© 2003 by ICES/CIEM International Council for the Exploration of the Sea/Conseil International pour l'Exploration de la Mer
Ocean thermal conditions in the post-smolt nursery of North American Atlantic salmon
a UMass/NOAA CMER Program Blaisdell House, University of Massachusetts, Amherst, MA 01003, USA
b Science Branch, Department of Fisheries and Oceans PO Box 566, St. John's, Newfoundland, Canada A1C 5X1
c Department of Fisheries and Oceans, Institut Maurice-Lamontagne 850 Route de la Mer, C.P. 1000, Mont-Joli, Quebec, Canada G5H 3Z4
*Correspondence to K. D. Friedland; tel: +1 413 545 2842; fax: +1 413 545 2304. e-mail: friedlandk{at}forwild.umass.edu.
The effect of climate on the post-smolt survival of North American Atlantic salmon is obscure owing to the difficulty in interpreting the only relationships thus far observed between the abundance of these stocks and climate, which focuses on winter conditions. Placing significance on winter post-smolt survival is contrary to conventional thinking that the spring period is more important, because that is when the post-smolts migrate to sea and transition to ocean life takes place. The pre-fishery abundance for North American stocks was compared to thermal conditions in potential post-smolt nursery areas during the period 1982–1999. Pre-fishery abundance was modeled as a reconstruction of one-sea-winter (1SW) and two-sea-winter (2SW) age salmon populations. Cohort abundance was compared to mean temperature and thermal habitat (sea surface area within a given temperature range) in five index areas. Stock size was negatively correlated with mean sea surface temperature during June. Correlations were comparatively stronger between stock abundance and thermal habitat, further asserting that June conditions – the first month at sea for most stocks in the region – may be pivotal to survival. These correlations suggest that post-smolt survival is negatively affected by the early arrival of warm ocean conditions in the nursery area. Hypotheses related to post-smolt migration, predation, and the availability of suitable prey are discussed.
Keywords: Atlantic salmon, post-smolt survival, sea surface temperature, thermal habitat
Received 26 June 2002; accepted 8 January 2003.
 |
Introduction |
|---|
 Top Introduction MethodsResultsDiscussionReferences |
|---|
Variability in Atlantic salmon stock size is associated with life-history events that occur during the first year at sea. After growing in freshwater as juveniles, salmon migrate to sea as post-smolts during spring. Their migrations are initiated by physiological changes mediated by environmental cues during freshwater residency (Whalen et al., 1999). The fish are challenged by a wide array of potential sources of mortality at sea; however, most fish are believed to be lost to predation during their first weeks in the ocean (Fisher and Pearcy, 1988; Holtby et al., 1990; Eriksson, 1994; Salminen et al., 1995). The mortality that occurs during this critical period may be controlled intrinsically by the growth rate of the fish and their resulting ability to escape predators; or, mortality rates may be independent of growth and may result from variation in predator abundance and distribution.
An emerging paradigm, which is based on observations of Atlantic salmon in the Northeast Atlantic, suggests that the controlling factor for survival and abundance is growth-mediated mortality. Friedland et al. (1998) observed that patterns of survival of two European salmon stocks, originating in southern Norway and western Scotland, were correlated with the thermal regimes present during the first weeks at sea of post-smolts. During the years when warm thermal regimes invaded the North Sea and southern Norwegian coastal area in a time frame coincident with post-smolt migration, survival was good. Alternatively, when cold conditions persisted during the migration period, survival was poor. Climate mediation of survival was suggested by the environmental correlate; however, the nature of the mechanism was further elucidated with growth data for one stock. Friedland et al. (2000) reported that post-smolt growth followed a pattern similar to the time series of survival and climate variation. When warm sea surface temperatures (SSTs) were present, survival was higher and post-smolt growth was also correlated. Thus, post-smolt survival of European Atlantic salmon appears to be controlled by the same mechanisms hypothesized for a wide range of marine species: faster growth during a critical period is associated with the ability to escape predation (Anderson, 1988; Pepin, 1991).
The mechanisms controlling survival of North American Atlantic salmon are obscure. Reddin (1988) and Ritter (1989) searched for environmental correlates by relating survival rate to thermal conditions during the post-smolt year. Such variables, as the northern extent of the 4°C isotherm in January and the extent of ocean surface covered by water warmer than 4°C and limited by latitude and longitude cutoffs, also in January, were positively correlated with survival. The working hypothesis offered was that distribution patterns of both post-smolts and predators might be responding to this winter climate signal and resulting in varying survival (Ritter, 1989). Reddin and Friedland (1993) computed a thermal habitat index whereby the habitat sum was weighted by temperature and associated catch rates. They also reported a correlation between winter conditions and stock size of North American stocks. Friedland et al. (1993) systematically searched for thermal habitat trends in other seasons that might be related to stock size, but thermal habitat variation in winter provided the strongest correlate to stock size. Though this temperature correlation appeared to be robust to various methods of computation and description, it was also baffling because the notion that the critical period for salmon was in winter was difficult to support with other biological data: Atlantic salmon are quite large by their first winter and have outgrown vulnerability to most predators affecting them earlier in the post-smolt year (Allen et al., 1972).
Post-smolt growth data for North American salmon is equivocal in its support of growth- and size-related control of post-smolt mortality. Studies on hatchery fish show that between-stock survival rates can be related to post-smolt growth (Friedland et al., 1996). However, no evidence has been produced yet that shows that inter-annual variation in survival is related to growth for North American stocks (Friedland, 1998). This may be related to the selection of stocks analyzed thus far; perhaps, the analysis of wild stocks over a wide geographic range will shed further light on this issue.
