How climate and post-smolt growth control marine mortality in Atlantic salmon; the potential effects of a changing climate on the marine survival of Atlantic salmon Kevin Friedland National Marine Fisheries Service Narragansett, Rhode Island
Things we may be concerned about, but cannot show a population level response, or things that are mortality factors for one group of stocks, but do not seem to be a factor for another.
There are climate and biological drivers that have been demonstrated for stock groupings, which form the basis of recruitment mechanisms.
Post-smolt growth plays a differential role in the marine survival of the two continental stock groups For southern tier North American stocks, post-smolt growth does not contribute to the pattern of variability in marine survival and adult recruitment. Whatever variability that occurs due to size mediated mortality is overshadowed by the variability associated with other sources of mortality. For southern tier European stocks, marine survival and adult recruitment correlate with post-smolt growth, particularly the growth that occurs during the summer-early fall period. Thus, marine survival is controlled by size mediated mortality occurring during this period.
Time series of recruitment and post-smolt growth signals Stock abundance (A) and composite post-smolt growth (B) time series for European and North American stock complexes. All data are normalized. 2 A Europe North America Recruitment 1 0-1 2 B Post-smolt growth 1 0-1 -2 1968 1974 1980 1986 1992 1998 2004 Smolt year Data drawn from: Peyronnet et al., 2007 McCarthy et al., 2008 Friedland et al., 2009 Hogan & Friedland, 2010
Relationship Between Proportionally Allocated Monthly Growth Indices and PC of Return Rate a) Burrishoole, 1SW b) Drammen, 1SW c) Drammen, 2SW d) Girnock, 2SW e) Greenland, 2SW f) Lagan, 1SW Friedland et al., 2009
Salmon recruitment is associated with decadal scale changes in the thermal regime of the North Atlantic. The climate forcing associated with this regime is the Atlantic Multi-decadal Oscillation (AMO). However, the two stock groups are responding to warming conditions in ecologically distinct ways at different junctures of their life history. For southern tier North American stocks, warming conditions during the spring associated with inshore and coastal ocean waters has been associated with declining marine survival. For southern tier European stocks, warming conditions during summer into fall in the Northeast Atlantic spatially associated with their summer nursery area has been associated with declining marine survival.
Correlation between recruitment and climate signals in the North Atlantic Bivariate scatter between European recruitment and the North Atlantic Oscillation (NAO) and the Atlantic Multi-decadal Oscillation (AMO) indices (A and B, respectively). Bivariate scatter between North American recruitment and the NAO) and the AMO indices (C and D, respectively). All data are normalized. European recruitment 2 1 0-1 A r=-0.10, p=0.52 B r=-0.67, p<0.01 North American recruitment 2 1 0-1 C r=-0.20, p=0.22 D r=-0.62, p<0.01-2 -1 0 1 2 NAO -2-1 0 1 2 AMO
Significant correlation between SST and Atlantic salmon recruitment Each panel is a contour plot the p-levels of correlations between recruitment and SST. Left hand panels are for North American recruitment, right hand panels for European, rows are for months May through December (A-H, respectively).
The absence of evidence to support a growth/size related recruitment mechanism in North America coupled with the changes in survival associated with changes in spring inshore thermal conditions suggest that recruitment is controlled by changes in predation pressure associated with the warmer conditions. Southern tier North American post-smolts are entering a warmer ocean, with different predator species composing the predator field and higher predator abundances. The impacts may be exacerbated by a phenological mismatch between the cues for post-smolt migration and ocean conditions.
