mise à jour du
17 septembre 2007
 Animal behaviour monographs
Ethology of the bicolor damselfish
Eupomacentrus partitus
Myrberg A


The ethogram
Research began with a qualitative description of the behaviour shown by members of the field and laboratory colonies. It soon became apparent that the various activities carried out by these fish were but a series of numerous, rather discrete, species-specific, behavioural units which were sufficiently stereotyped in form, duration and orientation, that their occurrence could easily be noted by trained observers. These units, in turn, provided the basic material from which arose all subsequent analyses, interpretations and experimental verification of certain points of the study.
There were three basic types of be units: motor patterns (i.e. locomotory movements and postures), colour patterns; and sounds. Although colour patterns varied in duration, motor patterns and sounds were considered instantaneous events since the duration of a given event was generally less than I s (there were exceptions, e.g. Chasing, Skimming, HeadStanding; but even these were usually only a few seconds in duration). All patterns, irrespective of type, entered the ethogram since all were important in reaching an understanding of the intricate mechanisms that controlled the behaviour of the subject species.
Behaviour patterns within the ethogram were not analysed, quantitatively, as to their respective variations in intensity or duration (as carried out, for example, by Rasa 1969, on the Lateral display of Pomacentrus jenkinsi). Though certainly possible, such analyses would have expended much time and effort for a minimum of return since patterns, at the level of our analysis, by and large, demonstrated Typical Intensity (Morris 1957; see discussion by Barlow 1963). I agree with the interpretations by Baerends (1957) and Nelson (l964a) that the elementary action patterns described in the ethogram, though complex, were relatively stereotyped and it appeared self-evident and natural to describe and record behaviour in terms of activities of that order. I do not claim that these action patterns (motor, sonic and colour) correspond to Fixed Action Patterns (Lorenz 1956) since descriptions were based largely on convenience and not upon the criteria of the F.A.P.
Any ethogram necessitates some type of ordering. This was done, mainly, by grouping action patterns together under various titles, each title having an apparent functional correlate (e.g. courtship, agonism, nest care, etc.). I realize fully the danger of such categorization since often morphological and contextual criteria result in hybrid classification (see discussion by Nelson 19 64b). Nevertheless, impressions gained during the qualitative phase formed a basis for classification and a test of the reality and significance of this classification was an important part of the subsequent quantitative phase of the study. Fortunately, the bicolor damselfish exhibited many discrete patterns. 1f it had shown but a few consistent patterns, such a classification would have been meaningless, except on the basis of given probabilities that a single event occurs within several or all functional categories.
Motor Patterns Patterns and Postures of Basic Locomotion
Fish of the field and laboratory colonies demonstrated the same basic locomotory movements and postures. Movements were often so rapid, however, that adequate descriptions were often accomplished only by reduced-speed cinematography and video-taping.
Forward Swimming was accomplished largely by the caudal fin with the pectoral fins being used mainly for changing direction. This mode of swimming, though common to the genus, Eupoinacentrus (Emery 1968) is contrasted with the 'pectoral swimming' of two other well-known genera of damselfishes, i.e. Abudefduf: Cummings 1968; Helfrich 1958; Stevenson 1963; and Chromis: Abel 1961; Myrberg, Brahy & Emery 1967; Turner & Ebert 1962. The caudal fin is generally used by the latter genera only during sudden and rapid manoeuvring. Fins were expanded to varying degrees during slow Forward Swimming, the degree of expansion depending on the swimming speed and the particular motor pattern being exhibited.
Backward Swimming was uncommon in bicolors, but a number of specific situations predictably brought it to the fore: (1) a territorial holder approaching an unfamiliar, small object at, or near, its territory (e.g. a new conspecific, a congeneric, a hand, a rod); (2) a female approaching closely a territorial holder outside its territory; and finally (3) rearward movements during Head-Standing and Fluttering. Backward Swimming was accomplished by the pectoral fins paddling alternately while the caudal fin and the posterior portion of the dorsal fin were used as rudders.
Hovering occurred as the fish remained motionless in the water column. The pectoral fins performed the major sculling movements while the rear portion of the dorsal and tip of the caudal fin provided apparent compensation. The median and pelvic fins were slightly expanded. The pattern was uncommon since the bicolor was an active fish; however, it occurred most frequently following Nipping activity. Certain motor patterns appeared similar to Hovering but maximal spreading of median fins and the degree of spreading by the pelvic fins (maximal-Fin-Spread; minimal-Lateraland Frontal-Display) readily distinguished them as being separate and distinct.
Fin-Spread (Fig. 6) was a basic locomotory posture often evident during an agonistic encounter or while carrying out braking movements, following a rapid manoeuvre. The median and pelvic fins were maximally spread (as were the pectorals during braking movements). The body was horizontal and occasionally, the rear portion of the body had a slight lateral bending. This posture is the same as noted in other fishes (e.g. anabantids: Miller & Hall 1968; badids: Barlow 1962a; centrarchids: Carter-Miller 1963; cichlids: Baerends & Baerends-van Roon 1950; cyprinodonts: Foster 1967). A detailed account of this motor pattern, applicable to the present species is found in Myrberg (1965).
