The sensory basis of fish schools: Relative roles of lateral line and vision

Abstract

Quantitative analyses of the effects of temporary blindfolding (BL) or lateralis section (LS) on the schooling performance of saithe (Pollachius virens) are reported. Comparison of effects of the sensory deprivations on school structure and dynamics allow determination of the relative roles of the lateral lines and vision in normal schooling. Extended 3-dimensional records were made of the positions of sensorily deprived individuals in normal schools as well as whole schools with sensory treatments. 1. Blinding had little effect on the position experimental fish took up with respect to their neighbors within the school (Fig. 2). In contrast, lateralis section resulted in a great increase in the frequency of neighbors at 90° bearing (directly alongside) (Fig. 2). 2. Both BL and LS fish exhibited different characteristic nearest neighbor distances (NNDs) than controls, with blinding increasing NND and lateralis section decreasing NND (Fig. 4). 3. Importance of the lateral line in transmission of a fright response and sudden velocity changes within a school is demonstrated by experiments in which schools were intentionally startled. In schools of LS fish there exist significant relationships between latency to startle and the distance and angle between a fish and the startling object. Such relationships are not seen in normal schools: with intact lateral lines, fish respond with similar latencies to objects in all directions (Fig. 5). 4. Fish which are blindfolded are able nonetheless to match short-term changes in velocity of their neighbors and actually show higher correlations than do controls (Fig. 8 b). Lateralis section does not reduce correlations between fish and neighbors (Fig. 8 d) probably because LS fish take up positions at which they can best determine neighbors' velocities by vision alone.