Zum Problem der Signalbildung, am Beispiel der Verhaltens-Mimikry zwischen Aspidontus und Labroides (Pisces, Acanthopterygii)


  • Wolfgang Wickler

    1. Aus dem Max-Planck-Institut für Verhaltensphysiologie, Seewiesen und Erling/Andechs
    Search for more papers by this author
    • 1

      Professor Dr. Konrad Lorenz zum 60. Geburtstag gewidmet.



Aquarienbeobachtungen bestätigen, daß verschiedene “Korallenfische” den Räuber Aspidontus für den Putzer Labroides halten. Sie verwechseln dieselben Merkmale wie wir. Untersucht wurde die Stammesgeschichte des Nachahmer-Tanzes: Er ist ursprünglich ein Epiphaenomen, Ausdruck eines Taxienkonflikts und nicht einheitlich motiviert. Wie er zu seiner Mimikry-Bedeutung kommt und wieviel daran Präadaptation ist, wird ausführlich besprochen.

Die Putzkunden müssen wahrscheinlich die Putzer erst kennen lernen und suchen nach Merkmalen, um den Putzer vom Nachahmer zu unterscheiden. Deshalb muß dieser jede Abweichung vom Vorbild vermeiden und ahmt alle Altersstufen und Lokalrassen genau nach. Die Lokalrassenbildung des Putzers kann man als Ausweichentwicklungen vor dem Nachahmer deuten.

An Mimikry-Fällen läßt sich sendeseitige Signalbildung gut verfolgen. Mimetische Verhaltensweisen und Signale entstehen aus “atelischen Beiprodukten” anderer Entwicklungsgänge; um zu wissen, welche das sind, muß man auch den vom Signal-Merkmal aus gesehen vorletzten Selektionsdruck kennen.


The evolution of animal communication usually involves co-adaptive changes of the signal-sender as well as of the signal-receiver. There are cases, however, in which only one of them, the sender or the receiver, profits by the communication, e.g. in mimicry. The model and the receiver can be taken as a stable communication system, and the evolution of mimicking characters gives the opportunity to study the evolution of a signal by a sender alone without co-adaptive interference from the receiver.

Here the fin-eating blenny Aspidontus taeniatus, which mimics the cleaner Labroides dimidiatus, has been studied with special reference to the evolution of its labroides-like swimming and “dancing” movements. The phylogenetic development of these movements can be traced back by comparing closely related, non-mimicking species. In completing the ethograms of 5 blennioid fishes it turned out that the labriform swimming-movement of Aspidontus and its dancing are to a large degree pre-adaptations with respect to their mimicry-function; that is they appear as by-products of the normal blenny behaviour in certain conflict situations. There is only a relatively small amount of true ritualization (signal-adaptive development) in the blenny's mimicking movements.

The same dancing movements and their derivatives are shown — with few exceptions — by all the species studied, though they usually function as special social signals (e.g. courtship or threatening movements). In Aspidontus, however, the movements of the dancing-complex have no special meaning, but instead occur in quite different conflict situations. So, for instance, the Aspidontus starts dancing if he is hungry and meets a large fish which he does not dare to attack immediately.

The fish which he meets very probably will have had some experiences with the cleaner Labroides and will know its dancing. He therefore may mistake the dancing blenny for a cleaner and will assume the cleaning position, thereby eliciting an attack from the blenny. Observations in aquaria show this to be true.

The development of the model-receiver communication system is discussed and more details concerning the evolution of other mimicking characters of the blenny are given.

Finally, the dancing of the cleaner is as unspecific as that of the blenny. The special meaning of the dancing-signal is introduced into the model-mimic-receiver communication system by the fact that, when the receiver meets the Labroides, he is nearly always cleaned. The external characters of the Labroides act as conditioned stimuli which are positively reinforced if the cleaner is willing to clean; remain un-reinforced if he is not; and are negatively reinforced by the mimic which therefore has to be less numerous than the model (which it actually is).

The Aspidontus even mimics some local colour-forms of the Labroides, which suggests that the deceived fish try to find differences between model and mimic according to the different experiences with them. This view is strengthened by the fact that mainly young fishes are deceived by the mimic. The conspicuousness of Labroides seems to have evolved under the selection pressure of being more easily learned by the fish-to-be-cleaned.