Salmon Feeding in the Labraor Sea
Abstract
Atlantic salmon returning from marine migrations to eastern Canada and western Ireland during 2002 and 2003 were analysed for tissue concentrations of bio-accumulated caesium 137 (137Cs). Salmon from Canadian and Irish waters demonstrated concentrations (0.20 ± 0.14 Bq kg−1 and 0.19 ± 0.09 Bq kg−1, mean ± s.d., respectively) suggesting similar oceanic feeding distributions during migration. Canadian aquaculture escapees had a similar mean tissue concentration (0.28 ± 0.22 Bq kg−1), suggesting migration with wild salmon. However, significantly higher concentrations in 1-sea-winter (1SW) escapees (0.43 ± 0.25 Bq kg−1) may alternatively suggest feeding within local estuaries. High concentrations in some Canadian 1SW salmon indicated trans-Atlantic migration. Low concentrations of Canadian multi-sea-winter (MSW) salmon suggested a feeding distribution in the Labrador and Irminger Seas before homeward migration, because those regions have the lowest surface water 137Cs levels. Estimates of wild Canadian and Irish salmon feeding east of the Faroes (∼8°W) were 14.2% and 10.0% (1SW, 24.7% and 11.5%; MSW, 2.9% and 0.0%), respectively. We propose that most anadromous North Atlantic salmon utilize the North Atlantic Gyre for marine migration and should be classified as a single trans-Atlantic straddling stock.Spares, A.D., Reader, J.M., Stokesbury, M.J.W., McDermott, T., Zikovsky, L., Avery, T.S., and Dadswell, M.J. 2007. Inferring marine distribution of Canadian and Irish Atlantic salmon (Salmo salar L.) in the North Atlantic from tissue concentrations of bio-accumulated caesium 137. – ICES Journal of Marine Science, 64: 394–404.
Introduction
Marine mortality has been suggested as a major contributing factor to the decline of wild Atlantic salmon populations (Cairns, 2001; Hutchinson et al., 2002; Lacroix and Knox, 2005). This mortality is proposed to happen within a salmon's first months at sea ( Holm et al., 2003). However, a recent study indicated good survival rates nearshore of Atlantic salmon post-smolts ( Lacroix et al., 2005). As a result, numerous researchers have proposed that research effort should focus on investigating factors that affect salmon survival offshore ( Hutchinson et al., 2002; Lacroix and Knox, 2005). Determining the cause of marine mortality requires a thorough understanding of the spatial and temporal distribution of Atlantic salmon during their oceanic feeding phase (Friedland, 1998; Jonsson and Jonsson, 2004). Unfortunately, Atlantic salmon migration routes and oceanic distribution patterns are poorly known (Hansen and Quinn, 1998; Holm et al., 2003).
The currently accepted model for open-ocean migration of Atlantic salmon is based on past tag returns (Templeman, 1967; Jensen, 1980a, b; Swain, 1980; Meister, 1984; Baum, 1997; Jacobsen et al., 2001), mixed stock discrimination studies ( Reddin et al., 1984a; Reddin, 1987; Reddin and Freidland, 1999, Jacobsen et al., 2001), and the distribution of commercial high seas fisheries (Christensen and Lear, 1980; Jensen and Lear, 1980; Hansen and Pethon, 1985; Reddin, 1986; Reddin and Dempson, 1986; Jakupsstovu, 1988; Scarnecchia et al., 1991; Vigfússon and Ingólfsson, 1993). Southern European stocks (<62°N) are proposed to migrate in relatively straight lines from natal rivers to feeding grounds off West Greenland (Went, 1973; Swain, 1980). Alternatively, northern European stocks (>62°N) are proposed to migrate to feeding grounds off the Faroe Islands and Iceland, and in the Norwegian Sea ( Hansen et al., 1993, Jacobsen et al., 2001), so may be subject to alternative causes of mortality. All North American stocks are thought to migrate directly to waters off West Greenland in the warmer months and to overwinter in the Labrador Sea and off the Grand Banks east to 44°W longitude (Reddin, 1986).
Environmental tracers, such as caesium-137 (137Cs), have been utilized over a range of spatial scales to infer migratory behaviour of marine fish and mammals ( Born et al., 2002; Tolley and Heldal, 2002), including Atlantic salmon ( Tucker et al., 1999). Globally dispersed during the later half of the 20th century as a nuclear fission byproduct, 137Cs formed a highly pronounced east–west gradient in the North Atlantic as a consequence of anthropogenic inputs from nuclear reprocessing facilities, Sellafield in the Irish Sea and La Hague in the English Channel, and fallout from the 1986 Chernobyl accident ( Tucker et al., 1999). Environmental concentrations decrease from point sources with distribution largely controlled by oceanographic processes (Marine Institute, 1999). Discharges of 137Cs from Sellafield and La Hague peaked in the 1970s and 1980s, but have since been reduced by more than two orders of magnitude ( Povinec et al., 2003), decreasing concentrations in the Irish Sea (Marine Institute, 1999) and the eastern Atlantic ( Povinec et al., 2003). However, bioavailability of 137Cs still exists from the water column and through remobilization from contaminated sediments ( Povinec et al., 2003).
Here, we infer the marine distribution of Atlantic salmon based on relative bio-accumulated 137Cs concentrations found in the tissues of migrant salmon returning to Canadian and Irish rivers, compared with known 137Cs levels in surface waters of the north Atlantic Ocean. On the basis of these data and past research, we propose an alternative hypothesis that Atlantic salmon utilize the North Atlantic Gyre for their marine migration (Figure1). Our information will be crucial to the development of hypotheses concerning marine migration patterns and the causes of marine mortality of Atlantic salmon.
Figure 1.
North Atlantic Gyre (open arrows; compiled from Krauss, 1986; Hansen and Osterhus, 2000; Lavender et al., 2000; Jakobsen et al., 2003) showing redistribution of mean surface seawater levels of 137Cs (mBq l−1) in the north Atlantic Ocean for the period 2000–2003 (compiled from RPII, 2002, 2003; NRPA, 2003, 2004; STUK, 2003, 2004; Dahlgaard et al., 2004; HELCOM, 2004; J. Smith, pers. comm.; I. Sværen, pers. comm.). Atlantic salmon sample sites for 2002 and 2003 are indicated (black dots, Canada; black triangles, Ireland, open squares, aquaculture sites).
Material and methods
Wild, adult salmon were collected opportunistically from fishers or researchers in Atlantic Canada and western Ireland. Aquaculture salmon, to be used as western and eastern Atlantic controls, were obtained from salmon farms in the Bay of Fundy (BF), Canada, and Killery Fjord, western Ireland, respectively. Wild Canadian parr and Irish smolts were collected from the same watersheds as adults, if possible. Returning aquaculture escapees were obtained from researchers in the Bay of Fundy, Canada.
Morphometrics and scale samples were taken at collection. Fork length (FL) of salmon was measured to the nearest 0.1 cm and wet body mass (WT) to the nearest 0.1 kg. When only heads were available, fork lengths and wet body weights were calculated from FL to WT and head length (HL) to FL ratios from whole body adult samples (n = 113 and 161, respectively). Scales samples were collected and stored as in Shearer (1992). If no scales were available (n = 71), otoliths were used for age determination. Heads of adult fish were removed and stored frozen in individual plastic bags before analysis.
Laboratory analysis
Frozen heads were thawed and head length (tip of snout to posterior edge of the opercle bone; Hubbs and Lagler, 1964) was measured to the nearest 0.1 cm, using calipers. Unknown head lengths from partial or unmeasured heads were estimated from HL:FL relationships for all adult samples. Otoliths were removed with non-contaminated Teflon forceps, cleaned, and stored in acid-washed polyethylene tubes. Head tissue weights were obtained to the nearest 0.1 g, using an electronic balance unless otherwise specified. Head wet weights (HWt0) were taken before initial drying, which was done at 175°C for 24 h. Dentary, maxillary, operculum, parietals, pectoral fins, and dorsal process of the shoulder girdle were removed and weighed together (HWt b ). Remaining bones and soft tissue were dried at 175°C for 48 h and subsequently ashed at 450°C for 12 h or until a fine, grey powder ash was obtained (Morinville, pers. comm.). Individual ashed samples were homogenized with a mortar and pestle, weighed to the nearest 0.001 g, and stored in glass vials at sub-zero temperatures until shipment for gamma ray spectroanalysis. Parr whole-body wet samples (BWt x ) were weighed to the nearest 0.01 g. Samples were dried at 175°C for 72 h, then frozen until shipment. Smolt whole-body samples followed parr procedures, but were subsequently ashed for 6 h at 450°C.
Scales were cleaned, mounted between two glass slides, and aged using a projection microscope. Ages were determined by consensus from readings obtained by two separate investigators. Samples lacking appropriate scales for analysis were aged from whole otoliths read with a dissecting microscope, using a two-investigator reading procedure as well. Scale and otolith annuli were counted according to guidelines outlined by Shearer (1992) and Chilton and Beamish (1982), respectively. Disagreements over presumed ages were resolved by re-reading samples with both investigators present. Sea age for samples without collected scales or otoliths was estimated using the FL:sea age relationship for all the adult samples collected.
