Brain window language of fireflies

Abstract

Before mating female fireflies of the species Photuris versicolor respond to the courtship flashes of conspecific males. After mating they become femme fatales (FFs) by answering the courtship flashes of males of other species. By analyzing the responses of females to artificial stimulus flashes we have identified the language of the firefly, based on brain windows and on biological oscillators with a period of about 2 seconds. The response latency of the famale is a function of the stimulus interval, and these stimulus-response relationships form six clusters or “dialogs”. Each dialog is composed of a narrow range of stimulus intervals and their associated narrow range of response latencies. These dialogs are related to the three response behaviors of females: responses to the short stimulus intervals of conspecific males and responses to medium and long stimulus intervals of heterospecific males. These interval classes are named courtship, mimicry and patrolling respectively. Each behavior uses one receive brain window and one send brain window. The receive window is a time interval during which a female will only accept the stimulus flash. The send window is a time interval during which the female can only initiate a response. Similar stimuli can produce different latencies depending on the context stored in the females' memory. This context or expectation is defined by the past history of stimulation. The context is used to dynamically adjust the brain windows depending upon the behavior. The six dialogs together form the brain window language of the firefly. The language is defined in terms of message quantum, distance and redundancy. The message quantum is the smallest item used in coding and it is a fraction of the biological oscillator period. The coding distance is a measure of how far the receive windows are separated from each other and determines the safety of coding. The coding distance measured in message quanta is large for courtship and small for mimicry communication. Language redundancy shows the efficiency of coding and it is greater than one. A computer model has been developed that uses the brain window language. The model is in excellent agreement with the experimental data.