Warning This Web page has been archived on the Web.

Archived Content

Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards, as per the Policy on Communications and Federal Identity.

Skip booklet index and go to page content

COSEWIC assessment and update status report on the Atlantic Cod in Canada

Biology

General

Despite having been fished for more than 500 years in Canadian waters, there remain large gaps in our knowledge of many of the most basic elements of the biology and ecology of this species.  Nonetheless, it is known that after hatching, a period of time that takes approximately 60 degree-days, the larvae obtain nourishment from a yolk sac until they have reached a length of 1.5 to 2.0 mm.  During the larval stage, the young feed on phytoplankton and small zooplankton in the upper 10 to 50 metres of the water column.  After a few weeks, the larvae swim, or ‘settle’, to the bottom, where they appear to remain for a period of 1 to 4 years throughout most of the species’ Canadian range.  These settlement areas are known to range from very shallow (< 10 m to 30 m) coastal waters to moderately deep (50 to 150 m) waters on offshore banks.  In addition to providing food, these settlement areas almost certainly provide habitat that provides the larval and juvenile cod protection from predators.  After this settlement period, it is believed that the fish begin to undertake the often-seasonal movements (apparently undirected swimming in coastal waters) and migrations (directed movements to and from specific, highly predictable locations) characteristic of adults.  Anecdotal reports suggest that very large adults (> 100 cm) may not migrate as extensively as smaller adults.

 

Reproduction:  Life history variation

The life history of cod varies a great deal (Myers et al. 1996; McIntyre and Hutchings in press).  Most life history traits, such as age and size at maturity, longevity, and size-specific fecundity differ greatly among populations, while some, such as egg size, appear to be similar throughout the species’ range.  As with most indeterminately growing organisms (those that continue to increase in size after maturity), fecundity (the number of eggs per female per breeding season) increases with body size.  In cod, as with most fish, the number of eggs per female generally increases with body mass as a power function.  Body size at a given age is a function of growth rate, a parameter that varies greatly among cod populations, being relatively slow in the north and fast in the south.  In turn, growth rate of cod is a function of temperature, food supply, density, and the proportional allocation of energy to reproduction.

In the relatively warm waters at the southern end of its Canadian range (Georges Bank, off the state of Maine) and in the Bay of Fundy, cod commonly attain maturity at 2 to 3 years of age (Trippel et al. 1997; McIntyre and Hutchings in press).  By contrast, cod inhabiting the Northeast Newfoundland Shelf, eastern Labrador, and the Barents Sea reproduce for the first time between 5 and 7 years of age (Myers et al. 1997b; Smedbol et al. 2002).  One consequence of these population differences in age at maturity is population variation in generation time.  However, when providing an estimate of generation time, one needs to be cognizant of the fact that generation time has almost certainly changed over time for Atlantic cod.  This can be attributed to apparent reductions in age at maturity in some populations (Trippel et al. 1997) and to reductions in longevity.  For example, in the early 1960s, it is estimated that more than 50% of the eggs produced by Newfoundland’s northern cod stock were produced by females 10 years of age and older (Hutchings and Myers 1994).  By the late 1980s, this age class is estimated to have contributed less than 10% of the eggs.  In the late 1990s and as recently as autumn 2000, females older than age 10 were not sampled by DFO surveys of the northern cod stock (Lilly et al. 2001).

Size at maturity can also differ significantly among cod populations.  On average, length at maturity typically ranges between 45 and 55 cm.  Smaller sizes at maturity have been reported in recent years for Eastern Scotian Shelf cod (33-37 cm; Paul Fanning, DFO, Bedford, NS, personal communication) and for cod in the genetically isolated population inhabiting Gilbert Bay, Labrador (31-42 cm; Morris and Green 2002).  At the other extreme, length at maturity has been reported to be as large as 65 cm for males and 85 cm for females in Ogac Lake, Baffin Island (Patriquin 1967).  The number of eggs produced by a single female in a single breeding season typically ranges from between 300,000 and 500,000 at maturity to several million for females greater than 75 cm in length.  There is recent evidence that size-specific fecundity, that is, the number of eggs produced per unit of body mass, differs significantly among cod populations in the Northwest Atlantic and within populations over time (McIntyre and Hutchings in press).  Egg diameter, which shows a weak, positive association with body size, ranges between 1.25 and 1.75 mm (Chambers and Waiwood 1996).

