Skrei fisheries are found from Romsdal to Nordland, and much less commonly south to Bergen or farther north in Finnmark. Georg Ossian Sars began his studies of the Lofoten cod fishery in 1864, where he had been posted to determine the causes of fluctuations in the yield, which had been disappointing in the preceding decade. He distinguished it from the Finnmark fishery by noting that it is a gaat-fiske or spawning fishery, but noted there are other spawning cod fisheries on the coast, at least from Stadt northwards. The Finnmark fishery took juvenile cod. He divided the parts of the life cycle of the "winter cod" which he was able to observe in the Vestfjord into three phases, the "coming in", the spawning, and the "going out".

Sars was told that cod eggs float in the water. Initially sceptical, since at that time all fish were thought to spawn on the bottom, he soon found that this was indeed the case, and is often credited with the discovery of planktonic fish eggs. The following year, 1865, he was able to hatch both wild caught eggs as well as some obtained by artificial fertilization and maintain the larvae alive in jars until the yolk sac was resorbed. He speculated that the dispersion of the eggs and larvae by winds and currents might have something to do with the well known "fishing periods" of which he was seeking the causes. This was at least an inkling of the now widely accepted view, principally associated with the name of Johan Hjort, that fish year class strength is determined during the first year of life.

He later (1866) found that the young fish fed initially on Calanus and Temora, that at a slightly later stage (mort, smågjed) they sought shelter (together with young haddock) under drift and under medusae (Cyanea, Aurelia), and ate the hyperiid parasites of the medusae, and that these young cod in turn were eaten by pollack and other fish, and by gulls. In subsequent years, Sars followed the changing habits of cod from the pelagic to the benthic stage, and described the change in diet, first to larger zooplankton such as euphausiids, and later to bottom living crustaceans and polychaetes.

He also concluded that different kinds of cod recognized by the fishermen, such as "algae-cod", "sandeel-cod" and so on, were all juveniles of the winter cod and owed their distinctive colour patterns and body forms to their habitats and diets. "I proved one important fact which will play a prominent part in all future investigations of the cod-fisheries, viz: that the small form of codfish, which, under different names (algae-fish, bottom-fish, &c.), is found on our northern and western coasts all the year round, is not, as was formerly thought, a separate variety of the codfish, but the offspring of the winter codfish." (report for 1870).

An important problem which Sars was unable to resolve concerned the whereabouts of the winter-fish outside the spawning season. He at first guessed that they lived in deeper water close to land, some perhaps in the Vestfjord itself, others on the outer banks; he seems initially to have come to this conclusion by analogy with earlier (eighteenth century) arguments concerning the migrations of the herring, arguments in which the notion of very extensive migrations (from and to the Arctic) had been demolished. Each subsequent year, his annual reports pushed the whereabouts of the winter cod outside the spawning season farther and farther offshore, but the idea we now have of extensive migrations in the Barents Sea escaped him. Sars later encountered a rich (but as it subsequently turned out short-lived. This fishery lasted from 1873 to 1882, and yielded about 600,000 fish in 1879., its peak year. The catch by 18 Nowegian vessels there in 1883 amounted to 3 fish!) cod fishery off NW Svalbad during the Norske Nordhavs Expedition in 1878. These fish were not in spawning condition; from their appearance Sars suggested they might have come from Lofoten, but they might equally have been a local colony, analogous to the stock of Icelandic cod found in West Greenland in the mid-twentieth century. Sars was also unable to age the cod since no method for doing so existed at that time, and did not manage to proceed with the problem of how many kinds of cod there were.

