29 January 2007
Spatial shifts in spawning habitats of Arcto-Norwegian cod induced
by climate change.
Svein Sundby and Odd Nakken Institute of Marine Research P.O. Box 1870 Nordnes N-5817 Bergen
Norway Abstract
The Arcto-Norwegian cod is an Arcto-boreal cod stock found in the colder part of the habitat range of Atlantic cod stocks. Only the West Greenland cod and the Northern cod off Newfoundland are found at lower temperatures. A general feature for the Arcto-Norwegian cod is that it tends to have good recruitment in warm years and poor recruitment in cold years. The present paper shows that the spawning intensity at the various spawning sites is, as well, influenced by climate variations. However, while the recruitment response to temperature is on an interannual time scale the response of changes in spawning site is on a multidecadal time scale. The Arcto-Norwegian cod spawns along Norwegian coastline from Bergen to Finnmark, a distances of more than 1500 km. On a multidecadal scale there have been two cold and two warm periods during the 20th century, cold in the 1900-20s, warm in 1930-50s, cold in the 1960-70s and warm since the mid 1980s. Time series 1900-1975 from the Norwegian Directorate of Fisheries on cod roe indices from the spawning sites along the coast shows that the southernmost spawning areas is more important during the cold periods while the northernmost spawning areas are more important in warm periods. After the mid 1970s the observations ended. However, qualitative observations show that there have been poor spawning fisheries in the southernmost spawning areas during the present warm period, and in 2004 spawning Arcto-Norwegian cod was again observed along the coast of East-Finnmark.
Introduction
The Arcto-Norwegian cod is an Arcto-boreal cod stock found in the colder part of the habitat range of Atlantic cod stocks. Only the West Greenland cod and the Northern cod off Newfoundland are found at lower temperatures (Sundby 2000) (Figure 1). A general feature for the Arcto-Norwegian cod is that it tends to have good recruitment in warm years and poor recruitment in cold years (Sætersdal and Loeng 1987;
Ellertsen et al. 1989). This feature has also been shown to be valid for the other cod stocks in the lower temperature range (deYoung and Rose 1993), while the cod stocks in the upper temperature range (like the North Sea cod and the Irish Sea cod) respond to a temperature change oppositely (Ottersen 1996; Planque and Fox 1998; Planque and Frédou 1999). This particular temperature response of Atlantic cod stocks indicates that temperature, in additions to its direct effect on recruitment (i.e. rapid growth through the most vulnerable life stages), is a proxy for other processes of importance to the recruitment mechanisms, for example, supply of prey items by advection of copepods (Sundby 2000).
Figure 1. Distribution of Atlantic cod stocks and mean annual temperatures (Sundby 2000).
For Arcto-Norwegian cod temperature has been documented to influence the entire life history in a number of ways: Maturation and egg production of the individuals depend on feeding conditions which in turn is related to prey abundance and
temperature (Kjesbu et al. 1998). Pelagic juveniles in the western, and warmer, parts of the Barents Sea grow faster than those in the eastern and colder areas (Bjørke and Sundby 1986; Suthers and Sundby 1993), a feature that has been reproduced by modelling larval and juvenile growth and transport (Vikebø et al. 2005). Year-class strength and individual size of 0-group cod are positively correlated with ambient temperature (Nakken 1994; Ottersen and Loeng 2000). Interannual spatial
temperature variations strongly influence the distribution and growth of young and immature cod (Nakken and Raknes 1987; and Ottersen et al. 1998).
