Science Advisory Report  2010/071

Impacts of Grey Seals on Fish Populations in Eastern Canada

Summary

  • Grey seals are found in three linked Canadian Atlantic marine ecosystems south of the Laurentian Channel. The southern Gulf of St. Lawrence (4T) is a shallow semi-enclosed sea that is very productive in summer but freezes in winter, and many fish populations, including cod, migrate and overwinter in the warm deeper waters off Cape Breton (4Vn). The cod stocks in the two Scotian Shelf ecosystems (4VsW and 4X), do not conduct extensive migrations and are larger at age than 4T cod. A small resident cod stock is also found in 4Vn but no analyses were reviewed for this stock. Grey seal is a highly mobile species foraging in waters from Georges Bank to the northern Gulf of St. Lawrence.  Over this broad geographic range, grey seals encounter different prey assemblages, with differing depth distribution and geographic extent. All of these factors are known to influence grey seal foraging behaviour and diet, and coupled with the different dynamics of the Gulf, Sable Island, and coastal Nova Scotia grey seal herds and their prey species, suggest that the influence of grey seals on prey populations such as cod varies by ecosystem.
  • There have been dramatic changes in these ecosystems over the past several decades. Groundfish resources and fisheries once characteristic of the southern Gulf have been replaced by small-bodied demersal fishes and invertebrate fisheries. Like the southern Gulf, the eastern Scotian Shelf ecosystem supported active groundfish fisheries until the early 1990s, which have now been replaced by fisheries on invertebrate species such as shrimp and crab.  On the western Scotia Shelf and in the Bay of Fundy, invertebrate fisheries have also increased to unprecedented levels, but some fishing for groundfish continues with cod as a bycatch in fisheries directed towards other species.
  • The size of the Canadian grey seal population is estimated by a model incorporating estimates of pup production, reproductive rates and information on removals. Trends in abundance from four population models were presented at the meeting. These models make slightly differing assumptions about adult mortality rates, the effects of increasing density on population growth rate and ice-related mortality of seal pups in the southern Gulf. When these assumptions are taken into account, the population is estimated to have increased from approximately 10,000 animals in 1960 to about 330,000-410,000 animals in 2010, depending on which model is used.
  • For management purposes the grey seal population is divided into three herds based upon pupping sites.  The herds feed in coastal and offshore waters of 3Pn, 3Ps, 4R, 4T, 4Vn, 4VsW, 4X and 5Z. The largest herd numbering some 260,000 to 320,000 seals, depending on model assumptions, occurs on Sable Island. The rate of increase of this herd has slowed from 12.8% during the 1980s to approximately 4% in the last 5 years. It is uncertain if this reduction in population growth rate is due to changes in age-specific reproduction, young of the year mortality rates, or both. The southern Gulf of St. Lawrence (Gulf) herd numbers around 55,000-71,000 animals, depending on model assumptions, and although apparently increasing, the rate of increase varies with mortality of young due to poor ice conditions. The coastal Nova Scotia herd is the smallest of the three, numbering around 20,000-22,000 animals. Although the majority of these animals are born on Hay Island along the eastern Shore, there is an increasing abundance and new breeding colonies being established in southwest Nova Scotia with a contribution by immigration from the Sable Island herd.  Seals from each of these herds range widely throughout the year while foraging and may contribute to colonization of new breeding sites.  Reasons for the large increase in the number of grey seals are not fully understood, but reduction in hunting and an increase in ice-breeding habitat in the Gulf are likely to have contributed.  Although little is known about historical abundance, current population size is the largest measured in the past several hundred years.
  • There are distinct cod stocks in each of the three ecosystems.  All stocks have shown declines of at least 80% in abundance and all remain low today.  Overfishing reduced the stocks in 4T, 4Vn, and 4VsW to low abundance by the early 1990s.  Overfishing also contributed to the lesser decline in the 4X stock up to the mid 1990s. Fisheries have been closed or greatly reduced on all these stocks for the past two decades.
  • Despite the severely reduced fishing mortality, survival of adult cod in 4T has remained at a very low level over this period, and the stock has continued to decline. If current levels of productivity and natural mortality were to persist, the stock is estimated to decline to levels near extirpation within 40-50 years. Similar elevated levels of natural mortality are observed in other large demersal fishes in 4T such as winter skate and white hake, which have also declined to very low abundance, and are continuing to decline.  The 4Vn resident cod stock has remained very low since the 1980s. The 4VsW cod stock fell rapidly in the late 1980s leading to collapse, followed by fishery closure in 1993.  Stock biomass remained low for over a decade but has recently shown an increase and improved survivorship of young cod.  The 4X cod stock also experienced high stock mortality and continued to decline after the mid 1990s when fisheries were restricted.