Gamma ray spectrometry
Ashed samples were weighed to the nearest 0.001 g and loaded into polyethylene cylinders for one-time counting using a high-purity germanium (HPGe) coaxial well gamma ray detector located at École Polytechnique, Montréal. Individual juvenile samples <15 g wet weight (n = 44) and below detection limits were pooled and counted just once. Counted samples were returned and stored at sub-zero temperature. Both minimum and maximum concentrated samples and a further nine randomly selected samples were recounted to validate results, using two HPGe gamma ray spectrometers at McGill University, Montréal.
Data analysis
Tissue concentrations of 137Cs (Bq kg−1 wet weight) for respective Canadian and Irish Atlantic salmon samples were regressed against wet body weight (WT). All samples were grouped by year of capture, continent of origin, oceanographic region of natal river drainage, natal river, and category (aquaculture, escapee, wild). Both generalized linear models (GLM ANOVA) and analysis of variance (ANOVA) followed by a post hoc Tukey test were used to examine overall trends between 137Cs burden, duration at sea (sea age, FL, WT), and grouping variables (SYSTAT, 2004). Outliers, as defined by SYSTAT (2004), were removed for statistical analyses based on sample means and for the plotting of box plots comparing 137Cs concentrations. However, outliers were considered to be critical data points and evidence for the development of a trans-Atlantic migration hypothesis. A significant trend between sea age cohorts and 137Cs concentrations (ANOVA, p < 0.001) allowed separation of 1-sea-winter (1SW) and multi-sea-winter (MSW) wild Canadian samples pooled by oceanographic region of natal river drainage. The pooled samples were then compared with pooled samples from endangered inner Bay of Fundy (iBoF) populations using a post hoc Tukey test. Any region with a sample size (n) of <3 was excluded from the analysis. Fork length and 137Cs burden for wild migrants was regressed, using both raw and log-transformed data. Log-transformed 137Cs burden data yielded an even lower value of r 2 value compared with a raw data regression. Therefore, raw data trend-line values (Figure2) were subtracted from originals [137Cs] to calculate Cs residuals to be used to de-trend data and to increase sample sizes for natal river comparisons within oceanographic river drainage regions: Atlantic coast (AT), BF, and Gulf of St Lawrence (GSL) in Canada and the west coast of Ireland (WC). Cs residuals for each sample were calculated using
where [137Cs] is the 137Cs concentration (Bq kg−1 wet weight) and FL is fork length (cm) at capture. Individual (n = 16) and pooled (n = 44) juvenile 137Cs burden data were compared using a Z-test. A two-sample Kolmogorov–Smirnov test was used to compare 137Cs concentration frequency distribution of wild North American samples collected in 2002 and 2003 with the 1995/1996 study of the Ste Marguerite River Atlantic salmon population (see Tucker et al., 1999).
Figure 2.
![Tissue concentration of 137Cs (Bq kg−1 wet weight) in relation to fork length (FL, cm) of wild adult Atlantic salmon sampled in 2002 and 2003 [r2 = 0.26, y = 0.6409 exp(−0.0199x)].](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/icesjms/64/2/10.1093_icesjms_fsl040/1/m_fsl04002.gif?Expires=1668299708&Signature=yBD8AiV-83B2YjzoDKs1fR3y4nOqjCgXcOJIrh3cE-VMYD1lnVU-mdoqrZY16WVXhVkgClpNt6eMO6-rZ50TjT9sGs1AGbeDS8ULiT52KNMt8YZXoDpwnzADLcw0somTHo3DmhtoGMT11CGgRReMVbH76iEs-u5IJHrWg8FWQ2bUDlKTvmrv-SJ38kHC-elIYbXwOiQkY~wlgsV~LFfpjkxBjrGpOGmeKMfSiGOPf2sCsD8U-AEdh4S7dPYMRm1pI8t-0NYtPEjI5IGW9VVTsQtmvTjFs6okctM~OPWlnbIy-zwMG1ExFAKq8cqrXND7INfv0vv2D5WD~sEkYxRTgA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Tissue concentration of 137Cs (Bq kg−1 wet weight) in relation to fork length (FL, cm) of wild adult Atlantic salmon sampled in 2002 and 2003 [r 2 = 0.26, y = 0.6409 exp(−0.0199x)].
Results
In all, 173 wild, adult salmon were collected in Atlantic Canada (n = 143) and western Ireland (n = 30) between September 2002 and November 2003. A further 12 aquaculture salmon controls were obtained from a farm in the BF, Canada (October 2002), and 10 from Killery Fjord in western Ireland (July 2003). Wild Canadian parr (n = 44) and Irish smolts (n = 16) were collected from the same watersheds as adult samples where possible. In all, 26 returning aquaculture escapee adults were obtained from research operations in watersheds of the BF, Canada (Table1).
Table 1.
Sampling locations and tissue concentrations of 137Cs (Bq kg−1 wet weight) in Atlantic salmon collected from Canada and Ireland in 2002 and 2003 (n = 282), assigned to either wild (W), aquaculture (A), or aquaculture escapee (AE) origin, and age groups of 0-sea-winter (0SW), 1-sea-winter (1SW), and multi-sea-winter (MSW) fish by scale or otolith examination.
| Country | Region/River | Origin | Stage | n | [137Cs] (Bq kg−1 wet weight) | CL (95%) | |
|---|---|---|---|---|---|---|---|
| Mean | s.e. | ||||||
| Canada | GSL | W | Parr | 44a | 0.30 | – | – |
| Canada | AT | W | 0SW | 3 | 0.47 | 0.08 | 0.37 |
| Canada | AT | W | 1SW | 26 | 0.21 | 0.04 | 0.08 |
| Canada | BF | W | 1SW | 12 | 0.20 | 0.03 | 0.07 |
| Canada | BF | A | 1SW | 12 | 0.09 | 0.01 | 0.03 |
| Canada | BF | AE | 1SW | 12 | 0.43 | 0.07 | 0.16 |
| Canada | iBoF b | W | 1SW | 8 | 0.21 | 0.04 | 0.10 |
| Canada | Black R. | W | 1SW | 2 | 0.21 | 0.01 | 0.13 |
| Canada | Gaspereau R. | W | 1SW | 4 | 0.20 | 0.08 | 0.25 |
| Canada | Big Salmon R.c | W | 1SW | 2 | 0.23 | 0.13 | 1.59 |
| Canada | GSL | W | 1SW | 35 | 0.28 | 0.03 | 0.05 |
| Canada | AT | W | MSW | 1 | 0.07 | – | – |
| Canada | BF | W | MSW | 7 | 0.17 | 0.02 | 0.04 |
| Canada | BF | AE | MSW | 14 | 0.15 | 0.02 | 0.03 |
| Canada | GSL | W | MSW | 60 | 0.14 | 0.01 | 0.02 |
| Canada | iBoF b | W | MSW | 4 | 0.15 | 0.02 | 0.06 |
| Canada | Gaspereau R. | W | MSW | 1 | 0.15 | – | – |
| Canada | Big Salmon R.c | W | MSW | 3 | 0.15 | 0.03 | 0.11 |
| Ireland | WC | W | Smolt | 16 | 0.53 | 0.04 | 0.08 |
| Ireland | WC | W | 1SW | 26 | 0.19 | 0.02 | 0.04 |
| Ireland | WC | A | 1SW | 10 | 0.23 | 0.04 | 0.08 |
| Country | Region/River | Origin | Stage | n | [137Cs] (Bq kg−1 wet weight) | CL (95%) | |
|---|---|---|---|---|---|---|---|
| Mean | s.e. | ||||||
| Canada | GSL | W | Parr | 44a | 0.30 | – | – |
| Canada | AT | W | 0SW | 3 | 0.47 | 0.08 | 0.37 |
| Canada | AT | W | 1SW | 26 | 0.21 | 0.04 | 0.08 |
| Canada | BF | W | 1SW | 12 | 0.20 | 0.03 | 0.07 |
| Canada | BF | A | 1SW | 12 | 0.09 | 0.01 | 0.03 |
| Canada | BF | AE | 1SW | 12 | 0.43 | 0.07 | 0.16 |
| Canada | iBoF b | W | 1SW | 8 | 0.21 | 0.04 | 0.10 |
| Canada | Black R. | W | 1SW | 2 | 0.21 | 0.01 | 0.13 |
| Canada | Gaspereau R. | W | 1SW | 4 | 0.20 | 0.08 | 0.25 |
| Canada | Big Salmon R.c | W | 1SW | 2 | 0.23 | 0.13 | 1.59 |
| Canada | GSL | W | 1SW | 35 | 0.28 | 0.03 | 0.05 |
| Canada | AT | W | MSW | 1 | 0.07 | – | – |
| Canada | BF | W | MSW | 7 | 0.17 | 0.02 | 0.04 |
| Canada | BF | AE | MSW | 14 | 0.15 | 0.02 | 0.03 |
| Canada | GSL | W | MSW | 60 | 0.14 | 0.01 | 0.02 |
| Canada | iBoF b | W | MSW | 4 | 0.15 | 0.02 | 0.06 |
| Canada | Gaspereau R. | W | MSW | 1 | 0.15 | – | – |
| Canada | Big Salmon R.c | W | MSW | 3 | 0.15 | 0.03 | 0.11 |
| Ireland | WC | W | Smolt | 16 | 0.53 | 0.04 | 0.08 |
| Ireland | WC | W | 1SW | 26 | 0.19 | 0.02 | 0.04 |
| Ireland | WC | A | 1SW | 10 | 0.23 | 0.04 | 0.08 |
aSamples (n = 44) pooled and counted only once by gamma ray spectrometry.
biBoF pools Big Salmon, Black, and Gaspereau River samples.
cAccording to Amiro (2003), the Big Salmon River is considered iBoF, whereas the Black, Gaspereau, and Irish River populations are excluded based on a high proportion of virgin, MSW salmon.
