Behav. Ecol. Sociobiol. 6, 67-79 (1979) Behavioral Ecology and Sociobiology 9 by Springer-Verlag 1979
Electric and Motor Responses of the Weakly Electric Fish,
Gnathonemus petersii (Mormyridae), to Play-back of Social Signals
Bernd Kramer
Fachbereich Biologie, Universitfit Konstanz, Postfach 5560, D-7750 Konstanz, Federal Republic of Germany Received January 12, 1979 / Accepted April 30, 1979
Summary. 1. Seven isolated G. petersii resting in their daytime hiding-places were stimulated via a dipole model (Fig. 1 a) with previously tape-recorded electric organ discharge (EOD) patterns in an attempt to de- termine whether G. petersii distinguishes two different intraspecific EOD patterns, rest and attack.
2. Rest pattern A was characterized by a broad distribution of EOD intervals, a low mean discharge rate (8 Hz, Fig. 3), and a long period of significantly positive autocorrelation (2 s, Fig. 4a). Accordingly, the spectrum of EOD rate fluctuations showed a low frequency range (0.005 to 0.12 Hz, Fig. 5a). Attack pattern B was a considerably different EOD interval distribution of high mean discharge rate (25 Hz, Fig. 3), showing a short period of significantly posi- tive autocorrelation (0.8 s, Fig. 4b), only. Here, the spectrum of EOD rate fluctuations was at a consider- ably higher frequency range (0.09 to 0.47Hz, Fig. 5b).
3. Play-back of attack pattern B elicited signifi- cantly more bodily startle responses from the experi- mental fish (Fig. 6) than did the rest pattern A (Ta- ble 1). Also the number of attacks directed at the dipole model was significantly greater during stimula- tion with attack pattern B (Table 2, Fig. 6).
4. The EOD responses of the experimental fish differed in several respects depending on which stimu- lation pattern was used. The modes of the pulse rate histograms as well as their spans were lower during play-back of rest pattern A than during stimulation with attack pattern B (average 12.3 vs 16.2 Hz, and average 47vs 56 Hz, respectively; Fig. 11). Short- term (0.2 s) EOD rate correlations were stronger when the fish were stimulated with rest pattern A than when they were Stimulated with attack pattern B (average correlations 0.67 and 0.61, respectively;
Figs. 10 and 11). Significant positive correlations were maintained for longer periods of time during rest pat-
tern stimulation than during attack pattern stimula- tion (average 1.94 and 1.24 s, respectively; Figs. 10 and 11). The spectra of EOD rate fluctuations of the stimulated fish were at lower frequency ranges during rest pattern stimulation than during attack pattern stimulation (average amplitude-spectrum peak frequencies 0.02 and 0.07 Hz, respectively;
Figs. 12 and 13).
5. Although maximal cross-correlations from the EOD rates to the stimulus pulse rates were weaker during rest pattern stimulation (average 0.2) than dur- ing attack pattern stimulation (average 0.33), signifi- cant cross-correlations were maintained for longer periods of time during rest pattern stimulation than during attack pattern stimulation (average 1.78 and 0.92 s, respectively). The lags of maximal cross-corre- lations were greater during rest pattern stimulation than during attack pattern stimulation (average 2.6 and 0.8 s, respectively; Figs. 14 and 15).
6. The results clearly showed that at least two specific EOD time patterns encode different' messages' in the intraspecific communication system of G, peter- sii.
Introduction
The observations by MShres (1957) and Szabo (1961) first suggested that the African group of weakly elec- tric fishes, the Mormyridae, might use their electric organ discharges (EODs) for communication. The ex- perimental analysis of this hypothesis began with in- vestigations showing that the fish responded to various kinds of stimuli, including electrical, with dis- charge rate changes (Harder etal., 1967; Stinger, 1972; Moller, 1970; Serrier, 1973; Bauer, 1974).