  • For 4T cod, a number of the potential causes of the increased adult mortality were examined, including unreported catch, emigration, disease, contaminants, parasites, poor fish condition, life history change and predation due to seals and other predators. A review of the weight of evidence for each cause supported a conclusion that predation by grey seals was likely the greatest contributor to increased mortality in large southern Gulf cod. Changes in the species and size composition of the southern Gulf fish community are consistent with changes in mortality similar to those observed for cod. Grey seal predation is also considered an important component of the high natural mortality of winter skate and white hake, two species at high risk of extirpation in the southern Gulf.
  • Determining the diet of grey seals relies mostly on indirect methods because there are limited opportunities to directly observe what they eat. The methods used are based on recovery of hard parts such as fish ear bones from stomach contents, intestines, and feces, and the analysis of blubber chemistry in seals and their prey. Each of these methods has strengths and weaknesses.  In addition to challenges in determining what a grey seal has eaten, it is also difficult to obtain a representative sample of the diet from grey seals because they range widely and their diet varies by sex, season, area and other factors. Analyses of the above data sources indicate a wide range of values for the percentage of cod in the diet of grey seals; an overall average of 2-7% in 4VsW, and in 4T, from 1% for females in summer to 24% for the only sample of males in winter. All of these methods make assumptions that may be violated to a greater or lesser extent, but it is believed that the estimates are reasonable representations of the diet in the areas that have been sampled.
  • In order to estimate consumption, the various estimates of diet must be extrapolated to the entire area and over the entire year. However, gaps in sampling greatly increase the uncertainty of the consumption estimates and may introduce bias. In particular, consumption of large cod may be underestimated because of current spatial and seasonal gaps in diet sampling during periods when cod are aggregated.  
  • Energetic models indicate that individual grey seals require approximately 1-2 tonnes of prey annually (3-6 kg/day), depending on seal age and sex, and energy content of the prey mixture in their diet.  This variability in energy content of prey affects the stock-specific estimates of cod consumed by seals (see paragraph 8). Our best estimate of grey seal consumption of cod in recent years lies within the range of 4,500 to 20,000 tonnes per year for 4T, and between 3,000 to 11,000 tonnes per year for 4VsW. These estimates themselves have high variance and their wide ranges reflect uncertainty attributable to the assumptions made to address gaps in sampling in 4T and the treatment of the diet data in 4VsW.  
  • In 4T, grey seals are considered a significant source of mortality for large cod (>35cm) and other adult, bottom-dwelling fish. Southern Gulf cod occur in dense aggregations during seasonal migrations, spawning and on the overwintering grounds. Satellite tracking indicates that some grey seals, in particular males, forage where these aggregations occur. Digestive tract samples from seals foraging on overwintering aggregations of cod contain a relatively high proportion of cod (about 24% in males and 10% in females, based on intestine samples), and a high proportion (58%) of these cod were greater than 35cm in length. Seals are also considered a source of high mortality on winter skate and white hake, species considered by Committee on the Status of Endangered Wildlife in Canada (COSEWIC) to be at heightened risk of extirpation. 
  • For 4VsW cod, seal predation is also noteworthy, but its magnitude compared to other sources of mortality varies with model assumptions. Most models leave a large portion of mortality unaccounted for and attribute only a small (less than 17%) portion of total cod mortality to seal predation.  Comparable information is not available for the mortality inflicted by grey seals on cod in 4X and 4Vn. 
  • Grey seals transmit a parasite, larval sealworm, which accumulates in the flesh of cod and other groundfish species resulting in increased processing costs and reduced marketability. In 2008-09, sealworm abundances in 4T cod increased dramatically, reaching levels greatly exceeding those reported in 4VsW and 4X cod in 2006. Infection appears to have little impact on the condition of 4T cod.