The spawning dynamics itself, however, have been considered to be invariant with varying temperature conditions. During the 1970s and 1980s spawning periods were monitored at selected sites at the main spawning areas in Lofoten (Ellertsen et al. 1989) and the timing and duration was remarkably constant between years, starting in the beginning of March, peaking during the first days of April and terminating in the beginning of May. This interannual constancy in spawning period has been
interpreted as an indication of that seasonal light cycle is the major trigger of the spawning. On a longer term, i.e. from the 1929 to 1982, it has been observed that spawning has changed to occur about two weeks later (Pedersen 1984). This finding was based on data from the Norwegian fishery statistics by investigating the time of
the season when the land-based fishing industry stopped receiving cod roe for processing. It has been discussed whether this long-term change in peak spawning is linked to the similar long-term change in age at maturation. From the early 1950s to the 1990s age at maturation decreased from about 10 to 7 years (Jørgensen 1990). The areas along the south-eastern border of the Lofoten archipelago in Northern Norway has been known, at least since the 9th century (Eigils saga), as the major spawning location of skrei, the Old-Norse term of the Arcto-Norwegian cod. Skrei
means “the wanderer” that comes migrating from far away (i.e. the Barents Sea) to spawn along the Norwegian coast, and the term was used to part it from the local coastal cod populations. The skrei has an appearance different from the coastal cod both by its colour and shape. Moreover, it has its specific otolith structure and age at maturity. In addition to the central Lofoten spawning district, there are a number of other spawning locations of skrei along the Norwegian coast from Western Norway to Finnmark (Sundby and Godø 1994). Figure 2 shows the locations of spawning areas in three categories: major, minor and occasional spawning areas along a 1500 km long coastline from Sotra off Bergen to Sørøya in West Finnmark. The two major
spawning areas are Lofoten and Møre. The minor ones are scattered at specific sites along the coast. The occasional spawning locations are mainly offshore areas where spawning are absent for a number of years and then intense spawning may occur for two to three years, as for example on Moskenesgrunnen and Eggagrunnen off Lofoten (Figure 2). So, even if the Lofoten spawning district definitely is the most important spawning area, less that 1/3 of the skrei spawn here (Sundby and Bratland 1987). Sæterdal and Hylen (1964) analysed the catch statistics for the skrei spawning fisheries for a 90 years period from 1860s to 1950s. They observed large interannual variations in the catches, but they also noticed that there had occurred a substantial long-term shift from higher catches in the southern locations before the 1930s to higher catches in the northern locations after this time. In the 1910s the catches in the southern locations, in fact, exceeded the catches in Lofoten in some years. The authors ascribed this development as “natural fluctuations” but they also indicated that changing fishing pattern might have influenced the change.
Figure 2. Spawning districts and locations of Arcto-Norwegian cod categorised in 1. major 2. minor and 3. occasional spawning locations. (After Sundby and Godø 1994).
During February and March 2004 and 2005 the fishing industry reported large numbers of mature and pre-spawning cod at the fishing grounds along the coast of East Finnmark. This feature occurred after a period of high temperatures for longer period of time, i.e. since the beginning of the 1990s. This situation motivated for a closer inspection of historical fisheries statistics to explore whether similar situations had occurred during earlier periods, particularly during the previous long-term trend of warm period during 1930-1950s. The present paper addresses the issue of the long-term changes/oscillations in the spawning intensity at the southern and northern skrei spawning locations along the Norwegian coast during the 20th century. We show that these variations are correlated to similar long-term changes/oscillations in the marine climate of the northeastern North Atlantic and discuss possible mechanisms behind this correlation.
Material and methods
A time series of an egg production index for skrei was published annually in the fisheries statistics of “Norges fiskerier” during the period 1904 - 1940. This egg production index is here presented in four spawning regions, the southern location “Møre”, the central spawning location “Lofoten”, and for the two northernmost
districts “Troms” and “Finnmark”. The tabulated egg production index, which we here will term FN, had this definition:
FN = R/NC
where R is the total amount of liters of cod roe delivered to the region over the entire fishing season, and NC is the corresponding number of 1000 cod delivered. During the
period 1935-1952 the fisheries statistics changed the definition of the index to be based on weight of the cod instead of numbers of cod. Hence, the new index, which we here will term FW, had this definition:
FW = R/WC
where WC tons of gutted cod delivered during the entire spawning season. We used
the overlapping period of 6 years from 1935-1940 for the two indices to derive the original index FN from FW. The regression line for the southern and northern
spawning districts (Møre, Troms and Finnmark) was:
FN = 2.49 FW + 14 and the correlation coefficient R2 = 0.93.
For the central spawning region Lofoten the correlation was somewhat different: FN = 3.27 FW - 3 and the correlation coefficient R2 = 0.82.