Experiments dealing more specifically with electric
0340-5443/79/0006/0067/$02.60
68 B. Kramer : Responses of Gnathonemus to Play-back of Social Signals c o m m u n i c a t i o n a n d social b e h a v i o u r were m a d e b y
B a u e r (1972), M o l l e r a n d B a u e r (1973), B a u e r a n d K r a m e r (1974), K r a m e r (1974, 1976a a n d b), Bell et al. (1974), R u s s e l l et al. (1974), K r a m e r a n d B a u e r (1976) a n d M o l l e r (1976). Specific E O D p a t t e r n s re- lated to p a t t e r n s o f social b e h a v i o u r were observed.
I n t e r a c t i n g fish e x h i b i t e d t i m e - r e l a t e d discharge rate changes. T h e p r e f e r r e d l a t e n c y r e s p o n s e o f Gna- thonemus petersii, a statistically s i g n i f i c a n t preference for a n E O D delay o f a p p r o x . 10-12 m s to f o r e i g n pulses, was i n t e r p r e t e d as a m e c h a n i s m a i d i n g the fish to separate their o w n signals f r o m f o r e i g n signals ( j a m m i n g a v o i d a n c e ) . S i m i l a r r e s p o n s e s h a v e b e e n f o u n d i n o t h e r m o r m y r i d species (Heiligenberg, 1976, 1977), n o t always c o m p a t i b l e w i t h t h e j a m m i n g a v o i d - ance h y p o t h e s i s ( K r a m e r , 1978).
I n a d e n e r v a t i o n e x p e r i m e n t the a b s e n c e o f E O D s i n a p a r t n e r was s h o w n to greatly increase the a m o u n t o f i n t e r n a l c o n s t r a i n t (i.e., d e p e n d e n c e o f the p r o b a - bility o f o c c u r r e n c e o f a given act o n earlier acts) i n the m o t o r p a t t e r n sequence exhibited b y a r e c i p i e n t G. petersii, c o m p a r e d w i t h the presence o f a n electri- cally i n t a c t conspecific ( K r a m e r , 1976c). F r o m these e x p e r i m e n t s it was e o n c l u d e d t h a t i n f o r m a t i o n was t r a n s m i t t e d b y conspecific electric signals w h i c h signif- i c a n t l y i n f l u e n c e d the b e h a v i o u r o f a r e c i p i e n t G.
petersii.
G. p e t e r s i i exhibits several types o f E O D p a t t e r n s r e l a t e d to m o t o r activity (Belbenoit, 1972; Bauer, 1974) a n d to v a r i o u s k i n d s o f social i n t e r a c t i o n s be- t w e e n p a r t n e r s ( K r a m e r a n d Bauer, 1976; K r a m e r , 1976b), as do several o t h e r m o r m y r i d s ( K r a m e r , 1978). I n a c o m m u n i c a t i o n system it m a y b e expected t h a t a set o f different signals serve the p a r t n e r s for c o m m u n i c a t i n g different messages.
This h y p o t h e s i s was tested b y p l a y - b a c k experi- m e n t s i n the p r e s e n t work. T w o different tape- r e c o r d e d E O D p a t t e r n s were p l a y e d b a c k via a d i p o l e m o d e l to the e x p e r i m e n t a l subjects, w h i c h were n o t s t i m u l a t e d otherwise. C o n t r a r y to the s t i m u l a t i o n ex- p e r i m e n t s cited, the d i p o l e m o d e l was devised to re- p r o d u c e the electric field o f a G. petersii as closely as possible i n the p r e s e n t work.
O n e o f t h e t w o s t i m u l a t i o n p a t t e r n s (A) was a resting discharge activity, the o t h e r p a t t e r n (B) was the E O D activity r e c o r d e d f r o m a n a t t a c k i n g G. p e t e r - sii which d i s p l a y e d high discharge rates.
M o t o r b e h a v i o u r r e s p o n s e s o f t h e e x p e r i m e n t a l subjects were f i l m e d by a video r e c o r d i n g system.
T h e E O D r e s p o n s e s were a n a l y z e d b y a u t o c o r r e l a t i o n a n d c r o s s - c o r r e l a t i o n , a n d b y spectral a n a l y s i s m e t h o d s . T h e r a t i o n a l e o f this a p p r o a c h was t h a t n u m e r i c a l m e t h o d s c a p a b l e o f h a n d l i n g large a m o u n t s o f d a t a were r e q u i r e d i n o r d e r to detect q u a n t i t a t i v e differences in p a t t e r n s o f r e s p o n s e s o b s c u r e d b y noise.
