possible in recent years because of the lack of abundance indices ... [and] catches have been low or non-existent and there is no sampling available ... [and] a survey ... in 1991 ... detected very few fish", the TAC remained at 20,000 t and was reduced only for 1993 and 1994 to 1,000 t. Reported catches remained near zero; the stock, such as remained of it, was not given the benefit of a zero TAC. Nevertheless, estimated discards of cod in the 2H shrimp fishery average 70 t annually; because these cod would likely be quite small the implied mortality in terms of numbers is perhaps quite high.

Current declines are unprecedented for Cod in Atlantic Canada. Over a longer term and fairly comparable to the presentation in figure 2 of Lear and Parsons (1993), data assembled from various sources (Steele et al 1992) to estimate the landings in 2J3KL shows (their figure 1) inshore catches around 75,000t and rising from the early 1800's, nearing 200,000t by the late 1800s, and continuing to rise slightly (with catches that are generally within plus or minus one-third of the decadal means) to approximately 200,000t around 1950. Lear and Parsons' figure shows inshore catches peaking at approximately 250,000t per year slightly earlier -- around 1920. Up to 1950 increases in foreign/offshore landings were roughly paralleled by inshore landings. After 1950 landings rose dramatically to (briefly) around 800,000t due to increases in the offshore fishery, while in the same time period (1950 to 1975 approximately) inshore landings fell by about 60%. The peak landings in the 1960s were not sustained and indeed when a TAC was first imposed in 1974 it was too high to be effective (Lear & Parsons 1993).

FIG. 5
near here

FIG. 6
near here

The research vessel average catches for 2J3KL, plotted (Figure 5) from data in Myers and Cadigan (1995) show a drastic decline after 1990 for all year-classes. Fishing is a major source of mortality: the fishing mortalities (F) calculated using the standard assumptions (Myers & Cadigan 1994) are plotted to show increases over time (Figure 6). This is fairly similar to the graph that results from Myers and Cadigan's (1995) correlated error model; the standard model data is presented here in the interest of simplicity. If F = 1.0 the implied removal by fishing is 63%; an F = 2.5 implies removal of 91%.

In 1977 DFO declared a rebuilding strategy for 2J3KL Cod, aimed at restoring biomass to between 1.2 and 1.8 million tonnes by 1984. This goal was not achieved. The reasons suggested (Rice & Evans 1986) included: insufficient time; effect of depressed stock size on recruitment; fishing mortality higher than planned; target biomass was unrealistically high; and "pure bad luck" which produced unexpectedly poor year-classes. In a recent summary (Lear & Parsons 1993) of the history and management of 2J3KL the lower biomass estimate (Rice & Evans 1986) is not mentioned but a contemporaneous estimate (Baird & Bishop 1986) of 1.2 million tonnes is. It was observed (Rice & Evans 1986) that in the numbers-at-age matrix the grouping together of all fish older than 14 (the 14+ group), coupled with the back-calculation method the number of younger fish in the previous year caused (under the determined survival rate) an over-estimate of the numbers in age-classes about 10 to 13. The previous solution to this technical problem was to ignore the 14+ (assume them to number zero), but this was judged unacceptable because the old fish "form a larger fraction of the spawning biomass than they do of the fishable biomass"; it was thence determined (Rice & Evans 1986), from modified calculations designed to overcome the deficiency in previous methods, that the biomass in 1984 was overestimated by the previous method and in fact was only about 0.5 million tonnes (about one-third of the target level). For 2J3KL the future was recently described as "grim" by the assistant deputy minister of DFO, Scott Parsons, (Whiffen 1997b), who went on to say

"the stock is at an extremely low level, that there are no signs of good recruitment, and without good recruitment, good year classes, the stock may further decline ... I hope it will recover, but the signs at the moment are discouraging".

Table 4
near here

Declines of Cod are seen in virtually all NAFO subareas (Table 4), and for other groundfish as well (DFO 1994). The report states that the population of 2J3KL Cod stock is "in a very depressed state, probably at an all-time low ... at dangerously low levels" (p. 21, DFO 1994). Most stocks are said to have biomasses that are either dangerously low or near or at historical lows (DFO 1994; DFO 1995; DFO 1996i). The 2GH cod stock was fished very heavily after a period of sustained inshore yields in the range of 11,000 t yr-1; the highest harvest was over 90,000t in the mid 1960s, after which catches declines precipitously, yet the TAC set from the mid 1970s to 1993 was 20,000 t, far above any reported catches during that period (Lear & Parsons 1993; DFO 1996b); it is difficult to understand how this treatment translates as management. The stock has effectively been wiped out. The temperature hypothesis has been applied to 2GH as an excuse (DFO/Anonymous 1997 pers. comm.), but only in vague and unpublished form and it is difficult to understand how, if "the climate changed in 1987" (DFO at April 1997 COSEWIC meeting), this could have acted retroactively for 2GH; and excuses using climate notwithstanding, there is no apparent rationale for the TAC of 20,000 t.

Among the most optimistic statements in DFO reports (DFO 1994; DFO 1995; DFO 1996i) was for the 3Ps stock: "Possible strong 89 yr-class, may contribute significantly in '95+". The 3Ps stock is generally regarded as being more resilient than others around Nfld.; at an FRCC (Fisheries Resource Conservation Council) meeting in 1995 (St. John's, Aug. 31, Holiday Inn) opinions (based on bycatches and sightings of cod) from fishermen in that area were virtually unanimous that the 3Ps stocks were at a state near or at the level which would warrant re-opening of the fishery. However, fishermen and participants from other areas reported much more dismal signs (the word 'desert' was applied by one participant to 4R, the west coast of Nfld., particularly the west coast of the Northern Peninsula) and urged caution, and there is little evidence of strong recruitment. Fisheries were re-opened in 3Ps and 4RS (TACs of 10,000t and 6,000t) in 1997. The data from the 1997 reopening are not yet available, but it seems unlikely that the FRCC condition (that the majority of the catch not come from a narrow age range) will have been met. Informal reports indicate that in 1997 'slinky' fish and 'jellied muscle' were still being found on the eastern and western coasts of Newfoundland; the DFO reports assessing the frequency of these, the 1997 3Ps and 4RS fishery results, and the population estimates, are awaited.

The summary (Table 4) of the DFO groundfish report (DFO 1994), the main thrust of which seems not to be disputed (FRCC 1994), indicates a net downward trend has (depending on stock) not been abated or not significantly reversed (DFO 1995; DFO 1996i) and stock levels in most cases represent historic lows.

W. Doubleday commented (DFO/Doubleday 1996 pers. comm.) that "cod stocks in southern Newfoundland [3Ps], the northern [3Pn-4Rs] and southern Gulf of St. Lawrence [4T(-4Vn:nov-apr)] have begun to rebuild to the point where the FRCC has recommended a limited reopening of the commercial fishery for these stocks for 1997". The background of the FRCC recommendation is dubious: the FRCC recommendation (FRCC 1996) was at odds with DFO reports (DFO 1995; DFO 1996i), and also bluntly at odds with sections of the same document (see section on FRCC in this document). The advice from DFO appears in comments: "potential for rebuilding can be squandered easily if fisheries are reopened while a stock is still vulnerable" (DFO 1996i).

For 3Ps, "Cod from the 1989 year-class are now mature and must be protected ...increase in 1995 survey due to one large tow ... considered to be at a low level of abundance" (DFO 1996i). "Given the uncertainties and the lack of a firm conclusion on current stock size in the inshore, it would be necessary to get more positive signs before considering reopening of the fixed gear fishery at historical levels... there is an unqualified risk of over-exploitation if these inshore components are restricted to localised areas such as Placentia Bay... although inshore information has not indicated a decline it has not been established that the indices are capable of indicating a decline were one to occur" (Shelton et al 1996b). The fishery in 3Ps was re-opened with a TAC of 10,000t in 1997; results of the data arising from that fishery are not yet available as of Jan. 1998.

For 3Pn-4Rs "stock remains very low ... recent year-classes (1991, 1992 and 1993) are poor" (DFO 1996i), FRCC takes the more optimistic view "weak 1991-1992 year-classes, 1993 likely to be better" (FRCC 1996). In 1997 the fishery was re-opened with a TAC of 6,000t; results of the data arising from that fishery are not yet available as of Jan. 1998.

For 4T-4Vn:n-a "biomass close to lowest level observed ... spawning biomass would increase slightly (5%) if there is no fishery in 1997" and "some signs of improvement in recruitment but it is too early to be certain ... biomass close to lowest observed" (DFO 1996i). FRCC recommended (FRCC 1996) reopening (again contradicting its own appendix in the same publication) for 1997, but this area was not re-opened.

For the two Gulf stocks the "prospects for 1997" are listed as "no change", and for 3Ps "to be determined in August 96". Therefore, any recommendation for re-opening a commercial fishery in 3Ps should have been based on very new information, but a year later it is no more clear what that information was. Note that a 5% increase is all that was expected for 4T-4Vn:n-a, and compare the statement "fishing capacity is so high now that the fishing mortality exerted on the southern Gulf of St. Lawrence cod in 1993, with the fishery being closed most of the year, is estimated to have been at the F0.1 level" (DFO 1994); this suggests that a 5% increase could easily be turned into a further decline if fishing is resumed in 4T. See CAFSAC (1989) for a discussion of the 1986 survey which later proved to have been optimistic; to use a single year of apparent improvement to justify increasing fishing mortality carries perhaps a greater risk now than it did then. Confirming the need for caution, at public meetings in 1995 the FRCC (1995) asked the question 'how do you decide when to re-open a fishery', and used as an example the then-recent survey in 3Ps that found generally low abundance except for a single haul (DFO 1996i). FRCC has suggested criteria for re-opening fisheries; these criteria include: 1. "spawning biomass must have increased significantly ... target level ... is halfway between the level at closing and the average level over the historical period" 2. "average [of 3 consecutive years] recruitment, as defined [by age] above, must also have recovered from its value when the fishery closed, again to a level half-way to the historical average" (FRCC 1995); it seems implausible that FRCC's criteria have been met so soon in the stocks concerned. Myers and Cadigan (1995) caution "assessment biologists should be humble about their knowledge of stock status ... estimates for any one year simply are not sufficiently reliable to draw firm conclusions because the error is so large", and recommends that "management strategies should be sufficiently conservative that an overestimate of abundance in ... two consecutive years does not lead to overexploitation".

Myers et al (1995a) suggest that, because significant depensation (a tendency of depressed populations to produce fewer recruits than required for replacement) was found in only 3 (two pink salmon and one herring stock) out of 129 fish stocks for which adequate data exist, that in general recovery of stocks from overfishing is possible. The analysis cannot respond to depensation or any inflection which occurs at a level below the minimum observed stock size in the data; in other words not all stocks have been thoroughly "tested" for depensation by driving them to extremely low population levels. For example, the depensatory responses of the pink salmon (Alaska) are delineated at levels as low as 1-50,000 (Sashin Creek stock) individuals; this is much smaller than the 20-30 million estimated Cod population of 2J3KL (Myers & Cadigan 1994). Thus, only limited comfort can be taken from the statistical non-significance of depensation for 126 stocks, while the 3 constitute a warning that depensation can happen. Depensation, however, is one of several types of population dynamics that can explain failure to recover. It is possible for populations to have multiple stable states, and something like this may be the case for Cod in Divisions 2GH. Cod in 2GH were fished heavily in the 1960s, ending in the early 1970s, and have shown no signs of recovery up to the present, nearly 30 years since their heavy exploitation and yearly catches less than 500 t since 1985 (Lear & Parsons 1993; DFO 1994; DFO 1995; DFO 1996i); (TACs were not based on assessment, and remained at 20,000 t from 1974 to 1992, but catches never came close to this TAC).

Various hypotheses have appeared to explain the decline of Atlantic Cod. These and other suggested causes of, or contributors to, the decline are discussed under "Limiting Factors".

Habitat go to title toc

The range of Cod extends over the continental shelves and slopes, with some seasonal movements: in most areas Cod tend to move inshore during spring and summer, and offshore during winter. Rose (1993), using acoustical gear located dense aggregations of 2J3KL Cod foraging, feeding and spawning in about 350m of water; Leim & Scott (1966) described spawning as usually in waters to 150m in the northern part of the range and to 90m in the southern. That Cod are now found in deeper water (the difference between Leim & Scott's and Rose's depth figures) may not reflect change in habit as much as improvement in our technical ability, but whether they could reflect both has not been asked.

A description of habitat type within the range is susceptible to bias due to limitation of sampling efficiency in un-trawlable areas.

Distribution of Habitat & Trend in Quality and Quantity of Critical Habitat go to title toc

The habitat required by Cod is extensive, to accommodate the adult seasonal migrations, feeding, as well as the larval dispersion/migration. The principal threats to the habitat are fishing activity, loss of materials (such as non-biodegradable net scraps which can continue to fish for years) overboard from ships, and mineral exploration. Mineral (mainly oil and gas) exploration is as yet limited and subject to separate environmental evaluation.

Adult cod are found in much less of their previous habitat (FRCC 1997) than previously. Cod are virtually absent in the offshore of 2J3KL and 2GH. They have also become more concentrated as their numbers have diminished (Myers & Barrowman 1995). While fewer fish might be expected to occupy a smaller area (in total), the increase in aggregation is not accounted for. The loss of spawning aggregations (acknowledged by DFO at the April 1997 COSEWIC meeting) and the evidence for loss of inshore stocks (Ames 1997; Myers et al 1997) is consistent with the loss of occupancy.

Habitat requirements necessarily differ according to life-history stage. Eggs, larvae and adult Cod have different trophic positions and are unlikely to all share the same habitat requirements. Little information is available on the habitat requirements of pelagic stages; the assumption is that the habitat is acceptable and that food availability and predation intensity determine survival. Leim & Scott (1966) list the food of Cod: larvae ('fry') consume copepods, barnacle larvae and other small crustaceans; of young adults consume shrimp, small lobsters, euphausiids, mysids, etc.; of adults over 50 cm consume predominantly fish (herring, capelin, sand lance, and others including younger Cod), as well as molluscs, tunicates, ctenophores, brittle stars, sea-cucumbers, marine worms and also sea birds. If high levels of trawling reduce the population or distribution of these prey items, then it will also influence the population or distribution of Cod.

Despite that it is widespread, fishing has been subjected to scant environmental evaluation, with few exceptions (e.g., ICES 1992; GuignŽ et al 1993; GuignŽ et al 1995). This is presumably because fishing is culturally accepted and the changes in technology have been incremental. The potential impact of fishing on the habitat has undoubtedly increased with the increase in power of ships and the introduction of mobile gear, and "it has been demonstrated that this gear [otter-trawl] could have an impact on bottom habitat" (FRCC 1997). Trawling involves heavy gear and disrupts the bottom; how it affects the productive capacity of the fishing grounds is not yet established although recent work (Auster et al 1995) both strongly points to a methodology for studying this question and gives convincing data to suggest that trawling does severely impact the benthic community. Their study (Auster et al 1995) provides compelling evidence that trawling does disrupt the benthic community; they reference other studies are detail recent work which monitors the decline in biodiversity with trawling. The study was able to use untrawled sites and monitor them as trawling continued, finding, for example, that in 1992 a site previously surveyed in 1987 which had until shortly before 1987 been inaccessible to mobile fishing gear:

"much of the thin mud veneer was missing, exposing more of the gravel base, most of the epifaunal species were not present, and the extensive sponge community was reduced to the occasional small colony attached to the large boulders. Evidence of boulders having been moved could be seen in the video images" (Auster et al 1995).

If the findings are repeatable they may in part explain the situation Cod are now in, may be consistent with observations of changes in the survey patterns (Warren 1997), and may show the way to reverse it and prevent a re-occurrence. Trawling may perturb habitat because it mixes sediments, damages benthic organisms on which Cod feed (directly or indirectly) and which may provide shelter to smaller fish. In the short term this disturbance could increase predation both on benthic invertebrates and on smaller fish -- perhaps that is the origin of the perception that trawling is good (DFO/Doubleday 1996 pers. comm.). Fine materials take time to re-settle, and during that time may experience measurable displacement due to bottom currents; if the bottom currents in an area tend to favour a small set of directions then re-transport would likely be unidirectional and the effects cumulative. If trawling intensity is sufficiently high, large quantities of low-density organic material may be advected away from certain areas and not remain available to sustain the base of the local food chain. In addition, settlement of large quantities of organic material in untrawled locations may be in excess of the processing capacity of the benthic communities in those locations. Vent-like communities (pogonophorans, etc.) are reported on western slope of the Grand Banks and may be sustained by winnowed organic matter from trawling on the banks (P. Schwinghamer, DFO, pers. comm.). Compared to that applied to mineral exploration, the scrutiny given to the effects of fishing is less than token. Although petroleum exploration and exploitation is likely (but not certain) to exert a greater effect than trawling within the immediate area of activity, trawling is potentially a larger source of effects on benthos and fisheries simply because of the number of trawlers.

General Biology go to title toc

General biology has to be viewed with the caveat that much of what we would like to know is not well known, and the possibility exists that we particularly lack the information needed for prognoses given the unprecedented conditions now prevailing. Harris (1993) makes a strong statement by producing a long listing of things "we do not know" about biology and behaviour of Cod, and productivity and behaviour of the ecosystem.

Reproductive Capability

Breeding Age, Breeding Frequency, Number of Young, Age-Sex Ratio, Population Structure, Reproductive Rate

The growth rate of Cod varies, and the size at maturity also varies, beginning at roughly 5 years depending on region. Cod are potentially long-lived: "up to 16 years not uncommon, [but] younger fish make up the bulk of commercial catches" (Leim & Scott 1966). Such old individuals are now virtually absent in 2J3KL (Myers & Cadigan 1995) and at least very rare anywhere else. Adult Cod are capable of spawning repeatedly, presumably each year, given adequate nutrition. A female Cod of 1m total length produces about 5 million eggs 1.2 to 1.8mm in diameter, spherical, and buoyant in waters of 30ppt salinity (Leim & Scott 1966).

Repeat-spawning fish produce larger eggs with higher mean viability than first-spawning fish: an experiment to evaluate the difference showed that for fish of the same weight the production of larvae by first-spawners was only 12% of that of second-time spawners (Trippel 1996 ms); much of the loss was due to egg non-viability and failure to fertilise, which point to a combination of physiological and behavioural factors. Larger fish produce many more eggs and over a more extended period, and these generally with larger size, and hence greater likelihood of survival. The shorter spawning window of younger fish is supported by field data (Hutchings & Myers 1993).

"Typical" species attributes (e.g. age distributions) do not apply to reduced population(s). Size- and age-at-maturity are reduced concomitantly (Trippel 1995) with population size in Cod, Haddock, Witch Flounder, American Plaice and other species in the Canadian Atlantic and elsewhere. Because of the recent findings of low efficiency of first reproduction, the earlier maturation of fish provides meagre solace. Further caution is advised when considering the reproductive potential of young fish: "young males more frequently exhibited gonadal condition indicative of ending their spawning season sooner than old males"; "in 9 of the 14 years of sampling since 1976 the very young males (age 4) exhibited a higher incidence of the unspent condition as opposed to those that were slightly older (ages 5-8). Moreover, in 9 years all age 4 males had not initiated spawning" (Trippel & Morgan 1994). Water temperatures in the ranges experienced over the past ten years have apparently not inhibited spermatogenesis: "Station 27 water temperatures ... below normal throughout water column since 1983 and especially cold during the period of gametogenesis prior to the 1991 spawning season ... the findings of this study indicate that spermatogenesis did proceed and was not seriously impeded under these cold water conditions" (Trippel & Morgan 1994). Overall, "the reproductive potential of small adults is not equivalent to that from adults of a range of sizes even when their biomass is equivalent, and to assume so in stock-recruitment relationships may seriously bias estimates of a stock's reproductive potential ... Future investigations of fish-stock recruitment theory are challenged to revise conventional methodologies and incorporate these findings" (Trippel et al 1997).

Species movement go to title toc

Eggs, Larvae, Juveniles

Eggs are buoyant, so they are subject to drift. Suthers et al (1989) explored growth and location of the nursery area for postlarval Cod off southwestern Nova Scotia and showed lower growth at the inshore locations which had been presumed to be nursery areas, compared to that of similar-aged postlarvae further offshore. However, mortality considerations may outbalance the importance of slightly more rapid growth (e.g., Brown et al 1989) such that the relative importance, to eventual recruitment, of one area compared to another may not be absolutely indicated by growth rate differences. Larval Cod are found inshore at some locations during the spring and summer, for example Bellevue beach, NF (Schneider et al 1995 unpublished).

The location of eggs and larvae is initially determined by spawning location, and subsequently influenced by currents and presumably larval migration behaviour. A spawning location probed off northern Newfoundland (Rose 1993) was in water considerably deeper (³350m vs ²55m) than spawning locations documented on Brown's Bank (Suthers & Frank 1989; Suthers et al 1989). Hutchings and Myers (1994a) explain differences in times of spawning of Cod on the banks in NAFO sub-divisions 3Ps and 3L in terms of the bathymetry and the temperature of water masses accessible to Cod, such that the relationship of annual temperature indices to (earliness of) spawning time is negative for groups (e.g. 3Ps) within reach, or positive for groups not (e.g. 3L) within reach of warmer slope waters.

The importance of the 'tail' area of the Grand Bank as a habitat for very young groundfish (Walsh et al 1995) makes the Cod population in that area more likely to be affected by overfishing (outside the current Canadian EEZ).

There has been evidence of inshore spawning in Random Island sound (K. Smedbol, pers. comm.) as well as of overwintering adults. Until recently, whether these reflect local populations with distinct properties, or merely aggregations of individuals which happen to have reacted similarly to environmental cues, cannot yet be established. Pepin & Carr (1993) examined variations in a 307 base-pair segment of mtDNA of Gadus morhua from locations from St. Anthony to the NE and SE Grand Banks, and found no consistent spatial pattern of genetic differences in areas north and east of Newfoundland, and with one genotype (A) predominating (71-88%) in all locations. This was in contrast to morphometric characters which were different for inshore and offshore juveniles in the same NAFO division. Because morphometry and some meristic characters can incorporate environmental influence, identifications of stocks based on spawning times and growth rates cannot be definitive. This is where the recent work on microsatellite DNA has shown relatively fine-scale differences: Cod are (where this has been examined) composed of a number of genetically distinguishable population components (Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; FRCC 1997; Ruzzante et al 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b).

Adults

Adult Cod migrate seasonally, inshore in spring and offshore in winter (Leim & Scott 1966; Lear & Parsons 1993; Rose 1993). Associated with this migration is sometimes a movement between stock areas which complicates the evaluation of fishing mortality for each stock.

Breeding range: Spawning is fairly widespread (even if less intense) according to reports of inshore spawning, but thought to be fragmented according to suitable habitat (not defined). The greatest concentrations (in time/space) occur as spawning aggregations, which (instead of being protected) have historically been, for 2J3KL, the source of the most lucrative fishery. There is a pattern in sex ratios in different areas which is appears to be linked with spawning condition (spent/unspent), and fishing on these sex-biased concentrations could affect the overall sex ratio in the stock (Morgan & Trippel 1996).

Breeding times are variable, but do not in all areas respond the same way to climate. An intriguing, contrary, relationship of interannual variability in spawning time with temperature in different areas (3Ps and 3L) is reported (Hutchings & Myers 1994a), which is explained by Cod moving into warmer slope water when shelf temperatures are low in 3Ps - an option not available for Cod in 3L. Thus variation in spawning time in 3L is accounted for by the effects of water temperature on gonad development, and in 3Ps by the effect of behaviour and warmer slope waters in 3Ps.

