Do abiotic mechanisms determine interannual variability in length-at-age of juvenile Arcto- Norwegian cod? Geir Ottersen, Kristin Helle, and Bjarte Bogstad Abstract: For the large Arcto-Norwegian stock of cod (Gadus morhua L.) in the Barents Sea, year-to-year variability in growth is well documented. Here three hypotheses for the observed inverse relation between abundance and the mean length-at-age of juveniles (ages 1–4) are suggested and evaluated. Based on comprehensive data, we conclude that year-to-year differences in length-at-age are mainly determined by density-independent mechanisms during the pelagic first half year of the fishes’ life. Enhanced inflow from the southwest leads to an abundant cohort at the 0- group stage being distributed farther east into colder water masses, causing lower postsettlement growth rates. We can not reject density-dependent growth effects related to variability in food rations, but our data do not suggest this to be the main mechanism. Another hypothesis suggests that lower growth rates during periods of high abundance are a result of density-dependent mechanisms causing the geographic range of juveniles to extend eastwards into colder water masses. This is rejected mainly because year-to-year differences in mean length are established by age 2, which is too early for movements over large distances. Résumé : La variabilité inter-annuelle de la croissance chez l’important stock arcto-norvégien de Morues franches (Gadus morhua L.) de la mer de Barents est bien connue. On trouvera ici la proposition et l’évaluation de trois hypothèses pour expliquer la relation inverse observée chez les juvéniles (âges 1–4) entre la densité et la longueur moyenne à un âge donné. L’étude de données détaillées nous ont amenés à conclure que les différences inter-annuelles des longueurs à un âge donné sont déterminées principalement par des mécanismes indépendants de la densité durant la première demi-année pélagique de vie des poissons. Des influx accrus du sud-ouest entraînent le déplacement à l’âge 0 d’une cohorte importante vers l’est dans des masses d’eau plus froide, ce qui diminue les taux de croissance après l’établissement au fond. Il ne nous est pas possible de nier les effets dépendants de la densité de la variabilité des quantités de nourriture sur la croissance, mais nos données nous laissent croire que ce n’est pas là le mécanisme principal. Une autre hypothèse veut que les taux réduits de croissance durant les périodes de haute densité soient le résultat de mécanismes reliés à la densité qui amènent les juvéniles à étendre leur aire géographique vers les masses d’eau plus froide vers l’est. Cette hypothèse n’est pas retenue, car les différences inter-annuelles des longueurs moyennes sont déjà en place dès l’âge 2, âge trop précoce pour permettre des déplacements sur de grandes distances. [Traduit par la Rédaction] Ottersen et al. 65 Introduction Controlled experimental evidence has established that growth of fish is often inversely related to stock density. A change in the population size of a predator is assumed to af- fect intraspecific competition for food and thereby consump- tion and growth. An inverse relationship between growth and stock abundance is therefore indicative of density-dependent growth. Density-dependent growth in fish populations has been reported since early in the 20th century and further evi- dence has accumulated as stocks have become overexploited (e.g., Raitt 1939; Marshall and Frank 1999). However, other studies have found no significant correlation between growth and stock abundance (e.g., Daan 1974). It should also be noted that even when such a correlation is observed in field data it could be caused by a third variable, e.g., tem- perature acting on both stock abundance and growth (Jørgensen 1992). Individual growth in fish depends on density-dependent factors, such as the availability of prey. However, growth is also the integration of a series of processes (feeding, assimi- lation, metabolism, transformation, and excretion), the rates of which are all controlled by temperature (Michalsen et al. 1998). Brander (1995) examined 17 North Atlantic cod stocks and found that most of the variability in growth among stocks could be attributed to variation in temperature. The Arcto-Norwegian (or Northeast Arctic) cod spawn in March–May in patchy areas off northern and mid-Norway, with spawning mainly taking place in the Lofoten region (Fig. 1). Spawning occurs at the same sites each year, but spawning depth varies considerably, following the thermocline between the cold water of the Norwegian coastal current and Can. J. Fish. Aquat. Sci. 59: 57–65 (2002) DOI: 10.1139/F01-197 © 2002 NRC Canada 57 Received 20 March 2001. Accepted 22 October 2001. Published on the NRC Research Press Web site at http://cjfas.nrc.ca on 14 January 2002. J16271 G. Ottersen, 1,2 K. Helle, and B. Bogstad. Institute of Marine Research, P.O. Box 1870, Nordnes, N-5817 Bergen, Norway. 1 Corresponding author (e-mail: geir.ottersen@bio.uio.no). 2 Present address: University of Oslo, Department of Biology, Zoology Division, P.O. Box 1050, Blindern, N-0316 Oslo, Norway.