From 1953 to 1970 the roe index was no longer tabulated in the statistics. Therefore, the index, Fw, was constructed from the tabulated values of tons of roe received and
tons of skrei received, and transformed to FN, based on the regressions above. From
1970 to 1976 the statistics no longer parted between skrei and coastal cod. Hence, the last 7 years of the time series is not reliable for the Møre district because the major part of the cod delivered here was coastal cod. In the northern districts, however, the major part of the landings are skrei, so the index can still be used for these regions onto 1976, although with less precision. In 1977 “Statistisk sentralbyrå” took over the publication of the fisheries statistics from “The Directorate of Fisheries”. From that time also the data on cod roe ceased to be published.
Climate data are taken from the Russian Kola section that represent well the
interannual variations in the thermal climate of the entire region of the spawning cod (Sundby 1994). The temperatures are annual means of the Atlantic water in the section, i.e. 50 –200 m depth station nos. 4-7, for the period 1900-2000.
Results
Figure 3 shows the annual values of the roe index, FN, for the southern spawning
district Møre, for the central district Lofoten, and for the two northernmost spawning districts Troms and Finnmark. The interannual variations are large, but in addition, there are long-term trends in the data, and these trends are opposite in the southern and the northern districts. The long-term trend for the central spawning district Lofoten is different from the other districts. However, it resembles mainly the pattern of the two northern districts.
0 50 100 150 200 250 300 1900 1920 1940 1960 1980 Year Roe index, Fn Møre Lofoten Troms Finnmark
Figure 3. Annual values of roe indices, FN, 1904 –1976 for the southern spawning district Møre, the
central spawning district Lofoten, and the two northern spawning districts Troms and Finnmark.
The long-term trends stand out more clearly by presenting the five-years running means. Figure 4 shows the five-years running mean for Lofoten and Figure 5 shows the similar data for the southern (Møre) and northern spawning districts (Troms and Finnmark). Lofoten 0 50 100 150 200 250 1900 1910 1920 1930 1940 1950 1960 1970 1980 Year Roe index, Fn Lofoten
0 50 100 150 200 1900 1910 1920 1930 1940 1950 1960 1970 1980 Year Roe index, Fn Møre Troms Finnmark
Figure 5. Roe index, FN, at the southern spawning district (Møre) and the two northern districts (Troms and Finnmark), 5-years running mean.
Figure 6 shows the temperature time series from the Atlantic water of the Kola section, annual means and 5-years running means. The running mean point of the temperatures is the mean of the 5 years prior to the point. This is done because the roe indices are assumed to be influenced by the past growth conditions for the mature cod. 2 3 4 5 1900 1910 1920 1930 1940 1950 1960 1970 1980 Year Temperature Annual mean 5-yrs run. mean
Figure 6. Annual mean temperature and 5-years running mean temperature for the Atlantic water masses of the Kola section in the Barents Sea.
The long-term increase of the roe index in the northern spawning districts from the 1920s to the 1930s and the decrease from the 1950s to the 1970s seem to be correlated to long-term trends of the sea temperature. The inverse trends of the roe index for the southern district Møre indicate a negative correlation with temperature. In the
northernmost spawning district Finnmark, where the temperature is lowest, the duration of the period of high roe index in the mid 20th century is about 10 years shorter than in the other northern district Troms.
Figures 7 – 10 show the linear regressions between 5-years running mean
temperatures and the 5-years running mean roe indices. The southern district Møre is the only spawning location where the correlation coefficient between the roe index and temperature is negative. For Lofoten the correlation is positive and for the northern districts the correlation coefficients are clearly positive.
y = -34.904x + 265.82 R2 = 0.1054 50 70 90 110 130 150 170 190 3 3.5 4 4.5 5 Temperature Roe index, Fn Møre Lineær (Møre)
Figure 7. Linear regression between 5-years running mean temperatures and the 5-years running mean roe indices for the southern spawning district Møre.
y = 74.462x - 124.78 R2 = 0.3778 100 120 140 160 180 200 220 240 3 3.5 4 4.5 5 Temperature Roe index, Fn Lofoten Lineær (Lofoten)
Figure 8. Linear regression between 5-years running mean temperatures and the 5-years running mean roe indices for the central spawning district Lofoten.