A most important legacy of the period in which the pioneering studies of Sars were being conducted is the series, still continuing, of landing statistics. These began in 1860 (Aresberetning Vedkommende Norges Fiskerier, Offisiele Statistikk), and for winter cod, initially recorded the numbers of fish (later given by weight) taken separately with lines, hand-lines, and gill nets, as well as the quantities of liver and roe and the numbers of codheads, and their prices. Trawling began in the 1920s. This information was published for more than 20 localities between Lister and Jaederen og Dalene in the south to Varanger and Vardø in the northeast, and figures are weekly and annual. This time series has formed the basis of all subsequent analyses of long term trends in this fishery, some examples of which are referred to below. An important result derived from the weekly data is the mean date of capture of skrei in the Vestfjord, which can be used as a proxy for the mean spawning date. Cushing (1969) analysed this information for the period 1894 to 1967, and calculated the standard deviation as only 4 days. Thus for a period of 64 years, winter cod always spawned in the Vestfjord in the same Julian week.

The physical oceanography of the Norwegian Sea, based mainly on results obtained by the Michael Sars (named for G.O.’s father) from 1900 to 1904, was synthesized as then understood by Helland-Hansen and Nansen (1909). In chapter VII of this monograph, an attempt was made to compare annual variations in the Atlantic Current with variations in the fisheries, in harvests, and in tree growth. It was stated that, "We have also been able to trace a certain relation between these variations and cosmic causes" (my italics), by which they meant the sunspot cycle. Cosmic causes were frequently invoked during the nineteenth and early twentieth century to account for fishing periods, especially in relation to the herring fisheries (see appendix). This monograph provides valuable background to the simultaneous biological studies of Hjort.

Hjort (1905) had noticed parallels between the yields of the Lofoten fishery and the conditions of the roes and liver; in years when yields were low, the liver might even sink in seawater, and the sexual organs be undeveloped. Helland-Hansen and Nansen wrote that "...there is a close relation between the yearly variations in the conditions of the cod (as well as some other food fishes), and also the time of its arrival at the banks, and the yearly variations in the mean temperature of the Atlantic water as found in the Sognefjord section of May of the preceding year." They did not pursue the matter much beyond that, and it was soon to appear that the procedures used to analyse the Sognefjord data were marred (Einar Lea, 1913). Nansen (1913) himself in a study of the so called "Irish Current", a slope-hugging poleward current, recognized that the the rôle of the Atlantic drift might have been exaggerated.

Johan Hjort, like Sars, was commissioned by the Norwegian government to investigate the causes of fluctuations in the fisheries, again following a series of years with poor yields. His great advantage was that he could use the Michael Sars, while G.O.Sars’ field studies in Lofoten had been done with rowing boats. Hjort (1914) and his colleagues made considerable advances in knowledge of the life history of cod in the first decade of the 20^th century. Studies of cod scales (Damas, 1909) showed that there were many more year classes than had previously been thought, that the relative abundance of these year classes was very variable, and that the usual age of first maturity for the Lofoten cod is between six and eight years. Hjort showed: i) (contra Victor Hensen, who taught that fish spawned promiscuously and everywhere) that most cod between Romsdal and Nordland spawned in very restricted areas with depths of about 50 fathoms, ii) by tagging experiments that cod are able to "in the course of 5-6 weeks, cover the distance from Lofoten to Finnmarken, a journey of between 800 and 900 km", and by length at age comparisons iii) that loddefisk (a Finnmark name for the juvenile cod taken there, which feed on capelin or lodde) and skrei belong to the same stock, as do larger fish in the open Barents Sea. Fluctuations in the quality of the liver were shown to be "simultaneous, and of like direction and degree" throughout northern Norway from Romsdal to Tromsø (and the trends not related to the sunspot cycle), which led him to the conclusion that large scale features of the climate and oceanography controlled the population ecology (a conclusion earlier reached by Sars vis a vis the spring herring of western Norway).

Hjort’s tagging results were later confirmed by Oscar Sund, who was also able to demonstrate the return of cod from the Tromsø district to Lofoten, and link this return to a favorable southward flowing current on the shelf noted by Jens Eggvin in oceanographic sections. The early work on aging cod was amplified by Gunnar Rollefsen who was the first to examine cod otoliths. Rollefsen’s sampling of skrei from Lofoten in 1932 revealed that the fish ranged from 6 to 22 years old with 10-15 year old fish most abundant, that the age of first maturity was between 6 and 15 and predominantly 10-11 years, and that the spawning fish suffered an annual reduction in numbers of 35-40%. It also emerged that the loddefisk from Finnmark, although not yet mature, could be as old as fifteen.