Table 1.
Sampling locations and tissue concentrations of 137Cs (Bq kg−1 wet weight) in Atlantic salmon collected from Canada and Ireland in 2002 and 2003 (n = 282), assigned to either wild (W), aquaculture (A), or aquaculture escapee (AE) origin, and age groups of 0-sea-winter (0SW), 1-sea-winter (1SW), and multi-sea-winter (MSW) fish by scale or otolith examination.
| Country | Region/River | Origin | Stage | n | [137Cs] (Bq kg−1 wet weight) | CL (95%) | |
|---|---|---|---|---|---|---|---|
| Mean | s.e. | ||||||
| Canada | GSL | W | Parr | 44a | 0.30 | – | – |
| Canada | AT | W | 0SW | 3 | 0.47 | 0.08 | 0.37 |
| Canada | AT | W | 1SW | 26 | 0.21 | 0.04 | 0.08 |
| Canada | BF | W | 1SW | 12 | 0.20 | 0.03 | 0.07 |
| Canada | BF | A | 1SW | 12 | 0.09 | 0.01 | 0.03 |
| Canada | BF | AE | 1SW | 12 | 0.43 | 0.07 | 0.16 |
| Canada | iBoF b | W | 1SW | 8 | 0.21 | 0.04 | 0.10 |
| Canada | Black R. | W | 1SW | 2 | 0.21 | 0.01 | 0.13 |
| Canada | Gaspereau R. | W | 1SW | 4 | 0.20 | 0.08 | 0.25 |
| Canada | Big Salmon R.c | W | 1SW | 2 | 0.23 | 0.13 | 1.59 |
| Canada | GSL | W | 1SW | 35 | 0.28 | 0.03 | 0.05 |
| Canada | AT | W | MSW | 1 | 0.07 | – | – |
| Canada | BF | W | MSW | 7 | 0.17 | 0.02 | 0.04 |
| Canada | BF | AE | MSW | 14 | 0.15 | 0.02 | 0.03 |
| Canada | GSL | W | MSW | 60 | 0.14 | 0.01 | 0.02 |
| Canada | iBoF b | W | MSW | 4 | 0.15 | 0.02 | 0.06 |
| Canada | Gaspereau R. | W | MSW | 1 | 0.15 | – | – |
| Canada | Big Salmon R.c | W | MSW | 3 | 0.15 | 0.03 | 0.11 |
| Ireland | WC | W | Smolt | 16 | 0.53 | 0.04 | 0.08 |
| Ireland | WC | W | 1SW | 26 | 0.19 | 0.02 | 0.04 |
| Ireland | WC | A | 1SW | 10 | 0.23 | 0.04 | 0.08 |
| Country | Region/River | Origin | Stage | n | [137Cs] (Bq kg−1 wet weight) | CL (95%) | |
|---|---|---|---|---|---|---|---|
| Mean | s.e. | ||||||
| Canada | GSL | W | Parr | 44a | 0.30 | – | – |
| Canada | AT | W | 0SW | 3 | 0.47 | 0.08 | 0.37 |
| Canada | AT | W | 1SW | 26 | 0.21 | 0.04 | 0.08 |
| Canada | BF | W | 1SW | 12 | 0.20 | 0.03 | 0.07 |
| Canada | BF | A | 1SW | 12 | 0.09 | 0.01 | 0.03 |
| Canada | BF | AE | 1SW | 12 | 0.43 | 0.07 | 0.16 |
| Canada | iBoF b | W | 1SW | 8 | 0.21 | 0.04 | 0.10 |
| Canada | Black R. | W | 1SW | 2 | 0.21 | 0.01 | 0.13 |
| Canada | Gaspereau R. | W | 1SW | 4 | 0.20 | 0.08 | 0.25 |
| Canada | Big Salmon R.c | W | 1SW | 2 | 0.23 | 0.13 | 1.59 |
| Canada | GSL | W | 1SW | 35 | 0.28 | 0.03 | 0.05 |
| Canada | AT | W | MSW | 1 | 0.07 | – | – |
| Canada | BF | W | MSW | 7 | 0.17 | 0.02 | 0.04 |
| Canada | BF | AE | MSW | 14 | 0.15 | 0.02 | 0.03 |
| Canada | GSL | W | MSW | 60 | 0.14 | 0.01 | 0.02 |
| Canada | iBoF b | W | MSW | 4 | 0.15 | 0.02 | 0.06 |
| Canada | Gaspereau R. | W | MSW | 1 | 0.15 | – | – |
| Canada | Big Salmon R.c | W | MSW | 3 | 0.15 | 0.03 | 0.11 |
| Ireland | WC | W | Smolt | 16 | 0.53 | 0.04 | 0.08 |
| Ireland | WC | W | 1SW | 26 | 0.19 | 0.02 | 0.04 |
| Ireland | WC | A | 1SW | 10 | 0.23 | 0.04 | 0.08 |
aSamples (n = 44) pooled and counted only once by gamma ray spectrometry.
biBoF pools Big Salmon, Black, and Gaspereau River samples.
cAccording to Amiro (2003), the Big Salmon River is considered iBoF, whereas the Black, Gaspereau, and Irish River populations are excluded based on a high proportion of virgin, MSW salmon.
Of the 11 samples recounted for tissue concentrations of 137Cs (test of analytical accuracy), there was no significant difference between original and re-count results (GLM ANOVA, p = 0.130, two outliers removed). Overall analysis of Atlantic salmon returnees sampled in 2002 and 2003 revealed a trend for exponential decline of 137Cs concentrations with increasing time away at sea (Figure3). There was no significant difference between tissue 137Cs concentrations and collection year (GLM ANOVA, p > 0.05), so 2002/2003 Canadian samples were pooled for subsequent analyses. There were significant differences between Canadian samples categorized by origin (wild, aquaculture escapee or aquaculture; GLM ANOVA, p < 0.05), but not between Irish wild and aquaculture samples. Analysis of wild returnee cohorts revealed that 0-sea-winter (0SW) salmon (n = 3) had significantly higher mean tissue 137Cs burdens than older cohorts (ANOVA post hoc Tukey, p < 0.001, n = 158, 16 outliers removed). Grilse (1SW, n = 84) had significantly higher mean 137Cs tissue burdens than 2-sea-winter (2SW) salmon (n = 61) (ANOVA post hoc Tukey, p = 0.015). There was no significant difference between 2SW and all older cohorts (Figure2; ANOVA post hoc Tukey, p > 0.535).
Figure 3.
Relationship between sea age and tissue concentrations of 137Cs (Bq kg−1 wet weight) for all wild Atlantic salmon samples collected in 2002 and 2003 (n = 157, 17 outliers removed). Outliers (crosses) removed for statistical analyses and plotting of this Figure are shown.
Tissue concentrations of 137Cs in wild Atlantic salmon returning to Canadian rivers in 2002 and 2003 ranged from 0.04 to 1.08 Bq kg−1 wet weight. Returning aquaculture escapees ranged from 0.07 to 0.82 Bq kg−1 wet weight. There was a significant difference between 1SW wild and escapee Canadian salmon (ANOVA, p = 0.008). Canadian aquaculture salmon (controls) had significantly lower tissue 137Cs concentrations than 1SW wild and escapee salmon (Table1; GLM ANOVA, p < 0.03). Standardized frequency distributions of tissue 137Cs concentrations for wild, returning Atlantic salmon sampled from the Ste Marguerite River, Quebec, in 1995/1996 ( Tucker et al., 1999) and all wild Canadian samples collected in 2002 and 2003 resulted in a significant difference (Kolmogorov–Smirnov two-sample test, p = 0.01).
