CCAMLR

In order to describe an aquatic ecosystem scientists require a variety of information. Some of this information can be obtained using underwater acoustics. As examples acoustics is often used to provide information about biomass densities, biological species compositions, size distributions, behavior in space and time, and links to other environmental information such as habitats. Examples of underwater acoustic instrumentation used by scientists to obtain such information are scientific multi frequency split beam echo sounders, multi beam echo sounders, omni directional sonars, 3D matrix sonars, and various types of trawl gear instrumentation. Data are often obtained from fixed installations on board dedicated research vessels, but are sometimes also obtained from various kinds of moored installations, acoustic landers, remote buoys, autonomous underwater vehicles (AUV’s), and commercial fishery vessels. In this paper a variety of underwater acoustic technologies are presented to serve as input for the discussion of what kind of underwater acoustic instrumentation is needed for both research vessels, potentially commercial fishing vessels, and other types of installations in order to obtain improved information about the marine ecosystem in the Antarctic.

Fishery for krill is a major economical activity in the Antarctic Ocean. The increased Norwegian fishery inherently gives Norway a responsibility to contribute to the management of the marine resources in the southern ocean. During the International Polar Year 2008, RV “G.O. Sars” spent 3 months in the Antarctic Ocean to do investigations on euphausiids and other key components of the ecosystem. Major acoustic activities were to acoustically identify krill and estimate specimen size, investigate behaviour, measure target strength in situ, and of course verify the acoustic measurements biologically. Krill were identified from the relative frequency response of a 6-frequency hull-mounted echo-sounder system, and specimen size was estimated acoustically by means of several acoustic scattering models implemented in an optimized framework in the post-processing system LSSS.

The spatial distribution characteristics of ice-fish in the near-bottom layer are analyzed using the acoustic data obtained in the South Georgia area during the Russian bottom trawl surveys 2002. Acoustic and trawl density samples obtained in the near-bottom layer covered by trawl survey are compared. It is shown the possibility to estimate the bottom trawl efficiency. The high heterogeneity of horizontal and vertical fish distribution in the near-bottom layer of the survey area is shown. Influence of this fish distribution heterogeneity on trawl surveys efficiency and reliability is analyzed. In addition, the author discuses the results of bottom trawl surveys carried out in the South Georgia area during season’s 2000 and 2002. It is shown, that observed inter-seasons and inter-vessels differences between fish biomass estimates from bottom surveys may be to a considerable degree stipulated by heterogeneity of fish distribution in the near-bottom layer relative to the fishing gear operation zone (Russian trawl HEK-4M and UK trawl FP120) and the trawl station location rather than by the variability of the stock state.

An anatomically detailed acoustic scattering model has been used to estimate the backscattered target strength (TS) of six mackerel icefish (Champsocephalus gunnari) of total length 26.8–35.0 cm, at 38 kHz. The model used computed tomography (CT) scans of fish to determine the sound speed and density throughout the fish and simulated the complex interaction of the incident sound wave with the acoustic impedance contrasts throughout the fish. A preliminary length to tilt-averaged TS relationship is derived from the model results: 148)(log4.7010-=LTSwhere L is the fish total length in cm and is applicable to fish of length 26–35 cm. Existing in situ TS estimates are for smaller fish (16–26 cm) and hence are not directly comparable, but extrapolation of the in situ and model results reveal a large different in TS (6 dB at 26 cm) between the two datasets.