Behaviour, adaptability

Cod are generalists, consuming benthic or pelagic organisms, with some dietary variation associated with season and location. Not enough is known about the frequency (importance) of inshore spawning, whether there are in fact genetically discrete populations such that replacement is in even greater peril than the aggregate numbers presently suggest. The genetic analysis by Pepin and Carr (1993) pointed to a genetic bottleneck in the history of the northern Cod stocks, which may mean that the population(s) is disadvantaged by lower genetic diversity than exists in the east Atlantic/North Sea. The geographic range in the northwest Atlantic is nevertheless large and encompasses a wide range of temperature, depth, community composition, and this suggests that considerable flexibility has existed until recently.

Limiting Factors go to title toc

Why are Populations Declining?

Collateral Concerns (biological indicators) go to title toc

The symptoms now exhibited by Cod populations are not restricted to decline rates, as the DFO comments attempt to suggest. Populations have virtually all reached historic lows since 1990. Age structures collapsed (e.g., Myers & Cadigan 1994), with negative implications for recovery (Trippel & Morgan 1994; Trippel et al 1997). Further, there are: earlier maturity (Trippel 1995); reduced weights at age (DFO 1994; DFO 1995; DFO 1996i); contracted distributions (Myers & Cadigan 1995); possibly restricted spawning seasons, and reduced reproductive success (Hutchings & Myers 1993; Trippel & Morgan 1994; Trippel 1996 ms; Trippel et al 1997).

STOCK STRUCTURES: Complexity of population structure increases the risk implied by all concerns (numbers, biomass, collapsed ages structures, low recruitment, etc.) for each component. Recent findings confirm long-held suspicions (Templeman 1962; Lear & Parsons 1993) that Cod exist as a number of independent population units: Cod are (where this has been examined) composed of a number of genetically distinguishable population components (Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; FRCC 1997; Ruzzante et al 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b). Further work on historic samples (if available) is needed to determine whether previously unrecognised components were lost, and if so when and in what circumstances. Some populations seem to have been lost already.

Movement of Cod has occurred even across the Atlantic, but this is on the basis of a single specimen (Gulland & Williamson 1962). If such fish migrated as a result of disorientation it may be doubtful that they would successfully reproduce. Stocks or sub-stocks are important irrespective of the mechanism which defines them (whether genetic or not). The evidence indicates that instead of a single population of millions of individuals, there are smaller populations some of which have very much fewer individuals; those smaller populations are at much greater risk, and there is evidence for extinction of local stocks in Newfoundland and the Gulf of Maine (Ames 1997; Myers et al 1997), and this is consistent with the disappearance of spawning populations/aggregations to the northeast of Newfoundland as acknowledged (~"not there any more") by DFO at the April 1997 COSEWIC meeting. Extinction of local populations could explain the failure to recover elsewhere, e.g. 2GH which has not recovered since heavy exploitation in the 1960s.

Some of the concerns listed may be indications of more complex causes of the current status. There are also parallels and contrasts: populations have changed in other species as well, and while some of those changes are positive (e.g. shrimps), in general groundfish species have declined (DFO 1994). The pattern, among species or phyla, of similarities and differences in population trends may prove instructive if the Cod declines are in part an ecosystem problem.

Somatic growth rate (length and weight-at-age) has apparently declined. Recruitment has declined (drastically). There is increased spatial concentration of the fewer Cod remaining (Myers & Barrowman 1995; Myers & Cadigan 1995). There are changes in timing of migration, earlier maturity (Trippel 1995). There has been greatly reduced frequency of strong year classes in the 1980s onward and since 1990 there has been no strong recruitment (DFO 1994; DFO 1995; DFO 1996d; DFO 1996c; DFO 1996h; DFO 1996n; DFO 1996j; DFO 1996e; DFO 1996m; DFO 1996i; DFO 1996b; DFO 1997a; DFO 1997c; DFO 1997d; DFO 1997b). Collapsed age structures are argued to have severely reduced stock rebuilding potential (Trippel & Morgan 1994; Trippel 1996 ms; Trippel et al 1997) out of proportion to the (already great) loss of biomass.

Reduced somatic growth rates may point to causes not directly associated with removals. Because individuals within a population compete for available resources, individuals remaining in a population which has been reduced arbitrarily by removals should grow better because of reduced competition for the same resources. Thus, it would be expected that, now that in most management areas Cod populations have been not only decimated (reduced by a tenth) but reduced to the order of a tenth, and sometimes less, individuals should demonstrate improved growth on account of reduced within-species competition for resources. Such an effect was presumed to have occurred in the mid-1980s for 2J3KL (Hutchings & Myers 1994b). However, not only has this not been seen since 1990, but the opposite has been reported for most stocks (DFO 1994). This should inform us that as well as facing high mortalities due to fishing, there may be some negative effect from lower in the food chain (Warren 1997), or in the general environment. General environmental hypotheses include temperature changes (DFO 1996f; DFO 1996g; Drinkwater 1996; DFO/Doubleday 1996 pers. comm.; DFO/Anonymous 1997 pers. comm.; Drinkwater & Mountain in press), habitat changes due to trawling (ICES 1992; Auster et al 1995), and other sources of environmental variation.

The lower weight-at-age may (alternatively) be thought to result from either size-selective fishing mortality with effects being propagated through successive ages (Krohn & Kerr 1996), or a complex interaction of better growth rates which induce an earlier maturity and, in consequence of a nonlinear relation of growth to age in which the inflection is determined by age of maturity, earlier reduction in somatic growth rates after maturity (D.H. Steele, Biology, M.U.N., pers. comm.). The questions relating to this arbitration are whether the residuals in a regression of weight on age are related to the number of spawning checks observed; and whether larval and juveniles (pre-recruited) fish show higher or lower growth rates than historically observed. Laboratory studies did not indicate a measurable tradeoff between somatic growth and spawning (Fordham & Trippel 1996 ms).

Indices of concentration for 2J3KL Cod have increased over the decade (Myers & Cadigan 1995): although the Cod population is reduced, it is even more clustered; it is not known why this should happen. However, the existence of an apparent age-related tendency to aggregate in certain areas is noted in division 3NO (Walsh et al 1995), so it is possible that the observed increase in concentration indices (Myers & Cadigan 1995) could include the effect of a demographic change; if not demographic, the change in concentration may be useful in helping to identify the barriers to rebuilding of the stocks.

The major limiting factor is that too many fish have been caught without having an opportunity to reproduce (and may still be caught to the extent that illegal and unregulated fishing occur within and without the current Canadian EEZ, and as bycatch in shrimp and other fisheries). This point was raised by fishermen and participants at the FRCC consultations in St. John's (Aug 31 1995) and scientists in the main agree. A minimum size (NAFO 1995) of 41 cm (length, for whole Cod) allows immature fish to be caught; as such, it places too much reliance on TACs as a management tool. If the minimum size subject to fishing mortality were set at the size which had reproduced twice, it would be impossible to fish out the stock. When immature fish are subject to fishing mortality, experience shows that it is possible to destroy a fish stock. Immature fish must be protected from fishing mortality, the more so when they are illegal to take and therefore tend to be discarded without reporting.

Complex Systems and Uncertain Prediction go to title toc

It is unusual to have a stock status report requested on a species for which as much data are collected as for Cod. There have been COSEWIC reports on species of commercial importance, e.g., the Longjaw Cisco, Coregonus aplenae, which once supported a commercial fishery but which is now presumed extinct (Campbell 1987); the Lake Sturgeon which once was fished at nearly 2 million t.yr-1 in the late 1800s across much of central Canada but now supports catches of only 2% of that or 40,000 t.yr-1 (Houston 1987). In both cases, total populations are (were) not well known although there were regulations concerning their exploitation.

By contrast, hundreds of competent people spend entire lifetimes gathering and interpreting data relevant to assessment of population size and understanding the biology and complex ecological relationships of Gadus morhua. It is well acknowledged however that the complexity of dynamic systems with many components (i.e. more than 3) and many links among components (i.e. dependencies of numbers of one component upon numbers of other components) renders prediction of future state extremely unreliable in the medium to long term (May 1976; Gomes 1993). These uncertainties increase in response to uncertainties about the structure of the system, and in response to uncertainties about the interactions strengths. The best current understanding indicates that these are unavoidable properties of complex systems. There is no doubt that marine fisheries belong to the set of complex systems (e.g. Gomes 1993) -- also termed chaotic (Lear & Parsons 1993). The unpleasant property of unpredictability can be difficult to understand from a background which does not include experience of complex systems, randomness, and probability. Thus there is potential for a rift of understanding to exist between the system (containing the potential for complex dynamical behaviour), the scientists (who understand that some such behaviour exists but know there are limits to accurately predicting it), and the management decision process (which permits input from persons without any experience of even nonlinear, let alone complex and chaotic systems). This rift of understanding may act undesirably as a filter such that management initially regards as reasonable levels of harvest which later turn out to be unsustainable.

Acknowledging that the system is complex, it must be acknowledged that many of its components and links are not sufficiently well understood to build a model. These (early life history, habitat requirements, low-level food chain productivity, interaction with benthos and the effect of continuous trawling on it, etc.) are receiving some (e.g., ICES 1992), but insufficient, attention. Management can act conservatively in respect of the strongest links in the system, even in the absence of an operable and reliable comprehensive prediction model which would include subtle, indirect and minor effects. Fishing effort is a dominant source of mortality in recruited size-classes of these stocks, and the data indicate that fishing effort has resisted regulation. The 'causes' nominated for responsibility in the current crisis are in most part claims that certain links which could appear in a complex model are real and strong. However, too little is known to allow confirmation or refutation of some of these claims.

Hypotheses on Possible Limiting Factors go to title toc

Various causes have been suggested for the decline in Cod stocks. At present many of these are un-tested hypotheses, which is why they don't appear under 'Limiting Factors' but instead are listed and discussed briefly below:

Temperature changes (DFO 1994) have been alluded to for 2J3KL in the general context of groundfish stock declines. Indices of environmental temperature have shown net decline over the past few decades, although there is considerable short-term variation in this signal, and the longer record of temperature and climatic indicators includes apparently similar conditions at previous times (p. 16, DFO 1994) when the Cod stocks were apparently healthy and productive. Myers et al (1993) in discussing the possible salinity-recruitment relationship use ice extent as a proxy for salinity and find slightly ambiguous support for the proposed larval-freezing mechanism which may underlie the salinity-recruitment relationship. The degree to which potentially explanatory variables are correlated generally increases the difficulty of identifying mechanisms from associations in data. The temperature data may be subject to variations in distribution of, as well as average value of, temperature. Lower water temperatures do not necessarily mean lower fish weights, because the conversion efficiency (food into fish weight) has been found to increase as temperature decreases (Brown et al 1989), which has the potential to increase biomass growth at lower temperatures, all else being equal.

There is no explicit analysis which has survived peer review and which provides statistically significant substantiation for a role of recent variation of temperature in the decline of these stocks, although W. Doubleday has written (1996 pers. comm.) "there is at least one study that suggests temperature has had a significant effect on SPR and could have contributed to the decline of the 2J3KL cod stock ( ... DFO Atl. Res. Doc. 96/81)". Doubleday's mention of this DFO paper is ironic, because this very section of the Sept. 1995 version of this report (given to Peter Shelton and others in DFO) suggested: "short-term variations in individual growth rate do not appear to have been investigated in relation to short-term temperature change. It may be useful to use the short-term noise in the temperature signal to attempt to explain variations in individual growth rate, concentration index of the population, or any other biological variable". Since then a study of that type has appeared (Shelton et al 1996a), but if anything the paper indicates how weak the link is with temperature: the paper considers the annual weight increment (see otolith protocols discussed in this report) and relates this (r2=0.41) to the average area of the cold intermediate layer (CIL) as a proxy for temperature. An r2 of 0.41 on its own would seem encouraging, but an identical analysis omitting CIL still produced r2=0.36; therefore, CIL on its own is describing 5% at most, a very small portion of the total variation; and, as conceded in the analysis results, virtually all the estimates are biased. The paper cannot support the claim that temperature contributed to the stock collapses.

On the other hand, there is explicit analysis of temperature (Hutchings & Myers 1994b) that results in the conclusion "we reject hypotheses that attribute the collapse of the northern cod to environmental [temperature] change", and it is pointed out that "harvests equivalent to those of the past decade were sustainable in the nineteenth and early twentieth centuries in a considerably colder environment". Elsewhere:

"... Between 1980 and 1989, there was a weak positive relationship between the cold intermediate layer temperature anomalies and cod year-class strength (i.e. abundance of 3-yr-old cod). However, it is unlikely that low recruitment levels caused by below-average temperature of the cold intermediate layer alone had a major impact on the recent collapse of the northern Gulf stock. Nevertheless, the poor state of the spawning stock (low abundance, poor fish condition, and less buoyant eggs) and harsh late winter and early spring conditions in the northern Gulf can limit the potential for high recruitment and rapid recovery of the northern Gulf cod stock"(Ouellet 1997).

The Harris report crisply put the climate hypothesis in perspective:

Change in natural mortality (2J3KL) (e.g. see p 21, para. 5, alternative [a] "factors other than fishing mortality", DFO 1994): Note that natural mortality is very difficult to separate from fishing mortality; but because fishing effort is recorded (as ship hours, etc.) it can be tempting to assume a constant relation of effort to F, and (since F + M = Z) to attribute changing Z to changes in M if effort (as recorded) has been constant. This is not sound unless compensation is made for the increased efficiency of fishing fleets with advanced technology and more experience. Myers & Cadigan (1995) found that overfishing was more capable of explaining the decline than a change in natural mortality, and particularly on account of the increase in the spatial concentration of cod which (they suggest) may have helped to maintain high catch rates even when the Cod population was declining.

Seals are frequently mentioned in the DFO report under the category multispecies considerations (DFO 1994; DFO 1995; DFO 1996i). There are two aspects to the 'seals' issue: their role in the decline, and their role after the decline. In the 1994 report's introduction (DFO 1994) seals are said to have consumed about 40,000 tonnes of Cod in the Gulf of St. Lawrence and eastern Scotian Shelf in 1993. Considerations which may modify this are alluded to. If this amount is doubled for 2J3KL it is still dwarfed by the commercial catches of the last few years. However, at the low population levels of Cod the portion of total mortality attributable to seals may be different (greater if the same number are eaten, or smaller if seals switch away from Cod when they are not abundant). Seals may be a consideration in recovery of Cod stocks, but they are not a plausible sole or major cause of the decline because they have been a part of the system for longer than the fishing fleets, and are unlikely to now be at historically high levels -- at the discovery of Cod by Europeans the seal populations could reasonably be assumed to have been unaffected by harvesting and therefore at least at their present numbers. Recent studies of seal diets are unavoidably constrained by the possibility that the alteration of species composition has affected the choice of food by seals; the impact of seals on Cod therefore may be either exaggerated or minimised. Increased concentration (Myers & Cadigan 1995) of the remaining cod may affect predation by seals. The observation that seals' feeding imposes higher mortality on Cod at such low levels of the Cod stocks applies equally well (constant demand by seals, constant demand from fishing) to the impact of fishing during the last years when fishing was still on-going. The complexities of indirect food-web effects (e.g. FRCC 1995) can permit a predator to increase the population of prey by also feeding on its other predators, or on predators on earlier stages. A role of top-level predators in ecosystem structure has recently been modelled, with the indication from the model that they have a positive effect (Pauly & Christensen 1996). A seal cull could be an expense without benefit, particularly unless the seal product is marketable.

Overfishing: there is no doubt that this has occurred, given the frequency with which, and degree to which, the F0.1 catch levels and TACs have been exceeded (DFO 1994). The stock size trajectories in many ICES areas (ICES 1994) are similar, and overfishing is fairly openly acknowledged in that analysis. It has been shown that cod became more concentrated over time in the years leading up to the moratoria, and that the concentration of fish may have made them easier to catch and thus increased fishing mortality (F) without concomitant increase in effort (Myers & Cadigan 1995).

There is precedent for commercial extinction of fish stocks due to overfishing, e.g. Haddock on the Grand Bank, (p 7, DFO 1994), and Gulf Haddock doesn't seem to get a mention after 1986 (Clay 1986). W. Doubleday disputes the role of fishing, however:

"Haddock on the Grand Banks probably were never a long-term sustainable stock. In favourable conditions a few strong recruitments occurred. The fishing did not drive the stocks to commercial extinction. Haddock on the Grand Bank are at the northern limit of their distribution and any perturbation of the environment, specially lower temperatures would tend to cause recruitment failure, which is essentially what happened" (DFO/Doubleday 1996 pers. comm.).

Yet, Haddock stocks were the subject of advice, presumably with the intention of sustainability, in the neighbouring zone, 3Ps: "Although CAFSAC continually advised that catches should be restricted to by-catch only through the 1980s, TAC ranging between 150 to 3,200 t were put in place from 1987-1992" (DFO 1994).

Overfishing combined with low maximum rates of population growth (or low maximum rate of increase, modelled as the a-parameter of Ricker's stock-recruit equation). The a-parameter has been fitted (Myers et al 1996) for a number of stocks in the northeast and northwest Atlantic and showed a positive relationship of a with temperature. The a-parameter is positively correlated with temperature, meaning that west Atlantic stocks in cooler water are generally less able to sustain high fishing levels than are European stocks (note that this is not the same as the cold-water hypothesis above). With regard to this point, W. G. Doubleday (pers. comm. 1996) has noted "We know northern stocks cannot support the levels of fishing that the North Sea can. We have always aimed at much lower management targets, as well". Certainly F0.1 has acceptance as a conservative harvest level, but as acknowledged (DFO/Doubleday 1996 pers. comm.), the F0.1 target was not attained; harvest rates were far higher (Fig. 6) and increased through the 1980s and early 1990s to over F=2.0 (removal of 86%) for several of the age classes above 5 years.

Changes in lower trophic levels of the food chain are suggested by the reportedly reduced growth rates observed for several stocks (DFO 1994). (In considering growth rates, either weight or length divided by age, note that they are susceptible to any errors or biases that might be incorporated in ageing; so, because the recent ageing methods are undisclosed (D.B. Atkinson, DFO, pers. comm.), and because it seems that the readings were not done blind (Wells 1981; Wells 1983; Wells 1986), the reduced growth rates should be viewed with caution.) Analysis of spatio-temporal structure inferred from research cruises shows a structural "breakdown" coincident with the decline in abundance, which is considered consistent with (Warren terms this conjectural, thus cautioning the reader against reading his result as a confirmation of any hypothesis) the expected consequence of declining availability of prey species (Warren 1997); this work seems to suggest directions in which to look. Changes in growth rates could be habitat-related, but data on Cod diets (at all life-history stages) and on the response of bottom community condition to trawling history would be required to evaluate this, and the data on diets are too few (Lilly 1995) and the benthic studies are at the stage of developing methodology (GuignŽ et al 1993; GuignŽ et al 1995). Unfortunately, there do not exist untrawled grounds which can act as controls, and historic data on benthic communities is insufficient (P. Schwinghamer, pers. comm.); the present moratorium could provide an opportunity to examine benthic communities in an un-fished regime.

Poor recruitment/ juvenile mortality/ discarding: the hypothesis of poor recruitment is not supported by 2J3KL research vessel surveys, which show approximately normal recruitment (Myers et al 1995b) assessed as numbers of age-3 fish. However, these recruits do not seem to show up as later ages, and therefore are experiencing mortality, but are not accounted for in the commercial catch data. Myers et al (1995b) ascribe these disappearances to juvenile mortality associated with fishing. A particular component of overfishing is indicated by discrepancies between the research survey data and the VPA (from commercial data): in all of the 6 Cod stocks they examined, high juvenile mortality was associated with high adult mortality, "consistent with the hypothesis of discarding" (Myers et al 1995b). Myers et al. recommend that "high priority should be given to the measurement of discarding levels and the extent to which catch misreporting is related to changes in fishing mortality".

Demographic effects of overfishing in 2J3KL combined with age-related changes in length of spawning season: Hutchings and Myers (1994b) postulate that the current spawning season, being dominated by 'short-season' younger spawners, provides less opportunity for a meaningful number of larvae to coincide with conditions conferring good survival and the consequent generation of a strong year-class. The biomass-specific reproductive (and stock-rebuilding) capabilities of younger fish are now known to compare very poorly with older fish (Trippel 1996 ms; Trippel et al 1997).

Salinity: a model (Sutcliffe et al 1983; Myers et al 1993; Shelton & Atkinson 1994) has been used to model recruitment variation in 2J3KL Cod, and appears to have been quite successful (Myers et al 1993), for the data up to that covered by the 1992 assessment for recruits (recruits are defined as 3-year-olds, and identified by their presumed hatch year; the year in which the assessment is made is thus about 4 greater than the latest year-class for which a recruitment estimate is available) of year-classes up to 1987 but the salinity term becomes non-significant (Shelton & Atkinson 1994) when spawner biomass is included and the 1993 assessments of recruitment (for years 1960-1988) are used. Shelton and Atkinson offer two explanations for the difference in the performance of the model on different groups of years of data: "either the [correlation] is spurious in the first place, or the 1993 assessment is considerably in error. ... A further, more complicated explanation is that salinity was an accurate predictor ... but is no longer ... the departure between the [salinity] model and the 1993 assessment estimates [meaning the best look back at a number of years, making use of data first available in 1993] commences in 1983. Although the 1993 assessment [cited] shows that spawner biomass was relatively constant at low levels between 1982 and 1988, and only declined sharply thereafter, it is possible other changes in the population influencing early stage survival commenced earlier." These are intriguing observations, and lead to questions about the basic biology of Cod early life history. For example, given the complex changes in demography, distribution and possibly spawning season and location, is it possible that early- and late-spawned eggs respond differently to salinity variation, and that this may account for the varying success of this model? Indeed, although Shelton and Atkinson's (1994) figure 1 shows an increasing departure between mean levels predicted by the Sutcliffe model and the 1992/3 estimates for 1960-1987/8, inflections in the predicted trajectory are mirrored by inflections in both sets of abundance estimates (using either the 1992 or 1993 estimates). This would be consistent with their suggestion that salinity may not have a direct effect, but may be aliased to some other process which does.

Learning: an allusion in Rose (1993) to alterations in migration patterns of Norwegian spring-spawning herring following depletion of adults in the late 1960s ("original patterns have yet to be re-established") is the basis for conjecture that following severe stock depletion some Canadian Atlantic Cod stocks may also lose the ability to follow the present or past migration routes. There are parallels in salmonids and wildfowl which home on cues which are part of individual experience. There is a considerable body of opinion that supports the existence of multiple stocks within 2J3KL (Lear & Parsons 1993; Myers et al 1997), and yet the DNA support for these until recently was minimal (Carr & Marshall 1991; Pepin & Carr 1993) with some suggestive observations (Ruzzante et al 1995). Since 1995 the microsatellite work (Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; FRCC 1997; Ruzzante et al 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b) of the Dalhousie group has shown that genetically distinguishable populations exist on a within-management-areas scale, which raises increased concern for loss of stock components. Rose's suggestion may be one mechanism whereby stocks remain distinct; if so, the elimination of local stocks could have near-permanent consequences, as acknowledged by a recent workshop (DFO/Rice 1997).

Summary (of Limiting Factors) go to title toc

In summary, the biological indicators include poor recruitment, poor growth, collapsed aged structures, reduced areas of occupancy. Collapsed ages structures imply reduced spawning season lengths and reduced contribution of successful larvae. Lack of old fish implies reduction in quality of reproductive products. No successful account has yet been made which explains both the collapses and the failure of populations to recover as expected. The observations suggest a broad-scale change that affects many species, but are groundfish more affected than pelagic fish? Fish more or less than invertebrates?