Materials and Methods
Seven Gnathonemus petersii (12-17 cm standard length) were used in this study. Before the experiments the animals were maintained together in a 360-1 aquarium which was planted and contained many porous-pot hiding-places. The water was power-filtered by synthetic fibre and peat, and the temperature was 26 ~ C. Electrical resistivity ranged from 2.5 to 2.9 k~2.cni, and pH was near 7.
A 12-h/12-h light-dark cycle from 0700 to 1900 hours was used both for the experimental and the home tanks. The fish were fed daily on live Chironomus larvae or Tubijex worms.
Prior to the beginning of an experiment the animals were habituated to the experimental tank (1.0 x 0.5 • 0.5 m, 250 1) for at least 4 days. When the animal under study had accepted its hiding-place for most of the time during day, a 5-day series of play-back experiments started.
Two experiments per day were pertbrmed at 0900 hours and 1700 hours. In random order, one of two different EOD activity patterns stored on magnetic tape was played back via a dipole model, with the restriction that no pattern was played back twice a day. The EOD activity pattern A had previously been recorded from a resting, isolated G. petersii; the EOD activity pattern B had previously been recorded from an animal vigorously attacking a Mormyrus rume while displaying attack-associated high discharge rates (Figs. 3-5).
The stimulus field was generated by a Grass $88 stimulus generator connected to the dipole model via a stimulus isolation unit (Grass SIU 5). No attempt was made to reproduce the EOD wave-form; instead, the wave-form produced by the dipole was a slightly' rounded' 0.5-ms monopolar square-wave pulse (Fig. 1 b).
Heiligenberg (1976) had shown in Brienomyrus niger that stimula- tion with square-wave pulses similar to those used in the present study yielded the same results compared with sine-wave pulses adjusted to resemble natural EODs more closely, except for the square-wave pulses appearing slightly more efficient. The stimulus pulses were triggered by tape-recorded EODs as schematically shown in Fig. 2.
The hiding-place was a 16.5-cm length of porous pot with a 6.5-cm inner diameter. As this pot was dosed at one end by coarse plastic mesh, the fish stayed in a rather precisely specified position within the pot, with its chin appendage protruding slightly out of the open end (Fig. 1 a). The polarity of the stimulating field was chosen such that the field of a conspecific in antiparallel position to the fish was reproduced (i.e., the electrical 'head' of the dipole was near the tail of the fish). A modified dipole with three electrodes was used, which gives a good approximation of the electric field produced by G. petersii (Boundinot, 1972). The field intensity produced by the dipole inside the porous pot at a distance of 5 cm from the dipole (which was the distance dipole- body wall of the fish facing the dipole) is shown in Fig. 1 c (measured with the fish not in place). The amplitude of the stimulating field was regulated such that it was equal to the p-p EOD voltage of fish A which was used first in the study. The same amplitude was used throughout the study. The stimulus amplitude perceived by the fish in its hiding-place was approx. 36 dB above threshold.
The threshold criteria were : ' discharge arrest' or ' sudden discharge rate increase with regularization' (concerning regularization, cf.
Moller, 1970).
Details of the experimental set-up are given in Fig. 2. A l"- video system (Grundig BK 300) with two cameras (Grundig FA 70) was used. One camera monitored the lateral and the bottom (mir- ror) view of the behaviour of the fish, and the other monitored the display of an electronic counter triggered by the EODs (the EODs were also recorded on the audio track of the vido tape).
Thus each video picture was 'labelled' by the display of the EOD total from start of experiment. In this way, the separate video and audio records could easily be synchronized.