  • Models that make a variety of assumptions were used to estimate the reductions in seal abundance that would be required to reverse the declines of 4T cod. Results of the models differed greatly, reflecting uncertainties about interactions between seals and cod.  Consequently, it is not possible to specify a level of reduction that would be necessary or sufficient to reverse the cod decline. Given one set of assumptions about how to fill data gaps, the proportion of mortality due to seal predation is estimated to be so low that total elimination of this mortality source would not be sufficient to allow recovery of cod.  Given other assumptions about how to fill data gaps that would produce mortality estimates consistent with the weight of evidence (paragraph 7), seal removal could reduce cod mortality enough to allow recovery, but the necessary reductions would be substantial.  This model estimates that in order to lower natural mortality of cod to 0.4, the number of seals foraging in the southern Gulf area would need to be reduced by 70% to 31,000 seals. Currently, there are estimated to be 104,000 seals that forage in the southern Gulf, including 36,000 seals from Sable Island, 5,000 seals from the eastern Shore and 63,000 seals from the southern Gulf. Given current productivity and in the absence of fishing, cod could recover at a natural mortality of 0.4.  Removals of this magnitude are also predicted to decrease natural mortality of winter skate such that their abundance could stabilize and perhaps increase but this number of removals would be insufficient to halt the ongoing population decline of white hake.  If particular seals specialize in predation of cod and it is possible to target these seals for removal (e.g., males in areas of high overlap with cod such as Cabot Strait in winter), then the required removals could be much lower. Likewise, targeting grey seals that feed in areas where winter skate or white hake aggregate (Northumberland Strait, Laurentian channel) would likely require considerably smaller removals for comparable reductions in their mortality rates.
  • For 4VsW, the models reviewed provided a wide range of results reflecting uncertainty about seal-cod interactions in recent years. Many of the model scenarios found that seal predation is not an important component of cod mortality, and do not predict a large response of the 4VsW cod stock to changes in seal abundance.  If an intervention were to be made, the consequences for cod of a reduction in the seal population would depend on the age and sex of the seals removed. As the Sable Island seal herd is about 5 times larger than the Gulf herd, an intervention to significantly reduce this population would need to be much larger than that described for 4T in paragraph 14.  Removal or contraception of adult females is the most effective intervention, followed by removal of pups and then adult males. A predator control or a contraception program on the order of 10,000 seals per year for 5 years would have a very low probability of having a detectable consequence for cod. 
  • The magnitude of removals of grey seals foraging in the southern Gulf (described in paragraph 14) was not considered to pose a risk of irreversible harm to the seal population.  By contrast, if productivities of cod, winter skate and white hake in 4T remain at their present levels, further declines would be expected in these populations, which COSEWIC has determined are at an elevated risk of extirpation.
  • Culling is widely practiced as a means to limit predation on livestock and wildlife and can be effective at reducing predator abundance. Culling also has been used to reduce seal species. Nevertheless, although widely practiced, the extent of seal population reduction and the response of targeted prey populations to culls have rarely been evaluated, such that their effectiveness is poorly understood. Further, results from other predator control programs indicate that unintended consequences in food webs, that will be difficult to predict, are nonetheless commonly observed. Thus, any intervention in the southern Gulf would first require a thorough investigation of the likely multi-species impacts of a cod-seal interaction in this ecosystem, and second would require a carefully designed program that would include clearly-stated objectives and rigorous monitoring of the seal and cod populations and the ecosystem to evaluate the consequences.
  • There are several protocols and monitoring requirements that should be followed to effectively design and evaluate a predator control program. Key among these are clear statements of objectives and expected benefits and use of performance measures that provide a quantitative interpretation of the extent to which objectives have been met and benefits realized. To be informative, the control program should be planned and conducted in an adaptive management framework.

This Science Advisory Report is from the Fisheries and Oceans Canada, Canadian Science Advisory Secretariat, Zonal Advisory Meeting of October 4-8, 2010 on Impacts of Grey Seals on fish Populations in Eastern Canada.  Additional publications from this process will be posted as they become available on the DFO Science Advisory Schedule at http://www.dfo-mpo.gc.ca/csas-sccs/index-eng.htm.

View complete PDF document
(48 pages; 955K)

Accessibility Notice:

This document is available in PDF format. If the following document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.