Trophic interactions affecting a key ecosystem component: a multi-stage analysis of the recruitment of the Barents Sea capelin

• Main Authors: Hjermann, Dag Ø., Bogstad, Bjarte, Dingsør, Gjert Endre, Gjøsæter, Harald, Ottersen, Geir, Eikeset, Anne Maria, Stenseth, Nils Christian
• Format: Report
• Language: English
• Published: ICES 2009
• Subjects: capelin; lodde; stock collapse; bestandssammenbrudd; VDP::Agriculture and fishery disciplines: 900::Fisheries science: 920::Resource biology: 921; Arctic; Barents Sea; Gam; Arctic cod
• Online Access: http://hdl.handle.net/11250/102938

In the Barents Sea, capelin is a key food item for the North-East Arctic cod stock. This capelin stock has had very unstable population dynamics since 1985, with recruitment failures leading to three major collapses of the stock (>90% reduction of the stock size), resulting in decreased growth and survival of cod. Here we analyze in detail how predation and harvest affects the recruitment of capelin, using data on three different stages (i.e., larvae, zero-group and 1-year-olds) through the first 1.5 years of the capelin's life. We demonstrate that both herring predation (on capelin larvae) and cod predation (both on spawners and on offspring) has had major negative effects on capelin recruitment. Mortality is furthermore demonstrated to be strongly density-dependent, and is lower when temperatures are high – probably due to higher food availability for capelin. Harvesting maturing capelin on the way to the spawning grounds did affect the production of larvae, at least during the first half of the 1980s. However, the reduced production of larvae appears to a large extent to have been compensated by decreased density-dependent mortality on later life-stages, resulting in only minor effects on the abundance as 1-year-olds. Altogether, our study points to the importance of trophic interactions in determining the dynamic structure in high-latitude marine ecosystems. Keywords: stock collapse, predation, trophic interactions, harvesting, density-dependent mortality, Generalized Additive Models (GAM)