What is glaringly obvious is the fact that despite the efforts to find them, there are no climate or growth signals associated with the survival of North American salmon during their early marine life. This may be because of the spatial and temporal scales of the data used in prior climate analyses, which were based on ocean data derived from the comprehensive ocean and atmospheric dataset (COADS) and, thus, limited to a spatial resolution of 2° boxes. Additionally, the analyses were intentionally restricted from examining inshore areas. We report on a comparison of temperature and thermal habitat trends with refined estimates of stock size of North American salmon. The spatial and temporal focus of our comparisons is on inshore areas during spring, thus emphasizing the habitats post-smolts would most likely utilize during their early marine residency. Through these modifications, we hope to address the limitations associated with earlier studies.
|
Methods |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
Study area
The study area covers most of the ocean entry and post-smolt nursery habitat of North American Atlantic salmon (Figure 1). Five index areas were established within the region; they were constrained by the availability of temperature data and our assumptions about the extent of post-smolt habitat. Salmon post-smolts are known to orient to the surface (Dutil and Coutu, 1988); thus, sea surface conditions should be indicative of changes within the depth range they utilize. At a minimum, the areas represent the spatial extent of transit habitat during post-smolt migration. For each index area, temperature and thermal habitat indices were computed for the period bracketing the post-smolt migration to sea and early ocean life. The time period when post-smolts enter the marine environment varies among stocks and will be reviewed for a series of index stocks from North America.
|
Pre-fishery abundance
Trends in abundance were derived from a run-reconstruction model of one-sea-winter (1SW) and two-sea-winter (2SW) age groups. Because the abundance estimates were originally used to generate forecasts for resource managers, the time chosen was immediately prior to the commencement of 1SW fisheries; thus, these abundance values have become known as pre-fishery abundance (PFA) estimates. The reconstruction accounts for all known removals and natural mortality occurring to stocks as prescribed by the ICES North Atlantic Salmon Working Group (ICES, 2002). We proceed by describing the development of PFA estimates and then show their link to the marine environment.
Estimates of total abundance are generated separately for the following geographic regions: Labrador, Newfoundland, Québec, Gulf of St. Lawrence, Scotia-Fundy, and USA. Returns of 1SW and 2SW salmon to each region were estimated by updating methods and variables used by Rago et al. (1993a) and reported in ICES (1993) and updated in ICES (2002). The returns for both sea-age groups were estimated using a variety of methods, depending on the data available for individual river systems and/or management areas. These methods included counts at monitoring facilities, population estimates from mark-recapture studies, and the application of angling and commercial catch statistics, angling exploitation rates, and measurements of freshwater habitat. The 2SW component of the multi-sea-winter (MSW) returns was determined using the sea-age composition of one or more indicator stocks. Of the total of 570 rivers, the number of assessed rivers in eastern Canada and the US has ranged from a high of 108 (19%) to a recent low of 72 (13%). Because the number returning to adjacent rivers within a wide area varies in a coherent manner, the use of one or more rivers within a geographic region to describe others should not be problematic.
In the context used in this study, "returns" are the number of salmon that returned to the geographic region, including home-water commercial fisheries, except in the case of Newfoundland and Labrador regions where returns do not include commercial fisheries. This was done to avoid double counting of fish when commercial catches in Newfoundland and Labrador are added to returns of all geographic areas to create the PFA of North American salmon.
The non-maturing component of 1SW fish of the PFA estimator for year i (NN1i), destined to be 2SW returns (excludes 3SW and previous spawners), is constructed by summing 2SW returns in year i+1 (NR2i+1), 2SW salmon catches in commercial and aboriginal peoples' food fisheries in Canada (NC2i+1), and catches in year i from fisheries on non-maturing 1SW salmon in Canada (NC1i) and Greenland (NG1i). In Labrador, aboriginal peoples' food harvests of small (AH_s, <2.7 kg) and large salmon (AH_l, =2.7 kg) were included in the reported catches for the period 1999–2001. Because harvests occurred in both Lake Melville and coastal areas of northern Labrador, the fraction of these catches that are immature was labeled as af_imm. This was necessary because non-maturing salmon do not occur in Lake Melville where approximately half the catch originated. However, non-maturing salmon may occur in coastal marine areas in the remainder of northern Labrador. Consequently, af_imm for the fraction of aboriginal peoples' harvests that were non-maturing was set at 0.05 to 0.1, which is half of f_imm from commercial fishery samples (range: 0.1–0.2). The equations used to calculate NC1 and NC2 are as follows:
|
| (1) |
|
| (2) |
Similar to 1998–2000, the commercial fishery in Labrador remained closed in 2001. In past reports, salmon returns and spawners for Labrador, which make up one-sixth of the geographical areas contributing to return estimates of maturing 2SW salmon for Canada (NR2), were based on commercial fishery data, but this time series ended in 1998. To estimate PFA, a value for Labrador returns for 2SW for 1998–2001 had to be constructed by developing an appropriate raising factor. This raising factor (RFL2) was derived by dividing PFA without Labrador into PFA with Labrador based on the time series 1971–1997, and was set to the low and high ends of the observed range of values (1.05–1.27). A natural mortality rate (M) of 0.03 per month is assumed to adjust the numbers between the fisheries on 1SW and 2SW salmon (10 months) and between the fishery on 2SW salmon and returns to the rivers (1 month) as subsequently shown:
|
| (3) |
Maturing 1SW salmon are, in some areas, a major stock component and measuring their abundance is thought to be important to provide measures of abundance for the entire cohort from a specific smolt class. The calculations are analogous to those for immature 1SW salmon. For the commercial catches in Newfoundland and Labrador, all small salmon are assumed to be 1SW fish, based on catch samples that show its percentage to be in excess of 95%. Large salmon are primarily MSW salmon, but some maturing and non-maturing 1SW may also be present. Estimates of fractions of non-maturing salmon present in the Newfoundland and Labrador catch were based on gonad samples taken from commercial catches. The "large" category in SFA 1–7 and 14B consists of 0.1–0.3 1SW salmon (ICES, 1993; Rago et al., 1993b), while in SFA 8–14A, fish are mainly maturing (Idler et al., 1981). These values were assumed to apply to the aboriginal food fishery catches in marine coastal areas of northern Labrador.