Comparisons of 137Cs concentrations in wild 1SW salmon grouped according to natal river drainage within Atlantic Canada oceanographic regions (AT, BF, and GSL) revealed no significant differences (ANOVA, p = 0.56, n = 54, 18 outliers removed). Tissue concentrations of iBoF fish also showed no significant difference from fish from the three other regions (Figure4a; ANOVA, p = 0.61, n = 54, 18 outliers removed). Comparisons of wild MSW fish showed that BF fish (n = 7) had significantly higher tissue concentrations than GSL fish (n = 58, ANOVA, p = 0.029). When iBoF fish were compared as a separate region, only BF fish (n = 3) had significantly higher tissue concentrations than GSL fish (n = 58, ANOVA post hoc Tukey test, p = 0.024). There was no regional difference between the iBoF and the BF or GSL populations (Figure4b; ANOVA post hoc Tukey, p > 0.257). River-by-river comparisons yielded no significant differences among 137Cs residuals within the AT (Figure5a; GLM ANOVA, p = 0.214, n = 20, five outliers removed), BF (Figure5b; GLM ANOVA, p = 0.132, n = 13, one outlier removed), or GSL (Figure5c; GLM ANOVA, p = 0.414, n = 77, 11 outliers removed) regions.
Figure 4.
Wild Canadian Atlantic salmon tissue concentration of 137Cs pooled by oceanographic regions of natal river drainage (AT, BF, and GSL) compared with the iBoF populations for (a) 1SW (18 outliers removed) and (b) MSW salmon (two outliers removed). Outliers (crosses) removed for statistical analyses and plotting of these Figures are shown.
Figure 5.
(a) River-by-river comparison within (a) Canada's AT region (four outliers removed), (b) the BF region (one outlier removed), and (c) the GSL (nine outliers removed) of wild Atlantic salmon 137Cs residuals. Rivers analysed include the LaHave (LH), Northwest (NW), Western Arm (WA), Big Salmon (BS), Gaspereau (GP), Hammond (HM), Harry's (HR), Margaree (MA), Merrigomish Harbor (MH), Miramichi (MR), and Restigouche (RG).
Tissue concentrations of 137Cs in wild salmon returning to Irish rivers ranged between 0.09 and 0.52 Bq kg−1 wet weight. Aquaculture controls ranged from 0.09 to 0.41 Bq kg−1 wet weight (Table1). There was no significant difference between 1SW wild (n = 26) and aquaculture (n = 10) Irish Atlantic salmon tissue 137Cs concentrations (GLM ANOVA, p = 0.108, one outlier removed). Irish adult Atlantic salmon returning to the Corrib River had significantly higher Cs residuals (mean [137Cs] = 0.24) than home-bound migrants caught in the inshore Achill fishery (mean [137Cs] = 0.14, GLM ANOVA, p < 0.001, n = 25, five outliers removed). There was no significant difference in tissue 137Cs concentrations between wild Canadian and Irish salmon (Figure6a; GLM ANOVA, p = 0.277, three outliers excluded). However, comparison of Canadian 2002 (n = 12) and Irish 2003 (n = 10) aquaculture salmon showed a significant difference (Figure6b; GLM ANOVA, p = 0.001).
Figure 6.
(a) Frequency distribution of tissue 137Cs concentrations (Bq kg−1 tissue wet weight) for (a) wild Atlantic salmon from Canada (2002 and 2003; open bars, n = 144, mean = 0.20) and Ireland (2003; solid bars, n = 30, mean = 0.19) and (b) aquaculture Atlantic salmon from Canada (2002; open bars, n = 12, mean = 0.09) and Ireland (2003; solid bars, n = 10, mean = 0.23).
Tissue 137Cs concentrations of Irish smolts ranged from 0.34 to 0.91 Bq kg−1 wet weight (Table1). Their mean tissue concentration of 0.53 Bq kg−1 wet weight was significantly higher than Canadian parr (0.30 Bq kg−1 wet weight, one-sample Z-test, p < 0.001). Tissue concentrations of Canadian grilse returning after only spending summer months at sea (0SW) ranged from 0.30 to 0.57 Bq kg−1 wet weight and were also significantly higher than parr (one-sample Z-test, p = 0.027), but not significantly different from Irish smolts (GLM ANOVA, p = 0.539, one outlier removed).
Discussion
Similar ranges of tissue 137Cs burdens in all sea age cohorts of wild, Canadian, and Irish salmon indicated that similar feeding grounds were utilized during marine migration. The significant relationship between duration at sea and declining tissue 137Cs concentrations indicated that different sea age cohorts (1SW, MSW) of salmon may utilize different marine feeding grounds, and therefore follow different migratory routes within an overall migration pattern. High mean tissue 137Cs concentrations in some Canadian 1SW salmon suggested that they were distributed east of the Faroes before initiating homeward spawning migration, whereas the lower mean tissue concentrations of MSW salmon suggested a more northern distribution in the Labrador and Irminger Seas, corresponding to regions of low surface water 137Cs levels. Caesium clearance rates in tissue decrease with increasing animal size (Rowan and Rasmussen, 1994), so concentrations in larger fish would be expected to be higher, but this is not shown by our results. Atlantic salmon actively select colder sea surface temperatures with increasing body size (Jensen, 1967; Reddin and Shearer, 1987; Jakupsstovu, 1988; Holm et al., 2003), and bio-accumulation of 137Cs is positively correlated to water temperature (Rowan and Rasmussen, 1994), so cooler SSTs could be a contributing factor to declining 137Cs concentrations with increasing sea age.
Overall 137Cs concentrations in Canadian aquaculture escapees (0.28 ± 0.22 Bq kg−1 wet weight) were similar to those in wild fish (0.20 ± 0.14 Bq kg−1 wet weight), although 1SW escapees may display an alternative behaviour to wild migration patterns. Between 21% and 40% of salmon caught around the Faroes during the 1990s were aquaculture escapees ( Hansen et al., 1999; Hansen and Jacobsen, 2003), suggesting that escapees may revert to wild migration patterns. This is likely the case for escapees maturing as MSW salmon, because they would have a longer period to revert to wild migration patterns and consequently demonstrate 137Cs levels as their wild counterparts (Table1). In contrast, significantly higher concentrations in 1SW escapees may suggest feeding within local estuaries (Jonssen et al., 1993; Carr et al., 1997). Even MSW escapees, depending on the timing of their escape, may move to adjacent river mouths to await spawning runs (Gausen and Moen, 1991), and subsequently only increase their already low 137Cs burdens slightly to levels resembling wild returnees, because most of their body mass would have been derived from farm feed. Timing of escapes for aquaculture fish, relative to specific life stages (i.e. post-smolt, 1SW, or MSW), may determine the ultimate path of marine migration.
Bio-accumulation of 137Cs in fish is greater in freshwater than in saltwater (Rowan and Rasmussen, 1994). This relationship was demonstrated by the increasing mean 137Cs concentrations from parr (0.30 Bq kg−1 wet weight) to smolts (0.53 Bq kg−1 wet weight) and the significant decline of tissue concentrations in returning 1SW and MSW salmon. High tissue concentrations of 137Cs (0.47 Bq kg–1 wet weight) in 0SW post-smolts returning to spawn can be explained by their relatively short time spent at sea. The biological half-life of 137Cs in large fish is hundreds of days to several years (Rowan and Rasmussen, 1994), so returning 0SW salmon should still display 137Cs tissue concentrations reflecting freshwater residence, rather than recently bio-accumulated 137Cs from marine sources.
Comparing the frequency distribution of tissue 137Cs concentrations of wild Canadian returnees (1SW and MSW) from 1995/1996 ( Tucker et al., 1999) with 1SW and MSW herein (2002 and 2003) revealed a dramatic decrease in tissue concentration range between these time periods. Although frequency distributions of both data sets were significantly different, a similar Poisson trend with a few extreme outliers was observed. In both data sets, extreme outliers indicated trans-Atlantic migrants that had fed for extended periods of time in regions of high 137Cs contamination. The overall decline of the North Atlantic 137Cs east–west gradient ( Povinec et al., 2003) and redistribution of radioisotopes via the North Atlantic Gyre towards polar regions and East Greenland (Figure1; Dahlgaard et al., 2004) was reflected in the changes observed between the salmon tissue 137Cs concentrations found by Tucker et al. (1999) and our results. Overall, our results agree with those of Tucker et al. (1999) and with their 1995/1996 observations inferring pan-Atlantic migration of salmon returning to the Ste Marguerite River in Quebec.
