Distribution and activity of pelagic fish - acoustic studies in the Baltic Sea

Sammanfattning: Fisheries agencies around the Baltic Sea use hydroacoustics to assess stock sizes of herring and sprat. These assessments rely on the assumption that the acoustic properties of Baltic clupeids are the same as North Sea herring. This may lead to biased results, as system-specific differences in salinity and fish fat content may influence acoustic target strength. The acoustic properties of the Baltic clupeids were explored and a new relationship between target strength and fish length was developed (paper I). The intercept of this new relationship is 3.4 dB higher than normally used by the agencies. Applying this new relationship would reduce the acoustic stock biomass estimates by approximately 50%.Diel variation in the vertical distribution of fish may influence acoustic abundance estimates. Fish body posture may influence target strength and if fish concentrate near the surface or bottom this may affect detectability by an echo sounder. A seabed-mounted, upward pinging echo sounder was used to study diel variation in vertical distribution, acoustic size distribution and abundance of fish (Paper II). Differences between day and night were substantial and it was concluded that night time acoustics are to be preferred, at least in our study area. The seabed-mounted echo sounder was also used to study fish swimming activity and vertical distribution in relation to light intensity and water temperature (paper III). Four phases of fish distribution were distinguished over the diel cycle (day, night, dawn and dusk). Acoustic tracking was used to estimate the swimming speed of individual fish. The speed varied among the diel periods and the greatest difference was observed between day and night with twice as high swimming speed during the day. Regression models were developed to investigate the effects of fish size and environmental factors (water temperature, light intensity at the sea surface and in situ, measured at the depth of the fish) on swimming speed. Fish size, light intensities and temperature were all significant variables in the models, with fish size being generally most important. These results have clear implications for fish bioenergetics models. Such models should account for seasonal, light-driven cycles in the activity-induced respiration estimates, in particular when modelling populations at high latitudes.Vertical and horizontal fish distributions were studied from spring through autumn during two consecutive years (paper IV). The seasonal dynamics in vertical distribution patterns were consistent between years. Prior to thermocline formation, fish of all sizes concentrated near surface where water temperatures were higher than in the underlying water mass. During the summer period of pronounced thermal stratification, larger fish were found deeper than small individuals (including young-of-the-year fish), which remained close to the surface. In the autumn, when the thermal stratification diminished, the small fish moved somewhat deeper while larger individuals dispersed throughout the water column. Fish showed clear horizontal patchiness, but horizontal distributions were not significantly related to wind directions.