A raising factor (RFL1) similar to the one for estimating returns of 2SW salmon was developed to include Labrador returns in the maturing component of PFA during the recent closure of the fishery, and estimates for returns to Labrador for 1998–2001 were set accordingly to the low and high ends of the range of observed values during 1971–1999 (1.04–1.59).
The maturing 1SW component for year i (MN1i) is constructed by summing maturing 1SW returns in year i (MR1i) in Canada and the USA and their catches from commercial and food fisheries in Newfoundland and Labrador (MC1i). The same natural mortality rate (M) of 0.03 per month as that for immature fish is assumed to adjust the numbers between the fishery on 1SW salmon and returns to the rivers (1 month):
|
| (4) |
|
| (5) |
The estimated PFA represents the extant population and does not account for the fraction of the population present in a given fishery area. The model does not take into account non-catch fishing mortality in any fishery, because such estimates are not available on an annual basis and rates are not well described for some fisheries. Commercial catches were not included in the run-reconstruction model for the West Greenland fishery (1993 and 1994), Newfoundland fishery (1992–2001), and Labrador fishery (1998–2001), as these fisheries were closed. As PFA estimates of 2SW salmon require data on returns to rivers, the most recent year for which an estimate is available is 2000.
Probability density function for PFA estimates
With the exception of the Greenland catch (NG1i), none of the input values in the above equations can be estimated precisely. Estimates of total returns (NR2i+1, NR1i, and MR1i) are based on a variety of regional extrapolation methodologies and estimates of sea-age composition of the Newfoundland, and Labrador catches (NC1, MC1, and NC2) are based on observed ranges of sea-age compositions within market categories and observed ranges of maturing and non-maturing fractions in the 1SW component of the catch. The implications of the uncertainties in deriving NN1 and MN1 were addressed using Monte Carlo simulation. For each year, the approximate distributions of NN1 and MN1 were estimated by drawing random values for NR1, MR1, NR2, NC1, NC2, and MC1 from their observed range and assuming that they were uniformly distributed and then substituting the new random values into Equations 3 and 4 to obtain a realization of NN1 and MN1. Their probability density functions were approximated by 2000 independent realizations. Upper and lower bounds on the distributions were represented by the fifth and 95th percentiles.
Thermal habitat indices
Temperature regime of the ocean sea surface was based on data extracted from version 2 of the optimum interpolation (OI) SST analysis (Reynolds et al., 2002). These data are produced weekly on a 1° grid and use in situ and satellite SST plus SST simulated by sea-ice cover. Before analysis, satellite data are adjusted for biases using the methods of Reynolds (1988) and Reynolds and Marsico (1993).
Mean monthly temperature was computed for each index area by year from the OI dataset for the period bracketing post-smolt migration for that index area. Mean temperatures by month and year for 1° boxes were computed from weekly temperature fields (four or five estimates). The mean monthly values averaged over the boxes within each index area were correlated with total, 1SW, and 2SW PFA using Pearson's product-moment correlation. Next, monthly average temperatures by box were classified according to 2°C intervals (with midpoints starting at 1.5°C), and areal extent (km2) of each class was computed by index area, and the year as an index of its availability as "thermal habitat". For practicality, only those temperature intervals that produced time series without missing values were used in the analysis. Again, mean monthly thermal habitat estimates were correlated with total, 1SW, and 2SW PFA using Pearson's product-moment correlation. Significance probabilities were recomputed using the Dunn–Sidak correction for multiple correlations within an index area–abundance grouping (Systat, 2002)
Some key months and years were characterized with SST maps showing the transition of water temperature change over a 1-month period. Particular isotherms are highlighted in these maps.
|
Results |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
Time frame of smolt migrations
North American salmon smolts generally migrate to sea in May in the southern part of the range and in June in the northern part. Many rivers have traps designed to intercept migrating smolts, making it possible to characterize the initiation and duration of these migrations (Table 1). In the most northern river system considered (Western Arm Brook, on the west coast of Newfoundland), smolts migrate mostly in June with a peak during the second and third week of the month. Campbellton River, located in the northeast coast of Newfoundland, tends to have an earlier run that peaks in late May. De la Trinité and Saint-Jean rivers show a similar contrast in migration period for rivers from the same province. For rivers further south, peak smolt migrations range from the second half of May in the Miramichi to the first half of May in the Connecticut River (the southern extent of salmon distribution). Data from four and three stocks were used to estimate the migration time frames for index areas 1 and 2, respectively; only one stock was used for each of the remaining index areas 3–5 (Figure 2).
|
|
Pre-fishery abundance
Over the past two decades, there have been significant changes in the total PFA of the salmon stock complex. The abundance of 1SW fish has been estimated as high as 864 000 individuals and has declined to approximately half a million in recent years (Figure 3). The decline in 2SW fish has been more dramatic, going from nearly 575 000 to only 100 000 fish in recent years. The pattern of change is similar in the two age groups (r=0.76, p<0.01). Total abundance of the cohorts, represented by summing PFA of 1SW and 2SW fish, was nearly 1.4 million fish in the 1980s. The size of recent cohorts are only slightly greater than 1SW PFA, reflecting the more dramatic decline in number of 2SW fish.
|
Thermal conditions in the nursery area
The pattern of vernal warming differed among the five index areas (Figure 4). Some areas experienced much cooler winter conditions with widespread ice in the surface waters. Average April temperatures were just above 0°C in areas 1 (Gulf of St. Lawrence) and 5 (Newfoundland shelf), whereas areas 2–4 (associated with the Gulf of Maine and Scotian Shelf) were in the range of 3–5°C during April. All the areas reached at least 10°C by July in all the years with the exception of area 5 (5–9°C). The range of mean SST varied by season and area: all areas showed the greatest variation during June, and areas 3 and 4 showed the greatest variation overall.
