Warning This Web page has been archived on the Web.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 status report on the Atlantic Salmon (Inner Bay of Fundy populations) in Canada

Population Sizes and Trends

Search Effort and Data Quality

A great deal of effort has been applied to obtaining reliable measures of population size and trends by DFO, especially by the iBoF Working Group at the Bedford Institute of Oceanography. In addition to analyses of commercial fisheries catches of iBoF Atlantic salmon spanning more than a century, extensive data have been compiled on adult and juvenile abundances for the two iBoF index rivers, the Big Salmon (Gibson et al. 2003c) and the Stewiacke (Gibson and Amiro 2003).  Surveys for juvenile densities have recently been conducted in 43 iBoF rivers (Gibsonet al. 2003a, Gibson et al. 2004) and allow the evaluation of population status relative to a “normal index of abundance” of 2.4 eggs/m2 (DFO 2003). 

Commercial Fisheries Data: Records of commercial landings provide the longest historic indicator of population trends, but are subject to numerous uncertainties, including unknown or unreported effort, inaccurate origin assignment, and vagaries of human events such as wars.  Nevertheless, Dunfield (1985, 1986) provides an exhaustive summary and adjusts for apparent inaccuracies; these data have been provided by personal communication and appear in Amiro (2003). 

Big Salmon and Stewiacke rivers:  For both rivers, maximum likelihood models were used to integrate multiple sources of data, including: recreational fishing catch and effort data; fishway and fence counts; shoreline observation and dive counts; redd counts; electrofishing; and mark-recapture experiments.  Although these data have some incompleteness and uncertainty, sophisticated Bayesian analyses provide heuristic abundance and trend calculations with confidence intervals. 

Juvenile Surveys:  The intensive electrofishing surveys of 117 sites in 36 iBoF rivers in 2000 and 246 sites in 43 rivers in 2002 allow an assessment of population decline throughout the iBoF.  A total effort of approximately 244,000 seconds of shocking time was applied over 215,000 m2 of habitat and resulted in the capture of 2,513 Atlantic salmon (Gibson et al. 2003a).  Fry were not found in 30 rivers, and parr were absent in 22.  A more limited survey was conducted for 16 rivers in 2003 with similar results (parr were absent from 5 of the 16 rivers, and age-0 parr were absent from 9) (Gibson et al. 2004).  While such surveys have limitations when fish are rare (i.e., limits to mark-recapture and depletion methods), these limitations have largely been corrected through statistical techniques, such as Bayesian methods to provide probability densities for catchability (Gibson et al. 2003).

 

Current Abundance

Fisheries: the commercial and recreational fisheries are closed and therefore provide no current abundance estimates.

Big Salmon River:  Gibson et al. (2003c) provide a maximum likelihood estimate for the year 2002 of 55 anadromous adults with an 80% Bayesian Credible Interval (BCI) of 18-133.  Hutchings (2003) suggests an effective population size (N­­­e) to census size ratio of 0.21 to 0.64 for Atlantic salmon.  A rough Ne estimate for Big Salmon River anadromous adults in 2002 is therefore 12 to 35 (3.8 to 85.2 across the 80% BCI).

StewiackeRiver: Gibson and Amiro (2003) provide a maximum likelihood estimate for the year 2001 of 2 anadromous adults with an 80% BCI of 2-4. 

Juvenile Surveys: The juvenile survey data are densities in selected areas and are not extrapolated to the entire habitat.  Thus, they do not provide abundance data.

The opinion of the iBoF Working Group is that in 2003 there were less than 100 wild anadromous adult breeders spread across all the rivers, with 50-75 being the most likely and 200 being an upper estimate (18 May 2004, J. Gibson and P. Amiro, personal communication).  Although an additional number of males breed as mature male parr, these individuals have not been included in the current calculations.

 

Fluctuations and Trends

Commercial Fisheries: the annual quantity of iBoF salmon harvested, in kilograms, from 1875 through 1984, is shown in Figure 9.  Though highly variable, landings are highest from 1875 through 1924, smaller from 1925 through 1973, and least from 1974 through 1984 (after which the fishery was closed).  An overall trend of decline is apparent.  Using an average weight of 3.1 kg per fish, Amiro (2003) calculates an overall average of 4,032 salmon captured per year, ranging from 6 to 9,611, across the 109-year history of the commercial fishery.