Head-Stand was another basic locomotory posture that was often evident during interspecific interaction (e.g. when large fishes such as snappers, grunts, groupers and parrotfishes approached the nest of a guardian male). The body was inclined, head downward; all fins were spread maximally; the actor might be standing still or moving slightly forward or backward. The angle of inclination varied from a slight pitch to over 60° from the horizontal, the angle depending on the distance between the object of attention and the actor. If an intruder turned away, the actor, then, rapidly regained a horizontal attitude and carried on other activities. If the intruder, however, moved within an apparent critical distance, the actor suddenly 'exploded' from the Head-Stand posture, lunging at its adversary with great speed. Such explosive swimming and the ram or bite that followed usually resulted in the large adversary shaking itself and then moving slowly away. Head-Stand was not often followed by explosive swimming but a large proportion of the occurrences ol explosive swimming were preceded by Head Stand. Head-Stand was rarely seen durinl intraspecific interactions; rather its general re striction to interspecific agonism placed it rathei clearly apart from the numerous agonistk patterns to be described later. The hydrodynamk aspects of this posture, applicable to the present species, are discussed by Myrberg (1965).
Wandering was slow or moderate swimminl about its environs by a fish that was neither interacting directly with other fishes nor showinl a recognizable motor pattern. This activity, though non-descript, provided, if nothing else, a slight pause in a hectic recording session, the latter including as many as 600 distinct patterns recorded from a single individual during a hr, observation period. The duration of Wandering though not considered in this report, was obtained during recording sessions since its significance to certain problems has been amply demonstrated (see Barlow 1964).
Maintenance Patterns
Included here are those motor patterns which appeared to aid fish by either removing foreign matter from the body (Chafing, Fin-Flicking, Defecating, Body-Shaking) or providing a unique type of muscular activity (Yawning, Tail-Down, Tail-Dig). These patterns appeared to occur independently of a direct social context, feeding or nesting. Yet, they certainly did not appear to form a single grouping based on a common causal factor(s). This was borne out later during the quantitative phase of the study.
Chafe was carried out against some hard object. A fish swam to a plastic tube (laboratory) or a frond of Sargassum (field) and just prior to reaching the object, it accelerated slightly. The fish then rubbed part of its body (usually the mid-lateral section) against the object with its fins fully expanded. The subsequent rebound from the object was either straight ahead or arching, the latter resulting in circling back to the object. Two to four Chafes often occurred together. Bicolors rarely Chafed against the sandy bottom or scoured rock, apparently preferring objects of a smoother texture.
Fin-Flick was often associated with Chafing. While swimming about, a fish occasionally folded and then expanded rapidly one or more of its fins, primarily the dorsal and pelvics. The different combinations of fins involved were difficult to observe and since the pattern was uncommon, we recorded simply the occurrence of the general pattern rather than attempting to dissect it into its possible variations.
Yawn involved a rapid opening of the mouth, often maximally, accompanied by an equally rapid expansion and subsequent lowering of the median and pelvic fins. Although the pattern was easily recorded from members of the laboratory colony, it was difficult to record in the field. It could be easily seen when using the full telephoto capability of the UTY camera, but recording sessions rarely employed that aspect of the camera, since it restricted the area of surveillance.
Tail-Down. A fish swimming near the substrate would suddenly stop, within a few millimetres of the substrate (also folding its fins at the same time), and then drop its caudal fin to the substrate. Following this manoeuvre, the fish would either regain a horizontal position and swim off or follow through with the next pattern.
Tail-Dig involved a rapid beating of the caudal fin at the substrate while the actor moved forward, body inclined about 20° (headup) for a distance of about 10 cm. During this short traverse, the anal, pelvics and spinus dorsal were folded against the body; the pectorals, though held close to the body, appeared to act as stabilizers, as did the rear portion of the dorsal fin. The sweeping action of the caudal fin resulted in substantial sand being blown from the area of traverse and it appeared initially to be a technique whereby food matter was cast into the water column. Yet we never recorded a bicolor turning around after completing Tail-Digging and then feeding or even 'inspecting his work'. Males dug in their territories and along their borders, but it was also recorded occasionally some distance away from such areas. The function of depressions left by such activity remained a mystery. They were never used for any clear purpose. I have observed Tail-Digging in numerous species of damselfishes and its wide-spread occurrence indicates an important functional basis. I wish to venture a suggestion that this pattern has some, as yet unknown, chemical function related to nest activity or territoriality in the various species that perform it. Centrarchids (Lepomis) also show a pattern quite similar to Tail-Dig (Carter-Miller 1963), and there it appears as a nest directed activity.
Body-Shake was rarely seen in bicolors. The pattern consisted of a lateral shaking of the entire body, two or three times in rapid succession. The fins were folded slightly. The pattern occurred so seldom during recording sessions that subsequent analysis of it was not possible. It did not appear, however, to have any function other than possibly removing foreign matter from the body.
A comparative summary of the apparent maintenance patterns of bicolors showed that all patterns were carried out by members of the laboratory and field colonies, irrespective of sex.