|
Trends in thermal habitat area estimates reflected those observed in mean SST. Between areas 1 and 5, temperature ranges could be characterized each month (Table 2). Thermal habitat estimates for area 1 (the largest area with approximately 325 000 km2) ranged from approximately 26 000 to 153 000 km2 for a given temperature range in each month (1.5 to 15.5°C). The standard deviation tended to be 30% of the mean thermal habitat area, which applied to other areas as well. Area 2 (approximately 150 000 km2) was characterized by thermal habitat ranges from 3.5 to 13.5°C. Areas 3 and 4 were of similar size to that of area 2, totaling approximately 170 000 and 190 000 km2, respectively. Area 4 could be characterized by five monthly thermal habitat ranges for the 3-month period, which was conspicuously less than the other areas. The greater inter-annual variations in temperature for this area produced more incomplete time series (Figure 4d). Area 5 (235 000 km2 and furthest north) contained the coldest thermal habitat, of the five areas, reflecting the strong influence by the cold Labrador Current.
|
Relationship of PFA with thermal conditions
The correlative comparison of mean SST and PFA (Table 3) yielded only a significant correlation (p<0.05) for 2SW fish in June in area 1. The sign was negative, indicating that high abundance was associated with cooler conditions over the time series. However, after the Dunn–Sidak correction for multiple correlations, this correlation proved not to be significant.
|
The correlative comparisons of thermal habitat and PFA (Table 4) yielded some highly significant positive correlations for both 1SW and 2SW fish and the 7.5°C thermal habitat area in area 1 in June that are consistent with the weak relationship identified in the mean SST data. The correlations for 2SW and the total were also significant after the Dunn–Sidak correction. Significant correlations were also found for 1SW (April 7.5°C and June 9.5°C in area 2; June 11.5°C in area 3) and 2SW (June 11.5°C in areas 2 and 3), but none of these were significant after correction. None of the correlations for areas 4 and 5 were significant.
|
The spatial and temporal properties of the correlations in area 1 during June were examined through the analysis of the SST data at its original 1° latitude and longitude resolution, and through the use of the weekly temperature and thermal habitat fields. The correlation coefficients between the abundance time series and mean June SST for each 1° box within area 1 were contoured to reveal spatial trends (Figure 5). Spatially disaggregating the data reduced the correlation for 1SW fish suggesting this correlation is dependent on spatial combination of the data (Figure 5a). In contrast, the correlation between 2SW abundance and SST remained strong, and the highest values were centered in the western portion of the Gulf of St. Lawrence, coincidentally located close to the major salmon rivers of New Brunswick (Figure 5b). The correlation for total PFA largely show the same pattern as for 2SW fish, but the inclusion of 1SW fish reduces the significance (Figure 5c). Spatial disaggregation of thermal habitat data would not be meaningful.
|
Correlation coefficients between PFA and weekly mean SST for area 1 were highest during latter half of June (week beginning on June 8 through week beginning on June 22; Figure 6a), whereas the 7.5°C thermal habitat area provided the strongest correlations during the week beginning on June 22 (Figure 6b). Thus, both weekly SST and thermal habitat data suggest that the critical period for salmon survival in area 1 is during the latter half of June, coinciding with the end of the peak smolt run in the area (Figure 2).
|
SST distributions during June suggest widely varying conditions between years producing large and small smolt recruitments. During 1986, a year associated with high abundance estimates, the Gulf of St. Lawrence and associated coastal waters remained within the 8–10°C temperature range for the entire month (Figure 7a–d). In contrast, during 1998, a year that produced fewer recruits, the Gulf of St. Lawrence had SSTs in excess of 10°C, and water of >12°C was present by the end of the migration period (Figure 7e–h). These distributions are consistent with correlations between mean SST and abundance, and thermal habitat and abundance.
|
|
Discussion |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
The results provide a first indication that climate variation during spring may play a pivotal role in the patterning of post-smolt survival, and thus stock abundance, for North American salmon stocks. The examination of post-smolt ecology for these stocks is clouded by the problem of interpreting their habitat use in the absence of detailed migration information that shows their exact locations at sea after they exit their home rivers. Recent intensive studies have shown that European post-smolts migrate to the Norwegian Sea (Holm et al., 2000) by utilizing favorable currents (Jonsson et al., 1993). North American post-smolts face unfavorable currents to reach their eventual overwintering areas to the north of Newfoundland and in the Labrador Sea since they travel against the prevailing Labrador Current circulation (Reddin and Short, 1991). For many of the most productive stocks comprising the North American stock complex, viz. the Miramichi River and other rivers emptying into the Gulf of St. Lawrence, the first weeks at sea are most likely spent in coastal habitats. There are indications that these inshore habitats may serve not only as migration transit areas, but also as nurseries during the post-smolt growth season (spring and summer), at least in some years (Friedland et al., 1999).
The correlation between stock abundance and climate during June, the time frame associated with the first weeks at sea for many stocks, and in the Gulf of St. Lawrence, the area producing a high proportion of the annual smolt cohort, is encouraging. The negative correlation with surface temperature, the positive correlation with the extent of the 7.5°C, and the association of poor recruitment with the presence of warm water suggest that survival is favored when Gulf waters are warmed up slowly and stay below 10°C over large areas during the first weeks at sea.
The strengths and weaknesses (Gargett et al., 2001) of correlations are well known. Therefore, we are compelled to use caution in interpreting our results. Nevertheless, we see the new information as an opportunity to develop mechanistic hypotheses on what might be happening to salmon in the sea. Many ongoing research programs, like surveys for post-smolts, could use this information to guide future activities. We also see the potential for several research approaches that are currently undeveloped and will explore four areas of hypothesis development in greater detail: how ocean conditions may affect growth of post-smolts, whether variations in climate have altered predation pressure, or alternatively prey abundance, and how migration may be affected by variations in climate.
