Search

Age-structured population dynamics models play an important role in fisheries assessments. Such models have traditionally been estimated using crude likelihood approximations or more recently using Bayesian techniques. We contribute to this literature with three main messages. Firstly, we demonstrate how to estimate such models efficiently by simulated maximum likelihood using Laplace importance samplers for the likelihood function. Secondly, we demonstrate how simulated maximum likelihood estimates may be validated using different importance samplers known to approach the exact likelihood function in different regions of the parameter space. Thirdly, we show that our method works in practice by Monte Carlo simulations using parameter values as estimated from data on the Northeast Arctic cod (Gadus morhua) stock. The simulations suggest that we are able to recover the unknown true maximum likelihood estimates using moderate importance sample sizes and show that we are able to adequately recover the true parameter values. Les modeles de dynamique de population structures d'apres l'age sont tres utilises dans les evaluations halieutiques. Ces modeles sont traditionnellement estimes en utilisant des approximations de vraisemblance grossieres et plus recemment des techniques bayesiennes. Nous ajoutons trois messages principaux a cette litterature. Nous demontrons d'abord comment estimer de tels modeles de maniere efficace par vraisemblance maximale simulee a l'aide d'echantillonnages d'importance de Laplace pour la fonction de vraisemblance. Deuxiemement, nous demontrons comment les estimations simulees de vraisemblance maximale peuvent etre validees a l'aide de differents echantillonnages d'importance connus pour s'approcher de la fonction de vraisemblance exacte dans differentes regions de l' espace des parametres. Troisiemement, nous montrons que notre methode fonctionne en pratique par des simulations de Monte Carlo en utilisant les valeurs des parametres estimes a partir de donnees sur le stock de morues (Gadus morhua)de l'est de l'Arctique. Les simulations laissent croire que nous sommes capables de retrouver les estimations veritables de vraisemblance maximale inconnues en utilisant des tailles moderees d echantillons d importance et montrent que nous pouvons retrouver adequatement les vraies valeurs des parametres. [Traduit par la Redaction], Introduction Assessment of fish stocks is a challenging field of statistical modelling. As described in Gudmundsson (1994), the main type of available data are catch-at-age, giving measurements of the catch [...]

: The Barents Sea stock of capelin (Mallotus villosus) has suffered three major collapses (>90% reduction) since 1985 due to recruitment failures. As capelin is a key species in the area, these population collapses have had major ecosystem consequences. By analysing data on spawner biomass and three recruitment stages (larvae, 0-group, and 1-year-olds), we suggest that much of the recruitment failures are caused by predation from herring (Clupea harengus) and 0-group and adult Northeast Arctic cod (Gadus morhua). Recruitment is furthermore positively correlated with sea temperatures in winter and spring. Harvesting of maturing capelin on their way to the spawning grounds reduced the abundance of larvae significantly, but this reduction to a large extent is compensated for later in life, as mortality is strongly density-dependent between the larval stage and age 1. Altogether, our study indicates a very high importance of trophic interactions, consistent with similar findings in other high-latitude marine ecosystems. Resume: Le stock des capelans (Mallotus villosus) de la mer de Barents a connu trois importants effondrements (reduction de >90 %) depuis 1985, a cause d'echecs du recrutement. Parce que le capelan est une espece essentielle dans la region, ces effondrements demographiques ont eu des consequences majeures sur l'ecosysteme. Apres une analyse des donnees de biomasse des reproducteurs et de celles des trois stades du recrutement (larves, poissons d'age 0 et poissons d'un an), nous croyons qu'une partie importante de l'echec du recrutement est causee par la predation par les harengs (Clupea harengus) et par les morues (Gadus morhua) d'age 0 et adultes. Il y a, de plus, une correlation positive entre le recrutement et la temperature de la mer en hiver et au printemps. La recolte des capelans en etat de maturation durant leur migration vers les frayeres reduit le nombre de larves de maniere significative, mais cette reduction est en grande partie compensee plus tard dans le cycle parce que la mortalite est fortement reliee a la densite entre le stade larvaire et l'age 1. Notre etude indique globalement une tres forte importance des relations trophiques, ce qui concorde avec des observations similaires faites dans d'autres ecosystemes marins de haute latitude. [Traduit par la Redaction] Introduction, It is well recognized that in addition to fishing, fish stocks are affected by interactions with other species (predators, prey, and competitors) and by climate, which therefore ideally should be [...]

The population cycles of rodents at northern latitudes have puzzled people for centuries (1,2), and their impact is manifest throughout the alpine ecosystem (2,3). Climate change is known to be [...]

Many methods exist to assess the fishing status of data-limited stocks; however, little is known about the accuracy or the uncertainty of such assessments. Here we evaluate a new size-based data-limited stock assessment method by applying it to well-assessed, data-rich fish stocks treated as data-limited. Particular emphasis is put on providing uncertainty estimates of the data-limited assessment. We assess four cod stocks in the North-East Atlantic and compare our estimates of stock status (F/Fmsy) with the official assessments. The estimated stock status of all four cod stocks followed the established stock assessments remarkably well and the official assessments fell well within the uncertainty bounds. The estimation of spawning stock biomass followed the same trends as the official assessment, but not the same levels. We conclude that the data-limited assessment method can be used for stock assessment and that the uncertainty estimates are reliable. Further work is needed to quantify the spawning biomass of the stock. [ABSTRACT FROM AUTHOR]