Of the many hypotheses posed to account for the decline in biological indicators and in Cod stocks, most are not tested and overfishing appears to be the strongest; fishing mortality levels have far exceeded the F0.1 target in many areas (Myers et al 1993; Hutchings & Myers 1994b; Myers & Cadigan 1995). Furthermore, the magnitude of mortality inflicted on smaller fish, either as bycatch (discarded), or unreported catch, has been proposed (Myers et al 1995b) as a potentially significant contributor to the failure of recruitment.

Although it is tempting to rule out fishing as a factor which could affect a large number of non-target species, intense fishing on one species can have large effects on bycatch species. Even species targeted in one fishery may be by-catch in another. Industry data have been known to be unreliable in some areas (Gulf of St. Lawrence); there is no reason to suppose that data will be generally much more reliable in other areas, and it is particularly unlikely that data on discards approaches the quality aimed for with targeted species. Non-landed discards (fish spilled from trawl windows, animals damaged by the passage of the gear) are not even directly evaluated. The ecosystem effects of fishing, in all its aspects, require greater attention.

Special Significance of the Species go to title toc

The economic contribution of Cod has driven much of the settlement of eastern Canada and remained until recently a mainstay of coastal communities. As is obvious from its high productivity, Cod has also been a major feature of the northeast Atlantic ecosystem.

Tens of thousands of people are or were directly employed in harvesting and processing Cod in eastern Canada. The contribution to the local economy has been claimed to be as low as 6% and as high as 70% (Steele et al 1992). Much public attention remains focused on the prognosis for the fishery. The compensation for just those directly involved in the fishery and processing sectors has amounted to many billions of dollars; the further cost to the Canadian economy has been much greater.

The situation with Cod is mirrored in the situation of other groundfish species (DFO 1994), suggesting a common cause. Climate has been nominated by DFO, but not shown by evidence thus far. Habitat changes due to other causes, notably the interaction of gear with the benthic community, needs to be addressed. If separate stocks with heritable behaviours, growth or migration patterns exist for other species as well as for Atlantic Cod, the situation may be worse for some individual stocks or some may have disappeared.

Utilization go to title toc

World trade in Cod amounts to hundreds of thousands of tonnes, and has been higher when higher fishing levels have been permitted or possible. Historically, Cod was salted and dried to a state that had a storage life of several years. The portability and durability of this product made it a staple even in some areas where Cod itself was unknown (e.g. the West Indies). Cod is now produced as either frozen blocks (for further processing), filleted on board (factory ships), or as fresh fillets. It is marketed as fresh and frozen, as partially prepared sticks and pieces, and has also been used to make artificial crab meat. There is a small market for cod cheeks, cod tongues, and ovaries ('britches'). Cod liver oil has been a tradition in home remedy for most of this century, and to provide this the cod livers are rendered with low heat to extract the oil. Much of the offal is discarded although it may be suitable for some grades of fish meal and fertilizer. Attempts to make leather from the skins of Cod have been technically successful (L. Gushue, pers. comm.) but have lacked market development and materials supply.

Recommendations/Management go to title toc

(DFO has tried to argue (DFO/Anonymous 1997 pers. comm.) that the very suggestion,here, of management changes implies that the Report does not take seriously the idea of extinction (see Addendum / Report sections / recommendations). Certainly, if all Cod stocks were already extinct the analysis of management might be an academic exercise. But, as there seems to be some opportunity to reduce the threat of extinction of Cod stocks, it is of obvious relevance to consider measures that address both the declines observed and the management structures and methods that are capable of some improvement.)

The ability to doubt is a fundamental and essential feature of science. To make the best use of science and to guard against recurrence of stock failures, managers must allow for the open discussion of any scientific matter, even where it may impinge on policy; surely guidelines can be formulated to preserve science without unduly impeding administration.

The assessment process in the past used independent analyses from research survey and commercial catch data (as described above, these were greatly divergent in general) and used the midpoint between the two estimates on the basis that neither one could be justified over the other (CAFSAC 1989); prevention of overfishing in the short term at least might be achieved by using the more conservative estimate, until it could be determined where the real value was, relative to each estimate.

The F_0.1 level was selected and agreed upon as a DFO target over a decade ago, but rarely met (Steele et al 1992). That scientific advice was so frequently modified (Steele et al 1992; DFO 1994) from F_0.1, to Advised TAC, to setting the TAC, and that TACs in turn were so often exceeded by declared catches (which are widely thought to be conservative estimates of total catch) is an indictment of the process, not of the science. The depletion of Cod stocks in ICES areas is similar, and similar pressures probably prevail in both; the same can be said for the decline of major fisheries worldwide. A limiting protocol is needed so that TACs are legally bound to not exceed a limit set, independently and openly, by science.

The controversy surrounding possible causes for the current situation shows that too little is known about the biology and ecology of Cod, and the ecology of the ecosystem. While more effort must be put into policing and management functions, these give no certain guarantee of protection in the long term. The cost of heavy exploitation of any complex system is that probability of failure (or change) of the system is increased; the increased risk must either be accepted, or must be mitigated by informed adjustments to management together with strict controls.

Minimum sizes, mesh sizes, and fishing practises must minimise, and if possible eliminate, fishing mortality imposed on immature fish. Although it has been remarked that the cod (juvenile) bycatch in the capelin trap fishery has some use as an early source of data on year-class strength, this may not be worth the cost until the stocks have recovered. Discards and by-catches need to be accurately reported. The motivations to report or not report discards or by-catches need to be acknowledged in regulations (for example, if a fishery is liable to be closed because of excessive by-catch of another species, this will be a strong counter-incentive to comply with reporting) and in relations with the fishing community and industry. There must be improved compliance with gear, fishing area, and quota regulations. Regulations and conservation for different fisheries should be integrated as much as possible to avoid loss of juvenile Cod in (e.g.) smaller-meshed capelin traps; mesh sizes in Cod traps should allow the live escape of fish below the minimum size, because they may not survive (mortalities result from expanding airsacs as the trap is hauled) if brought to the surface. Technical developments in fishing gear may be needed in order to minimise these losses.

The best protection for the Cod fishery is any measure that will protect fish until they have reproduced at least once, preferably twice or more because of the poor biomass-specific contribution of small/young fish to stock rebuilding and the poorer quality reproductive output of the first-time spawners (Trippel & Morgan 1994; Trippel et al 1997). The best way to do this would be a regulation that incorporated an incentive. The idea is to link the average size caught to either a fee paid or to the amount to be deducted from remaining TAC. For example, this could be done using a Target Weight (TW) which would represent the average weight of a fish that had reproduced once, twice, or whatever was determined as the ideal. A formula that either reduced remaining TAC by a greater amount for the same weight in smaller fish, or charged a quota fee (QC) which was greater for the same weight in smaller fish, or a bit of both, would provide that incentive. At some point along the fishing and processing path fish are measured anyway. Only two constants would need to be set by regulation: TW and UNIT_FACTOR (a charge per unit weight, like stumpage). The quota charge levied for some quantity of fish could be calculated as:

QC = UNIT_FACTOR * total_wt. * TW / mean_wt (eq. 1)

The QC could be applied as a bill in dollars (like a stumpage fee) for each quantity of fish landed (those with lower average weights would be paying more per Kg), or it could be converted to a TAC reduction, or both. In the case of a TAC reduction there is the difficulty that it would affect all harvesters; but the advantage that comes with that is some peer pressure that might be beneficial. For any harvesters with a set TAC portion allocated to them, the TAC reduction would affect them only. A stumpage fee might be the easiest to administer. The stumpage rate for each size category is amenable to being linked, according to a scale which could be re-evaluated from time to time as necessary, to the potential contribution of the catch to future recruitment, according to developed relationships of viable gamete and zygote production to size (Trippel et al 1997).

Approaches along these lines, combined with prohibition on discards (there is little point in discarding if fish are dead already), would link the conservation incentive with costs and provide better data on mortality of sizes liable to be discarded. At any rate, selective discarding of fish above the minimum size (high-grading) should be prohibited.

Further useful suggestions for alternative (i.e. numerical) quota regulation were proposed by fishermen, and are listed under "Evaluation".

To improve management, there must be improved understanding of other factors including: lower levels of the ecosystem (primary production, interactions among ecosystem components); ecosystem effects of trawling compared to effects of other types of fishing (e.g., baited hooks, traps, gillnets); survival of discards and the importance of such mortality and removal; the early life history of Cod and factors acting on that period, some of which may influence year-class-success.

Many small fish die wastefully as a result of being hauled up in cod or capelin traps, only to be discarded if they are too small or illegal to keep. Some trawls now have "'windows' in codends to 'spill' fish in excess of a certain target catch level" (CAFSAC 1989) which affects CPUE calculations and may cause mortalities among the 'spilled' fish. Research should address ways of quantifying and minimising these mortalities. Cause of death in hauled traps is often the expansion of gas sacs as fish become exposed to lower ambient pressures. Avenues to investigate could include, for example, reducing rates of ascent, reducing trap depths, incorporating effective escape panels, or use of screens or other devices to divert (without damage) fish larger than capelin out of traps (similar to the idea of the Nordmore grate).

Some protection of large fish could be very beneficial because their biomass-specific contribution to following generations is greater than for smaller fish. This could be assisted by limiting the speed of trawls relative to the bottom, or relative to the near-bottom water, which would improve escape probabilities according to swimming ability (in turn related to fish length). This would also help prevent mesh stretching which temporarily reduces the effective mesh size of trawls and causes fish smaller than target to be caught. Fishing vessels might discover and implement this and better measures if a stumpage-like Quota Charge were applied and aggressively indexed to protect certain size classes. DFO criticises this, but offers no reasoning:

"His knowledge of how the fishery operates is weak, and leads to errors of fact or interpretation, such as: ... it is realistic to regulate towing speed as a conservation measure" (DFO/Anonymous 1997 pers. comm.).

The structure of the management system is critical. Fundamentally, the determination of a TAC and conservation measures should be a purely scientific matter; only the reduction of TAC or allocations of parts of it to various sectors is justifiably the sphere of politics. DFO hints at management approaches being considered for the future:

"The former minor stocks therefore require new approaches which will ensure that they are managed with caution even if there is limited biological information available. This could be achieved through a close partnership between DFO and the fishing industry. The goal of the partnership would be to gather information and implement measures which have a high probability of ensuring conservation while knowledge is acquired" (DFO 1996i).

However, although management did put credence in industry data,

"Gulf of St. Lawrence ... Input from industry indicated that catch data from groundfish fisheries during the mid-1980s up to the closure in 1992 were unreliable" (DFO 1995).

The two quotes seem irreconcilable. When considerations outside biology are given a role in living resource management decision-making the tendency is toward over-exploitation. This tendency needs to be recognised in the design of the process. The theoretical arguments are based on resource access and economic discount rates. The resource access argument (Hardin 1968) was that, in terms of a single user among many (n), the benefit B associated with appropriation of a single unit of common resource accrues to the user while the cost (resource depletion, D) is evenly borne by the many; the net benefit is B-(D/n); as long as B>(D/n), there is a net profit and therefore a motivation to over-exploit. The discount-rate argument (e.g. May 1994) states that if a resource increases at less than the available interest rate on capital, then the purely financial motivation of industry will be to liquidate the entire resource; only fish which produce faster than the bank rate have a hope of being willingly conserved. When interest rates are high this is a problem for many natural resources.

For any management system to be effective it must at a minimum be: (i) sensitive to biological information, (ii) binding on all sources of anthropogenic mortality to fish in terms of regulating mortality and (iii) in terms of accurate reporting of mortality. The first component requires that TACs and other biological management decisions be made on a scientific basis, sensitive to input regarding conservation, but well insulated from political input and from pressure by limited-interest groups including industry. The second and third components require compliance over the entire range of space, time and types of activities affecting the resource; this requires a rethinking of the relationship of industry, management and government, and is a particularly great challenge for straddling stocks.

The proper and necessary role of politics is to ensure that conditions exist which allow proper management, and that the benefits of the resource are fairly allocated. Too much difficulty arises from uncertainty associated with data for straddling stocks, and the difficulty of obtaining cooperation (e.g. use of NAFO objection procedure; D. B. Atkinson, pers. comm.) for proper management of those stocks. While current diplomatic attempts to obtain cooperation on straddling stocks appear to have made some progress internationally, the past lack of cooperation from some NAFO contracting parties does not augur well for future conservation. It seems that either the fish must begin to observe the current EEZ and stay in either the Canadian zone or the zone susceptible to the NAFO Objection procedure, or the EEZ must be changed to conform to the movements of the fish. Thus, necessity may eventually require taking responsibility for the ocean beyond the continental shelf, or to equi-distance (with appropriate weighting for land area and population), because problems of ocean dumping, vessel piracy, unlicensed mining, and smuggling all add to this need to bring law to the open sea in many parts of the world. This idea is supported by the principle that authority for any resource should rest with those who will suffer most by its abuse.

Evaluation go to title toc

Because of the ecological, social and economic importance of Gadus morhua, and the fact that Canadian waters accommodate a significant portion of the world population, and because Canadian Atlantic Cod include at least several genetically distinct (Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; FRCC 1997; Ruzzante et al 1997; Taggart & Ruzzante 1997b; Taggart & Ruzzante 1997a) or otherwise unique stocks (Lear & Parsons 1993; Rose 1993; Myers et al 1997), any widespread threat to this species deserves serious consideration. It is also pointed out (Myers et al 1997) that some inshore stocks may have been more catchable than offshore stocks, but that (while and if) offshore (or more migratory) stocks were more abundant and mixed with inshore stocks for parts of the year, inshore stocks may have continued to be exploited even though they were reduced to a very low level or eliminated (the mathematics show that a second, more abundant stock can maintain the profitability of a fishery that also kept catching the first, less abundant, stock).

The hypotheses to account for the decline of the Cod population are discussed in detail above; there is virtually no doubt that overfishing is the sufficient and major cause of these declines, although exacerbating factors may exist.

Recovery is uncertain for straddling stocks (mainly 2J3KLNO) while foreign fishing continues in parts of the range to which the Cod population now seems to have contracted; unless the slope and shelf areas just outside the 200-mile limit can be protected the areas of juvenile concentration remain at risk. The recent drama of the Canada - EEC diplomatic conflict, and the history of the EU's attitude as a participant in NAFO, are reasons to be cautious before being optimistic. The number of analogous fishery depletions worldwide, together with the problems of vessel piracy (larger vessels in the southwest Indo-Pacific, smaller vessels in the west central Atlantic), indicate an approaching watershed in the law of the sea which as it presently stands limits jurisdictions to 200 nautical miles. Cooperative arrangements fail if even a few parties refuse to cooperate: unilateral measures are the unpleasant but sole alternative.

Recovery is uncertain if the cause of low weight-at-age is a reduction in the productive capacity of the system. Further work is needed on the effects of all types of biomass harvesting (lines, traps, nets, trawls, dredges) on the productive capacity and the habitat suitability for Cod, and also on the effects of removal of one species on the production and subsequent availability of others. At least, it is clear (ICES 1992) that fishing effects include • mortality on target and incidentally caught species, • increased availability to scavengers and predators of some of the biomass disturbed or discarded, • disturbance of the sea-bed, and • litter of lost and dumped gear as well as other debris. Mortalities of various animals in the path of beam trawls have been estimated at from 15 to 55%, and for those caught mortalities range from near zero to 100% (ICES 1992). The suspension of the bottom components also allows a winnowing (P. Schwinghamer, DFO, pers. comm.) that permits the current to displace a portion of the material. If currents are somewhat directionally stable this results in incremental transport of the bottom materials, with the slower-sinking components drifting the farthest with each increment and therefore experiencing the greatest cumulative displacement. This sorting process could create large-scale heterogeneity and in heavily trawled areas cause the bottom composition to shift toward domination by heavier particles and modify sediment geochemistry including sediment-water exchange of nutrients (ICES 1992) and hypoxia. Cod may avoid areas with reduced oxygen levels or face higher physiological costs as oxygen saturation declines; there is year-to-year variability in oxygen concentrations, and 1992 transects showed several instances of levels lower than long-term averages (Kiceniuk & Colbourne 1995). Work in the U.S. has shown strong evidence that bottom-contacting gears reduce benthic diversity and biomass (Auster et al 1995), but the overall effects of trawling on the near-bottom ecosystem in Canada remain substantially undetermined.

Some of the immediate effects of fishing on the benthos can be assessed with emerging technology. Acoustic gear has been used to show that the short-term effect of trawling is perceptible (GuignŽ et al 1993; GuignŽ et al 1995), and this may offer a way of economically collecting data to supplement grab samples, bottom photography, etc., to map the condition of the bottom community and eventually to conserve it in its most productive condition. Experimental control areas need to be protected from fishing so that accumulation rates of organic material, biological diversity, and productivity can be compared with fished areas. Such exclusions of fishing should also be evaluated for their function as refuges from, and as a buffer to, overfishing. Certainly, if areas can be shown to be particularly important for early-life-history stages then such areas should be protected from bottom-contacting fishing gear. The current moratorium may constitute a rare opportunity to evaluate the effect of [cessation of] fishing on benthic communities in grounds normally fished.

Lost gear (ICES 1992) may continue to inflict mortality, and lost gear seems to be fairly abundant: "unpublished Norwegian reports indicate the capture, using towed grapnels, of large numbers of 'ghost' gill nets in two separate areas ... old nets were still fishing and ... in some there was 'a relatively large amount of fish'. More fish were observed in nets on soft bottoms than over harder substrates. Nets found in deeper waters also contained more fish." An unpublished DFO document referred to (unfortunately this one was not referenced) in ICES (1992) reported a study along the 50m isobath on the northern edge of Georges Bank: in 236 grapnel tows, 19 gill nets were found and the recovered nets contained the remains of 94 fish (Cod, Hake, Dogfish, and unidentified remains); two experiments on the residence time of remains from time of capture in ghost nets showed means of 2 and 6 days. Trash twine and so on from fishing gear also shows up frequently on birds and in bird nests, indicating its prevalence.

Better control of discarding and better monitoring of catches is also needed; these have been often discussed and need little mention here save to point out that without accurate sample data any estimation is suspect. In achieving this end a truly cooperative arrangement between managers and all those dependent on the fishery should be taken into account. Long term biological, social, economic considerations need to be given higher rank than short-term commercial considerations. "Economic" analyses should not mean merely "commercial", but take into account all benefits and costs, direct and indirect, of each option.

The historic importance of the Cod fishery should justify the fundamental research required as a background to management, beyond that which presently exists.

Other Concerns, Observations and Recommendations in the Literature go to title toc

Many suggestions, recommendations, and observations on factors contributing to the current status of Atlantic Cod, particularly for 2J3KL, are to be found in a number of publications previously cited. Many of these are not discussed above but do merit note; a few are presented below.

Square mesh vs. diamond mesh in trawled gear (FRCC 1994); public consultations, Pubnico, p104): there are observations that because diamond (diagonal orientation of mesh in net construction) mesh stretches to form a slit, it is effectively a smaller mesh, and higher towing speeds result in selection for smaller sized fish. Unless this can be dismissed, it merits further investigation to determine the utility of either requiring that 'square' mesh construction be used, or regulating towing speed for each gear type.

"Quotas should be given in numbers of fish rather than tonnage" (FRCC 1994); public consultations, Carleton, Quebec, p 101/2). A similar suggestion was made by inshore fishermen (Earl Johnson, supported by others) at an FRCC meeting Aug 31 1995 in St. John's. A numerically- rather than weight-based quota would provide a strong incentive to avoid catching the smaller age-classes. This type of change would require, to avoid encouraging discarding or on-board high-grading, the inclusion of discards in quota (recommendation #24, Keats et al 1986). Numerical quotas on bycatch should apply to gear with small mesh (capelin traps, cod traps, shrimp trawls) to encourage responsible choice of gear and location, or such gear should not be permitted.

Marine protected areas: (FRCC 1994); public consultations, Moncton, p 102, and Charlottetown, p 103). The allusion is brief, but presumably the suggestion was to incorporate protected areas into a management plan. If some areas are particularly valuable for early growth and survival of Cod and other groundfish (Walsh 1992; Walsh et al 1995) then such areas or portions of them should be identified and given priority for protection, and monitored to determine the extent to which their protection enhances recruitment and the fishery in the long term. Protected areas would facilitate developing knowledge about the effects of fishing on the benthos, and (if indeed different benthic communities develop in protected areas) the effect of benthic condition on growth and survival of pre-recruits.

Acknowledgements go to title toc

Thanks are due to respected colleagues, many of whom are at DFO, for their occasional but generous advice and comments: Gary Blundell, Sarah Climenhaga, Joe Brazil, Bill Brodie, Joe Brown, Bob Campbell, Peter Schwinghamer, Joe Kiceniuk, Joanne Morgan, Bill Silvert, Ram Myers, Pierre Pepin, Jim Carscadden, Don Steele, Bill Threlfall, Steve Walsh, W.G. Warren, and others. Thanks are also due to the library staff of DFO (Northwest Atlantic Fishery Center, Vancouver, Ottawa); to several reviewers who provided valuable comments on this manuscript at several stages; to World Wildlife Fund for assistance; to the Deorksen Institute for encouragement, to the J.L.B. Smith Institute, Rhodes University and the Hugh Kelly Fellowship, and to my family for patience.

Author's Recommendation of Status

Preamble go to title toc

An objective and conservative guideline is useful as a point of commencement to initially determine the range of proper status designations. Guidelines for guarding against that which has not been experienced, and what is hoped to be completely avoided, must necessarily be deductively determined and cannot be exact. Guidelines should bias error to the side of conservation of that which cannot be replaced should the designation later prove to be wrong. A conservative bias is not only desirable but necessary, because the only cases where knowledge exists about extinction are where extinction has occurred. Speculation beforehand is better than knowledge too late. The sole counter-argument to the use of these guidelines is, as specified in the documents cited, where the declines are demonstrably part of a natural fluctuation. The onus is clearly on DFO to support that claim with evidence, and after 5 years the evidence is sparse, to say the least. These population(s) are at or near historical low(s), and the temperature hypothesis is not supported by analysis. As of Jan. 1997 the IUCN had received no information from DFO to support its contention that these are natural fluctuations.

The guidelines used where possible as a starting point are those in the discussion document produced by the Endangered Species Committee (ESC) (Environment Canada 1995b): together with the verbal descriptions issued by COSEWIC. (COSEWIC has never produced any list of criteria which cannot be used.) The ESC subsequent document on the legislative proposal (Environment Canada 1995a) defines species (section 3.2) as:

"Species -- for the purposes of the Act, species would refer to organisms that are recognised by most taxonomists as a valid and distinct entity known to have occurred in the wild in Canada for at least the last 50 years. Species that appear occasionally as accidental or casual visitors would not be considered. Species would also be defined to include subspecies and geographically distinct populations."

The ESC guidelines (Environment Canada 1995a) are consistent with the COSEWIC verbal guidelines, but add the advantage of objectivity and consistency because they include numerical guidelines (Environment Canada 1995b) for evaluation of either persistent declines in the size of the population, or low absolute numbers, or habitat features. The Legislative Proposal (L or without note) (Environment Canada 1995a) and the discussion document (D) (Environment Canada 1995b) equate the following terms from the ESC and current COSEWIC definitions:

ESC	COSEWIC
Extinct in the Wild in Canada	Extirpated
Critically endangered (extremely high risk of extinction in the immediate future)	(equiv. not given)L; subsumed under EndangeredD = facing imminent extinction/extirpationL
Endangered (high risk of extinction in near future)	Endangered = facing imminent extinction/extirpationL
Vulnerable (high risk of extinction in the medium-term future)	Threatened = likely to become Endangered if limiting factors not reversedL
(no equivalence given)	VulnerableL = sensitive to human activities or natural eventsL
Lower risk (none of the above) (includes a Not-at-risk designation)	Not at risk
Data deficient (insufficient information to make a listing)	Indeterminate

The declines (rate of change) in indices of abundance are of use for evaluation. The quantitative aspect of the ESC guidelines (Environment Canada 1995b) in this respect are summarised:

COSEWIC ESC

• 'Endangered' 'Crit. Endangered' = 80% decline in 10 years (= 14.8% yr-1).

