ALBERT

Notes: Summary 1. Gnathonemus petersii respond to each other's electric organ discharge (EOD) with an “echo” discharge of their own at a latency of about 12 msec. This response persists until interfish distances reach about 30 cm (Fig. 4). 2. Artificial electrical stimuli were used to further characterize the response. Response threshold, latency dependence on stimulus intensity, polarity characteristics (Figs. 5–7), and differential regional sensitivity indicate that “medium” electroreceptors in the anterior region of the animal underlie the response. 3. Response probability depends upon the delay of the stimulus after the last EOD and also upon the instantaneous EOD rate (Figs. 8, 9). The echo response in turn resets the EOD rhythm of the responding animal (Fig. 10). These results suggest that the echo input pathway terminates on the presumed mesencephalic command center.

Notes: Summary 1. The electric organ discharge (EOD) patterns seen in pairs of mormyrid fishes (Gnathonemus petersii) during displays related to aggression and establishment of dominance are described. 2. A new method of reliably separating the discharges of the two fish was used. In this method a fine wire was attached to the tail of one animal (Fig. 1). 3. Discharge patterns were examined at different stages during the characteristic sequence of overt behavioral events which usually occurred after an intruder was put into a tank in which another fish of the same species had been resident for 1 hour or more. The resident attacked the intruder immediately, the intruder being initially unresponsive. After a few minutes, however, the two fish entered into intense, mutual, antiparallel displays (Fig. 2). The displays occurred repeatedly during a period of 0.5–30 min. This period ended suddenly with one of the fish clearly dominant as shown by one-sided attacks and avoiding behavior by the submissive fish. 4. All attacks were accompanied by a smooth acceleration to a high discharge rate which was usually terminated abruptly (Fig. 3). Anti-parallel behavior was accompanied by similar accelerations in both fish. Interdischarge intervals during these high rates changed discretely between those of about 15 msec and those of about 9 msec (Figs. 3–9). Initial attacks before the antiparallel period usually produced no effect or a brief acceleration in the discharges of the attacked animal. Similar attacks when dominance was well established caused a slowing of the discharge rate of the attacked fish. 5. The echo response in which one fish responds to the EOD of another with a discharge of its own at a latency between 11 and 14 msec was seen at all stages of the encounter. This latency corresponded rather exactly to the gap in the interval histogram between the shorter intervals around 9 msec and the longer ones around 15 (Figs. 11, 12). This correspondence led to a degree of avoidance of near synchronous discharges during those attacks which did not cause either slowing or accelerations in the attacked animal (Fig. 11). A degree of synchrony avoidance also occurred during the mutually high discharge rates of antiparallel behavior. This resulted from the phase locking of the two discharge trains which was often present at these times and which was probably due to the echo responses (Fig. 13). 6. Several features of the individual discharge trains and of their interaction were examined during the period of antiparallel activity. This was done in order to see if some critical parameter could be detected which would allow one to predict the winner of the encounter and which might be used as a signal by the fish themselves. No single feature among those we examined was clearly and consistently related to the outcome of the encounter.