Studies on the feeding behaviour of flatfish

Abstract

The feeding tactics of seven species of flatfish have been described by sequential analysis of their behaviour. The species studied were turbot, brill, Z. punctatus. P. regius. plaice, flounder and sole. Their feeding behaviour was observed in the laboratory using five different prey species, namely mysids, shrimps, corophiids, gammarids and enchytraeid worms. Forty-eight different elements of behaviour were recognised; the bothids had the most diverse behavioural repertoire, exhibiting 43 elements, the soleids were least diverse exhibiting 24 different elements and the pleuronectids occupied an intermediate position with 30. Frequencies and durations of behavioural elements were analysed and transition matrices and flow charts were presented to demonstrate the quantitative behavioural differences between families and species. Elements of behaviour were categorised as water column activity, bottom activity or inactivity. The bothids, particularly turbot, performed more water column activity than the other tvo families. The proportion of bottom activity and inactivity was dependent on whether frequencies or durations formed the basis for comparison. The frequency -xiii of elements of bottom activity was higher than that of inactivity but the durations of elements of inactivity were higher than the durations of bottom activity. This relationship was attributed to the elements of bottom activity being high in frequency but short in duration whereas the reverse situation applied to the elements of inactivity. The tactics of the species differed considerably and were found to be dependent on prey species, the tactics of the bothids involved much more hunting and stalking because their prey were more mobile. In contrast, the tactics of the pleuronectids and soleids, whose natural prey are less mobile, could be described as hunting and cropping. The elements of behaviour exhibited made this very apparent. The observations on the feeding behaviour of turbot were used to determine the importance of various prey stimuli in prey recognition. The response to selectively presented models and food cues was assessed quantitatively by a simple scoring method. Turbot preferred moving prey with a ratio of verticalshorizontal components of body shape of about 1:10. These attributes correspond well with those of the natural prey which constitutes the fishes' diet. Appendage movements were also found to be important but the general characteristics of body shape were unimportant. Inconspicuous cryptically coloured and translucent models were preferred to conspicuous ones. Turbot were found to be visual feeders and olfaction was unimportant in recognising prey. These results are discussed in relation to other work on flatfish and sticklebacks.