• 'Endangered' 'Endangered' = 50% decline in 10 years (= 6.7% yr-1);

• 'Threatened' 'Vulnerable' = 50% decline in 20 years (= 3.4% yr-1);

• 'Vulnerable' not given/Lower Risk = decline less than 3.4% yr-1

These equate with the verbal designations (for 'Threatened', 'Endangered' , and 'Endangered' respectively) supplied by COSEWIC (R. Campbell, COSEWIC, pers. comm.), but definitions based on numerical values are a more objective starting point.

DFO objects (DFO/Doubleday 1996 pers. comm.; DFO/Anonymous 1997 pers. comm.) to any application of criteria based declines, but provides no evidence to show that declines observed for stocks of Atlantic Cod are irrelevant to a prognosis; the existence of a complaint to the IUCN is, however and of course, irrelevant because anybody can complain. Nevertheless, Environment Canada's wording seems likely be adopted in legislation (Environment Canada 1995a). The reductions in Cod numbers at the time of imposition of moratoria (Myers et al 1995b) in several management units or stocks (Table 3) reach 84% (4VsW) to 98.9% (2J3KL); these declines are clearly outside the observed range of natural fluctuation, and are conservative in the sense that the maxima used to calculate the reductions were relatively recent and within (not before) the era of heavy exploitation. Declines recently occurred in virtually all (whether 2GH declined since 1989 is not known) areas (not all of which show reduced condition factors and reduced weight at age). Previous fluctuations reported (Keats et al 1986; Steele et al 1992; Lear & Parsons 1993) often apply to limited areas, and must include effects of availability (Lear et al 1986; Lear & Parsons 1993), of market, and of infrastructure, and therefore cannot be considered solely due to population changes. Furthermore, historic data would not approach the quality of modern data (research surveys). Within the range of modern data, even if yield were considered to indicate population levels, fluctuations could not be ascribed to natural causes because the stock was under heavy fishing pressure during all the times when the data is the most abundant, and past declines were nowhere near the magnitude of the recent declines (possible exception of 3Ps, but there is some question about the recent data). There is, thus, little relief in the historical record.

Declines are not the only reasons for concern about Atlantic Cod: there are concerns about low recruitment, collapsed age structures, reduced area of occupancy, and more; these concerns are made more serious by the emergence of recent information (the last 2-3 years) that shows genetic distinctness even on within-management-area spatial scales. Further, there are indications that some local stocks have already been lost.

According to the ESC guidelines all six stocks reported on by Myers et al. (1995b) would (if treated separately) warrant at least the designation 'Vulnerable' (guideline is 3.4%, = COSEWIC 'Threatened'); of those, three would warrant at least the designation 'Endangered' (guideline is 6.7%, = COSEWIC 'Endangered'); and of those, two (including 2J3KL, a major management unit) would warrant at least the designation 'Critically Endangered' (guideline is 14.8%, = COSEWIC 'Endangered'). Some stocks are reported to have continued to decline since the moratorium, which could affect designations.

Until recently there was little genetic evidence on the existence of genetic uniqueness among stocks. Recognition of stock structures is now developing fairly quickly (Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; Taggart & Ruzzante 1997b; Taggart & Ruzzante 1997a) and confirms earlier perceptions based on tagging (Lear & Parsons 1993; DFO/Rice 1997) that there are, within the larger management areas (e.g. 2J3KL) components which can be differentiated genetically.

Is extinction unthinkable? go to title toc

Some Atlantic Cod stocks (in the sense of management areas) have certainly reached commercial extinction; this not irrelevant as argued by DFO, but on the contrary is an important warning of possible extinction. The recent microsatellite genetic, and other, evidence indicates a complex structure of unique sub-units, among which transfer is limited. It is recognised that (Rose 1993; DFO/Rice 1997) replenishment of populations by transfer of recruits from more abundant neighbouring populations cannot be relied on. As a complex mix of unique populations, Atlantic Cod is at higher risk, and 'severely fragmented', a criterion applied by the IUCN (WCMC 1996), would probably understate the added risk factor for Atlantic Cod. Some stocks may already be extinct (Ames 1997; Myers et al 1997).

DFO argues that biological extinction is unthinkable for a species with millions of individuals. The response is: why should the last 1 to 30% percent of the fish be less likely to disappear than the first 70 to 99% which have disappeared? Populations are not behaving in accord with any model used by DFO, or the current status would have been predicted. Populations in 2J3KL were predicted in 1992 at the beginning of the moratorium to rebound within 2 years, but have not.

DFO (e.g. DFO 1996g; DFO/Anonymous 1997 pers. comm.) has asserted [1] that temperature, climate, cold water, harsh conditions etc. are a major factor in the declines, and [2] that there are indications of improvement "good recruitment in 5Z and 4X; not elsewhere, but with condition factor recovering, the possibility is back" (DFO/Anonymous 1997 pers. comm.).

Claim [1], the temperature hypothesis, is not supported by evidence. The appeals in previous (DFO/Doubleday 1996 pers. comm.) comments ("There is at least one study ... that suggests temperature", dealt with under 'Limiting Factors/Hypotheses/temperature') and vague statements in more recent comments only demonstrate the destitution of the case for temperature/climate vs. the case for overfishing, e.g.:

"The idea of different marine "regimes" is gaining wide credibility [amongst whom is not stated]. The author misunderstands the literature, however (pg 63) [no explanation given]. All good studies [none referenced] are agreeing shifts in regimes are driven by major environmental changes. Both collapses of some stocks and increases in others are begun by the environmental signal, not by events in the fisheries [no references offered]" (DFO/Anonymous 1997 pers. comm.).

... because it lacks references or data this is a mere assertion which cannot be probed, so we must leave it and move on.

Claim [2] is an assertion of hope; it obviously cannot hope to extend to 2GH. Even if the DFO comment were taken at face value, that recruitment is good only in 5Z and 4X, then it is not encouraging, but although DFO admits lack of good recruitment "elsewhere", the claim for 5Z is not well supported by DFO data. The latest (1997) report says "All [Cdn + US] surveys appear to demonstrate similar year-class strengths with a decline in total numbers between 1990-92 and have remained at low levels since then. The 1997 Canadian spring 3+ indices decreased from the previous year ... USA [indices] ... remained constant... [yearclasses] since 1990 are well below average" (DFO 1997c). The similar claim for 2J3KL "it is undoubtedly not at risk of extinction" (DFO/Anonymous 1997 pers. comm.) firstly ignores the stock components that are now being delineated (DFO/Rice 1997) and secondly is severely at odds with the quote from Scott Parsons, assistant deputy minister of science in the Department of Fisheries and Oceans:

" 'The message from the (Northwest Atlantic Fisheries Organization) scientific council in June about the state of the northern cod stock is very grim indeed' said Scott Parsons, assistant deputy minister of science in the Department of Fisheries and Oceans. 'It's a message that says the stock is at an extremely low level, that there are no signs of good recruitment, and without good recruitment, good year classes, the stock may further decline' ... 'I hope it will recover, but the signs at the moment are discouraging' ... " (Whiffen 1997b) (emphasis added)

This quote illustrates that responses based on gut feelings or the appeal to large numbers can have no credence.

DFO cannot dismiss points by ad hominem attacks on authors of peer-reviewed work that does not support DFO dogma, e.g.:

"The discussion of limiting factors (pgs 42-47) is incomplete and selective. It relies almost exclusively on the publications and viewpoints of Myers and Hutchings. They are becoming widely known for their selective approach to interpreting what has happened to Atlantic groundfish. For example the quote from Myers et al (1995) that "high juvenile mortality was associated with high adult mortality, consistent with the hypothesis of discarding" (pg 45), overlooks the fact the high juvenile mortality is consistent with many other hypotheses [not identified] as well" (DFO/Anonymous 1997 pers. comm.).

(The DFO comment fails to provide any evidence to support its assertion that the discussion was incomplete; no new references offered, etc.)

Furthermore, because of the ecosystem importance of Cod (an importance which is supported by the previous abundance of the species, but has yet to be fully explored under commercial-extinction scenarios), a reduction on these scales is an ecological calamity even if biological extinction of stocks can be avoided. There is precedent for commercial extinction which was not followed by biological extinction, e.g. Pacific Sardine and Georges Bank Herring which for decades neither recovered nor became extinct, but may now be increasing (Myers et al 1995a).

But some stocks of Cod may already have been lost (Ames 1997; Myers et al 1997). Gulf of St. Lawrence haddock (Clay 1986) appear to have disappeared from mention, as appears to have occurred for the northern population of cod off Labrador (included in 2GH). For 2GH, according to DFO (DFO 1994; DFO 1995; DFO 1996i; DFO 1996b), the yields from 2GH have not exceeded 500t since 1985 (compare this with sustained yields from an inshore fishery of 11-14,000 t.yr-1 in the 1950s and up to 96,000t in the 1960s -- note also that the TAC remained at 20,000t long after yields fell to below 500 t). Until evidence allows narrowing down of the prognoses, we have ample reason for caution; extinction is a threat that must be taken seriously.

Further uncertainties go to title toc

Several uncertainties demand further caution. What is the etiology of the current stock(s) status -- fishing was certainly a major factor, but were there significant contributions from non-anthropogenic causes or indirectly anthropogenic causes? If and when there is any recovery, can it be linked to the moratorium, or to the suspension of trawling (i.e. linked to reduced fishing mortality, or to reduced benthic community damage from trawling)? What is the significance of physiological indicators (low weight-at-age)? (And is it still there in a blind re-evaluation of otoliths.) What is the "retrospective pattern" problem shown by the VPA for many stocks and are there better ways to deal with it? Is it caused by unreported discarding? Can target Fs be corrected by a factor derived from the known retrospective pattern?

There is uncertainty about the effectiveness of conservation measures for trans-boundary stocks (NAFO divisions 5Z, 2J3KL, 3N, 3O, 3Ps) until the recently-developed convention on transboundary stocks is put into effect. It is clear that the major cause of the decline has been human activity. The remaining numbers of Cod is a weak argument against designation -- numerous species have in the past become extinct (great auk, passenger pigeon) and are listed (monarch butterfly) despite their high numbers (millions) and high fecundity. The stock components that have been identified are of necessity far fewer individuals than the management areas they are in, and are thus much less amenable to DFO's 'millions' argument.

Low density is in itself a further impediment to recovery (i.e. collapsed age structures, or if depensatory or if capable of settling into a low-abundance regime). With the demonstrated existence of genetically unique local stocks (implying for each much lower population levels than for the entire 'family' of stocks) and their spawning site fidelity, the situation necessarily emerges as being more serious, i.e. that those stocks could vanish and be replaced either by something that is not Cod, or by Cod that are less adapted to local conditions. Among-stock exchanges that could replenish weaker stocks cannot be counted upon (Rose 1993; DFO/Rice 1997). The non-recovery of 2GH is a warning that stocks can either vanish or remain at low numbers; it seems impossible to discover which has happened, because we can't even attempt a genetic interpretation of the few fish found there, and cannot know whether they are the same stock as the previous 2GH, unless there are tissue samples kept from when 2GH was a healthy stock); 2GH was written off via high TACs that vastly exceeded declared catches, and which remained at 20,000 t even after catches fell below 500 t (the 2GH TAC even remained at 20,000 t for a year after 2J3KL was closed). The 2GH TACs were patently ineffective as conservation measures.

Fig. 7 near here

Fig. 8
near here

According to the quantitative guidelines and other available biological indicators of the collapsed populations, the observed reduction of numbers could place Gadus morhua in the COSEWIC endangered category for most management areas (Table 3, Figure 7, Fig. 8, Table 4). This applies to many management areas (DFO 1994). Where biomass estimates are available only Division 4X might merit a COSEWIC 'not at risk' designation (even though its 1993 spawner biomass was at an historical low). Cod in 4T, 4VsW, 5Zjm, and 2J3KL merit various at-risk designations. For Cod in management areas 2GH, 3NO, 3Ps, 3Pn4Rs, the descriptive information in DFO and other documents (Lear & Parsons 1993; DFO 1994; DFO 1995; Myers et al 1995b; DFO 1996i) is used; designations merited range from Lower Risk to at-risk designations.

Not enough is known about the synecology of Cod to determine the population prognosis under current moratoria, which have not been successful in eliminating illegal- or by-catch in waters outside the Canadian EEZ; and despite the TAC of zero tonnes there remains a fishing mortality equivalent to F0.1 in the southern Gulf of St. Lawrence [4T] (DFO 1994). The nose and tail of the Grand Bank are two such areas which are important for larval Cod (Walsh et al 1995) but which have been largely impossible to regulate.

Recommendation go to title toc

Stock-by-stock? Or assume one homogeneous stock?

History of this issue for this Report: The request for this report was originally framed as a single-unit consideration, and not as a stock-by-stock consideration. DFO then objected to the consideration of Atlantic Canada as a single unit (DFO/Doubleday 1996 pers. comm.), arguing for consideration by management areas. The Report was revised to accommodate this comment from DFO, because it is obviously more risk-averse to consider management areas as stocks rather than all areas as one stock, and this philosophy seems supported by DFO for other species as well (DFO/Vitols 1997) . Since then, DFO at the April 1997 COSEWIC meeting opposed consideration by separate stocks. The motivation is clear enough: under the assumption that there is only one large homogeneous stock (as argued by Jake Rice for DFO in the April 1997 COSEWIC meeting), if even a small handful are deemed certain to persist in some location, e.g. southwest Nova Scotia, then DFO can make the argument that biological extinction will be averted. With multiple stocks, that argument cannot be made.

The large-scale panmixis, "only one large homogeneous stock", scenario is not supported by the evidence (Bentzen et al 1996; Ruzzante et al 1996; DFO/Doubleday 1996 pers. comm.; DFO/Rice 1997; FRCC 1997; Taggart & Ruzzante 1997b; Taggart & Ruzzante 1997a), much of which is very new and developing fast.

The DFO arguments for a single designation to cover all Atlantic Cod in Canada referred to a recent workshop on Cod stock components (DFO/Rice 1997). (Although the workshop was held March 3-5, DFO had not brought copies of the papers/abstracts to the meeting. A number of requests were made for a copy of the proceedings document but, despite that it is dated on its cover May 1997, it remained unavailable until well into October 1997: six months later than DFO cited it in the meeting; and later than the date by which the revised Report was due at COSEWIC).

The sole question that must be answered in order to insist on consideration of all Atlantic Cod as a single stock is: "can the existence of biologically unique stocks be dismissed?" What did the DFO workshop on Cod stock components conclude? Can stocks be dismissed as an illusion in a sea of homogeneity, or are there multiple stocks? No, they cannot: Even within (let alone among) currently used management areas, the likelihood of multiple stocks (inshore and offshore, each of which might contain multiple components) greatly predominated over the likelihood of a single homogeneous stock, and the following statement appears on page 11:

p11 "in light of the tabulated information [on whether all the presented types of data supported a single- or multi-component stock structure, where the 'homogeneous' Model I(one) scored 36.5 'UN‑likely' vs. 17 'likely'], the Workshop concurred that a precautionary approach to assessment would have to recognise the complex stock component structure explicitly. More importantly, a precautionary approach to management would have to ensure that any harvesting should be carefully distributed among spawning components, so that none of the components would be consistently over-exploited. These approaches would require both assessments and management approaches which are much more complex than have been implemented in the past" (DFO/Rice 1997) (emphasis added).

Thus, individual management units, for example 2J3KL, not only may differ from other management units but appear to contain multiple stocks (Ruzzante et al 1995; Ruzzante et al 1996; DFO/Rice 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b). The document further acknowledges uncertainty about the possibility and beneficial result of transfer from stronger stocks to weaker ones:

p. 10 (under hypothesis/scenario/model II: underlying nature of stock structure is "small inshore components as well as large offshore stock components") "In this scenario, strong growth and increases in abundance of inshore stock components may contribute little to rebuilding of the offshore stock components. Only if the inshore stock components began to migrate to offshore banks when densities were high, would offshore stock components gain recruits. This would be a new and complex set of behaviours for the traditional inshore stock components, and Rose (1993[migration highway/Nature], 1994[etal Changes in distribution..) suggests several reasons why such migrations are not assured, even when inshore abundances are high" (DFO/Rice 1997)

The FRCC accepts the existence of multiple components:

"Each stock unit, however, is not a homogeneous entity, as it may include several components ("sub-populations"), representing the genetic diversity of that stock unit. Within its geographical range, each component usually has its own spawning site. The disappearance of spawning components (i.e. no fish spawning any more in a particular site), or the shrinking of the spatial distribution of a population (as observed for the northern cod, where fish are now concentrated in the most southern part, "3L" area, and in some bays) may be interpreted as a loss in the genetic diversity of the population.

"The delineation of sub-stocks, within a large stock unit, remains a complex issue. The 3Ps stock ... is composed of several sub-units, which can also be augmented by fish coming from the western Gulf (3Pn area) and from eastern Newfoundland (3L area). The whole 2J3KL area is considered as a single stock for management purposes but recent studies indicate that this unit may be composed of at least two components, one in the southern area (3L) and one in the north" (FRCC 1997).

The existence of biologically meaningful stocks must be accepted by COSEWIC.

Stock by stock: subdivisions to use go to title toc

The workshop (DFO/Rice 1997) found the possibility of multiple stocks necessary to acknowledge in management, but the fine-scale delineations to enable separate management and monitoring are not yet in place, nor are they likely to be for some time. It is therefore necessary to use for COSEWIC purposes larger-scale definitions until better definitions become available, in which case some well-defined and well-understood stock components can be revisited with a view to re-evaluating their status.

Of all possible subdivisions of Atlantic Cod, the most practicable option presently available to COSEWIC is by Management-Areas, because the data exists by management area. Management areas at present, even though they turn out to contain multiple stock components or populations and require revision later, more closely approximate populations than would a single unit comprising the entire east coast of Canada. With further information we would expect and hope that a better definition of populations may become possible and permit the designations to be updated. For any putative stocks within management areas, at present there would not be the possibility of establishing status because the data of relevance do not exist at scales not previously recognised to be important, so this option is not available to COSEWIC at present.

Designations should recognise that the management areas may contain stocks which are not all at the same level of health or risk, and so care should be taken to provide designation sufficient to address those components which are at greater risk until those are delineated and can be treated separately. Revision should as soon as possible address in a precautionary manner stocks for which evidence of discreteness is developing, even if that evidence is not complete. When better data become available, individual stock components that are healthy could be re-evaluated under COSEWIC authority and given relaxed designations, depending on the evidence available, to permit prudent exploitation. Therefore, the status recommendations are given in the by-management-area format.

Designation by geographic management units (as preferred by DFO in 1996)

The table below contains the designations arrived at for each area, with a short list of the main relevant indicators for each. Note that for some areas a series of biomass estimates or a rate of decline is available, and for some areas they are not. Where they are not, determination is made on the basis of the DFO summaries (Table 4) and other factors discussed in this document. Note that for all Cod stocks the DFO scientific reports describe populations as being at or near the lowest level observed (although for populations without a history of biomass estimation, e.g. 2GH, the % collapse cannot be calculated). Collapsed age structures with their implied reduction in recovery potential (Myers 1937; Hutchings & Myers 1993; Trippel & Morgan 1994; Myers et al 1995a; Myers et al 1995b; Trippel 1996 ms; Trippel et al 1997) are a serious concern in all stocks, as discussed above. Note also the very recent discoveries, by research cited in this Report, of genetic distinguishability of components within the management areas.

In the table, designations of the form "A or B" mean that, depending on the weight that the committee chooses to place on concerns about collapsed age-structures (a cause for concern and a common feature of over-exploited marine fish populations), and other factors listed in the table, either term could be used. If, in the opinion of COSEWIC, the factors listed do not increase concern, the first term could be used; if the committee agrees that the concerns listed are serious and add to the concerns based on numerical data (or official descriptions of it) then the second term would be in order.

AREA	Basis of Recommendation, Comment	Recommendation (A or B ESC)	Recommendation (A or B) COSEWIC)
2GH	Northern Labrador. Basis: history. Very low catches over long period, compared to (0.5 - 2% of) high catches 30 yr ago. Brief, large fishery in the 1960s; no biomass estimate; TAC of 20,000 until '92, reduced to 1,000t '93, not based on any assessment; yields since 1985 all less than 500t. If catches are assumed to roughly indicate biomass then implied change is 95% over 30 years (9.5% decline per year) this may underestimate decline, esp. 1992-3, but no assessments to allow other comparison. Recent surveys show "abundance and biomass ... very low" and "no large fish" (DFO 1997d).	Endangered or Critically Endangered	Endangered
2J3KL *	Northern Cod, S. Labrador & N. Grand banks. Basis: rate of change and low biomass. By 1992 stock had fallen to about 1-2% of its 1962 level; 1994 biomass about 1% (one percent) of early 1980s levels, or loss of 26% per year; collapsed age-structures are additional reason for concern. Virtual absence of fish in the offshore (DFO 1996e).	Endangered or Critically Endangered	Endangered
3NO	Southern Grand Banks. Basis: rate of change and low biomass; 1993 biomass 8.8% of 1967 level (7.7% decline per year). (Cod in this management area not included in DFO overview & stocks status reports)	Endangered or Critically Endangered	Endangered
3Ps	St. Pierre Bank. Basis: rate of change and low biomass, short period of time elapsed since recovery. By 1993 stock had fallen to 23% of its 1960 value (decline about 4.3% per year); results from 1997 fishery not yet available (as of Jan. 98); concern about lack of recruitment, possible increase in concentration.	Vulnerable	Vulnerable or Threatened
3Pn,4RS	Northern Gulf. Basis: low biomass and rate of change. By 1994 stock had fallen to 5.2% of its 1984 level (20% decline per year); no sign good recruitment ('95), recent year-classes poor ('96); results from 1997 fishery not yet available (as of Jan. 98)	Endangered or Critically Endangered	Endangered
(4T-4Vn:nov-apr) *	Southern Gulf. Basis: rate of change and low biomass; 1993 biomass 11.4% of previous (1956) level (decline 5.2% per year), and is still close to lowest observed. Potential to increase 5% in 1997 if no fishery.	Vulnerable or Endangered	Threatened or Endangered
4Vn:may-oct	Sydney Bight. Basis: rate of change and low biomass. "short-term future ... bleak; not one survey index shows any convincing sign of an increase ... absence of incoming year-classes of note" (DFO 1996n)	Vulnerable or Endangered	Threatened or Endangered
4VsW *	Banquereau/Sable Is. Basis: rate of change and low biomass. 1995 biomass near historic lows (decline 6.03% per year); recruitment poor since 1984.	Vulnerable or Endangered	Threatened or Endangered
4X *	Browns Bank. Basis: rate of change is low ("slight increase" noted '95, '96) and low biomass; collapsed age structures. Recent biomass changes not large, spawner biomass historical low in '93, '94 biomass historical low but this is 60% of 1970 level and only 2% per year decline since 1970.	Lower risk or Vulnerable	Not at Risk or Vulnerable
5Zjm (Trans-boundary stock) *	Georges Bank. Basis: rate of change and low biomass. 1993-4 biomass the lowest observed, very rapid decline (about 60%, yearly decline 29%) since 1991, which was near the 1978-1991 mean of 50,000 t (range about 42,000 t to 61,000 t); recruitment poor since 1990;	Vulnerable or Endangered	Threatened or Endangered

Table references: (DFO 1994; Bishop et al 1995; DFO 1995; DFO 1996i). For additional data and for 3NO: (Myers et al 1995b). Note: recent research (Bentzen et al 1996; Ruzzante et al 1996; DFO/Rice 1997; Taggart & Ruzzante 1997b; Taggart & Ruzzante 1997a) leads to the indication in some areas and the possibility in others that genetically distinguishable stock components exist within each management area, and any of these may be at greater or lesser risk compared to the average such component within the management area. Some stocks are suspected to have vanished .* see graph of recent biomass trends in Fig. 7 or 8.