B. Kramer : Responses o f G n a t h o n e m u s to Play-back o f Social Signals 69
Stimulating
Apparatus I Recording Apparatus
... I ] ~ ] ~ III ~LC,,---r--I ~ t i ~ + - - J I REorder
[ !1 "
l " " " l-].~,~U " ~
Counter
l l ,.L--~--T-I.,Stimulus
I ~ ,l, vM~~ ~ rGene,o~or I , i I I Recorder"
rl~ ,e_-I. Vide~ ,, rl3n I >
'f It.T J" C ... " ~ ) I ~ I
I
Fig. 2. Experimental set-up. The fish were stimulated with previ- ously tape-recorded E O D pulse patterns via a dipole model. The tape-recorded pulses triggered square-wave pulses from a stimulus generator (see Fig. 1 b) connected to the dipole via a stimulus isola- tion unit. Stimulus pulses and EODs were picked up in the water from two pairs of orthogonal electrodes fixed on the walls of the tank. The signals from both pairs of electrodes were differentially amplified, rectified, and summed electronically. The times o f occur- rences of stimulus pulses and o f EODs were discriminated by a
~ window' circuit; thus separate recording of the two series o f events on two tracks of a magnetic tape was possible. The behaviour of the fish and the coucurrent display o f an electronic counter triggered by the EODs were taped by a video recording system
100 - mY .J~'\
5o- \ /
% Z
~ . ~ ~ / / / / / / / / / / / / / / / / / / / / ~
o
~/////////////////////////,, ; "///111IIIII/I//////////11/11//1 -50
\
-lOO X ~
c ; ' I; ' 2'0 cm'
Fig. I a-c. Physical and electrical aspects o f the stimulus conditions.
a The daytime hiding-place of the fish with the dipole model 5 cm from the right body wall o f the animal. The dipole model was a plexiglass rod with three ~ 2.6-mm ball electrodes (according to Boudinot, 1972). b Wave-forms of the electric organ discharges (EOD) of the fish and of the stimulus pulse seen by the same pair of electrodes 1 m apart, straddling the hiding-place and the dipole. Superimposed on the E O D is a logic pulse o f 5-V amplitude triggered from the EOD. These pulses were stored on magnetic tape or were fed directly into a digital analyzer. The logic triggering point was delayed by 0.01 ms to the zero crossing o f the rapid potential change of the E O D from head positivity to head negativ- ity. e Intensity o f the stimulus field at the body wall o f the fish facing the dipole, measured inside the porous pot with the fish not in place. The reference electrode was on potential zero as far away as possible in the experimental 250-1 aquarium
The method of EOD recording was the same as described by Kramer (1974). The times of occurrences of EODs and o f stimulus pulses were taped on different tracks o f an instrumentation tape-recorder (Kyberna 601) at 9.5 cm/s for off-line analysis by a digital analyzer. In order to analyze the relationships between the two time series o f events, observations were made at discrete, equispaced time intervals which had to be strictly in register for both time series.
A method o f data acquisition that allowed for these require- ments was employed although a single-input digital analyzer (Nico- let 1074 with SH-71 and -72 plug-in modules) had to be used.
The principle o f this method (modified from that described by Kramer, 1974) was as follows: in parallel with the two series o f experiment-generated events, recorded on separate tracks o f the tape, a reference signal (short pulses separated by 0.2-s intervals) was recorded on a third track of the tape. On a fourth track, start pulses (one single pulse for each record) were stored. Succes- sive analysis of both records (of stimulus pulses and of EODs) by the digital analyzer was initiated by one specific start pulse from the tape (track four). The number o f events occurring during a pair of tape-recorded reference pulses was counted by the digital analyzer and the results were placed into successive memory ad- dresses.
The data stored in this device were output on BCD paper tape and read and the data transferred to magnetic tape files in the computer centre of the University o f Konstanz. The data stored on these tapes were used as the input to programs written in Fortran IV, which performed various forms of time series analyses.
The results were output on a digital plotter connected to the com- puter.
Numerical methods, described by Box and Jenkins (1976), were used to obtain autocorrelations and frequency spectra of pulse rate fluctuations, as well as cross-correlations of E O D pulse rates to stimulus pulse rates.
70 B. Kramer: Responses of
Gnathonemus
to Play-back of Social Signals Following the above authors, the most satisfactory estimateof the kth lag autocorrelation
Pk
isr k = - -
Ck
CO
where
1 N--K
ck= ~ ~ (Z~--~)(Z~+k--5),
k=0,1,2 ... Kt=l
is the estimate of the autocovariance 7k, 2- is the mean of the time series, and z t is an observation at time t.