Temperature conditions may affect post-smolts directly through the physiological effects of temperature on growth or indirectly through the development of annual bloom conditions and the resulting secondary production in the form of forage species. Because salmonid growth increases linearly with water temperature, given an adequate food supply (Brett, 1979), we are faced with the dilemma of making sense of the outcome that warmer conditions are associated with poorer survival. Perhaps, growth is faster in cold conditions because the fish can use a smaller food ration more efficiently with lower metabolism. High temperature may increase metabolic demand beyond food resources available to the fish and inhibit growth. Indeed, fish on low food rations have a lower thermal preferendum than fish on high rations (Despatie et al., 2001), suggesting that growth will be optimized at a lower temperature when food is limiting. Temperature, among a wide array of factors, affects the timing and development of spring blooms. In the Gulf of St. Lawrence, temperature has been identified as an important factor controlling the succession of phytoplankton communities (Levasseur et al., 1984; Therriault et al., 1990), but spring blooms in the Gulf typically do not develop until late June or early July, and may be controlled to a greater degree by other factors, such as freshwater runoff or advective losses of phytoplankton from the relatively thin surface mixed layer (Fuentes-Yaco et al., 1997; Zakardjian et al., 2000; Plourde et al., 2001). The late development of the bloom begs the question as to whether primary production in the same season is important to salmon, because post-smolts would have been feeding for many weeks before the bloom enters the food chain. As elsewhere, the development of the spring bloom varies by size fraction, which in turn may affect availability to different grazers and ultimately the food resources of post-smolts (Tamigneaux et al., 1999).
The onset of bloom conditions in adjacent coastal waters is earlier (often as early as April/May) than in the Gulf and, thus, may contribute more directly to post-smolt growth. However, Head (2000) suggests that bloom initiation on the Scotian Shelf is not dependent on temperature, and, besides, our results indicate that it is the climate variation in June that co-varies with salmon abundance. In addition, ocean color data indicate that the phytoplankton production on the Scotian Shelf has markedly increased from the 1970s to the 1990s, while salmon survival has decreased (Zwanenburg et al., 2002). Overall, there is little coherence in any argument suggesting that warmer spring conditions have enhanced post-smolt growth through changes in primary production.
The amount of foraging habitat for post-smolts in the surface mixed layer may be related to the dynamics of the cold intermediate layer (CIL), a persistent mass of cold (<0°C) water lying immediately below the surface mixed layer and characteristic of the Gulf and some adjacent waters. The thickness of the CIL is influenced by the severity of winter conditions, which affects the thickness of the surface mixed layer, which varies accordingly. CIL thickness is strongly negatively correlated with distribution, condition, and growth of demersal species, such as Atlantic cod (Castonguay et al., 1999; Dutil et al., 1999). It would be interesting to attempt to relate salmon mortality with CIL conditions in the various parts of its range.
Inordinate predation pressure on early life-history stages of ground fish caused by the release of fishing pressure on pelagic species (Atlantic mackerel Scomber scombrus and Atlantic herring Clupea harengus) is an emerging theme, as the scientists try to interpret ground-fish recruitment patterns (Swain and Sinclair, 2000). The abundance of small pelagics since the mid-1980s could also have adversely affected post-smolt survival through predation after their first entry to the marine environment. The association of more rapid warming and poor survival of post-smolts may be interpreted as an indication that some predators invade post-smolt nurseries behind specific isotherms (Hinch et al., 1995). Alternatively, post-smolts may shift their seaward distribution, owing to warmer conditions, and move to areas where predators are more common. Migratory pelagic predators like mackerel are often associated with thermal transition zones (Garrison et al., 2000). However, this linkage appears to be rather plastic and is not associated with a specific barrier as suggested by the June 7.5°C thermal habitat area responsible for the correlations observed in this study (D'Amours and Castonguay, 1992). Cod distribution in the Gulf of St. Lawrence changes seasonally, and is associated with movements into cold water (<4°C) in summer, spatially removed from post-smolt distributions (Castonguay et al., 1999; Swain, 1999). Predation by gannets is patterned in fixed fields reflecting their home range relative to the rookeries, thus a change in prey distribution caused by temperature conditions may affect the predation for a given set of rookeries (Montevecchi et al., 2002). However, even the most liberal application of gannet predation rates does not explain the mortality of post-smolts during the first summer. Seals breed in much of the post-smolt nursery areas and have increased in number in recent years, a trend consistent with the decline in salmon stocks (Stenson et al., 2002). However, warmer conditions should create less spatial overlap with post-smolts during the summer months. A review to find the "perfect" predator, i.e. one with an impact on post-smolts in the nursery areas that varies in concert with the June climate signal, is frustrating. This frustration may be interpreted as supporting the idea that predation pressure is near saturation, and that factors related to predation avoidance are more important.
Post-smolts undergo an important ontogenetic change in feeding that may be key to their survival during the first year at sea. On first entering the marine environment, they feed mainly on terrestrial insects and marine invertebrates and then make a transition to piscivory, usually concentrating on the 0-group of winter-spawned species. For example, Salminen et al. (2001) reported that post-smolts in the Baltic first feed on young of the year herring. Feeding data available for North American nurseries indicate that herring do not necessarily serve this function. Dutil and Coutu (1988) found that the dominant food item in the diet of Gulf of St. Lawrence post-smolts was sand lance (Ammodytes americanus) larvae, one of the dominant ichthyoplankton species over large areas in the Gulf and neighboring waters (Locke and Courtenay, 1995; Lazzari, 2001). Wright and Bailey (1996) suggested that coupling between hatching and the onset of spring secondary production is important for the growth and survivorship of sand lance (Ammodytes marinus) from the Shetland Islands area. Changes in availability of this species have been implicated as critical to the success of seabirds (Rindorf et al., 2000). Similarly, capelin larvae are also abundant in most Gulf areas and may provide another food resource for post-smolts as they switch to a piscivorous life style. Does June climate variation correlate with the availability of a preferred ichthyoplankton prey? Because most of these species are only assessed intermittently, further work is required before conclusions can be drawn.