Fig. 9
near here

Generally, the designations are based on the following causes for concern (the DFO comments on this list are contained in the Addendum under Prognoses):

1. that the rates of decline are severe and often meet the criterion for Critically Endangered

2. that most stocks reached historical lows in population and spawning biomass levels and are still not above those levels

3. that age structures are collapsed, negatively affecting the prognosis

4. that for virtually no stocks (possible exception 1992 year-class in 4X, Brown's Bank) is there more than a hint of good recruitment

5. high mortality on immature fish

6. that weight at age is said to be low for many stocks

7. that there have been changes in spatial distribution in 2J3KL with the few remaining fish becoming more concentrated, for reasons unknown

8. that in some areas mortality continues to be high due to limited fishing or unknown factors.

Also affecting the prognosis are key unresolved questions about ecological relationships, influences of climate, and about stock structures and components (delineation of components must continue). A prognosis cannot presume optimistic answers to these questions.

Designation treating Atlantic Canada as a single unit go to title toc

Designation in this format is recommended against. It is in nobody's best interest, and not in the best interest of conservation. Firstly, there is now strong evidence that different spawning aggregations in many cases are genetically distinguishable, so the existence of discrete populations must be acknowledged as a reality; Secondly, each of these stocks is at a much higher risk than would be the case under the panmixis scenario, and there is no certainty (DFO/Rice 1997) that the weakest stocks can be bolstered by transfer from neighbouring stocks; obviously a single designation cannot do this. Thirdly, not all elements are at the same risk. Fourthly, on practical grounds, an all-in-one-unit format requires that the most at-risk element be recognised in the designation, so it would mean increased restrictions where they might not be required.

If a single designation is insisted on, it must be the same that would apply to the most at-risk stock/population/component.

Further RECOMMENDATIONS go to title toc

At least two years of sustained population growth, and multiple estimates supporting a substantial recovery, should be verified before status designations are relaxed. This is philosophically in accord with FRCC's (1995) requirement for sustained improvement before re-opening a fishery. Presently-ongoing sentinel fisheries, together with research sampling and tolerance of small by-catches should be permitted to continue at reasonable levels and monitored in order to provide updated information on population sizes. In areas with stocks at highest risk the sentinel fishery should be replaced with tag-and-release programs rather than allowing mortality of fish caught. DFO and other legitimate researchers need to be allowed to continue such prudent sampling as helps to transform harsh experience into useful wisdom.

Work needs to be expanded on the existence, nature of maintenance, and possible routes of recovery of local populations or sub-stocks/stock components; on early life history ecology; on geographic areas of particular significance to very young Cod; on climatic and other influences on growth, reproduction and survival; on dietary and habitat requirements and tolerances of Atlantic Cod; on the role of the age of fish on its successful reproduction and its contribution to the populations; and on methods for reducing mortality on the youngest and oldest fish, so that fish reproduce at least once, preferably twice, and older fish remain as a buffer against poor recruitment.

Literature Cited (note: citations lacking journal etc. are MSS obtained from authors) go to title toc

Ames, E. P. 1997. Cod and haddock spawning grounds in the Gulf of Maine. . Ed.). Rockland, ME, USA: Island Institute. 33 pp.

Auster, P. J., Malatesta, R. J., Langton, R. W., Watling, L., Valentine, P. C., Donaldson, C. L. S., Langton, E. W., Shepard, A. N. and Babb, I. G. 1995. The impacts of mobile fishing gear on low topography benthic habitats in the Gulf of Maine (Northwest Atlantic): a preliminary assessment. NAFO SCR Doc. 95/21.16 pp.

Baird, J. W. and Bishop, C. A. 1986. Assessment of the cod stock in NAFO Divisions 2J3KL. NAFO SCR Doc.86/4750.

Bell, K. N. I., Pepin, P. and Brown, J. A. 1995. Seasonal, inverse cycling of length- and age-at-recruitment in the diadromous gobies Sicydium punctatum and Sicydium antillarum (Pisces) in Dominica, West Indies. Can. J. Fish. Aquat. Sci. 52: 1535-1545.

Benson, B. 1995. No Apologies . Evening Telegram, St. John's, Nfld. Sunday Sept. 3, p. 15,16.

Bentzen, P., Taggart, C. T., Ruzzante, D. E. and Cook, D. 1996. Microsatellite polymorphism and the population structure of Atlantic Cod (Gadus morhua) in the Northwest Atlantic. NAFO SCR Doc. 96/21. 20 pp

Bishop, C. A., Anderson, J. T., Colbourne, E., Lilly, G. R., Myers, R. A., Rose, G. A., Schneider, D. E. and Stansbury, D. E. 1995. Cod in NAFO divisions 2J3KL. NAFO SCR Doc. 95/601-15.

Bowcott, O. and Gooch, A. 1994. Anger boils over in the Bay of Biscay. Guardian Weekly, Manchester. August 14, p. 8.

Bowcott, O. and Jury, L. 1994. Navy seizes trawler in row over nets. Guardian Weekly, Manchester. August 14, p. 8.

Brown, J. A., Pepin, P., Methven, D. A. and Somerton, D. C. 1989. The feeding, growth and behaviour of juvenile cod, Gadus morhua L., in cold environments. J. Fish Biol. 35: 373-380.

CAFSAC 1989. Advice for 1989 on the management of cod in divisions 2J3KL. CAFSAC advisory document 89/1.

Campana, S. E. and Jones, C. M. 1992. Analysis of otolith microstructure data. Pp 73-100 in D. K. Stevenson & S. E. Campana (Eds.), Otolith microstructure examination and analysis. Vol. 117. Can. Spec. Publ. Fish. Aquat. Sci.

Campbell, R. 1987. Status of the Longjaw Cisco, Coregonus alpenae, in Canada. Can. Field-Nat. 101: 241-244.

Carr, S. M. and Marshall, H. D. 1991. Detection of intraspecific DNA sequence variation in the mitochondrial cytochrome b gene of Atlantic cod (Gadus morhua) by the polymerase chain reaction. Can. J. Fish. Aquat. Sci. 48: 48-52.

Clay, D. 1986. Biomass and population biology as observed from research surveys in the southern Gulf of St. Lawrence: cod, white hake, haddock, plaice, herring, and redfish. CAFSAC Res. Doc. 86/8362.

Cook, M. 1973. The head, guts and sound bone dance . . Ed.). Portugal Cove, Nfld.: Breakwater Books. 65 pp.

DFO 1994. Report on the status of ground fish stocks in the Canadian northwest Atlantic. Atlantic Fisheries Stocks Status Report No. 94/4. Department of Fisheries and Oceans. pp. 177.

DFO 1995. Overview of the status of Canadian managed groundfish stocks in the Gulf of St. Lawrence and in the Canadian Atlantic. Stock Status Report No. 95/3E. Department of Fisheries and Oceans. pp. 19.

DFO 1996a. Atlantic groundfish management plan [for] 1996. No. Fs 1-60/1996. Fisheries and Oceans Canada. pp. 232.

DFO 1996b. Divisions 2GH Cod. DFO Stock Status Report 96/44 E pp. 2.

DFO 1996c. Eastern Scotian Shelf Cod. DFO Stock Status Report pp. 6.

DFO 1996d. Newfoundland region overview. DFO Stock Status Report 96/43 E pp. 7.

DFO 1996e. Northern (2J3KL) Cod. DFO Stock Status Report 96/45 E pp. 6.

DFO 1996f. Oceanograpic conditions in the Gulf of St. Lawrence in 1995. DFO Stock Status Report 96/51 pp. 7.

DFO 1996g. Overview of Canadian Marine Fisheries Resources - Fisheries Res. Branch, Sci. Sector, DFO. No report series, no series number (from DFO website as 'View96.pdf') pp. 25.

DFO 1996h. Overview of the fisheries of the Scotian Shelf and Gulf of Maine. DFO Stock Status Report pp. 3.

DFO 1996i. Overview of the status of Canadian managed groundfish stocks in the Gulf of St. Lawrence and in the Canadian Atlantic. Stock Status Report No. 96/40E. Department of Fisheries and Oceans. pp. 11 + annex10.

DFO 1996j. Southern Scotian Shelf and Bay of Fundy Cod. DFO Stock Status Report 96/93 E pp. 4.

DFO 1996k. State of the Ocean: Northwest Atlantic. DFO Stock Status Report 96/41 E pp. 7.

DFO 1996l. State of the Ocean: Scotian Shelf, Bay of Fundy and Gulf of Maine. DFO Stock Status Report 96/63 E pp. 7.

DFO 1996m. Subdivision 3Ps Cod. DFO Stock Status Report 96/81 E pp. 7.

DFO 1996n. Sydney Bight Cod. DFO Stock Status Report pp. 4.

DFO 1997a. Cod in Sydney Bight. DFO Stock Status Report A3-02 pp. 5.

DFO 1997b. Cod in the Southern Gulf of St. Lawrence. DFO Stock Status Report A3-01 pp. 4.

DFO 1997c. Georges Bank Cod in 5Zj,m. DFO Stock Status Report A3-04 pp. 7.

DFO 1997d. Newfoundland region groundfish overview. DFO Stock Status Report A2-19 pp. 8.

DFO 1997e. Oceanographic conditions in the Gulf of St. Lawrence. DFO Stock Status Report G4-01 pp. 8.

DFO/Anonymous 1997 pers. comm. Comments on K.N.I. Bell, "Status of Atlantic Cod, Gadus morhua, in Canada" March 15 1997 (comments transmitted under letter from H. Powles to R. Cambell July 21, 1997). 11 pp.

DFO/Atkinson 1996 pers. comm. e-mail of Jan 19, 1996 in which my request for information on protocols for age evaluation from otoliths of Atlantic Cod was refused.

DFO/Doubleday 1997. May, 1997 interview with CBC Halifax about COSEWIC deferral and Cod Report.

DFO/Doubleday, W. G. 1996 pers. comm. (Comments on previous draft of manuscript, letter to Bob Campbell, COSEWIC, date-stamped Dec. 23 1996).

DFO/Rice, J. R., Workshop Chairperson 1997. Proceedings of the workshop on Cod stock components, March 3-5, 1997, St. John's, Nfld. Canadian Stock Assessment Proceedings Series 97/0620pp + Appendices not sequentially numbered, but by counting the doc. totals 101 pp.

DFO/Rowat, W. A. 1997. Lobbying / Comments on Mar. 15 1997 draft of Cod S. Rep., letter to COSEWIC members and/or their supervisors.

DFO/Vitols 1997. Letter dated May 22 1997 from Lisa Vitols, special assistant, office of the Minister Fisheries and Oceans, to J. Darling.

Drinkwater, K. F. 1996. Atmospheric and oceanic variability in the northwest Atlantic during the 1980s and early 1990s. J. Northw. Atl. Fish. Sci. 18:

Drinkwater, K. F. and Mountain, D. G. in press. Climate and oceanography. Pp [ca. 22 pages] in Northwest Atlantic grounfish: perspectives on a fishery collapse. American Fisheries Society.

Environment Canada 1995a. The Canadian Endangered Species Protection Act: a legislative proposal. Endangered Species Conservation, Environment Canada / Canadian Wildlife Service, Ottawa, Ontario, K1A 0H3. CW 66-143/1995. ISBN 0-662-61973-0. . Ed.). 18 pages pp.

Environment Canada 1995b. A National approach to endangered species conservation in Canada - Discussion document. Endangered Species (Committee), Environment Canada, Ottawa, Ontario, K1A 0H3. i + 32.

Flanagan, C. 1997. Crosbie denies allegations - politicians didn't interfere with scientist's research, he says. Evening Telegram, St. John's NF, 119: 155 Friday, Sept. 5, p. 15.

Fordham, S. E. and Trippel, E. A. 1996 ms. Feeding behaviour of cod (Gadus morhua) in relation to spawning.

FRCC 1994. Conservation - Stay the Course: 1995 conservation requirements for Atlantic groundfish. Report to the Minister of Fisheries and Oceans, Fisheries Resource Conservation Council, Ottawa, Canada K1A 0E6. pp. iv + 116 (+ 23 appendix).

FRCC 1995. Considerations on re-opening a closed fishery. A FRCC discussion paper prepared by the stock assessment subcommittee. FRCC95.TD.1. July 1995.

FRCC 1996. Building the Bridge. 1997 Conservation Requirements for Atlantic Groundfish - Report to the Minister of Fisheries and Oceans. FRCC.96.R.2 October 1996. vii + 142 + (appendix)82 pages. No. FRCC.96.R.2. Fisheries Resource Conservation Council. Published by Min. Public Works and Services Canada. ISBN 0-662-25083-4.

FRCC 1997. A groundfish conservation framework for Atlantic Canada -- Report to the Minister of Fisheries and Oceans. No. FRCC.97.R.3. Fisheries Resource Conservation Council. Published by Min. Public Works and Services Canada. ISBN 0-662-25861-4. pp. 50 + 15(appendix).

Gomes, M. C. 1993. Predictions under uncertainty. . Social and Eonomic Studies, No. 51. Ed.). Institute of Social and Economic Studies, Memorial Univ. of Nfld. xii + 202 pp.

GuignŽ, J. Y., Schwinghamer, P., Liu, Q. and Chin, V. H. 1993. High resolution and broadband processing of acoustic images of the marine benthos. Proc. Inst. Acoustics 15: 237-252.

GuignŽ, J. Y., Schwinghamer, P. and Siu, W. C. 1995. Quantifying the impact of trawling on benthic habitat using high resolution acoustics and chaos theory.

Gulland, J. A. and Williamson, G. R. 1962. Transatlantic journey of a tagged cod. Nature 195: 921.

Hardin, G. 1968. The tragedy of the commons. Science 162: 1243-1247.

Harris, L. 1990. Independent review of the state of the northern cod stock - final report. Prepared for the Hon. T. Siddon, Minister of Fisheries. Submitted February 1990.

Harris, L. 1993. Seeking equilibrium: an historical glance at aspects of the Newfoundland fisheries. In K. Storey (Ed.), The Newfoundland groundfish fisheries: Defining the reality, (pp. 1-8). : Inst. of Social and Econ. Res., M.U.N.

Hilborn, R. 1997. Statistical hypothesis testing and decision theory in fisheries science. Fisheries 22: 19-20.

Hollowed, A. B. and Wooster, W. S. 1992. Variability of winter ocean conditions and strong year-classes of Northeast Pacific groundfish. ICES mar. Sci. Symp. 195: 433-444.

Houston, J. J. 1987. Status of the Lake Sturgeon, Acipenser fulvescens, in Canada. Can. Field-Nat. 101: 171-185.

Hutchings, J. A. and Myers, R. A. 1993. The effect of age on the seasonality of maturation and spawning of Atlantic cod, Gadus morhua, in the Northwest Atlantic. Can. J. Fish. Aquat. Sci. 50: 2468-2474.

Hutchings, J. A. and Myers, R. A. 1994a. Timing of cod reproduction: interannual variability and the influence of temperature. Mar. Ecol. Prog. Ser. 108: 21-31.

Hutchings, J. A. and Myers, R. A. 1994b. What can be learned from the collapse of a renewable resource? Atlantic cod, Gadus morhua, of Newfoundland and Labrador. Can. J. Fish. Aquat. Sci. 51: 2126-2146.

Hutchings, J. A., Walters, C. and Haedrich, R. L. 1997. Is scientific inquiry incompatible with government information control? Can. J. Fish. Aquat. Sci. 54: 1198-1210.

ICES 1992. Report of the study group on ecosystem effects of fishing activities (Copenhagen 7-14 April 1992). ICES CM1992/G:11 Ref.: Session T pp. ii + 144.

ICES 1994. Report of the ICES Advisory Committee on Fishery Management. ICES Coop. Res. Rep. No. 210, parts 1 and 2. International Council for the Exploration of the Sea. pp. 147 (Part 1) 222 (Part 2).

Keats, D., Steele, D. H. and Green, J. M. 1986. A review of the recent status of the northern cod stock (NAFO Divisions 2J, 3K, and 3L) and the declining inshore fishery. A report to the Newfoundland Inshore Fisheries Association, on scientific problems in the northern cod controversy. . Ed.). St. John's: The Newfoundland Inshore Fisheries Association. 68 pp.

Kiceniuk, J. and Colbourne, E. 1995. [manuscript] Relating oxygen levels to the biology of aquatic organisms in the Newfoundland offshore waters. Author address: Science Branch, Dept. Fisheries and Oceans, St. John's, NF A1C 5X11-25.

Krohn, M. and Kerr, S. 1996. Declining weight-at-age in Northern Cod and the potential importance of the early-years and size-selective fishing mortality. NAFO SCR Doc. 96/5610 pp.

Lear, W. H., Baird, J. W., Rice, J. C., Carscadden, J. E., Lilly, G. R. and Akenhead, S. A. 1986. An examination of factors affecting catch in the inshore cod fishery of Labrador and eastern Newfoundland. Can. Tech. Rep. Aquat. Sci.

Lear, W. H. and Parsons, L. S. 1993. History and management of the fishery for northern cod in NAFO divisions 2J, 3K and 3L. Pp 55-90 in L. S. Parsons & W. H. Lear (Eds.), Perspectives on Canadian marine fisheries management. Can. Bull. Fish. Aquat. Sci., 226.

Leim, A. H. and Scott, W. B. 1966. Fishes of the Atlantic coast of Canada. Bull. Fish. Res. Board Can. 155: 485 pp.

Lilly, G. R. 1995. Did the feeding level of the cod off southern Labrador and eastern Newfoundland decline in the 1990s? Nfld. Regional Groundfish Stock Assessment Working paper 95/21

May, R. M. 1976. Simple mathematical models with very complicated dynamics. Nature 261: 459-467.

May, R. M. 1994. The economics of extinction. Nature 372: 42-43.

Morgan, M. J. and Trippel, E. A. 1996. Skewed sex ratios in spawning shoals of Atlantic cod (Gadus morhua). ICES J. mar. Sci. 53: 820-826.

Mork, J., Ryman, N., StŒhl, G., Utter, F. and Sundnes, G. 1985. Genetic variation in Atlantic Cod (Gadus morhua) throughout its range. Can. J. Fish. Aquat. Sci. 42: 1580-1587.

Myers, G. S. 1937. Fresh-water fishes and West Indian zoogeography. Ann. Rept. Smithsonian Inst. for 1937: 339-364.

Myers, R. A. and Barrowman, N. J. 1995. Changes in the concentration of 2J3KL cod from 1981 to 1994.

Myers, R. A., Barrowman, N. J. and Hutchings, J. A. 1997. Inshore exploitation of Newfoundland Atlantic cod (Gadus morhua) since 1948 as estimated from markÐrecapture data. Can. J. Fish. Aquat. Sci. 54: 224-235.

Myers, R. A., Barrowman, N. J., Hutchings, J. A. and Rosenberg, A. A. 1995a. Population dynamics and the potential for recovery of exploited fish stocks at low levels.

Myers, R. A. and Cadigan, N. G. 1994. Correlated error model results for 2J3KL cod in 1993. NAFO SCR Doc. 94/50: 9 pp1-9.

Myers, R. A. and Cadigan, N. G. 1995. Was an increase in natural mortality responsible for the collapse of northern cod?

Myers, R. A., Drinkwater, K. F., Barrowman, N. J. and Baird, J. W. 1993. Salinity and recruitment of Atlantic cod (Gadus morhua) in the Newfoundland region. Can. J. Fish. Aquat. Sci. 50: 1599-1609.

Myers, R. A., Hutchings, J. A. and Barrowman, N. J. 1995b. Why do fish stocks collapse? The example of cod in Atlantic Canada. Ecol. Applic. 7: 91-106.

Myers, R. A., Hutchings, J. A. and Barrowman, N. J. 1996. Hypotheses for the decline of cod in the North Atlantic. Mar. Ecol. Prog. Ser. 138: 293-308.

NAFO 1995. Processed length equivalents. NAFO/FC Doc. 95/10, Serial No. N2634 [Amendments to conservation and enforcement measures, 17th Annual meeting Sept. 1995].

Napier, D. 1996. Something Fishy -- there are still lots of ways for fishermen to get rich, most of them illegal. Saturday Night, 1996: November, p. 21-24.

Neilson, J. D. 1992. Sources of error in otolith microstructure examination. Pp 115-125 in D. K. Stevenson & S. E. Campana (Eds.), Otolith microstructure examination and analysis. Vol. 117. Can. Spec. Publ. Fish. Aquat. Sci.

NORDCO 1981. It were well to live mainly off fish. Report prepared for Government of Newfoundland and Labrador. February 1981. Jesperson Printing Limited, St. John's, Nfld.

Ouellet, P. 1997. Characteristics and vertical distribution of Atlantic cod (Gadus morhua) eggs in the northern Gulf of St. Lawrence, and the possible effect of cold water temperature on recruitment. Can. J. Fish. Aquat. Sci. 54: 211-223.

Pauly, D. and Christensen, V. 1996. Rehabilitating fished ecosystems: insights from the past. NAGA The ICLARM Quarterly 1996: 13-14.

Pepin, P. and Carr, S. M. 1993. Morphological, meristic, and genetic analysis of stock structure in juvenile Atlantic cod (Gadus morhua) from the Newfoundland shelf. Can. J. Fish. Aquat. Sci. 50: 1924-1933.

Rice, J. and Evans, G. 1986. Re-examining target spawning biomass for the cod stock in NAFO divisions 2J+3KL. NAFO Scientific Council Reports, Doc. 86/30

Ricker, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bull. Fish. Res. Board Can. 191: xviii + 382 p.

Rivard, D. 1982. APL programs for stock assessment. Can. Tech. Rep. Fish. Aquat. Sci. No. 1091. Gov't. of Canada, Fisheries and Oceans. pp. 146.

Rose, G. A. 1993. Cod spawning on a migration highway in the north-west Atlantic. Nature 366: 458-461.

Rousseau, D. L. 1992. Case studies in pathological science. American Scientist 80: 54-63.

Rowe, M. 1993. An overview of straddling stocks in the Northwest Atlantic - "Peace in our time?". In K. Storey (Ed.), The Newfoundland groundfish fisheries: defining the reality, (pp. 9-20). : Inst. of Social and Econ. Res., M.U.N.

Ruzzante, D. E., Taggart, C. T., Cook, D. and Goddard, S. V. 1995. Genetic differentiation between inshore and offshore Atlantic Cod (Gadus morhua L.) off Newfoundland: microsatellite DNA variation and antifreeze protein level. Working paper 95/1. NF Regional Groundfish Assessment Review, 24-28 April 1995, NWAFC, St. John's, NF.

Ruzzante, D. E., Taggart, C. T., Cook, D. and Goddard, S. V. 1996. Genetic differentiation between inshore and offshore Atlantic cod (Gadus morhua) off Newfoundland: microsatellite DNA and antifreeze level. Can. J. Fish. Aquat. Sci. 53: 634-645.