Figure 9.  Commercial Atlantic salmon catches (kg) in the Inner Bay of Fundy, for the years 1875-1984 from fishery districts in Albert and Westmorland Counties, New Brunswick.  From R.W. Dunfield (pers. comm. to P. Amiro, DFO, Bedford Institute of Oceanography).

Figure 9.  Commercial Atlantic salmon catches (kg) in the Inner Bay of Fundy, for the years 1875-1984 from fishery districts in Albert and Westmorland Counties, New Brunswick.  From R.W. Dunfield (pers. comm. to P. Amiro, DFO, Bedford Institute of Oceanography).

Big Salmon River: The number of anadromous adults returning to spawn from 1964 through 2002 is shown in Figure 10a.  The number of returning adults peaked in 1966 at 5,043 (80% BCI = 3,996 to 6,686).  This is in marked contrast to the 2002 estimate of 55 (18-133). The overall trend has been extremely negative.  Table 2 summarizes the rate of decline over 11 years (three generations based on 3.7 years per generation) as well as over 5-year, 10-year, 20-year, and 30-year periods.  There is a 90% probability that the mean population size declined by more than 94.1% over 11 years, and by more than 96.7% over the last 30 years.  In the past five years alone, rates of decline have exceeded 76%.

StewiackeRiver: The number of anadromous adults returning to spawn from 1965 through 2001 is shown in Fig. 10b.  The number of returning adults peaked in 1967 at 6,693 (80% BCI = 4,698 to 10,998).  This is in marked contrast to the 2001 estimate of 2 adults (2-4).  The overall trend has been extremely negative.  Table 2 summarizes the rates of decline.  There is a 90% probability that the mean population size declined by more than 99% over 11 years (three generations), and by more than 99.6% over 30 years.  In the last five years of the study alone, rates of decline have exceeded 92.4%.

Table 2.  Estimates of rates of decline for Atlantic salmon populations in the two iBoF index rivers. Minimum rates of decline at 90% confidence were measured over 11 years (3 generations), as well as over time periods of 5, 10, 20, and 30 years.  The 3-generation estimate (italicized) averages abundances across a generation length 4 years (3.7 rounded), while other estimates use a 5-year moving average.  Estimates for the Big Salmon Riverare calculated from Gibson et al. (2003c) and are based on the preferred model (Model #4).  Estimates for the Stewiacke River are provided by or calculated in Gibson and Amiro (2003).
RiverTime
Period
(Years)
Moving
Average
(Years)
StartEndMinimum Rate of
Decline
(90% Confidence)
Big Salmon1141988-19911999-200294.1%
 551992-19961997-200176.0%
 1051987-19911997-200194.7%
 2051977-19811997-200196.0%
 3051967-19711997-200196.7%
Stewiacke1141987-19901998-200199.0%
 551992-19961997-200192.4%
 1051987-19911997-200198.8%
 2051977-19811997-200199.0%
 3051967-19711997-200199.6%

Juvenile Surveys: The juvenile surveys in 43 iBoF rivers in 2002 demonstrate that the collapse of iBoF salmon populations is not limited to the Big Salmon and Stewiacke rivers (Figure 11).  Live Gene Bank stocking is supporting nine of the rivers.  Of the remaining 34 rivers that are dependent on natural wild production, fry were found in only 4 (3.4%), suggesting little or no adult reproduction in 2001, and parr were found in only 12 (35.3%).  Mean densities of fry and parr in the New Brunswick rivers were below 5.2 and 3.8 per 100 m2, while the conservation requirement for these stages is 29 fry and 38 parr per 100 m2.  Thus, fry were at 18% and parr were at 10% conservation levels.  In the Nova Scotia rivers, fry were totally absent and parr were below 7.1 per 100 m2, or 19% of conservation needs.  In the smaller 2003 survey, only three age-0 parr were found in ten rivers without LGB support, all of which came from the Point Wolfe River, suggesting that very few salmon spawned in these rivers in 2002 (Gibson et al. 2004).