Hjort’s (1914) conclusions have had a most powerful influence on fisheries research ever since their publication. He wrote that "...the numerical value of a year class is apparently determined at a very early stage, and continues in approximately the same relation to that of other year classes throughout the life of the individuals." He argued that variations in year class strength were not due to variations in the quantities of eggs spawned (based on analysis of roe weights), and that that they were most likely due to annual fluctuations in the amount of plankton available as food for the larvae, and to dispersal patterns. The good 1903 and 1904 year classes, both associated with a late skrei fishery in Lofoten, led him to suggest that later spawned larvae were better matched to their food supply, and thus laid the foundations of the match/mismatch hypothesis which has been so successfully exploited subsequently (Cushing, 1990).

Ottestad imposed his chosen cycles on the Lofoten cod time series. A more conventional approach is to analyse time series to discover what periodical signals they might contain. Fourrier analysis is the most widely applied means of doing this, but fails to identify nonlinear signals, for which purpose MESA (maximum entropy spectral analysis) is preferred. Robert Currie has pioneered this technique for the analysis of a great variety of biological, geophysical, and economic time series. In the vast majority of these time series, two signals dominate, with periods of about 19 and 10-11 years, with the former usually the strongest. These are generally identified with the luni-solar or nodal tide and the sunspot cycle respectively. Estimating the error on period lengths obtained by MESA is a bit of an art form, but is probably around ± 2 years in the present context, and for shorter multiannual cycles; thus Ording’s tree ring cycle of 17.5 years might possibly be equated with the nodal signal, and perhaps his 23 year cycle too. His 57 year cycle is plausibly attributed to the Gleissburg cycle (i.e., variation in the solar constant), but internal oscillations of the ocean-atmosphere system provide a possible alternative mechanism (Schlesinger and Ramankutty, 1994). The most recent attempt to clarify the role of the nodal tide in relation to the Nowegian cod is due to Harald Yndestad (1999), who has demostrated an 18.6 year signal in recruitment estimates.

The early results mentioned above, that skrei visit the same banks each year to spawn, and that they time these visits very precisely, provide a logistic framework for UVAC field studies; we know where and when to find the eggs and larvae. Hjort’s hypothesis, that year class strength is determined during the first few months of life, indicates that the causes of egg and larval mortality must be known if we are to improve our understanding of recruitment processes. Our hypothesis is that variation in UV is one of these causes.

Speculations about periodic forcing, although not always regarded as front page news, have persisted in the fisheries literature for about 150 years. There might be something in it! From the point of view of my own input to UVAC, one leading question is: do time series of UV radiation reaching the Vestfjord contain multiannual signals? If so, what are the possible mechanisms that transmit such signals to the cod? Is the link a direct effect on the eggs and larvae, or does it operate through complex and nebulous meteorological events?

Currie, R.G., Wyatt, T., & O’Brien, D.P., 1993. Deterministic signals in European fish catches, wine harvests, and sea-level, and further experiments. International Journal of Climatology, 13: 665-687

Cushing, D.H., 1969. The regularity of the spawning season in some fishes. Journal du Conseil, Conseil International pour l’Exploration de la Mer, 33: 81-92

Cushing, D.H., 1990. Plankton production and year-class strength in fish populations: an update of the match/mismatch hypothesis. Advances in Marine Biology, 26: 249-293

Helland-Hansen, B., & Nansen, F., 1909. The Norwegian Sea, its physical oceanography. Report on Norwegian Fishery and Marine-Investigations, II (2), 233pp

Hjort, J., 1914. Fluctuations in the great fisheries of Northern Europe. Conseil permanent international pour l’Exploration de la Mer, Rapports et Proces-Verbaux, 20:

Izhevskii, G.K., 1961. Okeanologicheskie osnovy formirovaniya promyslovoi produktivnosti morei (Pishchepromizdat), ("Oceanological principles as related to fisheries productivity in the seas")