There were no significant regional differences of mean 137Cs burdens in tissue from 1SW Canadian salmon originating from three Atlantic Canadian oceanographic regions based on natal river drainage, which indicated that all migrants had followed similar migration routes and/or fed in similar regions before homeward migration. Endangered iBoF river Atlantic salmon consist of 35 or more rivers from the Black River near the Saint John estuary in New Brunswick, around the inner bay to the Cornwallis River in Nova Scotia (Anon., 2001; Amiro, 2003). Genotype data group the Gaspereau, Black, and Irish Rivers with iBoF stock ( Verspoor et al., 2002), but Amiro (2003) states that these rivers do not belong to the iBoF complex based on a high proportion of MSW virgin fish. Inner Bay of Fundy salmon have been hypothesized to remain within the BF and Gulf of Maine during their marine migration (Amiro, 2003), but our results revealed no significant differences of mean 137Cs in tissue from 1SW or MSW salmon originating from the iBoF and the three other Canadian oceanographic regions. Atlantic salmon feeding and migrating solely within the BF and Gulf of Maine would be expected to bio-accumulate maximum tissue concentrations of 0.23 Bq kg−1 wet weight based on a bio-accumulation factor of 130 (Rowan and Rasmussen, 1994) relative to ambient ocean levels (1.7–1.8 mBq l−1; Table2). The range of tissue 137Cs concentrations measured for iBoF populations (0.10–0.43 Bq kg−1 wet weight) exceeded this maximum expectation, even after excluding Gaspereau and Black River samples (0.10–0.35 Bq kg−1 wet weight). Therefore, our results do not support the hypothesis that IBoF salmon remain in the BF and Gulf of Maine during their marine phase. Our data suggests that iBoF salmon exit the BF to follow migration routes and/or feed in regions similar to most other Canadian Atlantic salmon. Further evidence is provided by the recent work of Lacroix and Knox (2005), who demonstrated that most post-smolts exit the BF rapidly unless they are caught in the large mid-bay gyre.
Table 2.
Minimum–maximum, mean levels of 137Cs in surface water (mBq 137Cs l−1), wild Atlantic salmon tissue (Bq kg−1 wet weight) based on a bio-accumulation factor of 130 relative to seawater levels ( Tucker et al., 1999), and resident piscivorous fish tissue (Bq kg−1 wet weight) in the North Atlantic from 2000 to 2003.
| Region | Surface water [Cs137] (mBq l−1) | Estimated salmon tissue [Cs137] (Bq kg−1 wet weight) | Resident fish tissue [Cs137] (Bq kg−1 wet weight) | Source |
|---|---|---|---|---|
| Bay of Fundy | 1.7–1.8, 1.7 | 0.22–0.23, 0.22 | – | J. Smith (pers. comm.) |
| Grand Banks, Labrador Sea | 1.8 | 0.23 | – | J. Smith (pers. comm.) |
| West Greenland | 3.1–3.2, 3.2 | 0.40–0.42, 0.42 | – | Dahlgaard et al. (2004) |
| Irminger Sea | 1.8 | 0.23 | – | Dahlgaard et al. (2004) |
| East Greenland | 4.2–5.5, 4.9 | 0.55–0.72, 0.64 | – | Dahlgaard et al. (2004) |
| Denmark Strait, Iceland Sea | 2.4–3.5, 2.8 | 0.31–0.46, 0.36 | – | Dahlgaard et al. (2004) |
| Faroe Islands | 1.8 | 0.23 | 0.1–0.2, 0.2 a | Dahlgaard et al. (2004); NRPA (2004) |
| Norwegian Sea, Barents Sea | 1.8–5.7, 2.7 | 0.23–0.74, 0.35 | 0.1–1.2, 0.3 b | NRPA (2003, 2004); I. Sværen (pers. comm.) |
| Greenland Sea | 4.4–5.4, 4.9 | 0.57–0.70, 0.64 | – | Dahlgaard et al., 2004 |
| West coast of Ireland | 2.0–3.0, 2.5 | 0.26–0.39, 0.33 | 0.1–0.8, 0.3 b | RPII (2002, 2003) |
| Irish Sea | 6.0–31, 14 | 0.78–4.0, 1.82 | 0.1–10, 0.5b | RPII (2002, 2003); NRPA (2004) |
| North Sea | 2.1–10, 4.7 | 0.27–1.3, 0.61 | 0.1–2.1, 0.5 b | NRPA (2003, 2004) |
| Baltic Sea | 25–70, 49 | 3.3–9.1, 6.37 | 1.2–32, 9.3 c | HELCOM (2003); NRPA (2003, 2004); STUK (2003, 2004) |
| Region | Surface water [Cs137] (mBq l−1) | Estimated salmon tissue [Cs137] (Bq kg−1 wet weight) | Resident fish tissue [Cs137] (Bq kg−1 wet weight) | Source |
|---|---|---|---|---|
| Bay of Fundy | 1.7–1.8, 1.7 | 0.22–0.23, 0.22 | – | J. Smith (pers. comm.) |
| Grand Banks, Labrador Sea | 1.8 | 0.23 | – | J. Smith (pers. comm.) |
| West Greenland | 3.1–3.2, 3.2 | 0.40–0.42, 0.42 | – | Dahlgaard et al. (2004) |
| Irminger Sea | 1.8 | 0.23 | – | Dahlgaard et al. (2004) |
| East Greenland | 4.2–5.5, 4.9 | 0.55–0.72, 0.64 | – | Dahlgaard et al. (2004) |
| Denmark Strait, Iceland Sea | 2.4–3.5, 2.8 | 0.31–0.46, 0.36 | – | Dahlgaard et al. (2004) |
| Faroe Islands | 1.8 | 0.23 | 0.1–0.2, 0.2 a | Dahlgaard et al. (2004); NRPA (2004) |
| Norwegian Sea, Barents Sea | 1.8–5.7, 2.7 | 0.23–0.74, 0.35 | 0.1–1.2, 0.3 b | NRPA (2003, 2004); I. Sværen (pers. comm.) |
| Greenland Sea | 4.4–5.4, 4.9 | 0.57–0.70, 0.64 | – | Dahlgaard et al., 2004 |
| West coast of Ireland | 2.0–3.0, 2.5 | 0.26–0.39, 0.33 | 0.1–0.8, 0.3 b | RPII (2002, 2003) |
| Irish Sea | 6.0–31, 14 | 0.78–4.0, 1.82 | 0.1–10, 0.5b | RPII (2002, 2003); NRPA (2004) |
| North Sea | 2.1–10, 4.7 | 0.27–1.3, 0.61 | 0.1–2.1, 0.5 b | NRPA (2003, 2004) |
| Baltic Sea | 25–70, 49 | 3.3–9.1, 6.37 | 1.2–32, 9.3 c | HELCOM (2003); NRPA (2003, 2004); STUK (2003, 2004) |
aCod and haddock.
bFish tissue concentrations taken from observed mean levels in Atlantic salmon, cod, haddock, halibut, herring, mackerel, monkfish, plaice, pollack, sole, and whiting sampled within respective regions.
cCod, herring, flounder, and plaice.
Table 2.
Minimum–maximum, mean levels of 137Cs in surface water (mBq 137Cs l−1), wild Atlantic salmon tissue (Bq kg−1 wet weight) based on a bio-accumulation factor of 130 relative to seawater levels ( Tucker et al., 1999), and resident piscivorous fish tissue (Bq kg−1 wet weight) in the North Atlantic from 2000 to 2003.
| Region | Surface water [Cs137] (mBq l−1) | Estimated salmon tissue [Cs137] (Bq kg−1 wet weight) | Resident fish tissue [Cs137] (Bq kg−1 wet weight) | Source |
|---|---|---|---|---|
| Bay of Fundy | 1.7–1.8, 1.7 | 0.22–0.23, 0.22 | – | J. Smith (pers. comm.) |
| Grand Banks, Labrador Sea | 1.8 | 0.23 | – | J. Smith (pers. comm.) |
| West Greenland | 3.1–3.2, 3.2 | 0.40–0.42, 0.42 | – | Dahlgaard et al. (2004) |
| Irminger Sea | 1.8 | 0.23 | – | Dahlgaard et al. (2004) |
| East Greenland | 4.2–5.5, 4.9 | 0.55–0.72, 0.64 | – | Dahlgaard et al. (2004) |
| Denmark Strait, Iceland Sea | 2.4–3.5, 2.8 | 0.31–0.46, 0.36 | – | Dahlgaard et al. (2004) |
| Faroe Islands | 1.8 | 0.23 | 0.1–0.2, 0.2 a | Dahlgaard et al. (2004); NRPA (2004) |
| Norwegian Sea, Barents Sea | 1.8–5.7, 2.7 | 0.23–0.74, 0.35 | 0.1–1.2, 0.3 b | NRPA (2003, 2004); I. Sværen (pers. comm.) |
| Greenland Sea | 4.4–5.4, 4.9 | 0.57–0.70, 0.64 | – | Dahlgaard et al., 2004 |
| West coast of Ireland | 2.0–3.0, 2.5 | 0.26–0.39, 0.33 | 0.1–0.8, 0.3 b | RPII (2002, 2003) |
| Irish Sea | 6.0–31, 14 | 0.78–4.0, 1.82 | 0.1–10, 0.5b | RPII (2002, 2003); NRPA (2004) |
| North Sea | 2.1–10, 4.7 | 0.27–1.3, 0.61 | 0.1–2.1, 0.5 b | NRPA (2003, 2004) |
| Baltic Sea | 25–70, 49 | 3.3–9.1, 6.37 | 1.2–32, 9.3 c | HELCOM (2003); NRPA (2003, 2004); STUK (2003, 2004) |
| Region | Surface water [Cs137] (mBq l−1) | Estimated salmon tissue [Cs137] (Bq kg−1 wet weight) | Resident fish tissue [Cs137] (Bq kg−1 wet weight) | Source |
|---|---|---|---|---|
| Bay of Fundy | 1.7–1.8, 1.7 | 0.22–0.23, 0.22 | – | J. Smith (pers. comm.) |
| Grand Banks, Labrador Sea | 1.8 | 0.23 | – | J. Smith (pers. comm.) |
| West Greenland | 3.1–3.2, 3.2 | 0.40–0.42, 0.42 | – | Dahlgaard et al. (2004) |
| Irminger Sea | 1.8 | 0.23 | – | Dahlgaard et al. (2004) |
| East Greenland | 4.2–5.5, 4.9 | 0.55–0.72, 0.64 | – | Dahlgaard et al. (2004) |
| Denmark Strait, Iceland Sea | 2.4–3.5, 2.8 | 0.31–0.46, 0.36 | – | Dahlgaard et al. (2004) |
| Faroe Islands | 1.8 | 0.23 | 0.1–0.2, 0.2 a | Dahlgaard et al. (2004); NRPA (2004) |
| Norwegian Sea, Barents Sea | 1.8–5.7, 2.7 | 0.23–0.74, 0.35 | 0.1–1.2, 0.3 b | NRPA (2003, 2004); I. Sværen (pers. comm.) |
| Greenland Sea | 4.4–5.4, 4.9 | 0.57–0.70, 0.64 | – | Dahlgaard et al., 2004 |
| West coast of Ireland | 2.0–3.0, 2.5 | 0.26–0.39, 0.33 | 0.1–0.8, 0.3 b | RPII (2002, 2003) |
| Irish Sea | 6.0–31, 14 | 0.78–4.0, 1.82 | 0.1–10, 0.5b | RPII (2002, 2003); NRPA (2004) |
| North Sea | 2.1–10, 4.7 | 0.27–1.3, 0.61 | 0.1–2.1, 0.5 b | NRPA (2003, 2004) |
| Baltic Sea | 25–70, 49 | 3.3–9.1, 6.37 | 1.2–32, 9.3 c | HELCOM (2003); NRPA (2003, 2004); STUK (2003, 2004) |
aCod and haddock.
