Search

The Baltic Sea is characterised by a heterogeneous oceanographic environment. The deep water layers forming the habitat of Baltic cod (Gadus morhua callarias L.) are subjected to frequently occurring pronounced anoxic conditions. Adverse oxygen conditions result in physiological stress for organisms living under these conditions. For cod e.g. a direct relationship between oxygen availability and food intake with a decreasing ingestion rate at hypoxia could be revealed. In the present study, the effects of oxygen deficiency on consumption rates were investigated and how these translate to stock size estimates in multi-species models. Based on results from laboratory experiments, a model was fitted to evacuation rates at different oxygen levels and integrated into the existing consumption model for Baltic cod. Individual mean oxygen corrected consumption rates were 0.1–10.9% lower than the uncorrected ones. At the currently low predator stock size, however, the effect of oxygen-reduced consumption on the total amount of eaten prey biomass and thus predation mortalities was only marginal. But should successful management lead to higher cod stock sizes in the future, then total predation mortalities will greatly increase and thus improved precision of these estimates would be valuable for the assessment of prey stocks.

A new methodology is presented to reconstruct migration pathways of individual fish inhabiting ecosystems with moderate-to-strong gradients in temperature or salinity. The method uses measurements of ambient pressure, temperature and salinity obtained from electronic data storage tags attached to individual fish and is particularly applicable in areas with negligible tides. We demonstrate the method with Baltic cod. Hydrographic fields obtained from hydrodynamic modelling were used as a geolocation database to identify daily positions of Baltic cod by comparison with the environmental data collected by the tags. Using randomly distributed individual parameter perturbations in the range of the instrument precision of the tag we simulated a cod migrating through the Baltic Sea. The distance between the prescribed and geolocated positions of this artificial cod was on average 2.9 ± 4.7 (SD) km. Subsequently, the method was used to reconstruct migration routes of 10 real cod tagged in the Bornholm Basin of the Baltic Sea in early spring 2003. Differences were compared between the tag data and the geolocation database. The uncertainty in geolocation at recapture day was on average 75 ± 23 (SD) km, as shown by comparison between geolocated position and recapture position.