y = 91.291x - 238.66 R2 = 0.5441 0 20 40 60 80 100 120 140 160 180 3 3.5 4 4.5 5 Temperature Roe index, Fn Troms Lineær (Troms)
Figure 9. Linear regression between 5-years running mean temperatures and the 5-years running mean roe indices for the northern spawning district Troms.
y = 79.007x - 269.47 R2 = 0.581 0 10 20 30 40 50 60 70 80 90 100 3 3.5 4 4.5 5 Temperature Roe index, Fn Finnmark Lineær (Finnmark)
Figure 10. Linear regression between 5-years running mean temperatures and the 5-years running mean roe indices for the northern spawning district Finnmark.
After the last cool period in the northeast Atlantic during the 1960s and 1970s the sea temperature started increasing again from the early 1980s, but the time series on the roe indices were unfortunately been brought to end during the mid 1970s. The reports on the increasing skrei spawning fishery in eastern Finnmark during spring 2004 indicate that the stock might respond similarly to the warming in the mid 20th century.
Discussion
The Arcto-Norwegian cod is managed as one stock unit, but it has been distinguished between the proper Barents Sea and the Svalbard components that are geographically separated as immature fish (Trout 1956; Maslov 1972). It is assumed that fish from the two components mix during spawning. However, there are also indications through tagging experiments that fish from the eastern Barents Sea mainly spawn from Lofoten and northwards while the western component generally spawns more offshore and at the spawning areas to the south of Lofoten. (Trout 1956; Hylen et al. 1961; Godø et al. 1984; Godø 1984; 1986).
The recent observations of mature pre-spawning cod in the northernmost spawning district of Finnmark during spring 2004 indicate a repetition of the increased spawning intensity during the previous warm period. The fishery statistics clearly demonstrated that a similar period occurred in the previous warm period of the mid 20th century, from about 1930 to 1950s. Moreover, the fishery statistics shows that the more important spawning district of the two northern districts, Troms, (located closer to the central district Lofoten) has a similar interdecadal pattern to the warm period of
the mid 20th century, but this trend has been more difficult to qualitatively observe, since spawning takes place here annually, and the large interannual variations have disguised this long-term trend in spawning activity. Also, in the central spawning district Lofoten the long-term trend in spawning activity appears to co-vary with the temperature, but the pattern are here somewhat different with two peaking periods at the beginning and at the end of the warm period. The most remarkable result,
however, is the inverse response to the climate trends in the southernmost spawning district Møre. This indicates that the Arcto-Norwegian cod stock responds to climate change with a displacement in locations of spawning: When temperature increases the spawning is displaced northwards, while a reduction in temperature results in a
southwards displacement of spawning. The multidecadal temperature pattern has in the present paper been documented by the observations from the Kola Section in the Barents Sea through the 20th century. Recently, Sutton and Hodson (2005) showed an extended time series back to 1870 of multidecadal summer temperatures of the entire North Atlantic region. They confirm the multidecadal pattern of the Kola section during the 20th century and further show that the last three decades of the 19th century was warmer than normal.
Among the Atlantic cod stocks, the Arcto-Norwegian cod is, together with the
Northern cod and West Greenland cod stocks, located at the lower range with respect to temperature habitats as shown in Figure 1. Under such conditions it is likely that an increase in temperature results in a northwards expansion of the habitat since ambient temperatures are suboptimal in all the regions theirs habitats with respect to growth rates of larvae and juveniles (Otterlei et al. 1998) adults (Jobling 1981) and with respect to recruitment (Sundby 2000). However, the reduction in spawning intensity at the southern (and warmest) fringe of the spawning districts of the Arcto-Norwegian cod indicates that the expansion of the spawning district in the north is occurring on the cost of a reduction in the south. There are several possible mechanisms behind such a response:
1) Because of the long spawning migration of Arcto-Norwegian cod an expansion of its feeding habitat eastwards in the Barents Sea during warm periods might result in too long migration routes to the southernmost spawning areas., The resulting free pelagic drift of the offspring back into the Barents Sea would, in turn, result in settlement of young cod farther east and north in the Barents Sea where they are able to survive due to the higher temperature. A shift to colder climate would result in higher winter mortality of 1-group and 2-group cod as shown by Ponomarenko (1984). This would particularly important in coldest part of the Barents Sea, i.e. the northern and eastern Barents Sea. This would, in turn, lead to a more southwesterly distribution of the entire stock which would allow for a more southerly spawning migration.