bFish tissue concentrations taken from observed mean levels in Atlantic salmon, cod, haddock, halibut, herring, mackerel, monkfish, plaice, pollack, sole, and whiting sampled within respective regions.
cCod, herring, flounder, and plaice.
The distribution of Atlantic salmon at sea based on 137Cs concentrations in tissue suggests that salmon feeding west of the Faroe Islands (8°W), in the Labrador and Irminger Seas, or east of the Grand Banks would be expected to bio-accumulate levels <0.23 Bq kg−1 wet weight. Atlantic salmon feeding off West and East Greenland would be expected to bio-accumulate higher levels, ranging from 0.40 to 0.72 Bq kg−1 wet weight (Table2). West Greenland feeding grounds are only occupied by predisposed MSW salmon in summer and autumn (Jensen, 1980a, b; Jensen and Lear, 1980; Swain, 1980; Reddin, 1986). Migrants feeding off East Greenland are hypothesized to continue moving within the Irminger/East Greenland Current into the Labrador Sea, where 137Cs levels are considerably lower, and therefore dilute high concentrations bio-accumulated in the East Greenland Current. Atlantic salmon feeding east of the Faroe Islands (excluding the Baltic Sea) would be expected to bio-accumulate levels up to 4.0 Bq kg−1 wet weight. The highest 137Cs burdens in the eastern North Atlantic would be for salmon feeding in the North Sea or close to the Irish Sea, where maximum estimates range from 1.3 to 4.0 Bq kg−1 wet weight, respectively. These expected levels of 137Cs in salmon are consistent with results from resident piscivorous fish in the northeast Atlantic (up to 10.0 Bq kg−1 wet weight; Table2). On the basis of Irish aquaculture controls and expected and measured bio-accumulation levels, a conservative estimate for 137Cs concentrations in tissue of Atlantic salmon feeding east of the Faroe Islands (8°W) was ≥0.3 Bq kg−1 wet weight.
Of 73 wild Canadian grilse sampled, 24.7% had tissue 137Cs concentrations of ≥0.3 Bq kg−1 wet weight. Grilse from all Atlantic Canadian oceanographic regions based on natal watersheds were represented (AT, BF, and GSL). A tissue 137Cs concentration of 1.08 Bq kg−1 wet weight from a 1SW salmon caught in the Northwest River, Newfoundland, provided sound evidence of trans-Atlantic migration. Of 68 MSW salmon sampled, just 2.9% were inferred to have fed east of the Faroes before returning as maiden spawners. Of 141 Canadian wild Atlantic salmon sampled, 14.2% (24.7% 1SW, 2.9% MSW) were inferred to have spent a significant period feeding in the northeast Atlantic before homeward migration. Whether or not 1SW aquaculture escapees follow similar migration patterns to wild 1SW salmon, 50% of 1SW escapees sampled had tissue concentrations >0.3 Bq kg−1 wet weight. No escapees returning as MSW salmon had tissue concentrations >0.3 Bq kg−1 wet weight, suggestive of a similar migration to wild MSW salmon.
Our estimate of 14.2% for wild Canadian salmon undertaking a migration beyond the Faroes may be an underestimate when the decline of the trans-Atlantic 137Cs gradient ( Povinec et al., 2003) and range of northeast Atlantic resident piscivorous fish 137Cs burdens (0.10–10.0 Bq kg−1; Table2) are considered. Even migrants with tissue 137Cs concentrations ranging from 0.10 to 0.30 Bq kg−1 could have fed and/or migrated within northeast Atlantic waters. Tucker et al. (1999) estimated 43% of returning migrants to the Ste Marguerite River, Quebec, had fed beyond Iceland, and that 59% of 1SW and 36% of 2SW salmon fed in the northeast Atlantic. In a study based on discriminant scale analysis of 247 2SW Atlantic salmon sampled from the Faroese fishery during 1981 and 1982, Reddin (1987) concluded that 1.2% were of Canadian origin, similar to our estimate of 2.4% of MSW salmon hypothesized to have undertaken trans-Atlantic migration. Unfortunately, 1SW salmon, which are abundant off the Faroes during the winter fishery (Jakupsstovu, 1988), were not examined in Reddin's study.
Of 26 wild Irish grilse sampled during 2003, only 11.5% had tissue concentrations of 137Cs ≥0.3 Bq kg−1 wet weight. No Irish 2SW salmon sampled had tissue levels >0.3 Bq kg−1 wet weight. Recent studies of wild Norwegian Atlantic salmon returning during 2002 found mean 137Cs tissue concentrations of 0.2 Bq kg−1 wet weight (NRPA, 2004). These results suggest that most Irish and other European populations of Atlantic salmon follow the North Atlantic Gyre and spend a significant period feeding in the western Atlantic before migrating home again.
Our samples of aquaculture Atlantic salmon were intended as non-migratory controls, but this role may have been compromised because of the geographical origin of their feed. Feed for Canadian salmon aquaculture is obtained globally, the majority from Chilean and Peruvian anchovy fisheries (Saulier, pers. comm.). The mean tissue concentration of 0.09 Bq kg−1 wet weight we found for Canadian aquaculture salmon may not be the best indicator of average western North Atlantic tissue 137Cs concentrations for Atlantic salmon. On the other hand, Irish aquaculture salmon are fed 60–70% marine products, the majority consisting of herring, sandlance, capelin, and Norwegian pout from the North Sea, Norwegian Sea, and Barents Sea (Grøttheim, 2002), and probably were a better indicator of average eastern North Atlantic tissue 137Cs concentrations for Atlantic salmon. Decreasing levels of Cs137 in the North Atlantic ( Povinec et al., 2003) has resulted in indistinct regional levels (Figure1), making inferences about exact feeding grounds difficult. Many of our sample concentrations were low, and the detection limit for HPGe gamma ray spectrometers is only 0.1 Bq kg−1 wet weight (NRPA, 2004), so inferences based on individual 137Cs concentrations are not considered as powerful as inferences based on observed trends.
The currently accepted migration model has been contradicted by past tag returns and more recent research ( Shelton et al., 1997; Tucker et al., 1999; Hansen and Jacobsen, 2000; Holm et al., 2003). Two tagged smolts from Canada, one from the Matamec River in the GSL (Gibson and Côté, 1982), the other from the Saint John River in the BF ( Reddin et al., 1984b), were recaptured as MSW salmon off Norway in 1979. Two tagged smolts from the Liscomb River on Nova Scotia's AT were subsequently recaptured in the Faroese fishery (Reddin, 1987), as were two from the Penobscot River, USA (Baum, 1997). Conversely, salmon tagged as smolts in Scotland were recaptured along the Labrador coast, and adults tagged off the Grand Banks were recaptured in Ireland and Scotland (Redden et al., 1984b). A tagging study of ocean-migrating Atlantic salmon north of the Faroes resulted in four recaptures from GSL rivers (Miramichi, n = 3; Kouchibouguac, n = 1; Hansen and Jacobsen, 2000). Recent research trawls in the northeast Atlantic caught post-smolts following the North Atlantic Gyre northeast into the Norwegian Sea and Greenland Sea ( Shelton et al., 1997; Holm et al., 2000, 2003). This suggests that trans-Atlantic migration is not only undertaken by European stocks, but by North American stocks as well.