 

Reproduction:  Spawning behaviour

Atlantic cod typically spawn over a period of less than three months (Brander 1994; Chambers and Waiwood 1996; Kjesbu et al. 1996) in water that may vary in depth from tens (Smedbol and Wroblewski 1997) to hundreds of metres (Brander 1994; Morgan et al. 1997).  Although individuals are assumed to breed annually, Atlantic cod are described as batch spawners because of the observation that only 5 to 25% of a female's egg complement is released at any given time during her 3- to 6-week spawning period (Chambers and Waiwood 1996; Kjesbu et al. 1996).  Spawning intervals of 2 to 6 days appear typical of individual females held in captivity (Kjesbu 1989; Chambers and Waiwood 1996; Kjesbu et al. 1996).

The behaviour that immediately precedes the release of sperm and eggs was initially documented at nineteenth century Atlantic cod hatcheries in Newfoundland (Templeman 1958) and Norway (Dannevig 1930).  These observations, and those of Brawn (1961), describe a "ventral mount" in which the male, while grasping the female with his pelvic fins and matching her swimming speed, positions himself beneath the female with the urogenital openings of both fish opposite one another.

Based on the only two papers in the primary scientific literature that have reported cod behaviour during spawning (Brawn 1961; Hutchings et al. 1999), successful reproduction in Atlantic cod involves a complex repertoire of behaviours within and between sexes.  Spawning male cod appear to establish a dominance hierarchy, with rank determined by aggressive interactions, particularly chases of one male by another, and possibly by body size, larger individuals often being dominant over smaller individuals.  Agonistic interactions, continuing through the spawning season, may allow high-ranking males to defend territories.  Limited genetic data suggest that male fertilization success increases with male body size and/or behavioural dominance (Hutchings et al. 1999) and that eggs from a single reproductive bout can be fertilized by more than one male (Hutchings et al. 1999; Rakitin et al. 2001).  Based on these genetic data, and on the direct observation of spawning behaviour by cod from southern Nova Scotia and the Southern Gulf of St. Lawrence (S. Rowe and J.A. Hutchings, unpublished data), it is reasonable to assume that satellite males, adopting a 'sneak' form of mating, are a regular feature of cod reproduction and that satellite males are able to fertilize some of the eggs released by a female during a spawning.  The phenotypic and behavioural correlates of reproductive success in Atlantic cod are currently being examined in a series of spawning experiments being conducted at Dalhousie University by the author.

Hutchings et al. (1999) hypothesized that interactions between sexes are consistent with the hypothesis that females, and possibly males, exercise mate choice.  One prominent behaviour observed in large experimental tanks is the circling of individual females by individual males on or near the bottom (Hutchings et al. 1999; Skjæraasen et al. submitted).  Several factors associated with this circling behaviour are suggestive of female mate choice.  Firstly, these circling bouts are initiated and terminated by females.  Secondly, by restricting circling to occasions when they are positioned directly on the bottom, females can effectively prevent ventral mounts by circling and non-circling males.  Thirdly, circling provides females the opportunity to be in close physical contact with, and assess the quality of, several males prior to spawning.  Hutchings et al. (1999) have also hypothesized that females may be choosing males on the basis of the sounds produced by male gadids during spawning (Brawn 1961; Hawkins and Amorim 2000; Nordeide and Folstad 2000).  Acoustic communication in cod is facilitated by drumming muscles whose rapid vibrations against the air bladder are capable of producing low-frequency sounds audible to other cod.  Preliminary analyses suggest that the size of a male's drumming muscle, relative to that male's body size, may be positively associated with mating success and may differ among populations (S. Rowe and J.A. Hutchings unpublished data).

 

Survival

The high fecundity of Atlantic cod (ranging from several hundred thousand to several million eggs per female per breeding season) represents a life history adaptation by cod that allows them to adopt the reproductive strategy of releasing eggs directly into the water column and of providing these eggs with no protection, either through the construction of egg nests or through the provision of parental care.  This strategy of maximizing the production of eggs the sizes of which approach, or attain, the physiological minimum for survival has been interpreted as an adaptive response to environments in which egg size confers no consistent, inter-generational advantage to survival in early life (Hutchings 1997).