2) The shift in spawning areas might be a response to inverse variations in two
subpopulations of the Arcto-Norwegian cod with separate feeding areas and spawning areas: one western Barents Sea and Svalbard component spawning at the Møre
spawning area as indicated by Godø et al. (1984), and a eastern Barents Sea
component spawning in Lofoten, Troms and Finnmark. In cold periods the transport of offspring towards the coast of Svalbard is relatively stronger than the transport wind the eastern branch into the Barents Sea. This has been shown by Vikebø et al. (2005) who modelled growth and transport of larval cod from the spawning area in Lofoten under a reduction of the Atlantic Currents due to a reduction in thermohaline circulation and a cooler climate. The cooler climate resulted in a larger reduction of
the current branch into the Barents Sea compared to the branch flowing along the west coast of Svalbard. Hence, a larger fraction of the offspring will be transported and settled to the west coasts of Svalbard during cool periods. As the Svalbard component tends to use the Møre coast as its spawning area it might explain the increased
spawning intensity at Møre during cool periods, while in warm periods a larger fraction of the offspring will be directed into the Barents Sea, and the offspring fraction that happen to be transported along the west coast of Svalbard might to a larger extent be subjected to loss from the natural habitat of the continental shelf because of stronger currents that might transport the larvae into the deep regions of the Fram Strait and the Arctic Basin.
3) Another possible mechanism for the reduction of the western component and the southernmost spawning districts during warm periods could be linked to predator-prey interrelationships; i.e. increased competition and/or predation during the early free-drifting pelagic stages. For example, herring has been observed to be a substantial predator on cod eggs (Melle 1985) and the abundance of herring has shown parallel fluctuations with the interdecadal climate fluctuations (Toresen and Østvedt 2000), and high herring abundance would result in relatively higher egg predation in the Møre district than in the northern districts because Møre is a major spawning area for herring. A reduction in prey abundance for the adult stages during warm periods might also reduce the adult western stock component. For example, the highly abundant Arctic prawn (Pandalus borealis) in the western Barents Sea and along the Svalbard coasts might be reduced in warm periods and the capelin might be displaced eastwards into the Barents Sea (Gjøsæter 1998).
It should be emphasized that three suggested mechanisms are not contradicting, and it might as well be a combination of them that causes the long-term shift in the
spawning areas. Fish stocks respond to climate variability and climate change in a multitude of ways, and the type of response depends on the periodicity of the climate variations. The amplitudes of the interannual to decadal temperature variations are large in the Northeast Atlantic. At the Kola section of the Barents Sea the decadal-scale peak-to-peak mean annual sea temperature is 1.5 oC. Phenomenon like individual growth rates (Jobling 1981; Ottersen and Loeng 2000), year-class
formation (Sætersdal and Loeng 1987; Ellertsen et al. 1989), and spatial distributions (Nakken and Raknes 1987; Ottersen et al. 1998; Michalsen et al. 1998) has been shown to have immediate responses to the interannual temperature variations (Nakken 2002). In other words, these rapid responses are linked to processes of the population ecology. The interdecadal-scale peak-to-peak temperature signal is typically less than half of the decadal-scale signal, eg. for the Kola Section during the 20th century it is 0.7 oC. This more moderate and long-term amplitude influences fish stocks in a different way like the spawning behaviour, as shown in the present study. Changes in spawning behaviour seems to linked to other kinds of ecosystem processes, e.g. changes in the life history of fish stocks, species interaction, trophic transfer, and evolutionary processes of the ecosystem. These are processes related to the system ecology. Such processes will work on longer time scales than those of population ecology. However, their implications on coastal societies could potentially be larger than interannual to decadal variations. In the study and assessments of climate change it is important to be aware of these different kinds of responses. The presently
described response in spatial change of spawning behaviour shows that it is linear and
The temperature increase in the Barents Sea over the next 50 years due to global warming is predicted to be ~1.5 oC (ACIA 2005). If the response in spawning behaviour still is linear, also within such a range of temperature variation, we would expect that the Arcto-Norwegian cod couldmigrate farther into the eastern parts of the Barents Sea and that the southernmost spawning districts might even less important in future than today.
