Tilt angle distribution and swimming speed of overwintering Norwegian spring spawning herring

Ingvar Huse 0 Egil Ona ] 0 0 I. Huse, and E. Ona: Institute of Marine Research , P.O. Box 1870, N-5024 Bergen , Norway [tel: When plankton production in the feeding areas decreases in the fall, adult Norwegian spring-spawning herring migrate into two fjords in northern Norway. In these wintering areas the herring occupy deeper water. Lacking the ability to refill the swimbladder they are constantly negatively buoyant. This leads to diVerent adaptive behaviour during the day and at night, behaviour which is reflected in swimming angle. Split-beam tracking methods and still-frame photography have been used to study the herring behaviour inside the dense wintering schools. Negative buoyancy seems to be controlled through constant swimming at speeds between 0.25-0.42 ms"1 because these are suYciently high to generate lift when the pectoral fins are used as spoilers. During the day, when the layers aggregate, the average swimming angle is close to horizontal while positive average swimming angles of up to 40) were recorded at night. A bimodal distribution of tilt angles, with one positive and one negative component, indicating a ''rise and glide'' swimming strategy was also observed at night. Vertically undulating split beam tracks confirmed this particular type of swimming behaviour. As adult herring are directional targets at the echo-sounder frequency used for acoustic assessment of the stock, the possible impact of the observed tilt angles on average acoustic target strength is discussed. Introduction Since 1987, the entire adult stock of spring spawning Norwegian herring has been overwintering in two fjords in northern Norway, Ofotfjord and Tysfjord (68)20*N, 17)E). The herring start to enter the fjords in October when their principal food, Calanus finmarchicus L., disappears from the surface layers on the feeding grounds after the summer bloom. They stay in the fjords until the end of January when they start their spawning migration. During the winter the herring barely feed. This period can thus be looked upon as an exercise in predator avoidance and energy conservation. The main predators are cod, saithe and killer whales, which are all predominantly visual feeders on herring. Therefore, in addition to avoiding predators by schooling during the hours of daylight, the herring also prefer to be at depth during the day, in order to avoid the surface-orientated killer whales and to be in waters with the lowest possible illumination to also avoid predatory fish. At night, at least part of the population migrates to the upper water layers. Herring schools are therefore typically observed at depths of 100400 m during the day and from 50400 m at night. Energy conservation means minimizing swimming and basic metabolism. Basic metabolism in this nonfeeding situation is mainly a function of temperature. As the temperature profile within the vertical distribution range is quite homogeneous (Rttingen et al., 1994), little can be gained in energy terms through vertical migration. Energy expenditure for herring in this particular situation therefore seems to be mainly related to swimming activity. Herring swim in order to maintain position both vertically and within the school when schooling. Being physostomous, with no gas glands or other known mechanisms of refilling the swimbladder except by surfacing, herring are neutrally buoyant at shallow depths and will expend increasing amounts of energy as their depth increases. The depth of neutral buoyancy will vary with the body density of the individual fish and the density of the surrounding water. However, at a depth of 100 m, for example, the swimbladder volume will be only 1/11, and at 200 m only 1/21, of the surface volume, thus contributing only marginally to the density of the fish. Herring at these depths are therefore likely to be negatively buoyant and must adopt special swimming strategies in order to maintain a particular depth with a minimum expenditure of energy. Their density may also increase throughout the wintering season as lipid stores are consumed for energy and during gonad development (Rttingen et al., 1994), increasing body density relative to the water masses. Echo integration is the principal method of estimating the abundance of Norwegian spring spawning herring. Since spawning stock estimates are made during overwintering, it is important to gain knowledge of the behaviour of herring in order to evaluate possible eVects of behaviour on acoustic target strength. As dorsal average target strength is very sensitive to changes in tilt angle (Nakken and Olsen, 1977; Foote, 1980), our studies were carried out to evaluate potential bias in stock estimates stemming from such changes in this rather special overwintering situation. Materials and methods The investigations were carried out in the wintering area during two surveys in December 1993 (R/V Michael Sars) and January 1994 (R/V Johan Hjort). A steel T-frame holding a 12) beam width, 38 kHz split-beam transducer, a horizontally mounted Photosea-1000 camera with an optional time lapse unit with an Osprey flash gun, and a Simrad FS-3300 Scanning Sonar, was lowered by cable into the herring schools (Fig. 1). The ships were drifting with only navigation lights on during the observation sessions, and vessel movement was logged continuously through the DiVerential Geographic Positioning System (DGPS) navigational system. The scanning sonar was mounted so that a vertical observation field, 1.7#180) parallel to the camera direction was created, and the exact depth of the rig was measured by the depth sensor in the sonar. Photographs were triggered manually during all dives except for a few dives in January when an automatic timer controlled the photography. Photographs were never taken less than 20 s apart and mostly at longer intervals. The herring were observed to scatter when the photographic flash went oV but the distribution was normally re-established in less than 10 s. The scanning sonar was used to position the T-frame within the herring shoal and to observe the herring concentration in front of the camera for manual photography. The camera was loaded with 250 frames of film for each dive, and the film (Kodak Tmax, 400ASA) was developed directly after the dive. Herring tilt angle measurements were made from the photographs, using a high resolution digitizing table. Only fish which were visually evaluated as being in a plane perpendicular to the optic axis and occupying the central third of the picture were selected for measurement. The vertical axis of the picture was determined from the image of a heavily weighted nylon line that hung permanently in front of the camera. Surface illumination readings were logged using a Li-Core LI-1000 light meter with a sensor calibrated in Micro Einstein (ME). The sensitivity of the light meter was 0.0001 ME and logged readings were 15 min averages of readings sampled every 5 s. The swimming speed of individual he (...truncated)