Although other researchers have suggested that Atlantic salmon undertake trans-oceanic migrations using the North Atlantic Gyre current system (Jensen, 1967; Stasko et al., 1973; Reddin et al., 1984b; Tucker et al., 1999; Friedland et al., 2001), ours is the first study that incorporated migrants from both sides of the Atlantic and used similar temporal scales to infer an overall migration model based on empirical data. Changes in the dumping regime of nuclear wastes from Sellafield into the Irish Sea (Kenny, 2003) has created other environmental tag options. 99Tc has been increasing since the 1990s ( Dahlgaard et al., 2004), and may reform another east–west gradient of radionuclides. To prove that Atlantic salmon utilize the North Atlantic Gyre for marine migration, further research on exact routing of salmon is required.
Ocean temperature selection based on natal latitudinal origin has been demonstrated in Atlantic ( Jacobsen et al., 2001) and Pacific salmon (Brannon, 1984), and if migration within the North Atlantic Gyre is the correct Atlantic salmon migration model, serious consideration should be given to the need to implement an international management scheme based on latitudinal and geographic origin of Atlantic salmon populations relative to the North Atlantic Gyre (Bisbal and McConnaha, 1998; Potter et al., 2004). Certain Atlantic salmon populations, such as those from southern latitudes, need protection from fishing on oceanic feeding grounds when migration timing and SSTs suggest their presence. National management plans aiming to control salmon catches within EEZs may provide a measure of success, but to increase marine survival, managing human activities in international waters is crucial. We propose that anadromous Atlantic salmon (other than Baltic Sea stocks; Jutila et al., 2003) be classified as a trans-Atlantic straddling stock, making international cooperation essential to ensuring their survival at sea.
Acknowledgements
We thank the Annapolis Valley, Listuguj (J. Casey, T. Metallic and V. Metallic), Membertou, Pictou Landing, and Red Bank (G. Hare) First Nations, the Atlantic Salmon Federation, the Canadian Department of Fisheries and Oceans, the Galway Fishery, Margaree Salmon Association, Nova Scotia Department of Inland Fisheries, NovaScotiaFishing.com, NS Salmon Association, Marine Institute of Ireland, Parks Canada, Pictou County Rivers Association, and the International St Croix River Commission for assistance in sample collection. We thank G. Morinville, McGill University, for the laboratory gamma ray spectrometry analysis, and the Atomic Energy of Canada Ltd, G. Kanisch, I. Sværen and M. Holm, I. Osvath, J. Smith and R. Nelson, J. Gwynn, M. Kelly, and S. Nielsen for providing current 137Cs and Atlantic salmon data. ADS and JMR were supported by fellowships from Acadia University, the Atlantic Salmon Federation, the Canadian Wildlife Federation, Mountain Equipment Co-op (JMR), the Nova Scotia Salmon Association, PADI Foundation (JMR), and PADI Project Aware (ADS). Research was funded by grants from Acadia University (MJD), the Canadian National Sportsmen's Shows (ADS and MJD), and the New Brunswick Wildlife Trust Fund (ADS and MJD).
References
Anon
Canadian species at risk
,
2001
Ottawa
Canadian Wildlife Service, Environment Canada
pg.
38
.
Population status of inner Bay of Fundy Atlantic salmon (Salmo salar), to 1999
,
2003
pg.
2488
Canadian Technical Report of Fisheries and Aquatic Sciences
.
Maine Atlantic Salmon: a National Treasure
,
1997
Hermon, Maine
Atlantic Salmon Unlimited
pg.
224
, .
Consideration of ocean conditions in the management of salmon
,
Canadian Journal of Fisheries and Aquatic Sciences
,
1998
, vol.
55
(pg.
2178
-
2186
)
, , , , , .
Regional variation of caesium-137 in minke whales Balaenoptera acutorostrata from West Greenland, the Northeast Atlantic and the North Sea
,
Polar Biology
,
2002
, vol.
25
(pg.
907
-
913
)
. , et al.
Influence of stock origin on salmon migratory behavior
,
Mechanisms of Migrations in Fishes
,
1984
New York
Plenum
(pg.
103
-
111
)
.
An evaluation of possible causes of the decline in pre-fishery abundance of North American Atlantic salmon
,
2001
pg.
2358
Canadian Technical Report of Fisheries and Aquatic Sciences
, , , .
The occurrence and spawning of cultured Atlantic salmon (Salmo salar) in a Canadian river
,
ICES Journal of Marine Science
,
1997
, vol.
54
(pg.
1064
-
1073
)
, .
Age determination methods for fishes studied by the groundfish program at the Pacific Biological Station
,
Canadian Special Publication of Fisheries and Aquatic Sciences
,
1982
, vol.
60
pg.
102
, .
Distribution and abundance of Atlantic salmon at west Greenland
,
1980
, vol.
176
(pg.
22
-
35
)
Rapports et Procès-verbaux des Réunions Conseil International pour l'Exploration de la Mer
, , , .
Levels and trends of radioactive contaminants in the Greenland environment
,
Science of the Total Environment
,
2004
, vol.
331
(pg.
53
-
67
)
.
Ocean climate influences on critical Atlantic salmon (Salmo salar) life history events
,
Canadian Journal of Fisheries and Aquatic Sciences
,
1998
, vol.
55
Suppl. 1
(pg.
119
-
130
)
, , , , .
Forecasts of Atlantic salmon transoceanic migration: climate change scenarios and anadromy in the North Atlantic
,
2001
Fisheries in a changing climate. American Fisheries Society Annual Meeting
August 20–21
Phoenix, Arizona
, .
Large scale escapes of farmed salmon (Salmo salar) into Norwegian rivers threaten natural populations
,
Canadian Journal of Fisheries and Aquatic Sciences
,
1991
, vol.
48
(pg.
426
-
428
)
, .
Production de saumonneaux et recaptures de saumons adultes etiquettes a la riviere Matamec, Cote-Nord, Golfe du Saint-Laurent, Quebec
,
Naturaliste Canadienne
,
1982
, vol.
109
(pg.
13
-
25
)
.
Radioactive contamination in the Norwegian marine environment
,
2002
, .
North Atlantic-Nordic Sea exchanges
,
Progress in Oceanography
,
2000
, vol.
45
(pg.
109
-
208
)
, . .
Distribution and migration of Atlantic salmon, Salmo salar L., in the sea
,
The Ocean Life of Atlantic Salmon, Environmental and Biological Factors influencing Survival
,
2000
Oxford, UK
Fishing News Books
(pg.
77
-
87
)
, .
Origin and migration of wild and escaped farmed Atlantic salmon, Salmo salar L., in oceanic areas north of the Faroe Islands
,
ICES Journal of Marine Science
,
2003
, vol.
60
(pg.
110
-
119
)
, , .
The incidence of escaped farmed Atlantic salmon, Salmo salar L., in the Faroese fishery and estimates of catches of wild salmon
,
ICES Journal of Marine Science
,
1999
, vol.
56
(pg.
200
-
206
)
, , .
Oceanic migration of homing Atlantic salmon, Salmo salar.
,
Animal Behaviour
,
1993
, vol.
45
(pg.
927
-
941
)
, .
The food of Atlantic salmon, Salmo salar L. caught by long-line in northern Norwegian waters
,
Journal of Fisheries Biology
,
1985
, vol.
26
(pg.
553
-
562
)
, .
The marine phase of the Atlantic salmon (Salmo salar) life cycle, comparisons to Pacific salmon
,
Canadian Journal of Fisheries and Aquatic Science
,
1998
, vol.
55
Supp.1
(pg.
104
-
118
)
HELCOM
Concentrations of the artificial radionuclide caesium–137 in Baltic Sea fish and surface waters
,
2004
, , .
Spatial and temporal distribution of post-smolts of Atlantic salmon (Salmo salar L.) in the Norwegian Sea and adjacent areas
,
ICES Journal of Marine Science
,
2000
, vol.
57
(pg.
955
-
964
)
, , , , , . .
Migration and distribution of Atlantic salmon post-smolts in the North Sea and North East Atlantic
,
Salmon at the Edge, 7–23
,
2003
Oxford, UK
Blackwell Science
pg.
307
, .
Fishes of the Great Lakes region
,
1964
Ann Arbor, Michigan
University of Michigan Press
pg.
213
, , , .
A synthesis of the joint meeting. Causes of marine mortality of salmon in the North Pacific and North Atlantic Oceans and in the Baltic Sea
,
2002
March 14–15, 2002
Vancouver, BC
North Pacific Anadromous Fish Conference Technical Report 4
, , , .