Post-smolts represent a relatively minor component of the food web and, therefore, may not be affected by fluctuations in predator and prey levels as much as they are controlled by individual demands of swimming and migration. They may have strong migrational behaviors that select for temperature ranges and thus influence swimming expenditure to avoid undesirable thermal conditions. The association of warm conditions in June with poor survival may reflect increased swimming demanded of these cohorts as they seek optimal temperatures. Patterns of survival for salmon stocks throughout the North Atlantic illustrate that stocks with shorter migrations to the feeding grounds have higher survival rates (Bley and Moring, 1988; Friedland, 1994). The relationship between migration and survival may be an important scalar on the stock level as well.
|
Acknowledgements |
|---|
We thank Richard Reynolds and Diane Stokes for their help in providing SST data and our colleagues in the ICES North Atlantic Salmon Working Group for their tireless work.
|
References |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
-
Allen K.R, Saunders R.L, Elson P.F. (1972) Marine growth of Atlantic salmon (Salmo salar) in the northwest Atlantic. Journal of the Fisheries Research Board of Canada 29:1373–1380.[Web of Science]
Amiro P. G. and Jansen H. (2000) Impact of low-head hydropower generation at Morgans Falls, LaHave River on migrating Atlantic salmon (Salmo salar). Canadian Technical Report of Fisheries and Aquatic Sciences, 2323. 25 pp.
Anderson J.T. (1988) A review of size dependent survival during pre-recruit stages of fishes in relation to recruitment. Journal of Northwest Atlantic Fisheries Science 8:55–66.
Bley P. W. and Moring J. R. (1988) Freshwater and ocean survival of Atlantic salmon and steelhead: a synopsis. Maine Cooperative Fish and Wildlife Research Unit, Orono, ME. Biological Report. 88 pp.
Brett J.R. (1979) Environmental factors and growth. In Hoar W.S, Randall D.J, Brett J.R (Eds.). Fish Physiology(Academic Press, London) Volume VIII: pp. 599–675.
Caron F. and Courtemanche D. A. (2001) Rapport d'opération de la rivière Saint-Jean 2001. Société de la faune et des parcs du Québec, Direction de la recherche sur la faune. 66 pp.
Caron F., Gauthier C., Lamy R. (2000) Rapport d'opération de la rivière de la Trinité en 2000. Société de la faune et des parcs du Québec. Direction de la recherche sur la faune. 86 pp.
Castonguay M, Rollet C, Fréchet A, Gagnon P, Gilbert D, Brêthes J.-C. (1999) Distribution changes of Atlantic cod (Gadus morhua L.) in the northern Gulf of St. Lawrence in relation to an oceanic cooling. ICES Journal of Marine Science 56:333–344.
Chaput G., Hardie P., Hayward J., Moore D., Shaesgreen J. NSPA1. (2002) Migrations and biological characteristics of Atlantic salmon (Salmo salar) smolts from the Northwest Miramichi River, 1998 to 2000. Canadian Technical Report of Fisheries and Aquatic Sciences, 2415. 66 pp.
D'Amours D and Castonguay M. (1992) Spring migrations of Atlantic mackerel, Scomber scombrus, in relation to water temperature through Cabot Strait (Gulf of St. Lawrence). Environmental Biology of Fishes 34:393–399.[CrossRef][Web of Science]
Dempson J. B., Furey G., Bloom M. (2001) Assessment of the status of the Atlantic salmon stock of Conne River, SFA 11, Newfoundland, 2000. Canadian Stock Assessment Secretariat Research Document, 2001/030. 45 pp.
Despatie S.-P, Castonguay M, Chabot D, Audet C. (2001) Final thermal preferendum of Atlantic cod: effect of food ration. Transactions of the American Fisheries Society 130:263–275.[CrossRef]
Downton P. R., Downton D. G., Johnson R. W. (2001) Status of Atlantic salmon (Salmo salar L.) in Campbellton River, Notre Dame Bay (SFA 4), Newfoundland in 2000. Canadian Stock Assessment Secretariat Research Document, 2001/031. 73 pp.
Dutil J.-D, Castonguay M, Gilbert D, Gascon D. (1999) Growth, condition, and environmental relationships in Atlantic cod (Gadus morhua) in the northern Gulf of St. Lawrence and implications for management strategies in the Northwest Atlantic. Canadian Journal of Fisheries and Aquatic Sciences 56:1818–1831.
Dutil J.D and Coutu J.M. (1988) Early marine life of atlantic salmon, Salmo salar, postsmolts in the Northern Gulf of St. Lawrence. Fisheries Bulletin 86:197–211.
Eriksson T. (1994) Mortality risks of Baltic salmon during downstream migration and early sea-phase: effects of body size and season. Nordic Journal of Freshwater Research 69:100.
Fisher J.P and Pearcy W.G. (1988) Growth of juvenile coho salmon (Oncorhynchus kisutch) off Oregon and Washington, USA, in years of differing coastal upwelling. Canadian Journal of Fisheries and Aquatic Sciences 45:1036–1044.
Friedland K.D. (1994) Marine survival of restoration stocks. In Calabi S and Stout A (Eds.). A Hard Look at Some Tough Issues(Silver Quill Books, Camden) pp. 223–239 371 pp.
Friedland K.D. (1998) Ocean climate influences on critical Atlantic salmon (Salmo salar) life history events. Canadian Journal of Fisheries and Aquatic Sciences 55:suppl. 1, 119–130.