Ruzzante, D. E., Taggart, C. T., Cook, D. and Goddard, S. V. 1997. Genetic differentiation between inshore and offshore Atlantic cod (Gadus morhua) off Newfoundland: a test, and evidence of temporal stability. Can. J. Fish. Aquat. Sci. 54 (in press):

Schneider, D. C., Methven, D. A. and Ings, D. 1995 unpublished. Distribution and recruitment of demersal juvenile cod (0,1, and 2 group) in the coastal zone, NAFO divisions 3K and 3L. Working paper #20, Newfoundland Regional Groundfish Stock Assessment Review (April 1995, DFO St. John's)

Scott, W. B. and Scott, M. G. 1988. Atlantic fishes of Canada. . Can. Bull. Fish. Aquat. Sci., No. 219. Ed.). 731 pp.

Shelton, P. A. and Atkinson, D. B. 1994. Failure of the Div. 2J3KL cod recruitment prediction using salinity. DFO Atlantic Fisheries Research Document No. 94/66. pp. 1-14.

Shelton, P. A., Lilly, G. R. and Colbourne, E. 1996a. Patterns in the annual weight increment for 2J3KL cod and possible prediction for stock projection. DFO Atl. Fish. Res. Doc. 96/8127 pp.

Shelton, P. A., Stansbury, D. E., Murphy, E. F., Brattey, J. and Lilly, G. R. 1996b. An assessment of the cod stock in NAFO Subdivision 3Ps. DFO Atl. Fish. Res. Doc. 96/91

Sinclair, M. 1988. Marine populations : An essay on population regulation and speciation. . Books in recruitment fishery oceanography: Washington Sea Grant Program, Ed.). Seattle: University of Washington Press. x + 252 pp.

Steele, D. H., Andersen, R. and Green, J. M. 1992. The managed commercial annihilation of Northern Cod. Newfoundland Studies 8: 34-68.

Stevenson, D. K. and Campana, S. E. (Ed.) 1992. Otolith microstructure examination and analysis. Vol. 117. vi+126 p. Can. Spec. Publ. Fish. Aquat. Sci.

Sutcliffe, W. H., Loucks, R. H., Drinkwater, K. F. and Coote, A. R. 1983. Nutrient flux onto the Labrador Shelf from Hudson Strait and its biological consequences. Can. J. Fish. Aquat. Sci. 40: 1692-1701.

Suthers, I. M. and Frank, K. T. 1989. Inter-annual distributions of larval and pelagic juvenile cod (Gadus morhua) in southwestern Nova Scotia determined with two different gear types. Can. J. Fish. Aquat. Sci. 46: 591-602.

Suthers, I. M., Frank, K. T. and Campana, S. E. 1989. Spatial comparison of recent growth in postlarval Atlantic cod (Gadus morhua) off southwestern Nova Scotia: inferior growth in a presumed nursery area. Can. J. Fish. Aquat. Sci. 46(Suppl.1): 113-124.

Taggart, C. T. and Ruzzante, D. E. 1997a. Random Island regions and the inshore/offshore stock question. Pp 97-98 (Appendix pages not sequentially numbered as printed, but by counting the doc. totals 101 pp) in J. R. DFO/Rice, Workshop Chairperson (Ed.) Proceedings of the workshop on Cod stock components, March 3-5, 1997, St. John's, Nfld. Canadian Stock Assessment Proceedings Series 97/06, Canadian Stock Assessment Secretariat, DFO.

Taggart, C. T. and Ruzzante, D. E. 1997b. An integrated approach to stock identification: preliminary results of the high Prioirity Project on cod stock mixing in the Gulf of St. Lawrence. Pp 88-90 (Appendix pages not sequentially numbered as printed, but by counting the doc. totals 101 pp) in J. R. DFO/Rice, Workshop Chairperson (Ed.) Proceedings of the workshop on Cod stock components, March 3-5, 1997, St. John's, Nfld. Canadian Stock Assessment Proceedings Series 97/06, Canadian Stock Assessment Secretariat, DFO.

Templeman, W. 1962. Divisions of cod stocks in the Northwest Atlantic. ICNAF Redbook 1962, part III: 79-129

Trippel, E. 1995. Age at maturity as a stress indicator in fisheries. BioScience 45: 759-771.

Trippel, E. 1996 ms. Egg viability and offspring production of young Atlantic cod (Gadus morhua): implications for stock-recruitment and life history analyses.

Trippel, E. A., Kjesbu, O. S. and Solemdal, P. 1997. Effects of adult age and size structure on reproductive output in marine fishes. Pp 31-62 in R. C. Chambers & E. A. Trippel (Eds.), Early life history and recruitment in fish populations. London: Chapman and Hall.

Trippel, E. A. and Morgan, M. J. 1994. Age-specific paternal influences on reproductive success of Atlantic cod (Gadus morhua L.) of the Grand Banks, Newfoundland. ICES mar. Sci. Symp. 198: 414-422.

Unland, K. 1996. Canada working to keep cod, other fish off endangered list. The Globe and Mail, Toronto. Saturday, October 19, 1996, p. 1, A4.

Walsh, S. J. 1992. Factors influencing distribution of juvenile yellowtail flounder (Limanda ferruginea) on the Grand Bank of Newfoundland. Neth. J. Sea Res. 29: 193-203.

Walsh, S. J., Brodie, W. B., Bishop, C. A. and Murphy, E. F. 1995. Fishing on juvenile groundfish nurseries on the Grand Bank: a discussion of technical measures of conservation. Pp 54-73 in N. L. Shackell & J. H. M. Willison (Eds.), Marine protected areas and sustainable fisheries. Wolfville, N.S.: Science and management of protected areas association, Acadia University.

Warren, W. G. 1997. Changes in the within-survey spatio-temporal structure of the northern cod (Gadus morhua) population, 1985-1992. Can. J. Fish. Aquat. Sci. 54 (Suppl. 1): 139-148.

WCMC 1996. (World Conservation Monitoring Centre) Assessing biodiversity status and sustainability. B. Groombridge & M. D. Jenkins (Eds.). WCMC Biodiversity Series, No. 5. N. M. Collins (Ser. Ed.). Cambridge: World Conservation Press. 104 pp.

Wells, R. 1980. The determination of the ages of cod from otoliths collected in NAFO Divisions 2J, 3K and 3L. NAFO SCR Doc. 80/IX/1565 (supposed to be 28).

Wells, R. 1981. Age Determination of Cod From Otoliths Collected in Divisions 2J, 3K and 3L. NAFO Sci. Counc. Stud. 1: 77-101.

Wells, R. 1983. An examination of age compositions estimated for cod of the Flemish Cap in the period 1977-82. NAFO SCR Doc. 83/VI/2613.

Wells, R. 1986. On the validity of age determination of cod from Canadian research vessel cruises to Flemish Cap, 1977-85. NAFO SCR Doc. 86/9017.

Whiffen, G. 1997a. Cod study dead in the water: official - decommissioning of support vessel could mean Canada behind in research technology. Evening Telegram, St. John's, Nfld. Monday June 9 1997, p. 1.

Whiffen, G. 1997b. Future grim for northern cod / No political interference: Parsons. Evening Telegram, St. John's, Nfld. Friday September 5, 1997, p. 1.

Whiffen, G. 1997c. Key to rebuilding fishery -- leave fish alone (FPI boss says groundfish crisis due to several factors but 'natural mortality' [not (omission corrected next day)] primary cause. Evening Telegram, St. John's, NF. Fri. Sept. 5 1997 (sub-headline corrected Sat. Sept. 6), p. 4.

Winters, G. H. 1986. Aide-memoire on 2J3KL assessment: Non gratum anus rodentum? CAFSAC Res. Doc. 86/not listed

ADDENDUM: Issues and responses to comments go to title toc

This Addendum to the Report aims to respond to several comments that have been raised in relation to my draft report since it was first submitted to COSEWIC last year. Some of these issues are likely to come up in discussion at the April 98 meeting. While some issues are explored in detail in the Report itself, the most important or contentious issues are highlighted here in order to give the reader a clear review of the debate. Members may request further information or clarification from the author via e-mail <ihkb@giraffe.ru.ac.za>.

It is not the intention of these comments or the Report to malign the scientists or individuals employed by DFO. Indeed, much of the science supported by DFO is positively referred to. In this Report "DFO" means the organisation as it comes across in its official documents, its policies, and in the press, and it does not necessarily mean its scientific activities, branches or scientists; otherwise, if those latter are meant, the wording should make that clear.

Background: Timing and delays go to title toc

Some questionable things have happened and these contributed to delays and a decision to defer consideration of the Report from the April 1997 COSEWIC meeting.

This Report was commissioned by COSEWIC in 1994. Several drafts had been prepared, commented on by the subcommittee chair and DFO (late 1995), and revised before mid-1996. In 1996 the Report was sent for re-review, with comments due by the end of October; in November I was told that outstanding comments (~"none of which was from DFO", which at the time seemed consistent with the fact that two DFO scientists had forwarded comments) could be awaited no longer, and that a revision was desired by Dec. 31 so that it could be sent to COSEWIC members. The revision was shipped by the author on Dec. 22, ahead of deadline. The subcommittee chair was leaving for a long holiday about that time. On Dec. 23 DFO date-stamped a lengthy list of late comments. But the Report was not sent out to COSEWIC members immediately, and was instead returned to the author in late February 1997 with the very late comments (16 page) of DFO (by then nearly 4 months late). A revision was prepared on very short notice (the March 15 revision) and sent to COSEWIC for distribution. (Properly, the Report should have been distributed and a separate list of responses/amendments/rebuttals compiled to the DFO comments.) Inevitably, the upset timing resulted in members having less time to review the Report. A letter (DFO/Rowat 1997) dated April 7, 1997 was sent to the agencies of COSEWIC members, arguing against a listing (the letter was not copied to the Report's author). Consideration was deferred from the April 1997 COSEWIC meeting.

Following the April 1997 COSEWIC meeting, the feedback and comments reached the author only in mid-August. More importantly, key documents did not become available until mid-October -- notably the proceedings of the Cod Stock Components Workshop, a DFO document cited at the April 1997 COSEWIC meeting in support of a single all-Canada designation; 6 months after meeting, 5 months after the date on its cover (see Population basis of designations in this Addendum).

DFO has agreed to not object on the basis of timing this time around (from Dr. Howard Powles, who has recently taken over the file).

Population basis of designations go to title toc

Population homogeneity or heterogeneity?

The evidence is clear that stock structures are complex and the understanding of their delineations is growing. Uniqueness has been genetically demonstrated (see below), often associated with overwintering and spawning site fidelity. This implies a heightened risk of loss of these unique, discrete, populations. For example, stocks of herring which were eliminated (e.g. Downs Herring, Georges Bank herring) failed to return even though there were other herring that presumably could have replaced them under the panmixis scenario (if true); it is suspected that inshore populations of cod may already have been lost in 2J3KL (Myers et al 1997), and it is believed (Ames 1997) that inshore New England Atlantic Cod (Gadus morhua) stocks which were fished out also failed to return (or be replaced). These observations are consistent with Sinclair's member/vagrant hypothesis (Sinclair 1988).

The importance of this new data for COSEWIC cannot be understated; COSEWIC has a mandate to consider populations. Given the past few years' history there can be no argument against the national significance of the populations of Atlantic Cod, on account of both the biodiversity and the economic importance. On the basis of the evidence described and cited in this Report, to attempt to evaluate Atlantic Cod under the assumption of a large panmictic population would be unjustified and therefore irresponsible.

Reluctance is evident on the part of DFO to consider individual stocks for designation; declared preference is to instead consider only a single designation for all cod in Atlantic Canada. Firstly, this reflects a change of mind by DFO since the comments of 1996 (DFO/Doubleday 1996 pers. comm.), presumably for strategic reasons. Secondly, the key question is not what it implies for other designations (which is argument to the consequences and as such irrelevant), but whether the individual stocks have biological reality. Despite that the DFO delegation at the April 1997 COSEWIC meeting appeared to cite it as finding against the heterogeneous/unique-stocks scenario, a workshop concluded that:

"There was full agreement that, to different degrees in different Regions, there is substantial structural heterogeneity in all the stocks as presently defined" and that "a precautionary approach to assessment would have to recognise the complex stock component structure explicitly. More importantly, a precautionary approach to management would have to ensure that any harvesting should be carefully distributed among spawning components, so that none of the components would be consistently over-exploited. These approaches would require both assessments and management approaches which are much more complex than have been implemented in the past" (DFO/Rice 1997) (original is all in bold).

Peer-reviewed work based on otolith microchemistry and on microsatellite DNA (Ruzzante et al 1995; Bentzen et al 1996; Ruzzante et al 1996; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b) also supports the existence of biologically different stocks and populations, each with fidelity to spawning site/time and usually to overwintering location. The recent work seems to be confirming much of the early understanding of cod stocks as determined from tagging (Lear & Parsons 1993). It is clear that stock structure is now being acknowledged, as work proceeds to delineate particular components, and that stock behaviours may, under the multiple-component scenario, be unique enough to preclude exchanges whereby weaker components could receive migrants from neighbouring stronger components:

p. 11: "strong growth and increases in abundance of inshore stock components may contribute little to rebuilding of the offshore stock components. Only if the inshore stock components began to migrate to offshore banks when densities were high, would offshore stock components gain recruits. This would be a new and complex set of behaviours for the traditional inshore stock components, and Rose (1993, 1994) suggests several reasons why such migrations are not assured, even when inshore abundances are high" (DFO/Rice 1997).

Arguments have been advanced that it would be 'unacceptable' to individually provide a designation for each of the NAFO subareas. No basis is offered for 'unacceptable', so it is presumably used as a synonym for 'inconvenient'. If convenience were the issue, the report should never have been commissioned and COSEWIC would obviously have no purpose.

The committee must address the question in terms of the best available definitions of individual stocks or populations, which is always the smallest definable geographic units which include them. For the moment, those are the NAFO divisions, possibly divided into inshore/offshore units, although each quite possibly subsumes multiple unique components whose status could vary and which should be addressed in revisions. Geographic or other identifiers used in the Report to delineate stocks should be updated as delineation of stock components develops.

DFO's shifting position on Populations go to title toc

Prior to 1997, the Report had treated Cod in Atlantic Canada as a single unit, somewhat uncomfortably, and new work was beginning to show strong results indicating heterogeneity. W. Doubleday (DFO/Doubleday 1996 pers. comm.) had, reasonably, objected that in the 1996 version:

The criticism was justified was therefore accommodated in the revision for the April 97 COSEWIC meeting.

However, at the April 97 meeting, a lack of population structure was asserted by DFO -- in full contradiction of W. Doubleday's earlier statement -- and DFO argued for the position ('one-stock') that designation should be on the basis of a single homogeneous population rather than by DFO management areas.

The 'one-stock' position turned out to also contradict the conclusion of DFO's Cod Stock Components Workshop, which had convened shortly before, even though the same workshop was cited as support of the 'one-stock' position, and even though the workshop had been chaired by a DFO participant at the meeting.

Note that although the Workshop document is dated May 1997 and a copy was repeatedly requested, it did not materialise until mid October 1997 -- in fact later than the planned completion date for the current Report revision; this of course delayed the revision). Population structure is a critical issue and was recognised as such at the meeting, but the wrong or at least very incomplete impressions were given at the April meeting (e.g. DFO, Jake Rice: "It is exactly that concern that brought everyone together - 72 page report that goes into some detail about this - abstracts on genetics - 28 participants. ... None of results led people to conclude that there were populations so isolated that they would be individually at high risk."). These -- misinterpretations -- were not corrected but only reinforced by text that glosses over it in the July '97 comments, which are here repeated in what seems to be their totality on this issue:

"All the material on genetic uniqueness among stocks (e.g. pg 63-4) has been overshadowed by recent research in this area, presented at a workshop on Cod Stock Components (February 1997); again this is too new to have been published but the results were generally met favorably by workshop participants.

"In particular, Taggart, Campana, and Beacham's contributions showed that notwithstanding whatever small local populations one may conjecture have been lost, a biologically meaningful pattern of genetic diversity remains in 1994-1996 in both places which have been studied: inshore offshore areas of 2J3KL and Cabot Strait" (DFO/Anonymous 1997 pers. comm.) .

The "overshadowed..." comment seems to imply that the Report was wrong in supporting by-stock designations to recognise unique components; BUT that implication itself is wrong because the workshop supports the Report's stock-by-stock approach, i.e. that there must be explicit recognition of stock components. According to the workshop proceedings, Taggart and others' work supports the existence of unique stocks, while Beacham needs to see whether differences persist inter-annually (DFO/Rice 1997). These (and other contributions of the same Workshop) do NOT paraphrase as "abstracts on genetics - 28 participants. ... None of results led people to conclude that there were populations so isolated ...". Note that the comments (DFO/Anonymous 1997 pers. comm.) nowhere explicitly address whether the evaluation should be on the basis of one stock or a number of them.

As a matter of conduct, COSEWIC might question whether it has received in proper measure the complete and accurate description of the findings of the Workshop. A rebuke would seem to be in order if the Committee determined that important information had been effectively withheld by an incomplete or inaccurate representation. The material was patently crucial. COSEWIC oversees a national responsibility for conservation. Important decisions are taken. Information is necessary to a proper decision. Jurisdictions have the responsibility to provide all pertinent information. COSEWIC will have to ask itself whether the record shows that information was, or was not, properly presented at the meeting, and in the written comments of DFO.

The Workshop (DFO/Rice 1997) came out strongly in support of heterogeneity in stock structure. It rated homogeneity (one stock) much less likely given the available evidence than heterogeneity (multiple stocks), and then concluded, "a precautionary approach to assessment would have to recognise the complex stock component structure explicitly".

Is it "Ludicrous", "Unrealistic", etc. to consider cod at risk? go to title toc

These are a few of the words applied by DFO (DFO/Doubleday 1997; DFO/Rowat 1997) to the report. The report was commissioned by COSEWIC. DFO is a member of COSEWIC. The DFO reaction therefore implies that only one conclusion of the report was thought of as permissible -- but, in that case, why commission a report at all?

If it is indeed ludicrous (DFO/Doubleday 1997), as DFO claims, to consider a numerous (though much reduced) and fecund (yet evidently compromised) species as being at risk, regardless of declines or of having reached historically low levels, regardless of having collapsed age structures (which have clear implications for reproductive success), regardless of changes in distribution and concentration, regardless of seriously lacking recruitment for long periods, regardless of the apparent disappearance of local stocks (Ames 1997; Myers et al 1997), and regardless of major stocks (2GH, and more recently offshore 2J3KL) having virtually or actually vanished without recovery for decades, then that should have been decided within COSEWIC at the outset and this report should as such never have been commissioned.

If this report was expected or required to conform to expectations of a nil result, then that should have been decided within COSEWIC at the outset and this author should have been told so that he could have declined the commission.

If it is normal, prior to a meeting, for institutional members of COSEWIC to lobby (DFO/Rowat 1997) at high levels against a report, without copying the author on communications and with no open statement to the effect that this lobbying was in progress, then that should be a policy endorsed by COSEWIC at the outset and the prospective authors warned so that they can decline to have anything to do with the process.

None of those things were decided or endorsed, so it is clear that the report has to be taken as a response to a legitimate COSEWIC question.

DFO comments of ~970721, detailed responses: go to title toc

Particular comments are responded to below. Generally, in its written comments to COSEWIC, DFO advances a multi-pronged attack on the process: firstly, to say that COSEWIC concern should not be triggered at until there are very few cod left (like saying a slippery slope is no problem as long as you are not yet at the bottom); secondly, to say that the criteria of decline should be irrelevant; thirdly, to suggest that we are witnessing a natural fluctuation (laughable notion in this instance) which would exempt the declines from attention under the guidelines (to support 'natural fluctuation' a lot of talk is spent on climate, but the evidence is limited, circumstantial, and the connection of these declines to climate change has not survived peer review). Fourthly, the argument is offered that things now improving "~to the point where limited reopenings" (DFO/Doubleday 1996 pers. comm.) even though the FRCC recommendation is overtly contradicted in the same volume that delivered it (FRCC 1996), and even though in Sept. 1997 Scott Parsons is quoted (Whiffen 1997b) as saying the future of northern cod was grim.

Ad hominem go to title toc

Ad hominem argument is the standby of those who have little else. The Report, and cited works, are accused by DFO of being selective. Here are examples from the most recent comments, although the previous comments (DFO/Doubleday 1996 pers. comm.) were likewise tainted.

"incomplete and selective opinions"; "The literature is represented incompletely and selectively"; "The discussion of limiting factors (pgs 42-47) is incomplete and selective. It relies almost exclusively on the publications and viewpoints of Myers and Hutchings. They are becoming widely known for their selective approach to interpreting what has happened to Atlantic groundfish"; "The list of references is strongly biased towards a small number of sources"; "They extract results from the assessments which the author did not consult, and interpret them further, depending on the interests of the secondary authors. This practice leads to incomplete or distorted perspectives, such as giving Keats (1986) credit for identifying the retrospective pattern, which was noted in CAFSAC and NAFO documents in 1949,1983,1984, and 1985, before Keats used the DFO findings in his report." (DFO/Anonymous 1997 pers. comm.)

In the press the ad hominem arguments are extended to conspiracy theory, expounded by Scott Parsons in response to allegations by scientists that political actions intruded on fisheries management: "It's a carefully orchestrated campaign of misinformation by a very few individuals who have been quite effective in what they are doing ... a couple of individuals were quite prominent in the early media assault ... half of those who signed [a petition] are colleagues of those people" (Whiffen 1997b). Isn't it more plausible to think there is a sincere concern about how (and why) resources have been mismanaged?

Note that, although DFO has said "Myers and Hutchings ... are becoming widely known for their selective approach", this is only since their work came to inconvenient conclusions, previous to which their work was officially respected within DFO, and even won the Templeman Award.

No depensation? go to title toc

Depensation is the tendency for a population to produce fewer births than deaths at low population levels. DFO argues that depensation is not a factor for Cod. Indeed, the Report acknowledges that a careful and respectable analysis (Myers et al 1995a) of a number of stocks showed that only a few percent showed detectable depensation at the levels yet seen. The worry is that, when a stock is at an historical low, behaviours previously not detectable might emerge. The 2J3KL stock generally, and especially the offshore component, has failed to show recovery; the 2GH stock is in a similar situation.

How can it be ruled out that some stocks that are still failing to recover are not already showing a depensatory behaviour?

Criteria based on declines go to title toc

The Report does not, of course, rely solely on decline values, but uses them as a starting point. DFO has complained (1) that the Report used IUCN criteria for declines, (2) that the Report relied on these alone, (3) that in calculating declines the Report relied on the maximum observed population size, (4) that these criteria are unsuitable for highly fecund marine species.

Re (1): It is disingenuous to describe these as IUCN criteria. They are Environment Canada criteria proposed for future use (Environment Canada 1995a; Environment Canada 1995b). Criteria provide an objective starting point, a desirable feature in any assessment of anything. Furthermore, DFO had a part in developing the criteria as well as an opportunity (not taken) to be present at the IUCN meeting which listed Atlantic Cod as Vulnerable.

Re (2): It is equally disingenuous to claim that the Report relied on the declines alone; other factors are discussed with a view to checking and/or modifying the preliminary indication, and for some management areas there are no data with which to apply the decline-based criteria in any case (e.g. 2GH).

Re (3): DFO erroneously casts the Report's recommendations as resulting from a ratio of the most recent to the highest abundance. This is not what was done in main, although Myers was quoted on this. What has been done wherever possible is to use the recent data series, which is DFO data but which is not available for all stocks. The lack of better data is something for DFO alone to apologise for.

Re (4): The suitability may be contested by DFO but at present these are the criteria we have and which were developed after input by hundreds of scientists, and DFO was a part of the process that generated the criteria in the first place. The outcome of the review is by no means a given.