The high-fecundity strategy adopted by Atlantic cod is an evolutionary response to the exceedingly high mortality associated with such a reproductive strategy.  Based on estimates of fecundity, weight-at-age, and age-specific abundance of northern cod, Hutchings (1999) estimated that survival from birth until the age of 3 years averaged 1.13 × 10 - 6, or approximately one in one million, for the cohorts of cod born from 1962 to 1988.  Between the ages at which cod first become vulnerable, or are recruited, to the commercial fishery (varying between 1 and 3 years for Canadian stocks of cod, being younger in the south) and the age at death, the annual mortality probability of cod, independent of age and size, has been estimated to be 18% (Pinhorn 1975).

Prior to the closure of most fisheries to targeted or directed fishing in the early 1990s (July 1992, for northern cod; January 1994, for northern Gulf cod; September 1993, for all other stocks except Western Scotian Shelf/Bay of Fundy cod and Georges Bank cod, neither of which ever closed), fishing was the dominant source of mortality for Atlantic cod.  At one extreme, it is estimated that fishing removed annually more than 70% of Newfoundland’s northern cod available to be caught in the late 1980s and early 1990s (Baird et al. 1992; Hutchings and Myers 1994).  Fishing remains a primary source of mortality for parts of the Newfoundland & Labrador, Laurentian North, and Maritimes Populations (Smedbol et al. 2002).  In some areas, most of the mortality experienced by Atlantic cod, at all life stages, can probably be attributed primarily to predation by fish and marine mammals, and secondarily to predation by invertebrates and birds (Bundy et al. 2000; McLaren et al. 2001).  Nonetheless, fishing still remains a sizeable source of mortality in many areas, a result of directed fishing quotas and bycatches.

With the exception of cod along one section of the south coast of Newfoundland (NAFO Division 3Ps), cod populations have grown little since the initial closure of the targeted fishing activities almost a decade ago.  Although contrary to predictions of rapid recovery made in the early 1990s (detailed by Hutchings et al. 1997), slow rates of population growth following collapses in population size are not atypical when one compares recovery rates for collapsed marine fish stocks worldwide (Hutchings 2000), even when one accounts for reductions in fishing mortality (Hutchings et al. 2001a; Denney et al. submitted).  The reasons for slower-than-expected rates of population growth are not known with certainty, although they may include one or more of the following:  management failure to reduce to fishing mortality to nil; changes to species community composition with concomitant changes to competitive interactions and predator-prey relationships; reduction in critical habitat; reductions in fertilization rates and/or social interactions necessary for successful reproduction with reductions in fish density.

The relatively slow rates of population growth may be a product of what is known in the fisheries literature as ‘depensation’ (Myers et al. 1995), and what is known throughout the ecological literature as ‘The Allee Effect’.  Both refer to situations in which per-capita rates of population growth decline, rather than continually increase, when population sizes fall below some ‘threshold’ level of abundance.  The existence of an Allee Effect has been suggested as one explanation for the relatively slow recovery of Atlantic cod and other marine fishes (Shelton and Healey 1999; Frank and Brickman 2000; Hutchings 2000, 2001a,b; De Roos and Persson 2002).  It is worth noting, however, that the observed population growth rates of cod may not be unduly slow, if one incorporates stochastic, or unpredictable, variation in the parameters used to model population growth for Atlantic cod, rather than basing one’s predictions on deterministic models (in which all model parameters are fixed).  Hutchings (1999) undertook such a modelling exercise for northern cod.  Based on his findings, and given the relatively slow rates of individual growth (weight-at-age) and comparatively old age at maturity experienced by northern cod, the observed rate of recovery of northern cod (indeed lack thereof) may not be unexpected.

If resource managers had been consistent in their setting of quotas for the collapsed cod fisheries, and had permitted no targeted or incidental fishing of cod in the past 10 years, the recovery rates for some cod stocks would almost certainly have been greater than what has been observed.  One example will serve to illustrate this point.  Between 1981 and 1989, the annual quota for northern cod never fell below 200,000 tonnes (1 metric tonne is equal to 1,000 kilograms) (Lilly et al. 2001).  A number of fishing industry observers and government resource managers have assumed that considerably smaller quotas would not harm the recovery of this stock.  In 1999 and 2000, the quotas for northern cod were 9,000 and 7,000 tonnes, respectively (Lilly et al. 2001), approximately 97% less than the quotas of the mid-1980s.  Despite their low levels, these quotas have resulted in rates of fishing mortality sufficiently high to negatively affect stock recovery (these are presented below; see LIMITING FACTORS AND THREATS), providing clear evidence that small quotas can negatively affect the recovery of fish stocks when the stocks themselves are at historically small sizes.