Friedland K.D, Dutil J.-D, Sadusky T. (1999) Growth patterns in postsmolts and the nature of the marine juvenile nursery for Atlantic salmon, Salmo salar. Fishery Bulletin 97:472–481.[Web of Science]
Friedland K.D, Haas R.E, Sheehan T.F. (1996) Post-smolt growth, maturation, and survival of two stocks of Atlantic salmon. Fishery Bulletin 94:654–663.[Web of Science]
Friedland K.D, Hansen L.P, Dunkley D.A. (1998) Marine temperatures experienced by post-smolts and the survival of Atlantic salmon (Salmo salar L.) in the North Sea area. Fisheries Oceanography 7:22–34.[CrossRef][Web of Science]
Friedland K.D, Hansen L.P, Dunkley D.A, MacLean J.C. (2000) Linkage between ocean climate, post-smolt growth, and survival of Atlantic salmon (Salmo salar L.) in the North Sea area. ICES Journal of Marine Science 57:419–429.
Friedland K.D, Reddin D.G, Kocik J.F. (1993) Marine survival of North American and European Atlantic salmon: effects of growth and environment. ICES Journal of Marine Science 50:481–492.
Fuentes-Yaco C, Vezina A.F, Larouche P, Gratton Y, Gosselin M. (1997) Phytoplankton pigment in the Gulf of St. Lawrence, Canada, as determined by the Coastal Zone Color Scanner. Part II: Multivariate analysis. Continental Shelf Research 17:1441–1459.[CrossRef][Web of Science]
Gargett A.E, Li M, Brown R. (2001) Testing mechanistic explanations of observed correlations between environmental factors and marine fisheries. Canadian Journal of Fisheries and Aquatic Sciences 58:208–219.
Garrison L.P, Michaels W, Link J.S, Fogarty M.J. (2000) Predation risk on larval gadids by pelagic fish in the Georges Bank ecosystem. I. Spatial overlap associated with hydrographic features. Canadian Journal of Fisheries and Aquatic Sciences 57:2455–2469.
Hayes F. R. (1953) Artificial freshets and other factors controlling the ascent and population of Atlantic salmon in the LaHave River, Nova Scotia. Fisheries Research Board of Canada Bulletin, 99. 47 pp.
Head E. (2000) Interannual variations in hydrography and spring bloom dynamics, and their effect on Calanus finmarchicus distribution and reproduction on the Scotian Shelf in the late 90s and 2000. ICES CM, 2000/M: 10.
Hinch S.G, Healey M.C, Diewert R.E, Thomson K.A, Hourston R, Henderson M.A, Juanes F. (1995) Potential effects of climate change on marine growth and survival of Fraser River sockeye salmon. Canadian Journal of Fisheries and Aquatic Sciences 52:2651–2659.
Holm M, Holst J.C, Hansen L.P. (2000) 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 57:955–964.
Holtby L.B, Andersen B.C, Kadowaki R.K. (1990) Importance of smolt size and early ocean growth to interannual variability in marine survival of coho salmon (Oncorhynchus kisutch). Canadian Journal of Fisheries and Aquatic Sciences 47:2181–2194.
ICES. (1993) Report of the Working Group on North Atlantic Salmon. ICES CM, 1993/ACFM: 10.
ICES. (2002) Report of the Working Group on North Atlantic Salmon. ICES CM, 2002/ACFM: 14.
Idler D.R, Hwang S.J, Crim L.W, Reddin D.G. (1981) Determination of sexual maturation stages of Atlantic salmon (Salmo salar) captured at sea. Canadian Journal of Fisheries and Aquatic Sciences 38:405–413.
Jonsson N, Hansen L.P, Jonsson B. (1993) Migratory behaviour and growth of hatchery-reared post-smolt Atlantic salmon Salmo Salar. Journal of Fish Biology 42:435–443.[CrossRef][Web of Science]
Lazzari M.A. (2001) Dynamics of larval fish abundance in Penobscot Bay, Maine. Fishery Bulletin 99:81–93.[Web of Science]
Levasseur M, Therriault J.-C, Legendre L. (1984) Hierarchical control of phytoplankton succession by physical factors. Marine Ecology Progress Series 19:211–222.[Web of Science]
Locke A and Courtenay S.C. (1995) Effects of environmental factors on ichthyoplankton communities in the Miramichi Estuary, Gulf of St. Lawrence. Journal of Plankton Research 17:333–349.
Marshall T. L., Jones R., Anderson L. (1999) Follow-up assessments of Atlantic salmon in the Saint John River drainage, N.B, 1998. Canadian Stock Assessment Secretariat Research Document, 99/109. 42 pp.
Meister A.L. (1969) Conservation of the Atlantic salmon in the State of Maine. Atlantic Salmon Journal 3:11–17.
Montevecchi W.A, Cairnes D.K, Myers R.A. (2002) Predation on marine-phase Atlantic salmon (Salmo salar) by gannets (Morus bassanus) in the Northwest Atlantic. Canadian Journal of Fisheries and Aquatic Sciences 59:602–612.
Mullins C. C., Caines D., Lowe S. L. (2001) Status of Atlantic salmon (Salmo salar L.) of Lomond River, Torrent River and Western Arm Brook, Newfoundland, 2000. Canadian Stock Assessment Secretariat Research Document, 2001/039. 56 pp.
Pepin P. (1991) The effect of temperature and size on development, mortality and survival rates of the pelagic early life stages of marine fishes. Canadian Journal of Fisheries and Aquatic Sciences 48:503–518.
Plourde S, Joly P, Runge J.A, Zakardjian B, Dodson J.J. (2001) Life cycle of Calanus finmarchicus in the lower St. Lawrence Estuary: the imprint of circulation and late timing of the spring phytoplankton bloom. Canadian Journal of Fisheries and Aquatic Sciences 58:647–658.
Rago P. J., Meerburg D. J., Reddin D. G., Chaput G. J., Marshall T. L., Dempson B., Caron F., Porter T. R., Friedland K. D., Baum E. T. (1993a) Estimation and analysis of pre-fishery abundance of the two-sea-winter population of North American Atlantic salmon (Salmo salar), 1974–1991. ICES CM, 1993/M: 24.