Figure 10.  Estimated adult returns to two iBoF index rivers (logarithmic scale)


Figure 10.  Estimated adult returns to two iBoF index rivers (logarithmic scale).Solid lines are estimated returns; dashed lines are the 10% and 90% posteriors defining the 80% Bayesian Credible Interval (BCI) measure of confidence.

(a)  Estimated returns to the Big Salmon River, 1964-2002, from Gibson et al. (2003c). (b) Estimated returns to the Stewiacke River, 1965-2001, modified from Gibson and Amiro (2003) to combine large and small returning salmon.


Figure 11.  Box plots showing the density of juvenile Atlantic salmon (fry and parr) in 41 inner Bay of Fundyrivers from a 2002 electrofishing survey at 233 sites.


Figure 11.  Box plots showing the density of juvenile Atlantic salmon (fry and parr) in 41 inner Bay of Fundyrivers from a 2002 electrofishing survey at 233 sites.

Points are median densities.  Boxes show inter-quartile spread.  Whiskers are drawn to the minimum and maximum densities.  Live Gene Bank (LGB) supported (light grey) are rivers into which juvenile Atlantic salmon have been released since 1996.  N is the number of sites electrofished in each river.  For comparison, the dashed lines show the DFO ‘conservation requirements’ for either fry (29 /100 m2) or parr (38 /100 m2) densities, but note that the plots combine these stages.  These reference points can therefore only indicate failure of a river to meet conservation densities (i.e., when the sum of fry and parr densities fail to meet either conservation requirement). The conservation requirements are based on ‘Elson norms’ developed by DFO for use in evaluating stock status.  From Gibson et al. (2003a) (conservation requirements added).


Figure 12.  Comparison of densities of Atlantic salmon and other species in iBoF rivers supported and unsupported by the Live Gene Bank.

Figure 12.  Comparison of densities of Atlantic salmon and other species in iBoF rivers supported and unsupported by the Live Gene Bank. The mean number of fish captured during the first pass while electrofishing at 172 sites on iBoF rivers during 2002 is shown.  LGB supported rivers are those into which juvenile Atlantic salmon have been released since 1996.  Sites where non-salmonid species were not recorded are not included.  From Gibson et al. (2003a).

Amiro (2003) estimates from recreational catch data that the iBoF populations at maximum numbers may have reached 46,614 adults (see also National Recovery Team 2002, DFO 2003).  He estimates an abundance of less than 500 in 1998, and less than 250 in 1999.  There were probably fewer than 100 adults in 2003.  Based on current rates of decline and the estimate of only 100 adults in the most recent year for which spawning data are available, even the most conservative approximations of risk would suggest extinction in the wild in the very near future.  For example, under the assumption that the iBoF is a single breeding population comprised of 100 individuals and that it declines at the minimal 90% confidence rate of the healthiest index river (94.1% per 3-generations), there will be fewer than 2 individuals within 15 years.

Various studies suggest a quasi-extinction threshold for salmon at roughly 100 females (N = 200) (Myers et al. 1995, Botsford and Brittnacher 1996, see also McElhany et al. 2000).  Below this number, for salmon, there is an extremely high risk of complete extinction through demographic stochasticity, inbreeding and other small-population factors.

 

Rescue Effect

As a unique DU, there is no likelihood of rescue, since neighbouring regions harbour genetically dissimilar Atlantic salmon. Even if neighbouring populations of salmon were sufficiently adapted to breed and spawn surviving progeny within the iBoF, the fact remains that the three neighbouring regions – the oBoF, the Scotian Coast, and Maine – have extremely depressed populations (DFO 2003, National Research Council 2004).  The US NMFS listed a Maine DPS as Endangered under theEndangered Species Act in 2001, and it remains supported by its own Live Gene Bank program and is unlikely to recover in the foreseeable future (National Research Council 2004).  Similarly, the oBoF and Scotian Coast regions would likely meet COSEWIC criteria for Endangered status, given their demographic trends.  For instance, preliminary analysis of decline rates for the best two index populations, the Saint John (oBoF) and the LaHave (Scotian Coast) rivers, show three-generation (11 years, 1993-2003) decline rates of 88.5% and 73.0% respectively (data from R. Jones, DFO; calculations by P. Amiro, DFO; 20 May 2004).  These decline rates, while not as severe as those of iBoF index populations, exceed COSEWIC’s decline criteria forEndangered status.