Seasonal differences in the origin of Atlantic salmon (Salmo salar L.) in the Norwegian Sea based on estimates from age structures and tag recaptures
,
Fisheries Research
,
2001
, vol.
52
(pg.
1659
-
1677
)
, , , , .
Near surface circulation in the northern North Altantic as inferred by Lagrangian drifters: variability from the mesoscale to interannual
,
Journal of Geophysical Research
,
2003
, vol.
108
C8
(pg.
3251
-
3302
)
. , .
Exploitation and migration of salmon in Faroese waters
,
Atlantic Salmon: Planning for the Future
,
1988
London
Croom Helm
(pg.
458
-
482
)
.
Atlantic salmon caught in the Irminger Sea
,
Journal of the Fisheries Research Board of Canada
,
1967
, vol.
24
(pg.
2639
-
2640
)
.
Recaptures of salmon at West Greenland tagged as smolts outside Greenland waters
,
1980
, vol.
176
(pg.
114
-
121
)
Rapports et Procès-Verbaux des Réunions du Conseil International pour l'Exploration de la Mer
.
Recaptures from international tagging experiments at West Greenland
,
1980
, vol.
176
(pg.
122
-
135
)
Rapports et Procès-Verbaux des Réunions du Conseil International pour l'Exploration de la Mer
, .
Atlantic salmon caught in the Irminger Sea and at East Greenland
,
Journal of Northwest Atlantic Fisheries Science
,
1980
, vol.
1
(pg.
55
-
64
)
, , .
Migratory behaviour and growth of hatchery-reared post-smolt Atlantic salmon Salmo salar
,
Journal of Fish Biology
,
1993
, vol.
42
(pg.
435
-
443
)
, .
Factors affecting marine production of Atlantic salmon (Salmo salar)
,
Canadian Journal of Fisheries and Aquatic Sciences
,
2004
, vol.
61
(pg.
2369
-
2383
)
, , , .
Differences in sea migration between wild and reared Atlantic salmon (Salmo salar L.) in the Baltic Sea
,
Fisheries Research
,
2003
, vol.
60
(pg.
333
-
343
)
.
Fearing Sellafield
,
2003
Dublin
Gill and Macmillan Ltd
.
The North Atlantic Current
,
Journal of Geophysical Research
,
1986
, vol.
91
(pg.
5061
-
5074
)
, .
Distribution of Atlantic salmon (Salmo salar) postsmolts of different origins in the Bay of Fundy and Gulf of Maine and evaluation of factors affecting migration, growth, and survival
,
Canadian Journal of Fisheries and Aquatic Sciences
,
2005
, vol.
62
(pg.
1363
-
1376
)
, , .
Survival and behaviour of post-smolt Atlantic salmon in coastal habitat with extreme tides
,
Journal of Fish Biology
,
2005
, vol.
66
(pg.
485
-
498
)
, , .
Mid-depth recirculation observed in the interior Labrador and Irminger seas by direct velocity measurements
,
Nature
,
2000
, vol.
407
(pg.
66
-
69
)
Marine Institute
Marine chemistry. Radionuclides
,
Ireland's Marine and Coastal Areas and Adjacent Seas: an Environmental Assessment
,
1999
Ireland
Marine Institute
(pg.
112
-
114
)
.
The marine migration of tagged Atlantic salmon (Salmo salar L.) of USA origin
,
1984
ICES Document CM 1984/M: 27
NRPA
Radioactivity in the marine environment 2000–2001. Results from the Norwegian National Monitoring Program (RAME)
,
2003
Østerås
Norwegian Radiation Protection Authority
Strålevern Rapport 2003: 8
NRPA
Radioactivity in the marine environment 2002
,
2004
Østerås
Norwegian Radiation Protection Authority
Results from the Norwegian National Monitoring Program (RAME) Strålevern Rapport 2004: 10
, , , , , , , et al.
Estimating and forecasting pre-fishery abundance of Atlantic salmon in the Northeast Atlantic for the management of mixed-stock fisheries
,
ICES Journal of Marine Science
,
2004
, vol.
61
(pg.
1359
-
1369
)
, , , , .
Temporal and spatial trends in the distribution of 137Cs in surface waters of northern European Seas – a record of 40 years of investigations
,
Deep-Sea Research II
,
2003
, vol.
50
(pg.
2785
-
2801
)
. , .
Ocean life of Atlantic salmon (Salmo salar L.) in the northwest Atlantic
,
Atlantic Salmon: Planning for the Future, pp
,
1986
London
Croom Helm
(pg.
483
-
511
)
.
Contributions of North American Atlantic salmon (Salmo salar L.) to the Faroese fishery
,
Naturaliste Canadienne
,
1987
, vol.
114
(pg.
187
-
193
)
, , .
Identification of North American and European Atlantic salmon (Salmo salar L.) caught off west Greenland in 1982–83
,
1984
ICES Document CM 1984/M: 12
, .
Origin of Atlantic salmon (Salmo salar L.) caught at sea near Nain, Labrador
,
Naturaliste Canadienne
,
1986
, vol.
113
(pg.
211
-
218
)
, .
A history of identification to continent of origin of Atlantic salmon (Salmo salar L.) at west Greenland, 1969–1997
,
Fisheries Research
,
1999
, vol.
43
(pg.
221
-
235
)
, .
Sea-surface temperature and distribution of Atlantic salmon (Salmo salar L.) in the northwest Atlantic
,
1987
, vol.
1
(pg.
262
-
275
)
American Fisheries Society Symposium
, , .
Inter-continental migrations of Atlantic salmon (Salmo salar L.)
,
1984
ICES Document CM 1984/M: 11
, .
Bioaccumulation of radiocaesium by fish: the influence of physiochemical factors and trophic structure
,
Canadian Journal of Fisheries and Aquatic Sciences
,
1994
, vol.
51
(pg.
2388
-
2410
)
RPII
Marine Monitoring
,
2002
Radiological Protection Institute of Ireland
RPII
Marine Monitoring
,
2003
Radiological Protection Institute of Ireland
, , .
Effects of the Faroese long-line fishery, other oceanic fisheries and oceanic variations on age of maturity of Icelandic north-coast stocks of Atlantic salmon (Salmo salar)
,
Fisheries Research
,
1991
, vol.
10
(pg.
207
-
228
)
.
Atlantic salmon scale reading guidelines
,
1992
, vol.
188
(pg.
1
-
4
)
ICES Cooperative Research Report
, , , , .
Records of post-smolt Atlantic salmon, Salmo salar L. in the Faroe-Shetland Channel in June 1996
,
Fisheries Research
,
1997
, vol.
31
(pg.
159
-
162
)
, , , .
Migration-orientation of Atlantic salmon (Salmo salar L.)
,
1973
, vol.
4
(pg.
119
-
137
)
International Atlantic Salmon Foundation Special Publication Series
STUK
Surveillance of environmental radiation in Finland
,
2003
Radiation and Nuclear Safety Authority (STUK). Årsrapport 2002. TOUKOKUU 2003. STUK-B-TKO4
STUK
Surveillance of environmental radiation in Finland
,
2004
Radiation and Nuclear Safety Authority (STUK). Årsrapport 2003. KESÄKUU 2004. STUK-B-TKO5
.
Tagging of salmon smolts in European rivers with special reference to recaptures off west Greenland in 1972 and earlier years
,
1980
, vol.
176
(pg.
93
-
113
)
Rapports et Procès-Verbaux des Réunions du Conseil International pour l'Exploration de la Mer
SYSTAT
Systat statistical software
,
2004
.
Atlantic salmon from the Labrador Sea and off West Greenland, taken during A. T. Cameron cruise, July–August 1965
,
1967
, vol.
4
(pg.
4
-
40
)
International Committee of North Atlantic Fisheries Research Bulletin
, .
Inferring ecological separation from regional differences in radioactive caesium in harbour porpoises Phocoena phocoena
,
Marine Ecology Progress Series
,
2002
, vol.
228
(pg.
301
-
309
)
, , , .
Detecting pan-Atlantic migration in salmon (Salmo salar) using 137Cs
,
Canadian Journal of Fisheries and Aquatic Sciences
,
1999
, vol.
56
(pg.
2235
-
2239
)
, , , , .
Restricted matrilineal gene flow and regional differentiation among Atlantic salmon (Salmo salar L.) populations within the Bay of Fundy, eastern Canada
,
Heredity
,
2002
, vol.
89
(pg.
465
-
472
)
, . .
Quota purchase
,
Salmon in the Sea and New Enhancement Strategies
,
1993
Oxford, UK
Fishing News Books
(pg.
249
-
263
)
.
Movements of salmon Salmo salar (L.) to and from Irish waters
,
Journal of Fisheries Biology
,
1973
, vol.
5
(pg.
659
-
671
)
© 2007 International Council for the Exploration of the Sea. Published by Oxford Journals. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
© 2007 International Council for the Exploration of the Sea. Published by Oxford Journals. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
Source: https://academic.oup.com/icesjms/article/64/2/394/2182595
0 Response to "Salmon Feeding in the Labraor Sea"
Post a Comment