J. Fish zyxwvuts Biol. zyxwvutsr ( zyxwvuts 1990) 36,6 1 5-6 1 6 zyxwvut Comparison of the social behaviour of an endemic Lake Malawi cichlid in the field and laboratory R. L. ROBINSON AND G. F. TURNER University College of North Wales, School of Biological Sciences, Bangor, Gwynedd LL57 2U W, U,K. (Received 17 October 1989, Accepted30 November 1989) Key words: behaviour; Lake Malawi; Hemiiilapia oxyrhynchus Many ethological studies on cichlid fishes have been carried out in the laboratory, especially those concerning communication and parental roles. Some comparisons between field and laboratory behaviour have been made for generalized riverine species. Behavioural plasti- city is to be expected in such species (Bayliss, 1974; Fernald, 1977; Neil, 1966; Schwanck, 1987) which are frequently likely to find themselves in small crowded pools-conditions similar to those found in laboratory aquaria. Here we report some preliminary comparisons of the field and laboratory behaviour of a species which is confined to, and has evolved within, the 30 000 km2 of Lake Malawi. Hcmitilapia oxyrhynchus is morphologically and behaviourally highly specialized for feed- ing on epiphytic algae, to the extent that the genus is considered monotypic. Like all other L. Malawi haplochromines, it is a maternal mouthbrooder, and it is strongly sexually dichromatic when breeding (Fryer & Iles, 1969). Field observations were carried out by snorkelling on Chembe beach, in front of the Cape Maclear field station. It was found that individuals forage, mainly on epiphytic algae but occasionally on plankton, in large groups, and maintain a characteristic three-spot pattern, These groups showed a marked diurnal rhythm, starting to move into shallow water (1-3 m) at 10.00 hours and steadily increasing in numbers throughout the day to a peak at l6.0& 18.00 hours, before retreating to deeper water (> 3 m) to spend the night asleep. During the earlier part of the day, they could be observed feeding on the Vallisneria and Potamogeton beds on patches of sand between rocks at Otter Point, Thumbi West Island, and at the Ilala Gap. It is likely that they migrate from those areas onto Chembe beach later in the day. Males in breeding colours wandered around with the feeding shoals, and did not seem to hold fixed territories. Aggression was never observed and individual territories were non-existent. Individual H. oxyrhynchus, caught from the same locality by beach seining, were marked and acclimatized to laboratory conditions for a few days, and observed in three laboratory situations. Two individuals were kept in a 9Ox30x30cm glass tank, two in a zy 1-5 x 1.8 x 0.4 m deep concrete pond, and six in a similar sized pond. All tanks were pro- vided with a large number of Vallisneria plants which were regularly replaced. For the first zy 3-5 days the fish were non-aggressive. In groups of two they were pale, in the group of six blotchy. Generally, in the laboratory, fish were much more aggressive, spent less time feeding, and more time motionless than in the field. They never formed feeding groups. In each group of two, one fish became dominant, and adopted a black eye bar and faint vertical stripes, while the subordinate showed the normal wild pattern of blotches, es- pecially after an attack by the dominant. In the pond with six fish, a complicated mixture of dominance hierarchy and territoriality was established. One fish, PG, which was always pale, ranged over most of the area, chased all the others and was rarely challenged. G was subordinate to PG, but chased the others; it was normally pale or blotchy with faint vertical stripes, but when chased became blotchy. Fish P controlled a small territory in the corner, 615 0022-1 I12:90/040615+02 S03.00/0 0 1990 The Fisheries Society of the British Isles