Acknowledgement
The work was supported by the INTEGRATION project under DEKLIM by the German Research Council and by the ECOBE project by the Norwegian Research Council. Thanks are extended to PINRO, Murmansk, Russia that provided the climate time series of the Kola Section.
References
ACIA 2005.
Bjørke, H. and Sundby, S. 1986. Abundance indices and distribution of post larvae and 0-group cod. In: Contribution to the third Soviet-Norwegian symposium in Murmansk, 26.-30. May 1986: "The effect of oceanographic conditions on distribution, and population dynamics of commercial fish stocks in the Barents Sea”. 127-144.
DeYoung B. and Rose G. 1993. On recruitment and distribution of Atlantic cod (Gadus morhua) off Newfoundland. Canadian Journal of Fisheries and Aquatic Sciences, 50 (12): 2729-2740.
Ellertsen, B., Fossum, P., Solemdal, P. and Sundby, S. 1989. Relations between
temperature and survival of eggs and first feeding larvae of the North-East Arctic cod (Gadus morhua L.). Rapp. P.-v. Reun. Cons. int. Explor. Mer, 191:209-219
Gjøsæter, H. 1998. The population biology and exploitation of capelin (Mallotus villosus) in the Barents Sea. Sarsia, 83 (6):
Godø OR., Nakken O. and Raknes A. 1984. Acoustic estimates of cod off Lofoten and Møre in 1984. ICES CM./G:47. 10p.
Godø OR. 1984. Migration, mingling and homing of North-east Arctic cod from two separated spawning grounds. Pp 289-302 in OR Godø and S Tilseth (eds.): The Proceedings of the Soviet-Norwegian Symposium on Reproduction and Recruitment of Arctic cod. Institute of Marine Research, Bergen.
Godø OR. 1986. Dispersion and mingling of cod from various nursery and feeding areas along the Norwegian cost and in the Barents Sea. Pp. 663-672 in M Alton (ed.): A workshop on comparative biology, assessment, and management of gadoids from the north Pacific and Atlantic Oceans. Northwest and Alaska Fisheries Center, Seattle, Wa.
Hylen A, Midttun L. and Sætersdal G. 1961. Torskeundersøkelser i Lofoten og Barentshavet i 1960. [Investigations on cod in Lofoten and in the Barents Sea in 1960]. Fiskets Gang 5: 101-114. (In Norwegian).
Jobling M. 1981. Temperature tolerance and the final preferendum - rapid methods for the assessment of optimum
Jørgensen, T. 1990. Long-term changes in age at maturity of Northeast Arctic cod (Gadus morhua L.). J. Cons. Int. Explor. Mer 46: 235-248.
Kjesbu OS, Witthames PR, Solemdal P, Walker MG. 1998. Temporal variations in the fecundity of Arcto-Norwegian cod (Gadus morhua) in response to natural changes in food and temperature. Journal of Sea Research 40(3-4): 303-321.
Maslov, NA. 1972. Migration of the Barents Sea cod. Fish. Res. Bd. Canada. Translation Series no 2129: 1-43.
Melle W. 1985. Predasjon på torskens egg og larver i Lofoten [Predation on cod eggs and larvae in Lofoten]. Cand scient thesis, University of Bergen. 142 pp.
Michalsen K, Ottersen G, Nakken O. 1998. Growth of North-east Arctic cod (Gadus morhua L.) in relation to ambient temperature. ICES Journal of Marine Science 55:863-877.
Nakken O. 2002. Understanding environmental controls on fish stocks and progress towards their inclusion
Nakken O, Raknes A. 1987. The distribution and growth of Northeast Arctic cod in relation to bottom temperatures in the Barents Sea, 1978-1984. Fisheries Research 5:243-252.