Are declines irrelevant? (The "millions" argument) go to title toc

DFO argues that declines are irrelevant for highly fecund species; however, this is merely an assertion. Although a move was made to have some of IUCN's criteria altered to better reflect what DFO considers reality, there is no way of knowing what the outcome will be, so it is irrelevant that DFO has made the move. DFO and many other agencies had input into formulating the guidelines that have become accepted, and while in any system there is occasional tuning, it is by no means assured that the rules will be changed either substantially or in DFO's favour.

Certainly the Report does not claim that a danger results from a fluctuation that has been seen before; however, the low population levels reached in the past few years for most stocks are historic lows for those stocks. This means that history of recovery from higher levels gives no assurance of recovery from the unprecedentedly low levels.

DFO makes the "millions" argument. This simple argument asserts that rates of decline don't matter because the numbers remaining are a guarantee against extinction.

i. That is like saying that although most of the stock was lost, the last 1%, 10%, is safer than the first 99%, 90% that has vanished.

ii. DFO's assertion implies a claim of authority, that the future is predictable by DFO. (But it was not DFO who predicted the collapses, instead it was DFO who ignored the warnings) If predictability is assumed, then it has to be asked: why were the declines not predicted by DFO, why did the rebound predicted by DFO in 1992 not follow, etc.?

iii. The "millions" argument applies much less well when individual stock components are recognised, and the weight of evidence supports that these exist (DFO/Rice 1997). The claim is at odds with what has happened (non-recovery) in major stocks, especially 2J3KL-offshore and 2GH.

Nature of declines go to title toc

Unique vs. natural fluctuation

Guidelines of the IUCN state "natural fluctuations will not normally count as a continuing decline, but an observed decline should not be considered to be part of a natural fluctuation unless there is evidence of this" (WCMC 1996). As of Jan. 1997 the IUCN (which had listed Atlantic Cod as vulnerable throughout its range) had received no information from DFO to support its contention that these are natural fluctuations.

The observed reductions in biomass or numbers cannot be called natural fluctuations because in most cases they are (according to DFO data) unprecedented: that's what a "historic low" is. The DFO stock reports are replete with historic lows.

Allusions (DFO/Anonymous 1997 pers. comm.) to fishery variations in the previous century are not evidence of population collapses. The case cannot be made that difficulties in catching fish in a non-mechanised fishery in the last century are evidence of natural fluctuation in abundance; landings fluctuate with availability, market conditions, accessibility, infrastructure, etc. Access to fish with more simple gear prior to 1900 was much more limited than it is today. No biomass can be estimated from the type of data mentioned.

This seems from one statement to be conceded, i.e., the extent of the declines observed does not represent a natural fluctuation:

"No one asserts that the EXTENT of the decline is a natural fluctuation within the period of detailed scientific record keeping" (emphasis DFO) (DFO/Anonymous 1997 pers. comm.),

but "within the period of ..." appears to be wiggle room, because the same document elsewhere claims that the collapses have precedent not within, but outside the era of scientific records, as in: "The ICES-GLOBEC Backward-Facing Working Group Report documents a period late in the 19th Century with many similarities" (DFO/Anonymous 1997 pers. comm.). That ploy constitutes relocation of the hypothesis to a region where it cannot be falsified, which disqualifies it as a scientific hypothesis, because it then would be like saying: "we call it as a natural fluctuation but the data is not available to test the existence of a precedent". It matters little whether the supposed precedent is in 18xx A.D. or 300 B.C.; if the hypothesis can't be tested then it belongs in the dogma category. As such, it need not be regarded as a valid hypothesis and can be consigned to the parts bin of 'pathological science' (Rousseau 1992).

A similar ploy is contained in previous comments: "There are natural fluctuations on longer time scales than observational data and it is not all that improbable that we are observing "natural", i.e. "environmentally induced" fluctuations at this time which are outside our previously observed range" (DFO/Doubleday 1996 pers. comm.). Note the appeal to "longer time scales than observational data"; a claim of faith that although there is no evidence to support them, there must have been such large fluctuations; then the assertion "not at all improbable" (see Hindsight vs. Foresight: Assessing Assertions).

Fishing vs. climate/cold water go to title toc

climate

Hypotheses, evaluations and decisions need to take into account the consequences that might follow an incorrect conclusion (Hilborn 1997). The chief hazard inherent in unsupported appeals to 'climate' or 'acts of God' is not only that they tend to legitimise inaction and support apathy (after all, one cannot change the weather): if they are wrong the effect may be to delay recognition of the real factor(s) and delay action to deal with them. Even if we were to accept that climate variations were associated with poor recruitment, the central issue is that the poor recruitment did not translate into a reduction of TACs, the overestimation of biomass supported higher TACs, the underestimation of fishing mortality also supported higher TACs; all of which is the responsibility of the decision-making levels of management, and all of which led to the overfishing which demolished stocks.

Environmental change invoked by DFO has no more to support it now than in the past. Almost as spectacular as the bizarre claim that fishing increases biomass (DFO/Doubleday 1996 pers. comm.), is the following claim for 2GH, an area with a abysmal management record that set TACs at 20,000t for many years, while yields never met that figure and while seven surveys all indicated "very low" abundance and biomass (DFO 1996b), that the collapse there is unrelated to overfishing:

"2GH ... The large catches 30 years ago coincided with an environmental regime when cod were abundant in many northern areas. There is no cause to assume such extreme northern stocks would be at higher levels at present, whatever level of harvest had been taken" (DFO/Anonymous 1997 pers. comm.).

Environmental variation is of interest to fisheries ecologists, but the magnitude of its effects has generally been difficult to ascertain, and fisheries worldwide are not generally higher in hotter places (often the reverse: there is a good reason why the Caribbean historically imported salt cod from Newfoundland). As with the 'salinity' model, there is always the danger that what seems to fit well with a given set of data may cease to work when new data for succeeding years are added (Shelton & Atkinson 1994). It may well be that salinity, temperature or other variables serve merely as proxies to some deeper underlying process (e.g. the interesting paper by Hollowed & Wooster 1992), but that process is far from understood, far from being citable as a cause of poor recruitment.

So there is a huge leap to be made, from trying to correlate variations in year-class strength in the time series of an un-collapsed fishery, to trying to explain a collapse by environmental variation. DFO has not made this leap, except as a leap of faith. The climate cause is an unproven and discredited hypothesis elevated to the level of official dogma (as implied by the otherwise pointless 'state of the ocean' documents for public consumption, e.g. DFO 1996f; DFO 1996k; DFO 1996l; DFO 1997e). It is used to explain what happened 20 to 30 years ago in 2GH (DFO/Anonymous 1997 pers. comm.), as well as virtually every recent collapse. The 'climate' dogma is a cover for mismanagement, and is a sign of instransigence. The supposed climate event has yet to be given specific attributes of time, place, and degree so that it could become a testable hypothesis. Those who find DFO's story less than compelling are called 'selective'. The steadfast adherence to hypotheses not supported by (in this case rejected by) rigorous testing, the failure of the adherent to carry out the critical tests which could support but also could reject the hypothesis, the smallness of the effect with the difficulty in measuring it, and the subsequent dismissal of every opposing view, are hallmarks of what has been termed "pathological science" (Rousseau 1992). The denigration of non-adherents as biased and selective is an extra shameful touch.

Everything, undoubtedly, must in principle have some effect on everything else; but any large effect should be easy to see in the record of other fluctuations, and the absence of that published result shows that either the effect is probably pretty small or the data are really not good enough to give adequate resolution. Climate can't be the catcher's mitt for anything that lacks the data for an alternative explanation; it is an hypothesis that must be framed in falsifiable terms in order to be taken seriously. That's not the way it comes across from DFO ... it comes across as an item of faith in the catechism of the bureaucracy, an excuse or perhaps a distraction. Seeking facts to support a favoured answer is not what the scientific process is all about. Mere plausibility is a good point from which to proceed to test the hypothesis, but plausibility alone is insufficient to elevate hypothesis to fact. Why, if climate is supposed to have caused the collapse, has no TAC has ever been reduced on the basis that climate looked like turning worse?

The Harris report crisply put the climate hypothesis in perspective:

"(p80) ... widely held belief that water temperature was an important controlling factor in some aspects of the northern cod life cycle ... unfortunately, support for this contention rests for the most part on anecdotal or intuitive evidence. That in itself does not refute its potential validity, but it does put the burden of proof on the scientist to define such a relationship, if one in fact does exist. ... Oceanographers can easily enumerate suspected mechanisms, but difficulties arise when attempts are made to link a specific force to a particular result." "(p81) ... the marine environment ... contains considerable natural variability ... one ... manifestation ... is the 200 to 300 year long record of northern cod catch statistics. These data show wide variations from year to year well before sophisticated modern technology would have had any influence on fishing success. It is a safe conclusion that environmental variability played some part in these differences, but it will never be known for certain whether it influenced stock recruitment, animal behaviour, or simply availability -- or possibly all three." (Harris 1990) (emphasis added).

'Climate' remains an hypothesis and should not be woven into explanations of declines until there is sufficient rigorous analysis that survives peer review.

fishing go to title toc

That fishing was the main cause of the fishery collapse is widely accepted, even by the CEO of FPI (Fishery Products International):

'' "The societal calamity that Atlantic Canadians are living through resulted from a failure of the whole fishing system," says the head of the country's largest fish processing company. "As Canadians, we have to be accountable for the overall situation we face today," said Vic Young, chief executive officer of Fishery Products International. "More importantly, we must accept responsibility for ensuring it never happens again" ... Young said that Canada has caught too many fish in its own waters and has been unsuccessful in preventing foreign fleets from overfishing offshore. "The amount of resource caught by traps, gillnets and otter trawls, foreign and domestic, has been far more than the resource could sustain" said Young. "In the end, it was the fish themselves which imposed the moratorium." In addition to human causes of the collapse, he added, changes in environmental and oceanographic conditions, harsh ice conditions and growing seal populations had an effect. "Natural mortality, however, has not been the fundamental cause of our fish crisis," he said. It has been caused by our lack of knowledge about the resource, too much fishing, trying to create too many jobs in a fishery which has been used as the employer of last resort." ... The greatest challenge, he said, will be to leave the fish alone long enough for a healthy recovery to take place. "Caution must prevail during the critical recovery period," he said. ... '' (Whiffen 1997c).

Many warnings were given, and generally went un-heeded. Mention of those documentary warnings (e.g. Keats et al 1986) provokes objections from DFO that the Report gave

"... Keats (1986) credit for identifying the retrospective pattern, which was noted in CAFSAC and NAFO documents in 1949,1983,1984, and 1985, before Keats used the DFO findings in his report" (DFO/Anonymous 1997 pers. comm.)

It is ironic that DFO seems to want credit for identifying retrospective pattern, but does not want the credit for doing nothing about it. (If DFO actually wishes to see previous works cited, why not give the citations?)

Fishing was far higher than targeted levels, often several times targeted levels. This point is objected to in the contorted argument:

"We do not contest the F0.1 target was not attained. However, this was a rebuilding target, not a status quo target. Missing a rebuilding target does NOT necessarily have "disastrous consequences". A conservation target was not consistently exceeded. The author should differentiate a management target from a conservation limit..." (DFO/Doubleday 1996 pers. comm.)

... heavy breathing aside (six kinds of limits or targets embedded in a spectacularly uninformative comment), the collapses of the stocks showed that all meaningful targets were missed. For a careful comparison of the various hypotheses, see (Hutchings & Myers 1994b; Myers et al 1995b; Myers et al 1996; Myers et al 1997).

indirect effects of fishing go to title toc

DFO scorns the idea that Cod habitat might have been degraded. Nevertheless, it is a candidate hypothesis which should be treated as seriously as 'climate', scientific evidence strongly suggests that habitat gets degraded by trawling, and it is not unreasonable to consider that contact of the bottom by very heavy gear towed by thousands of horsepower cannot be good for the bottom community which supports Cod (which usually feed off the bottom, as do most groundfish). I.e. there is a need to investigate this, but instead of doing so DFO merely asserts;

"factors often considered to be associated with increased risk of extinction are not operating for Atlantic cod. In particular, habitat has not been severely degraded or reduced (the statements in the Bell report on this issue are not based on a full review of the evidence on trawl damage to bottom habitats)" (DFO/Anonymous 1997 pers. comm.)

This DFO assertion is again not supported by any facts or references offered. Wiggle room is there because DFO does not exactly say the references are wrong. Indeed, it mirrors the earlier DFO comment:

"It HAS been established how trawling affects benthic productivity. Areas of small cod are NOT, usually, trawled, so exposing juveniles to increased predation is a false conjecture. The evidence is that modest levels of trawling INCREASES benthic productivity, and makes MORE food available" (emphasis in original) (DFO/Doubleday 1996 pers. comm.).

No references were offered to support the DFO view that trawling "INCREASES benthic productivity, and makes MORE food available", nor was explanation ventured as to how this might happen, and it would probably be generally regarded as a rather silly statement.

As well as the studies reviewed in the Report (and in ICES 1992) recent study and extensive review (Auster et al 1995) provides compelling evidence that trawling does disrupt the benthic community; that study references many others and details recent work which monitors the decline in biodiversity with trawling. The study was able to use untrawled sites and monitor them as trawling continued, finding, for example, that in 1992 a site previously surveyed in 1987 which had until shortly before 1987 been inaccessible to mobile fishing gear:

Auster concludes that

"Impacts of fishing gear have to be understood not simply in terms of removal of the targeted species but, more importantly, in terms of their impact on ecosystem productivity. ... Unfortunately, our current understanding of the impacts of mobile fishing gear is often more correlative than causal, particularly in the Gulf of Maine. ... There is an immediate need to provide information on the extent and magnitude of these effects with a directed program of study. ... Unfortunately ... in the Gulf of Maine ...no areas exist which can act as true reference sites or non-impacted controls. One approach to this problem is the designation of marine reserves which would provide reference sites in selected biogeographic regions .. experiments should be long-term in order to understand natural vs gear induced changes in habitat" (Auster et al 1995).

Judging from the DFO statements, there is no interest in any such program.

Mandate of COSEWIC & risk of extinction go to title toc

A legalism. The mandate is presented in such a way as to suggest the whole matter should never have come up. This rests on the argument, driven solely by assertion (the "millions" argument), that a species with many individuals cannot (an assertion) become extinct, and that COSEWIC has a mandate which ties its hands and prevents it from acknowledging the risk of extinction in such a situation. Plainly, species with many individuals have become extinct (passenger pigeon and Great Auk for just a couple of examples) and species with many individuals are listed and protected (Monarch butterfly). The IUCN has red-listed Atlantic Cod, re-evaluated the listing, and confirmed it.

Similarly legalistic (though quite beside the point) is:

"DFO agrees that a reduced age composition in the stock and the catch is a matter for concern. That is an influential signal in the assessment of stocks. Nonetheless, it is not a criterion recognized by IUCN or COSEWIC as associated with a high risk of extinction" (DFO/Anonymous 1997 pers. comm.)...

... but this paraphrases as: 'although DFO agrees collapsed age structures are an indicator of risk, COSEWIC should ignore them because IUCN hasn't explicitly listed them'. COSEWIC and IUCN both respond to indicators which affect a prognosis. The prognosis is affected (negatively) by collapsed age structures. It is even more serious given the complex population structure now being confirmed for Cod.

REPORT SECTIONS: Recommendations, Otolith, References, etc go to title toc

Section: Recommendations/Management Options

"The entire section on "Recommendations/Management Options" (pg 49-54) illustrates that even the author is not really thinking about a stock with any measurable risk of biological extinction. The author is thinking about more effective management regimes for harvesting cod. None of these proposed measures are new, and the new commitment to CHPs (Conservation Harvesting Plans) includes many provisions to improve fishing practices." (DFO/Anonymous 1997 pers. comm.)

This says that once a stock has "any measurable risk of biological extinction" we should stop thinking about better management that could be applied if we are lucky and don't lose a stock. The argument is of the same type that says COSEWIC has no business assessing a commercial species. How Canada and DFO got into this mess is what needs to be explored in order to avoid getting into it twice.

Otolith protocols

My comments about otolith protocols are repeatedly objected to by DFO, but the objection is empty. DFO's recent comments cites "lack of documentation of aging methods" (DFO/Anonymous 1997 pers. comm.). Responsibility for lack of information on this issue is entirely at the feet to DFO, because although this documentation has been requested of DFO (only DFO knows what DFO protocols have been since the publications of R. Wells), it has been refused, e.g.: "In response to the remark on the withholding of ageing methods, nobody is withholding how DFO measures fish length" (DFO/Doubleday 1996 pers. comm.).

The October 1995 version of the Report contained only a general description of otolith procedures, taken from the general literature; this was objected to by DFO. In turn I suggested I simply replace the offending section with a description of what the DFO otolith protocols, and I made that an official request (letter Jan. 20, 1996), which was refused. The refusal has been propagated since then by W. Doubleday and J. Rice (orally, at the April COSEWIC meeting). Jake Rice's retort (April 1997, COSEWIC meeting) that the methods were described by Wells was to dodge the point, because Wells wrote only up to 1986: the protocols since then have not been divulged.

It is clear that objectivity of measurement is desirable and that evaluations should be done blind wherever possible, and it usually is possible, for otoliths (Campana & Jones 1992; Bell et al 1995). If otoliths are not read blind, it is easy for operator bias to creep in, even unwittingly; so if there is a belief that environmental changes are affecting fish, what surprise is there if the results of subjective data come to support that claim? If otoliths are not read blind, it is not even surprising if the data turn out to appear to be consistent. Wells' papers do not anywhere state that readings are normally done blind, and the papers in fact suggest the opposite: that the readings were generally not done blind:

(p78) "size ranges by age-group ... overlap to such an extent that the use of fish length or otolith radius is not a reliable indicator of age, and, if used in conjunction with an interpretation of the otolith pattern, may in fact be a source of bias", and he "...despite the overlap in the otolith radius for different age-groups of cod, it is clear that, on the average, the otolith radius increases with age. The fact that many expert otolith readers not only maintain a fixed magnification but also prefer a particular microscope when interpreting otolith sections implies that otolith size assists the reader in determining the age" (Wells 1981).

Wells seems to have felt, consistent with what most would deem good practise, that it would be desirable to exclude the otolith size from the information presented to the otolith reader. He continues:

[although] "the occurrence of dominant year-classes and the similarity of growth patterns from the results of ageing would lead to confidence that otoliths are being consistently interpreted ... comparison of age composition and mean length-at-age data reported by different laboratories for the same cod stock would indicate whether there were substantial discrepancies or biases" and suggests "occasional exchange of otoliths within and between laboratories may act as a safeguard against overconfidence in age determination" (Wells 1981).

In 1983 there is a note about discrepancies in interpretation:

"At the June 1982 meeting of STACFIS it was apparent that age compositions of cod from the Flemish Cap as estimated by the Soviet and Canadian researchers differed substantially. An exchange of otoliths has shown that the discrepancies arose from differences in the interpretation of ages from otoliths and were not merely the result of sampling variation. The Canadian data are presented here in order to expose any internal inconsistencies and thus contribute to the definition and solution of the aging problem" (Wells 1983).

Finally, in 1986 (Wells 1986), is: "A blind test was set up", the clear implication of which is that in normal operations readings are not blind, and, for otolith exchanges also, "sampling details are known to the reader". Blind tests were done and repeated "When it became apparent that the 1979 [sic] readings were not in very close agreement with the original readings determined in 1979".

No information suggests that readings were routinely done blind, nor that they since then have changed to being done blind. No data on the occurrence of, or results of, suitable checks on the process since Wells' time has been forthcoming. Atkinson, Doubleday and Rice have all had opportunities to inform about the protocol, but have not. DFO's refusal to divulge methods since Wells' time is in itself yet another suggestion that the otolith protocols do not meet accepted standards.

It would seem from the information available that the otoliths are probably not read blind in all DFO laboratories. There is no excuse for not reading otoliths blind. If otoliths are not read blind, it is easy for operator bias to creep in, even unwittingly; if there is a belief that environmental changes are affecting fish, what surprise is there if the results of subjective data come to support that claim? If otoliths are not read blind, what surprise is there if the resulting data turn out to appear to be consistent with itself, or with the climate hypothesis! This calls into question any DFO conclusions that rely on otolith data, and this includes the weight-at-age data, the SPA, maturity-at-age, etc.

• A blind re-evaluation of a suitable randomly chosen subset of the old otoliths needs to be done, • current procedures should be openly reviewed, and • future readings should of course be done blind.

DFO comment on references go to title toc

See also the section 'ad hominem'. DFO grumbles that the Report does not rely on DFO's in-house publications, notably Stocks Status Reports (SSRs), sufficiently. This is a ridiculous complaint because in fact the Report relies very heavily on DFO documents and data. Is DFO saying that the summary reports it prepares do not reflect accurately the content of the source reports and data that they summarize? Should this complaint be taken seriously?

DFO has an official spokesperson policy and also exercises executive approval of documents and statements; this can only be expected to encourage strongly the authors of any report to give equal time to climatic factors and such like.

However, DFO has an official spokesperson policy and also exercises executive approval of documents and statements; this can only be expected to encourage strongly the authors of any report to give equal time to climatic factors and such like. This alone adds a taint that has to be borne in mind when reading the documents. A number of 1996 and only five 1997 Stock Status Reports (SSRs) were sent by DFO in answer to a request for the information that DFO felt had not been sufficiently used. But SSRs are not like the assessment documents, which at least have authors and are written with some attention to references. SSRs are not like peer-reviewed papers. They are usually anonymous (sometimes with a contact person identified), contain few references, in general are poorly documented publications restricted to summarising.

Stock Status Reports (SSRs)

These anonymous papers appear to be written for a non-scientific audience and are fairly brief. While some contain quality information, others have serious deficiencies. For example: DFO 1996. Divisions 2GH Cod. DFO Stock Status Report 96/44 E. 2 pp.

- Firstly, why does the report 96/44 E on 2GH give "N/A" for TACs 77-91? The TAC was 20,000 t long before 1992, in fact since 1974, so "N/A" is definitely wrong. In any case the same document says the quotas began in 1974 with 20,000t, which makes "N/A" look doubly silly. Similar 'N/A' or '--' are in other SSRs (e.g. DFO 1996e). Prior to the TAC the fishery was unregulated and catches far in excess of 20,000t had been occurring, so instead of N/A it should say > or >>20,000t or 'unlimited', and so "N/A" is equally misleading for the earlier period.

- Secondly, why does it say "the current management plan recognises that recovery of this stock is likely dependent on events in neighbouring stocks?" What stocks? 2J3KL? If so, 2GH has very poor prognosis; how does the statement square with the fact that 2GH collapsed long, long before 2J3KL?

- Thirdly, what is the purpose in saying "the 1995 and 1996 TACs of 200 metric tons were established to allow for test fisheries"? This is abundantly stupid, because the 95/6 TACs represent a further reduction from previous TACs; by what means could high TACs prevent a test fishery?

The 1996 2GH SSR shows the low standard that can apply for SSRs.

VPA and R/C plots (figure 4 in Report)

DFO says again that the Report's author doesn't understand VPA. Given the history, the question should be directed to DFO. Remarks about size distributions in the different catches are irrelevant, unless is shown how they could create a trend over time. The VPA process is sensitive to data quality. Discarding is rarely quantified in the assessments, yet we know it happens; so, there, is at least one impediment to data quality. Also, as admitted by DFO, there have been problems with industry reporting of catches: "Input from industry indicated that catch data from groundfish fisheries during the mid-1980s up to the closure in 1992 were unreliable" (DFO 1995). Even during the 1997 re-opening of 3Ps there were reports of dumping to avoid triggering penalties consequent to exceeding trip limits -- and it should be obvious that this would not be reported and that this must affect the quality of data used in assessments. The point of the R/C plots was that as an index that should remain constant over time, it doesn't -- and DFO has not explained why not.