The standard errors of autocorrelation estimates were de- termined as the square roots of the estimated large-lag variances
1 q
var[rk]~---~l+2 2
r2;k>q
N ( k=l 3for lags k greater than some value q beyond which the theoretical autocorrelation function is assumed to be zero.
The Fourier coefficients of a set of data points z t whose number was odd ( N = 2 q + l ) was calculated according to the method described by Box and Jenkins (1975):
2 u a , = ~ Z ztc~
t = l 2 N
b i = - - ~. z, s i n 2 ~ f ~ t , i = l , 2 , . . . , q g t = l
where f~ =
i/N
is the ith harmonic of the fundamental frequency1/N.
From the coefficients the amplitudes of the frequencies f~ were calculated
A(f)=~(a~+b{),
N i=1,2 ... q.As the smallest period of a sinusoidal component is two obser- vation intervals (i.e., 0.4s), the highest frequency that could be detected was 2.5 Hz in the present study.
The kth lag cross-correlation coefficient
p~y(k)
was estimated by ,,,c~(k)
r~ytt9 = , k=0,1,2,...
s~s,
where the estimate of the cross-covariance coefficient is
1 n k
%(k)=g~_(x,-x)(y,+~-y),
k=O,l,2 ....where K, y are the means of the x series and y series, respectively, with their associated standard deviations s~ and sy, x t and y, are observations of the x series and the y series at time t, and n is the number of observations.
Non-parametric statistical tests such as the Wilcoxon matched- pairs signed-ranks test and the randomization test for matched pairs were used according to the method described by Siegel (1956).
All values for P given in this paper refer to two-tailed regions of rejection of the sampling distributions.
R e s u l t s
Time Series Analysis o f the Stimulus Patterns
T h e fish w e r e s t i m u l a t e d w i t h p u l s e s e q u e n c e s p r e v i - o u s l y r e c o r d e d f r o m c o n s p e c i f i c s . F i g u r e s 3-5 de- scribe the p r o p e r t i e s o f t h e t w o s t i m u l u s p a t t e r n s .
S t i m u l u s p a t t e r n A h a d b e e n r e c o r d e d f r o m a rest- ing a n i m a l , s t i m u l u s p a t t e r n B f r o m a n a t t a c k i n g a n i - mal. R e s t p a t t e r n A is a b r o a d d i s t r i b u t i o n o f E O D i n t e r v a l s w i t h t h r e e w e a k m o d e s (Fig. 3a). A t t a c k p a t - t e r n B is c h a r a c t e r i z e d b y a b i m 0 d a l E O D i n t e r v a l d i s t r i b u t i o n w i t h a v e r y s h a r p g a p s e p a r a t i n g the t w o m o d e s (Fig. 3 a ; cf. a l s o K r a m e r a n d B a u e r , 1976).
T h e d i s t r i b u t i o n s o f d i s c h a r g e r a t e s o f p a t t e r n s A a n d B a r e s h o w n in Fig. 3 b. P a t t e r n A is c o n f i n e d to a r a t h e r n a r r o w r a n g e o f d i s c h a r g e rates b e t w e e n 2.5 a n d 40 H z w i t h the p e a k a t 7.5 Hz. D i s c h a r g e r a t e s o c c u r r i n g d u r i n g a t t a c k p a t t e r n B r e a c h 5-112.5 Hz, while t h e d i s c h a r g e r a t e m e a s u r e d m o s t o f t e n was 15 Hz.
I n s i g h t i n t o the k i n d s o f t i m e r e g u l a t i o n o f dis- c h a r g e r a t e c h a n g e s m a y be g a i n e d b y a u t o c o r r e l a t i o n m e t h o d s a n d b y s p e c t r a l a n a l y s i s o f the s e q u e n t i a l p u l s e r a t e d a t a ( e x a m p l e s are s h o w n in Figs. 8 a n d
9).