Correlation Between Penobscot 2SW Return Rates and Predator Abundance Merluccius bilinearis (a), Urophycis chuss (b), Squalus acanthias (c), Lophius americanus (d), Hemitripterus americanus (e), and Helicolenus dactylopterus (f). light gray, P<0.05; medium gray, P<0.01; and, dark gray, P<0.001. Friedland et al., in press
Phenological Mismatch, Gulf of Maine Change in temperature in the rearing areas not in synchronization with temperature in ocean 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 0 0 5 5 8 10 12 8 10 12 Air temperature, C Sea surface temperature, C Friedland et al., in press
Phenological Mismatch, Gulf of St. Lawrence Change in temperature in the rearing areas not in synchronization with temperature in ocean Friedland et al., 2003
North Atlantic Temperature Range Average annual temperature range in the North Atlantic area in degrees centigrade
The growth/size related recruitment mechanism in Europe is supported by the overlap in time and space of the growth and thermal response of the stock complex. Southern tier European post-smolts enter a coastal ocean that does not vary greatly, and though mortality is high during the early marine phase, the accumulated mortality during summer, controlled by size, is more variable and thus controls recruitment. Northeast Atlantic thermal conditions have shifted the community structure in the region, thus impacting salmon, along with other species.
Change in the Northeast Atlantic Foodweb Friedland et al., 2009
Looking Forward Where might anticipated climate change affect Atlantic salmon: Change in terrestrial thermal conditions may render some rearing habitat unsuitable for salmon. Smolt migration timing may take on further phenological mismatch. Change in ocean conditions may make marine survival unsustainable. Also recognizing there may be many other areas of concern: Changing precipitation may affect stream productivity and thus juvenile recruitment. Changing oceanic acid balance may impact salmon post-smolts and adults.
SRES - Special Report on Emissions Scenarios CO 2 concentration, ppmv 900 800 700 600 500 400 A1B A2 B1 300 1980 2000 2020 2040 2060 2080 2100 Year
Contrast in Emissions Scenarios Annual TAS and TOS for CCCma model, average of five runs 8 7 A1B A2 B1 10 A1B A2 B1 Air temperature, C 6 5 4 3 2 Sea surface temperature, C 9 8 7 6 1 2020 2040 2060 2080 2100 Year 5 2020 2040 2060 2080 2100 Year
Climate Projections Studies Ensemble approach: use the output of more than one climate projection model Delta method: since each model is based on a different set of parameters, correct for the bias between models by developing a set of delta or anomaly values to project over time
Climate Model Ensemble BCCR-BCM2.0, Norway CCCma-CGCM3.1(T47), Canada CSIRO-Mk3.5, Australia GFDL-CM2.1, USA MIROC3.2, Japan MPI-ECHAM5, Germany MRI-CGCM2.3.2, Japan NCAR-CCSM3.0, USA
Sample Locations for Emissions Scenarios TOS TAS Sea Surface Temperature Surface Air Temperature
Thermal Ecology of Atlantic Salmon Atlantic salmon water temperature tolerance in degrees Celsius by life history stage (http://www.krisweb.com/). Life Stage Optimum Range Minimum Maximum Literature Cited Spawning 5-8 C 4.0 C 10-12 C DeCola (1970), Danie et al. (1984), Beall and Marty (1983) Egg/Alevin 4-7.2 0.5 12 DeCola (1970), Peterson et al. (1977), Elliot et al. (1998) Early Fry 8-19 0.5 23.5-27.7 Danie et al. (1984), Jensen et al. (1991) Parr Feeding 15-19 3.8 22.5 DeCola (1970, Elson (1975), Danie et al. (1984), Elliott (1991) Parr Survival 0.5-20 0 27-32 Garside (1973), Elliot (1991), Elliott and Elliott (1995) Smolt (Migrating) 7-14.3 5 19 Bakshitansky et al. (1979), LaBar et al. (1978), Jonnson and Rudd-Hansen (1985), Duston et al. (1991), Shepard (1991) Adult (Migrating) 14-20 8 23 Elson (1969), Danie et al. (1984), Shepard, S.L. (1995)