Norges fiskerier 1910 – 1976. Norges offisielle statistikk. Fiskeridirektøren. Fishery statistics.
Otterlei et al 1998.
Ottersen G. 1996. Environmental impact on variability in recruitment, larval growth and distribution of Arcto-Norwegian cod. Unpublished dr. scient. thesis, Geophysical Institute, University of Bergen, Norway. 136 p.
Ottersen G. and Loeng H. 2000.. Covariability in early growth and year-class strength of Barents Sea cod, haddock and herring: the environmental link. Symposium on Recruitment Dynamics of Exploited Marine Populations: Physical-Biological Interactions. ICES Journal of Marine Science 57:67-85. Ottersen G, Michalsen K, Nakken O. 1998. Ambient temperature and distribution of north-east Arctic cod. ICES Journal of Marine Science 55:67-85.
Ottestad, P. 1942. On periodical variations in the yield of the great sea fisheries and the possibility of establishing yield prognoses. Fiskeridirektoratets Skrifter Series Havundersøkelser 7(5): 1-11. Pedersen, T. 1984. Variation of peak spawning of Arcto-Norwegian cod during the period 1929-1982 based on indices estimated from fisheries statistics. In Dahl, E., Danielssen, DS., Moksness, E. and Solemdal, P. (eds.), The propagation of cod (Gadus morhua L.). Flødevigen rapportserie 1, 1984: 301-316.
Planque B. and Fox CJ. 1998. Interannual variability in temperature and the recruitment of Irish Sea cod. Marine Ecology Progress Series, 172: 101-105.
Planque B, Frédou T. 1999. Temperature and the recruitment of Atlantic cod (Gadus morhua). Canadian Journal of Fisheries and Aquatic Sciences 56:2069-2077.
Ponomarenko IY. 1984. Survival of bottom-dwelling young cod in the Barents Sea and factors determining it. In: Godø OR, Tilseth S, editors. Proceedings of the Soviet-Norwegian Symposium on Reproduction and Recruitment of Arctic Cod. Leningrad , 25-30 September 1983. Bergen: Institute of Marine Research. p 192-212.
Sundby, S. 1994. The influence of bio-physical processes on fish recruitment in an arctic-boreal ecosystem. Dr. Philos. Thesis, University of Bergen, Bergen, Norway, 1994. 189 pp.
copepod populations. Sarsia 85: 277-298.
Sundby, S. og Bratland, P. 1987. Kartlegging av gytefeltene for norsk-arktisk torsk i Nord-Norge og beregning av eggproduksjonen i årene 1983 - 1985 (Spatial distribution and production of eggs from Northeast-arctic cod at the coast of Northern Norway 1983-1985). Fisken og Havet,
1987(1): 1-58 (In Norwegian, English abstract).
Sundby, S. and Godø, OR. 1994. Life history of the Arcto-Norwegian cod stock. Contribution to ICES Coop. Research Report 205 (25 pp + 22 figs).
Suthers, I. M. and Sundby, S. 1993. Dispersal and growth of pelagic juvenile Arcto-Norwegian cod (Gadus morhua L.), inferred from otolith microstructure and water temperature. ICES Journal of Marine Sciences, 50: 261-270.
Sutton, RT. And Hodson, DLR. 2005. Atlantic Ocean forcing of North American and European summer climate. Science, 309: 115-118.
Sætersdal G, and Hylen A. 1964. The decline of the skrei fisheries. A review of the landing statistics 1866-1957 and an evaluation of the effects of the postwar increase in the total
exploitation of the Arctic cod. Fiskeridirektoratets Skrifter Series Havundersøkelser 13(7): 56-69. Sætersdal G, Loeng H. 1987. Ecological adaptation of reproduction in Northeast Arctic cod. Fisheries Research 5:253-270.
Trout GC. 1956. The Bear Island cod: migrations and movements. Fisheries Investigations Series 2, 21(6). 51 p.
Vikebø F., Sundby S., Ådlandsvik B., and Otterå OH. 2005. Impacts of a reduced THC on growth, distribution and recruitment of Arcto-Norwegian cod. Fisheries Oceanography (In press).