F i g u r e 4 s h o w s t h e a u t o c o r r e l a t i o n f u n c t i o n s o f t h e p u l s e r a t e s o f t h e t w o s t i m u l u s p a t t e r n s . I n rest p a t t e r n A , p u l s e r a t e r e m a i n e d a b o v e o r b e l o w t h e m e a n f o r t e n successive o b s e r v a t i o n s (i.e., 2.0 s) since n i n e coefficients were s i g n i f i c a n t l y p o s i t i v e ; a f t e r t h a t t i m e n o c o r r e l a t i o n was f o u n d (coefficients w e r e n o t s i g n i f i c a n t l y d i f f e r e n t f r o m zero). S u p e r i m p o s e d o n this m o d e l was a n a l t e r n a t i n g d i s c h a r g e r a t e c h a n g e b e t w e e n a d j a c e n t o b s e r v a t i o n p e r i o d s (i.e., f r o m o n e 0.2-s o b s e r v a t i o n p e r i o d t o t h e n e x t one) since t h e s e c o n d coefficient was h i g h e r t h a n the first one. T h e c a u s e for this was a n a l t e r n a t i n g E O D i n t e r v a l - l e n g t h r e g u l a t i o n b e t w e e n a d j a c e n t E O D i n t e r v a l s (cf.
K r a m e r , 1974). So t h e d i s c h a r g e r a t e o f t h e n e x t o b - s e r v a t i o n b u t o n e w a s m o r e s i m i l a r to the first o b s e r - v a t i o n t h a n t h e o n e b e t w e e n these t w o o b s e r v a t i o n s . I n a t t a c k p a t t e r n B, p u l s e r a t e s r e m a i n e d a b o v e o r b e l o w t h e m e a n f o r a c o n s i d e r a b l y s h o r t e r t i m e only. A f t e r f o u r o b s e r v a t i o n p e r i o d s (i.e., 0.8 s), p u l s e r a t e was n o l o n g e r c o r r e l a t e d w i t h r e s p e c t to the first o b s e r v a t i o n since o n l y t h r e e coefficients were signifi- c a n t l y d i f f e r e n t f r o m zero. C o n t r a r y t o p a t t e r n A , in p a t t e r n B, p u l s e r a t e s o f a d j a c e n t o b s e r v a t i o n p e r i o d s were v e r y s t r o n g l y c o r r e l a t e d since t h e a u t o - c o r r e l a t i o n f u n c t i o n s t a r t e d w i t h a v e r y h i g h coeffi- cient o f 0.76.
A d i f f e r e n t a p p r o a c h t o t h e a n a l y s i s o f a series o f o b s e r v a t i o n s o f a v a r i a b l e w h i c h c h a n g e s in t i m e is a n e s t i m a t i o n o f the a m p l i t u d e s o f s i n u s o i d a l c o m - p o n e n t s o f u n k n o w n f r e q u e n c i e s b u r i e d in noise.
T h e s p e c t r u m o f the r e s t - p a t t e r n d i s c h a r g e r a t e f l u c t u a t i o n s s h o w e d h i g h a m p l i t u d e s ( a b o v e 1/3 o f t h e p e a k a m p l i t u d e ) in a f r e q u e n c y r a n g e o f 0.005 to 0.12 H z ; t h e p e a k was a t 0.041 H z (Fig. 5a). T h e spec- t r u m o f a t t a c k p a t t e r n B s h o w e d 1/3 p e a k a m p l i t u d e s w i t h i n a c o n s i d e r a b l y h i g h e r f r e q u e n c y r a n g e f r o m 0.09 to 0.47 H z ; the p e a k w a s at 0.23 H z (Fig. 5b).
im % 8 8
0.~
0.08
0.06
v)
z0.;- ~STIMULATION WITH PATTERN A
+o, b
0.2-
0.1-
10 20 30 40 50 60 70 80 90 100 t10 8.81-- 9 ELECTRIC ORGAN DISCHARGE RATE [HZ] D,08" I
|
O
o0,04 l
{ = 24.8 Hz o,o
j ~ , ~ _N=1491
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 INT~RDISCHARGE INTERVAL (MS)
Fig. 3a and b. EOD patterns used for 3-rain stimulation periods o f the experimental fish via the dipole model, a g O D interval histograms.