Thermal Changes in Rearing Habitat Climate model ensemble monthly maximum air temperature deviations North America Europe Surface air temperature, C 8 7 6 5 4 3 2 1 0-1 -2 BCCR CCCma GFDL NCAR CSRIO MPI MIROC MRI Surface air temperature, C 8 7 6 5 4 3 2 1 0-1 -2 BCCR CCCma GFDL NCAR CSRIO MPI MIROC MRI -3-3 -4 2020 2040 2060 2080 2100-4 2020 2040 2060 2080 2100 Year Year
Thermal Changes in Rearing Habitat Monthly maximum air temperature projections based on ensemble deviations North America Europe 24 23 Observed Projected 22 21 Observed Projected 22 20 Air temperature, C 21 20 19 18 Air temperature, C 19 18 17 16 17 15 16 1880 1920 1960 2000 2040 2080 14 1880 1920 1960 2000 2040 2080 Year Year
Thermal Changes in Rearing Habitat Daily maximum air temperature projections, CCCma only 100 In excess of 23 C In excess of 27 C North America 100 In excess of 23 C In excess of 27 C Europe 80 80 Number of days 60 40 Number of days 60 40 20 20 0 2020 2040 2060 2080 2100 Year 0 2020 2040 2060 2080 2100 Year
Phenology of Smolt Migration Projected deviation of date of arrival of 10 C air and ocean temperatures 0.2 0.0 10 C air 10 C SST Change in date, months -0.2-0.4-0.6-0.8-1.0 2020 2040 2060 2080 2100 Year
Thermal Conditions and Stock Abundance Linear and non-linear models of pre-fishery abundance North America Europe 1.0 0.8 Allometric, r 2 =0.38 Linear, r 2 =0.30 3.0 2.5 Allometric, r 2 =0.43 Linear, r 2 =0.33 Pre-fishery abundance, 10 6 0.6 0.4 0.2 Pre-fishery abundance, 10 6 2.0 1.5 1.0 0.5 0.0 8.5 9.0 9.5 10.0 10.5 Sea surface temperature, C 10.5 11.0 11.5 12.0 12.5 13.0 Sea surface temperature, C
Shift in Thermal Conditions in the Post-smolt Nursery Areas Climate model ensemble SST deviations Sea surface temperature, C 6 5 4 3 2 1 0-1 -2 BCCR CCCma GFDL NCAR CSRIO MPI MIROC MRI North America Sea surface temperature, C 6 5 4 3 2 1 0-1 -2 BCCR CCCma GFDL NCAR CSRIO MPI MIROC MRI Europe -3 2020 2040 2060 2080 2100-3 2020 2040 2060 2080 2100 Year Year
Historical and Projected Sea Surface Temperature SST projections based on ensemble deviations North America Europe Sea sruface temperature, C 13 12 11 10 9 8 Observed Predicted Sea sruface temperature, C 15 14 13 12 11 Observed Predicted 7 1880 1920 1960 2000 2040 2080 10 1880 1920 1960 2000 2040 2080 Year Year
Projected Stock Abundance Abundance expressed as both number and percent maximum PFA Pre-fishery Abundance, 10 3 Percent maximum PFA North America 400 Allometric 350 Linear 300 250 200 150 100 50 0 40 35 30 25 20 15 10 5 0 2020 2040 2060 2080 2100 Year Percent maximum PFA Pre-fishery Abundance, 10 6 Europe 1.2 Allometric Linear 1.0 0.8 0.6 0.4 0.2 0.0 45 40 35 30 25 20 15 10 5 0 2020 2040 2060 2080 2100 Year
Looking Forward Conclusions Temperatures in the rearing habitats are likely to increase and may produce episodic and sustained periods of thermal stress for juvenile salmon. There is an expectation of greater change in transitional ocean temperatures than rearing area air temperatures in North America that will likely exacerbate the phenological mismatch associated with smolt migrations. Ocean thermal conditions in key post-smolt nursery areas are expected to continue to change, making marine survival unsustainable for segments of the stock complexes from both North American and Europe.
Acknowledgements My thanks to my co-authors and the scientists from the ICES North Atlantic Salmon Working Group