Spatial orientation in salmon

Abstract

Salmon migrate to sea and return after several years to spawn in their natal stream. The fish are probably guided in the final stage of their journey by the odour/taste characteristics of their home water. But we have no viable hypothesis to explain how salmon orientate during many thousands of kilometres of oceanic travel and later find their way back to their home estuary. The purpose of this work was to determine which external stimuli may be used as a reference for spatial orientation in mid ocean. A group of eighteen young chinook salmon (Oncorhynchus tschawytscha Walbaum) were collectively trained to associate food with water flowing from a particular direction (270°) and then individually transferred to an unfamiliar hemispherical test arena in which the presence of cues/clues that may be used as a reference for direction could be controlled. Directional behaviour was observed by infra red video camera by day and at night and recorded on a video tape recorder for subsequent computer analysis. Appropriate controls were established and the directional preference of each fish before, during and after the test were identified to quantify the orientation response to the presence or absence of external stimuli. The mean of means orientation fox twelve fish was 264°. Preliminary study showed a response to a wide range of cues/clues including; daylight, overhead and underwater landmarks, light, sound, odour. Loss of a cue caused disorientation (with respect to the direction associated with food) for a short period, but the fish soon reverted to the learnt direction. Thereafter the test was changed and all known cues were first removed, the response to the re-introduction of a cue was observed. When the geomagnetic field was neutralised using Helmholtz coils the five fish tested showed random orientation. During a subsequent test the North/South component of the horizontal geomagnetic field was artificially shifted 90° clockwise. The mean of means orientation of twelve fish in this test changed from a bimodal mean of 258°/78° to a bimodal mean of 354°/174°. After restoration of the normal magnetic field the orientation reverted to 274°/94°. Medical studies show a loss of memory after administration of general anaesthetic. Five of the fish, used previously, and another group of naive fish were anaesthetised with MS222 and separately transferred to the test arena. All of these fish showed disorientation (with respect to the learnt direction) some for periods up to 17 days. Two months later all these fish had reverted to approximately 270°. By inference the preferred orientation was 'learnt' and not 'innate'. It could not be asserted with certainty that the test arena was devoid of 'local' cues. So after two years the remaining twelve fish were transferred by air, in light-proof containers, to Christchurch which is 800 km. south of Auckland and retested in a completely new re-designed test arena to establish the preferred orientation in the absence of any of the 'local' cues present in Auckland. The directional preference of five fish was tested in artificial light and at night in a light-proof room. A further five were tested at night in a test arena (sealed with transluscent plastic to prevent sight of overhead cues) situated outside the test building. The mean of means bimodal orientation for ten fish was 273°/93° which falls within the confidence limits (218° <-> 285°) established in Auckland. In artificial light and in 'no visible light' conditions the learnt direction remained virtually unchanged. Six of these fish were subsequently used to measure the orientation response to changes in the anistropic radiance of sunlight. Each fish had been exposed to sunlight in Auckland for two days prior to the move to Christchurch. Individual fish were transferred to a test arena (sealed with transluscent plastic cover) situated in a field, outside the light-proof room, and re-tested in sunlight and at night. The mean axis of orientation (273°/93°) previously recorded changed to a mean of means axis of orientation of 001°/181°. A strong unimodal orientation of 013° (r=0.36) was also recorded. Coincidently, the map direction from Christchurch to Auckland is 016°. These fish apparently imprinted on the Sun (Azimuth, altitude or daylength) within a 2 day period at the age of two years; and by implication detected either displacement or angular error. Radial statistics provide the best available measure of orientation but the resultant mean says very little about animal behaviour contributing to this mean. It is moreover a view over one particular period of time. This study showed that orientation behaviour changes during a day and between days. So the Christchurch data were later re-analysed using a computer program to link all of the direction observations for one fish to create a pseudo track. The resultant diagram suggests that bimodality is indicative of a foraging behaviour with the fish moving forwards and backwards along the bimodal axis (270°/090°). However, the unimodal track, judged random by the Rayleigh test, showed significant directivity. Radial statistics apparently record two distinctly different forms of orientation behaviour. Subsequent analysis by a computer program showed that the fish maintained direction by equalising the number of turns (in degrees) left and right of the preferred mean direction. This behaviour is unlikely to be conscious inertial guidance but it is based on perception of preference for direction. The view of the aerial world as seen from below water in a river and at sea was observed and photographed using Scuba gear. The 97.6° cone (Snell's window) formed by refraction of light at the sea/air interface means that the sun (as an object) is not seen from beneath the waves during early morning and late afternoon at times determined primarily by the altitude of the sun and the latitude of the observer, and also by variability in sea conditions. The sun was clearly visible in Snell's window from a depth of 10m. between midday on the day of the autumn equinox but was not visible at midday on the winter solstice (latitude 37° South). Light refraction does not seem to affect anistropic light radiance. Riverine fish maintain mid-water position by reference to fixed external cues/clues. Reliance on the sun as an unchanging, angular reference will inevitably cause a change in the geographical position of the observer as the sun's position changes North/South with season. Subject to the qualification that this behaviour of a small sample is not representative of the whole species, it was concluded that:- 1. The geomagnetic field was used as a reference for directional orientation. 2. The sun position was located by detection of anisotropic radiance. 3. The complementary nature of the angular relationship between gravity, geomagnetic field and the sun ecliptic provides a tri-coordinate frame of reference which is sufficient for spatial orientation. 4. Mean position and mean direction were maintained by equalising the number of turns made either side of a preferred position or direction.