The R/C plots nevertheless generated some interesting comments in the previous letter (DFO/Doubleday 1996 pers. comm.); perhaps the anonymous writer of the latest comments missed them?

Prognoses go to title toc

DFO claimed (at 1997 COSEWIC meeting) that there were indications of recovery in all areas. However, the populations in the southern areas are still near their lowest on record, and the nature of whatever indicators of recovery were alluded to for the northern areas is a puzzle. Certainly, a front-page story titled "Future grim for Northern Cod" in September quoted a senior DFO person indicating the opposite for 2J3KL:

"The message from the (Northwest Atlantic Fisheries Organization) scientific council in June about the state of the northern cod stock is very grim indeed" said Scott Parsons, assistant deputy minister of science in the Department of Fisheries and Oceans. "It's a message that says the stock is at an extremely low level, that there are no signs of good recruitment, and without good recruitment, good year classes, the stock may further decline" (Whiffen 1997a).

2GH is perhaps in worse shape, and should probably be considered nearly extinct: "[cod] have almost disappeared from Labrador" (DFO 1996g).

Causes for concern / DFO comment / author reply go to title toc

The Report listed (numbered statement) the following causes for concern, and each is followed by DFO comment and the Report's reply:

1. that the rates of decline are severe and often meet the criterion for Critically Endangered

DFO comment: "true - but does not imply any threat of extinction"

Reply: Is this saying that a population can become extinct without declining? Obviously a population that is has not declined is less close to extinct than one that has. See "millions" argument, above.

2. that most stocks reached historical lows in population and spawning biomass levels and are still not above those levels

DFO comment: "true - concern for commercial but not biological extinction"

Reply: The comment is a sidestep; what it boils down to is true, i.e.: commercial extinction is a milestone on the way to biological extinction; it is up to COSEWIC to decide whether to agree with DFO that such milestones can be ignored. A stock that has not collapsed is obviously at less risk than one that has. Comment also predicated on the dubious notion that the last (e.g.) 10% are more difficult to lose than the first 90%.

3. that age structures are collapsed

DFO comment: "true, but rebuilding in southern stocks, stable in northern ones" and "DFO agrees that a reduced age composition in the stock and the catch is a matter for concern. That is an influential signal in the assessment of stocks. Nonetheless, it is not a criterion recognized by IUCN or COSEWIC as associated with a high risk of extinction"

Reply: First statement is without evidence, a mere assertion. Rebuilding in southern stocks is not that certain either, later surveys even in 5Z (the most southern stock) showing less positive indication than previous: "estimates for 1995 and 1996 are lower than those reported last year ... lower Cdn survey index ... increased exploitation ... revision of USA [data] from estimates used in last assessment" (DFO 1997c). And, certainly, offshore 2J3KL cannot be called stable; if it were stable, why would S. Parsons call its future "grim" and say it "may further decline"? Same for 2GH and probably others.

Reply: Second statement is argument based on a technicality that itself is an assertion; neither group lists collapsed age structures as criteria not to be used; both groups accept any relevant data, and collapsed age structures are very relevant. Collapsed age structures have tremendous significance to the total production of eggs (Hutchings & Myers 1994b) and for the prospects for survival and growth of those eggs (Trippel & Morgan 1994; Trippel et al 1997), They are certainly an important consideration in evaluating the prognosis.

4. high mortality on immature fish

DFO comment: "mortality on young fish has decreased substantially in last 2 years"

Reply: It would be good to see data/papers on this, but none are cited. There AREN'T many young fish, is what the reports say, i.e. low recruitment can happen either because few eggs are produced, or because few survive (high mortality) to show up in samples. Doesn't seem so easy to make the claim that mortality has decreased.

5. that weight at age is said to be low for many stocks

DFO comment: "weight at age is stable or improving for ALL stocks"

Reply: assertion, and unlikely to be true for "ALL" stocks, e.g. 2GH, 2J3KL offshore; show the data (and confirm whether or not otoliths are read blind).

6. that for virtually no stocks (possible exception 1992 year-class in 4X, Brown's Bank) is there more than a hint of good recruitment

DFO comment: "good recruitment in 5Z and 4X; not elsewhere, but with condition factor recovering, the possibility is back"

Reply: If the DFO comment is true, that recruitment is good only in 5Z and 4X, then 'virtually no' remains accurate. But although DFO admits lack of good recruitment "elsewhere", the claim for 5Z is not well supported by DFO data: the latest (1997) report says "All [Cdn + US] surveys appear to demonstrate similar year-class strengths with a decline in total numbers between 1990-92 and have remained at low levels since then. The 1997 Canadian spring 3+ indices decreased from the previous year ... USA [indices] ... remained constant... [yearclasses] since 1990 are well below average" (DFO 1997c).

7. that there have been changes in spatial distribution in 2J3KL with the few remaining fish becoming more concentrated, for reasons unknown

DFO comment: "true, but widespread along coast ... false to say 'for reasons unknown', suggest we are mystified. We may not have eliminated all but one hypothesis, but we have some very plausible candidate explanations, based on general knowledge of cod biology"

Reply: "Widespread along the coast" is another sidestep, i.e. true but incomplete: Cod 'have virtually disappeared from parts of their range' and 'almost disappeared from Labrador' (DFO 1996g). The word "mystified" was not used in the Report. 'Hypothesis' is not equal to 'knowledge', and 'candidate explanation' does not even rate as 'hypothesis'. The reasons are indeed unknown, and the question is important. Is it possible that, for fear of finding the wrong answer, the question is not being seriously addressed?

8. that in some areas mortality continues to be high due to limited fishing or unknown factors.

DFO comment: "mortality has stabilised in all stocks; will have thorough review in January 1998. It is true that natural mortality seems to be higher than in previous years for many stocks, but this simply suggests the 'fishing only' story is wrong, and Fig. 6, which was borrowed from a published paper pushing this approach, is also wrong and misleading"

Reply: without data (none cited) the claim "has stabilised in all stocks" is only another assertion. Offshore 2J3KL is obviously not "stabilised" in any way that is knowable from the limited surveys that have been done. Same for 2GH.

Hindsight vs. Foresight: Assessing assertions go to title toc

Hindsight vs Foresight

Foresight and accurate prediction are convincing, but, as the saying goes, "hindsight is 20-20". Both within and outside of DFO, criticisms of management were made, and fishery collapses were predicted; yet DFO at the official level resisted or denied the criticisms all the way to 1991, and, yet, in 1992 had to close down first the 2J3KL fishery, then others. Then there was more and more talk about climate. If one bothered to count the percentage of DFO pages that were spent on climate by year, a dramatic increase would be expected after 1992. The effort to associate (not implicate, because 'implicate' implies a rigorous analysis) climate with fishery collapses, and by extension to serve as alibi for bad management, has been an exercise in hindsight.

DFO does not even acknowledge the risk of extinction of the offshore component(s) of Northern Cod (2J3KL): "No stocks are at risk of extinction in the near, middle, or long term" (DFO/Anonymous 1997 pers. comm.). But note that some inshore stocks of Atlantic Cod may already have disappeared (Ames 1997; Myers et al 1997). Why would DFO's assertions now be any more reliable than its previous rejection of warnings? Apparently we are supposed to just take their word for it.

DFO is not saying "we have made some great mistakes and have reviewed the situation with the help of outside scientists, and, for [referenced] reasons and based on [referenced] information arrive at the conclusion ...." Instead, DFO asserts 'climate', 'millions' and 'no discrete stocks' (which turned out to be not supported by the Cod Stock Components Workshop that was cited as support) ... then calls 'selective' and 'biased' anybody whose papers disagree with the DFO position. DFO asserts that things are getting better in some areas, that things are improving, and from these assertions further asserts that there is no risk of extinction in the near or distant future. Clearly, DFO expects to be regarded as an unquestionable authority whose mere assertions carry more weight than others' peer-reviewed analyses.

Assessing assertions (saying it don't make it so) go to title toc

An assertion is something that is stated. If evidence is referred to in connection with it, it could be considered an argument. Thus, an assertion is an argument without evidence. An argument is not merely a spaghetti of statements filling a paragraph, e.g. about climate, Azores or Aleutian highs or lows, and fish mixed in there somewhere; an argument must join the statements by some thread of logic. Failing that, an empirical statistical test might be in indication of some connection between the statements or observations. An empirical test if significant is still no more than, to use a quote from a respected scholar "only a correlation"; it only suggests that something -- what exactly it is need not always be known until further investigation -- might causally link the observations. Examples of interesting correlations abound in fishery science, including many ones that work well with data covering a certain time but do not work if data are added, e.g. the salinity hypothesis (e.g. the salinity hypothesis Shelton & Atkinson 1994). For Cod, the 'climate' hypothesis is still at the spaghetti stage.

Then how does one examine an assertion? With an assertion, all there is is a statement and the person who made it -- we are entitled to ask "on what evidence?", but if no evidence is forthcoming we can discard the assertion, or we might be kind enough to give the assertion a second chance and ask "says who?" Some authorities, having a record of accurate prediction, might merit respect such that their assertions are given weight, i.e. "the source is a proven authority." But no true scientist is an authoritarian; most scientists will concede no more than little extra thought before rejecting an unsupported idea, even from an authority with a good record. Thus, in a scientific context, assertions carry no weight at all.

If however one wanted to ask whether DFO had the record and stature that merited some extra thought or help in finding a way for the assertions to make sense, one would have to ask "what's their record?" How often has DFO been right when they said others were wrong? Here are examples relevant to the Cod issue:

Recall that in 1986 DFO was confronted by a report that pointed out problems in management (Keats et al 1986), but that report was discounted and was paid little attention. DFO now grumbles (DFO/Anonymous 1997 pers. comm.) that it knew about retrospective pattern before Keats et al; yet the point made by Keats et al was not simply that there was a pattern, but that DFO ignored this pattern in determining TACs. As apparently it continued to do so in most areas;

Recall that in 1986 an internal DFO document pointed to problems in the assessments (Winters 1986). The document was not officially recognised (no CAFSAC Res. Doc. number assigned).

Recall that in 1989 the scientists at DFO advised a large cut in the 2J3KL TAC to 125,000 tonnes, but DFO (executives/Ministry officials) assigned a TAC of 235,000 tonnes.

Recall that in 1989/90 NIFA brought a court action to require an environmental impact assessment of the northern cod stock and an interim injunction to stop offshore trawlers from fishing between January and May on the spawning stock. DFO argued against (Steele et al 1992), with J. Rice refuting his own earlier conclusion that spawning stock and recruitment were correlated, stating "the apparently strong relationship between level of recruitment and level of spawning biomass is highly dependent on the recruitment estimates from the year-classes produced in the early 1960's. Without these earlier (less reliable and less accurate) values referred to (above), recruitment appears to vary substantially, with spawning biomass influencing recruitment weakly". DFO won the case, but the fishery collapsed shortly thereafter;

Recall that in 1990 the Harris report (Harris 1990) asked for reductions that DFO declined to approve ; Harris said the choice was simple: reduce considerably now or reduce fishing much more severely later;

Recall that in 1992 the DFO, an organisation with hundreds of scientists and long experience with Cod, officially predicted that the moratorium should last only two years, by which time recovery was predicted to have been substantial enough for fishery resumption. Instead, not two but over five years later, a headline reads "future grim for northern cod" quoting Scott Parsons (Whiffen 1997b). That the situation in actuality turned out so much less brightly than predicted by the recognised experts is on its own a strong cause for concern.

The history reflects that Canada made "a botch of" managing Atlantic Cod (Crosbie, quoted in Flanagan 1997). It is gambler's logic to hope, after betting wrongly so many times, that the next bet will be the winner. DFO's assertions must be read in the light of DFO's record.

A designation is necessary because of historic low population levels, failure to realise projected recovery rates, rates of decline, contracted geographic distributions, the virtual disappearance of stocks (2GH and offshore 2J3KL for example). Concern is increased by the discovery/recognition that genetically distinguishable stock components or populations exist, that they represent biodiversity which could easily be lost or may have been lost already, and that they may indeed be unique enough to preclude effective transfer of recruits from stronger to weaker stocks (DFO/Rice 1997).

AREA	PRESENT STOCK STATUS	OUTLOOK
2GH	TAC NOT based on any assessment., stock thought to be low	Any rebuilding depends on appearance & survival of relatively strong year-classes
2J3KL	Probably all-time low, size-at-age, age-at-maturity decreasing ... stock at dangerously low level	Recovery not until production and survival of significant numbers of new recruits
3Ps	'89 is last strong year-class; difficulties interpreting research vessel data; inter-annual migrations between 3Ps and adjacent stocks (3Pn4RS, 3NO) complicate interpretation	Possible strong '89 yr-class, may contribute significantly in '95+
3Pn,4Rs	Stock abundance is clearly very low, possibly at its lowest level since data has been available (1974)...appears situation has probably deteriorated in '93...fish growth is low	Outlook not encouraging, no indication of significant recruitment, physiological condition is precarious
Southern Gulf (4T)	Stock abundance and biomass remain very low. Fishing mortality has been reduced and is close to the F0.1 target in '93	Prospects for future recruitment are poor. Fishing effort needs to remain at or below '93 levels AND recruitment has to improve to achieve recovery..
4Vn	Stock at very low level...TAC not taken since '89; length-, weight-at-age continue downward trend of past decade; F likely above F0.1 prior to '93 ...4Vn is an area of mixing between 4T, resident and 4VsW stocks not well defined	No sign of good recruitment since '87 year class; ..no signs of increase in the short term
4VsW	Fixed gear have not caught allocation since late '80s. The conjunction of overfishing, low recruitments, and to some extent, increasing seal population has placed this stock in jeopardy.	A number of years of good recruitment will be required before the fishable and spawning stock biomass can support a fishery.
4X	Spawning stock 93 is lowest recorded; landings lowest since 69;; sharp decline in spawning stock biomass; F has been well above 2*(F0.1) in '90-'93; frequency of good year-classes has declined since 70's; if TAC of 13,000t is taken in '94, F will be about 0.6 [i.e. if fishery takes 45%];	Indications of recruitment are poor. A lower exploitation rate would prolong the contribution of the '90 yr-class to the fishery and promote rebuilding.
5Zjm (Trans-boundary stock)	RV surveys agree and indicate historically low levels; last strong yr-class ('87) has been depleted, '90 year-class moderate but subsequent are below average; SPA [Sequential population analysis] indicates spawning stock declined rapidly in '90s, presently at lowest level estimated; age-3 mortality increased sharply since 89 & have been well above 2*(F0.1) 90-93. Catches have been declining. Although ITQs were intended to reduce capacity, only slight decline in fishing mortality seen since '90. ...exploited at exceptionally high levels	Recruitment will be below average. Urgent measures to protect the spawning stock are required.

AREA	Comments	Prospects for 1996
2GH	Catch negligible since 1990	No change
2J3KL	1994 biomass is lowest observed. Year-classes weak since 1987, improving in '93+'94. Biomass at an all-time low. Adult biomass very low. No signs of recovery.	Stock extremely low
3Ps	Cod from 1989 year-class are now mature and must be protected. No evidence of strong year-class after 1990; [apparent] increase in 1995 survey due to one large tow. Considered to be at a low level of abundance. Older fish have disappeared; growth rates have declined.	Uncertain
3Pn,4Rs	Cod condition, growth and recruitment have been poor; some improvement in condition in fall '94. Biomass estimates are very low; no signs of good recruitment. All indications are that the spawning biomass is very low.	No change
Southern Gulf (4T)	Recruitment has been poor in late 1980s and early 1990s; there is no indication of improvement in recruitment. Biomass close to lowest observed. Spawning biomass would increase slightly (6-10%) if there is no fishery in 1996.	No change
4Vn	Catches continue to decrease. Little or no recruitment since 1989. Stock in a very depressed state; no signs of improvement. Unit redefined as Nov.-Dec. catches from 4T stock.	No change
4VsW	Fishery closed since September 1993. Recruitment has been poor since 1984 [sic: 1984]. Adult population at historic low levels. Predation by grey seals is significant. Growth and condition of the fish are poor.	No change
4X	Exploitation rate more than twice the target. The 1990 and 1992 year-classes are about average. Stock showing slight increase in 1995, from the low levels observed in 1993 and 1994 (lowest in time series). F0.1 for 1996 around 6,300 t.	Reduction in catches
5Zjm (Trans-boundary stock)	Fishing mortality very high. '91+'92 year-classes below average; '93+'94 lowest observed. Stock biomass lowest observed. Need to develop stock rebuilding strategy. F0.1 for 1995 around 2,500 t.	Status report previously released

AREA	Comments	Prospects for 1997
2GH	Catch negligible since 1990, no recent surveys	No change
2J3KL	Biomass remains very low. Year-classes weak since 1986-87. ...No signs of recovery.	No change
3Ps	(from 1995 SSR) Cod from 1989 year-class are now mature and must be protected. No evidence of strong year-class after 1990; [apparent] increase in 1995 survey due to one large tow. Considered to be at a low level of abundance. Older fish have disappeared; growth rates have declined.	To be determined in August 96
3Pn,4Rs	Stock remains very low. Recent year-classes (1991, 1992 and 1993) are poor. Some improvements in condition in the fall of 1994 with large improvement in fish condition during 1995	No change
Southern Gulf (4T)	Recruitment has been poor in late 1980s and early 1990s. There are some indications of improvement in recruitment but it is too early to be certain. Biomass close to lowest observed. Spawning biomass would increase slightly (5%) if there is no fishery in 1997.	No change
4Vn:m.-o.	Fishery closed since September 1993. Recruitment has been poor since 1987. Adult population continues to be low. Short-term prospects are bleak.	No change
4VsW	Fishery closed since September 1993. Recruitment has been poor since 1987 [sic: 1987]. Adult population near historical low. Fish condition and growth poor. Predation by seals is significant. Environment continues to be cold.	No change
4X	(From Fall 1995 assessment). Biomass increasing from 1994 historical low. Exploitation rate in 1994 at 47% and above reference. 1992 year-class strong; that of 1993 is weak. F0.1 for 1996 [sic:1996] is 7,400 t (TAC is 11,000t)	To be determined in Sept 96
5Zjm (Trans-boundary stock)	Adult biomass increased slightly from 1994 historical low. Exploitation rate below F0.1. Recruitment since 1990 has been poor.	To be determined in Spring 97

AREA	Comments
2GH	N. Labrador. "Cod ... have almost disappeared from Labrador" (DFO 1996g). Catch negligible since 1990, survey "1991 detected very few fish" (DFO 1996i), and in 1996 survey both abundance and biomass "very low... show no large fish" (DFO 1996i).
2J3KL	Northern Cod. Biomass remains very low. Year-classes weak since 1986-87. ...No signs of recovery. State is "grim" - S. Parsons, ADM, DFO (Whiffen 1997b). Genetically distinguishable stock components, N/S (Bentzen et al 1996) and in/offshore (Ruzzante et al 1996; DFO/Rice 1997; Ruzzante et al 1997; Taggart & Ruzzante 1997a; Taggart & Ruzzante 1997b) have been identified, implies a greater risk of loss of populations, biodiversity.
3Ps	St. Pierre Bank. No updated information since 1996 (&1997 fishery re-opening).
3Pn,4Rs	N. Gulf. No updated information since 1996 (&1997 fishery re-opening). "now at lowest levels on record ... decline ...arrested ... [but] recruitment and growth rates have been poor" (DFO 1996g).
Southern Gulf (4T)	S. Gulf. Adult biomass has increased modestly since ... 1993 [but is] due mainly to [individual] growth. Recruitment ... well below average. Stock recovery will require increased recruitment" ~~(DFO 1997b).~~
4Vn:m.-o.	Sydney Bight. "Since 1987 ... very few fish over 11 years ... no signs of good or even average recent recruitment. The biomass increase observed in the last few years is due to the growth of older fish which are surviving the fishery closure. A retrospective pattern is seen in the biomass estimates, which means that the most recent biomass figures tend to be overestimated. Thus, the recent status of the stock is probably poorer than is shown in the following figure" "Only the offshore survey index (age 1) shows any indication of an improvement since ... 1993 ... potential contribution of these small, less than 15cm, fish to future recruitment has not been established. ... Until there is substantial recruitment to the reproductive stock, there can be no thought of reopening the fishery." (DFO 1997a).
4VsW	Banquereau/Sable Is. No updated information since 1996. "at lowest level observed" (DFO 1996g).
4X	Browns Bank. No updated information since 1996.
5Zjm (Trans-boundary stock)	Georges Bank. "All [Cdn + US] surveys appear to demonstrate similar year-class strengths with a decline in total numbers between 1990-92 and have remained at low levels since then. The 1997 Canadian spring 3+ indices decreased from the previous year ... USA ... remained constant" (DFO 1997c).

Status of Atlantic Cod, Gadus morhua, in Canada

Introduction/description go to title toc

Distribution (and Subpopulations) go to title toc

Management structures, entire range of Gadus morhua go to title toc

Europe

NAFO

Canada go to title toc

Legal Regulatory Framework (Canada)

Scientific Input to Management Decisions (Canada) go to title toc

Political-Level Input to Management Decisions (Canada) go to title toc

Recent structure and FRCC

Population Size and Trend go to title toc

Habitat go to title toc

Distribution of Habitat & Trend in Quality and Quantity of Critical Habitat go to title toc

General Biology go to title toc

Reproductive Capability

Breeding Age, Breeding Frequency, Number of Young, Age-Sex Ratio, Population Structure, Reproductive Rate

Species movement go to title toc

Eggs, Larvae, Juveniles

Adults

Behaviour, adaptability

Limiting Factors go to title toc

Why are Populations Declining?

Collateral Concerns (biological indicators) go to title toc

Complex Systems and Uncertain Prediction go to title toc

Hypotheses on Possible Limiting Factors go to title toc

Summary (of Limiting Factors) go to title toc

Special Significance of the Species go to title toc

Utilization go to title toc

Recommendations/Management go to title toc

Evaluation go to title toc

Other Concerns, Observations and Recommendations in the Literature go to title toc

Acknowledgements go to title toc

Author's Recommendation of Status

Preamble go to title toc

Is extinction unthinkable? go to title toc

Further uncertainties go to title toc

Recommendation go to title toc

Stock-by-stock? Or assume one homogeneous stock?

Stock by stock: subdivisions to use go to title toc

Designation by geographic management units (as preferred by DFO in 1996)

Designation treating Atlantic Canada as a single unit go to title toc

Further RECOMMENDATIONS go to title toc

Literature Cited (note: citations lacking journal etc. are MSS obtained from authors) go to title toc

ADDENDUM: Issues and responses to comments go to title toc

Background: Timing and delays go to title toc

Population basis of designations go to title toc

Population homogeneity or heterogeneity?

DFO's shifting position on Populations go to title toc

Is it "Ludicrous", "Unrealistic", etc. to consider cod at risk? go to title toc

DFO comments of ~970721, detailed responses: go to title toc

Ad hominem go to title toc

No depensation? go to title toc

Criteria based on declines go to title toc

Are declines irrelevant? (The "millions" argument) go to title toc

Nature of declines go to title toc

Unique vs. natural fluctuation

Fishing vs. climate/cold water go to title toc

climate

fishing go to title toc

indirect effects of fishing go to title toc

Mandate of COSEWIC & risk of extinction go to title toc

REPORT SECTIONS: Recommendations, Otolith, References, etc go to title toc

Section: Recommendations/Management Options

Otolith protocols

DFO comment on references go to title toc

Stock Status Reports (SSRs)

VPA and R/C plots (figure 4 in Report)

Prognoses go to title toc

Causes for concern / DFO comment / author reply go to title toc

Hindsight vs. Foresight: Assessing assertions go to title toc

Hindsight vs Foresight

Assessing assertions (saying it don't make it so) go to title toc

Tables go to title toc

Figure legends go to title toc go to figures page