Rago P. J., Reddin D. G., Porter T. R., Meerburg D. J., Friedland K. D., Potter E. C. E. (1993b) A continental run reconstruction model for the non-maturing component of North American Atlantic salmon: analysis of fisheries in Greenland and Newfoundland-Labrador, 1974–1991. ICES CM, 1993/M: 25.
Reddin D.G. (1988) Ocean life of Atlantic salmon (Salmo salar L.) in the Northwest Atlantic. Atlantic Salmon: Planning for the Future. Proceedings of the Third International Atlantic Salmon SymposiumOctober 21–23, 1986Biarritz, France(Croom HelmIn Mills D.H and Piggins D.J (Eds.). , London) pp. 483–511.
Reddin D.G and Friedland K.D. (1993) Marine environmental factors influencing the movement and survival of Atlantic salmon. In Mills D (Ed.). Salmon in the Sea(Fishing News Books, London) pp. 79–103.
Reddin D.G and Short P.B. (1991) Postsmolt Atlantic salmon (Salmo salar) in the Labrador Sea. Canadian Journal of Fisheries and Aquatic Sciences 48:2–6.
Reynolds R.W. (1988) A real-time global sea surface temperature analysis. Journal of Climate 1:75–86.[Medline]
Reynolds R.W and Marsico D.C. (1993) An improved real-time global sea surface temperature analysis. Journal of Climate 6:114–119.[CrossRef][Web of Science]
Reynolds R.W, Rayner N.A, Smith T.M, Stokes D.C, Wang W. (2002) An improved in situ and satellite SST analysis for climate. Journal of Climate 15:1609–1625.[CrossRef][Web of Science]
Rindorf A, Wanless S, Harris M.P. (2000) Effects of changes in sandeel availability on the reproductive output of seabirds. Marine Ecology Progress Series 202:241–252.[Web of Science]
Ritter J. A. (1989) Marine migration and natural mortality of North American Atlantic salmon (Salmo salar L.). Canadian Manuscript Report of Fisheries and Aquatic Sciences, 2041. 146 pp.
Salminen M, Erkamo E, Salmi J. (2001) Diet of post-smolt and one-sea-winter Atlantic salmon in the Bothnian Sea. Northern Baltic. Journal of Fish Biology 58:15–35.
Salminen M, Kuikka S, Erkamo E. (1995) Annual variability in survival of sea ranched Baltic salmon, Salmo salar L: significance of smolt size and marine conditions. Fisheries Management and Ecology 2:171–184.
Stenson G.B, Hammill M.O, Kingsley M.C.S, Sjare B, Warren W.G, Myers R.A. (2002) Is there evidence of increased pup production in northwest Atlantic harp seals, Pagophilus groenlandicus. ICES Journal of Marine Science 59:81–92.
Swain D.P. (1999) Changes in the distribution of Atlantic cod (Gadus morhua) in the southern Gulf of St. Lawrence – effects of environmental change or change in environmental preferences? Fisheries Oceanography 8:1–17.[Web of Science]
Swain D.P and Sinclair A.F. (2000) Pelagic fishes and the cod recruitment dilemma in the Northwest Atlantic. Canadian Journal of Fisheries and Aquatic Sciences 57:1321–1325.
Systat. (2002) Systat 10 Reference Manual(SPSS Inc, Chicago).
Tamigneaux E, Legendre L, Klein B, Mingelbier M. (1999) Seasonal dynamics and potential fate of size-fractioned phytoplankton in a temperate nearshore environment (Western Gulf of St. Lawrence, Canada). Estuarine, Coastal and Shelf Science 48:253–269.[CrossRef]
Therriault J.-C, Legendre L, Demers S. (1990) Oceanography and ecology of phytoplankton in the St. Lawrence estuary. Coastal and Estuarine Studies 39:269–295.
Whalen K.G, Parrish D.L, McCormick S.D. (1999) Migration timing of Atlantic salmon smolts relative to environmental and physiological factors. Transactions of the American Fisheries Society 128:289–301.[CrossRef]
Wright P.J and Bailey M.C. (1996) Timing of hatching in Ammodytes marinus from Shetland waters and its significance to early growth and survivorship. Marine Biology 126:143–152.[CrossRef]
Zakardjian B.A, Gratton Y, Vezina A.F. (2000) Late spring phytoplankton bloom in the Lower St. Lawrence Estuary: the flushing hypothesis revisited. Marine Ecology Progress Series 192:31–48.[Web of Science]
Zwanenburg K.C.T. (2002) Decadal changes in the Scotian shelf large marine ecosystem. In Sherman K and Skjoldal H.R (Eds.). Large Marine Ecosystems of the North Atlantic, Changing States and Sustainability(Elsevier, New York).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. B. Dempson, V. A. Braithwaite, D. Doherty, and M. Power Stable isotope analysis of marine feeding signatures of Atlantic salmon in the North Atlantic ICES J. Mar. Sci., September 25, 2009; (2009) fsp227v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Chaput, C.M. Legault, D.G. Reddin, F. Caron, and P.G. Amiro Provision of catch advice taking account of non-stationarity in productivity of Atlantic salmon (Salmo salar L.) in the Northwest Atlantic ICES J. Mar. Sci., January 1, 2005; 62(1): 131 - 143. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. D. Friedland, G. Chaput, and J. C. MacLean The emerging role of climate in post-smolt growth of Atlantic salmon ICES J. Mar. Sci., January 1, 2005; 62(7): 1338 - 1349. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Lajus, J. A. Lajus, Z. V. Dmitrieva, A. V. Kraikovski, and D. A. Alexandrov The use of historical catch data to trace the influence of climate on fish populations: examples from the White and Barents Sea fisheries in the 17th and 18th centuries ICES J. Mar. Sci., January 1, 2005; 62(7): 1426 - 1435. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||