Stimulus pattern A (black histogram) is the discharge activity o f a resting G. petersii (3-ms bins). Stimulus pattern B (white histogram) is the discharge activity of a G. petersii attacking and pursuing a partner (1-ms bins). Abscissa: discharge interval (ms); ordinate:
relative occurrences o f intervah o f a specific bin width (black ordinate is for black histogram). N, total number o f intervals; ~, mean discharge rate. b E O D rate histograms. Abscissa: EOD rate (2.5-Hz bins); ordinate: relative occurrences of a specific E O D rate
ct u_o P<0.001
zo ~ 0 . 0 1
B'
O 8 16 24 32 40 48 56 64
ORDER OF AUTOCORRELRTION
-y o.o,
0 8 16 24 32 40 48 56 64
ORDER 8F RUTQCORRELRTION
Fig. 4a and b. Autocorrelation analysis o f the pulse rates o f rest and attack stimulation patterns A (a) and B (b). Note that the autocorrelation function o f attack pattern B starts with a high coefficient but declines more steeply than does the autocorrelation function o f rest pattern A. The positive pulse rate correlation of attack pattern B 'dies o u t ' within 0.8 s ( ~ 3 significant coefficients P<0.001), whereas pulse rates of rest pattern A are positively correlated for 2 s ( ~ 9 significant coefficients). Superimposed on this pulse rate regulation is a short-term (0.2 s) pattern o f pulse rate change in rest pattern A, since the second coefficient is higher than the first
FREQUENCY [ HZ
b
FBEOUENCY C HZ
Fig. 5a and b. Amplitude spectrum analysis of the pulse rate fluctu- ations of attack patterns A (a)and B (b). Ordinate: amplitudes o f sinusoidal components in arbitrary units; abscissa: frequency in Hz. Note that the attack stimulation pattern B shows significant amplitudes at a considerably higher frequency range than does the rest stimulation pattern A
72 B. Kramer: Responses of Gnathonemus to Play-back of Social Signals
Fig. 6a--e. Motor responses of G. petersii on pIay-backs of social signals via a dipole, a Startle responses were observed significantly more often during stimulation with attack pattern B compared with rest pattern A. A startle response was a rapid forwards and backwards movement during which the fish never totally left its porous-pot hiding-place, b Attack responses on the dipole model were observed significantly more often during stimulation with attack pattern B. e In many instances a Lateral Display accompanied by a high discharge rate followed the attack responses in the same manner as observed during social behaviour
Motor Responses of the Fish to the Stimulation with the Dipole Model
Various kinds o f m o t o r responses were observed at onset o f stimulation o f the fish resting in their hiding- places (Fig. 1 a). The m o s t frequent response observed was a bodily startle response (Fig. 6 a) with discharge rate increase. A n o t h e r kind of response was an attack response on the dipole, which was sometimes followed by a lateral display with high discharge rate (Fig., 6b and c; cf. K r a m e r and Bauer, 1976). Various other kinds of behaviour were also observed during stimula- tion (e.g., swimming restlessly a b o u t the aquarium, probing the dipole model intensely with the chin ap-
pendage for m o r e or less sustained periods of time and swimming in rapid bouts near the b o t t o m o f the aquarium). N o effort was m a d e to quantify these complex behaviours; the n u m b e r o f startle responses and o f attacks on the dipole model, however, could easily be determined f r o m slow-motion play-backs of the video tapes.
Table 1 shows the outcomes of the startle response analysis separately for each fish. In each animal a n d every paired observation the rate of s_tartle responses per second, displayed by the animal inside its hiding- place, was found to be higher during stimulation with attack pattern B t h a n during stimulation with rest pattern A. The finding was significant individually for all o f those five out o f seven animals for which sufficient observations were available (P < 0.05, two- tailed r a n d o m i z a t i o n test for m a t c h e d pairs). Pooling of the data yielded a P ~ 0 . 0 0 1 (randomization test for matched pairs, two-tailed).
Except for two animals which never attacked the dipole model, all other animals (i.e., five) attacked the dipole model either significantly m o r e often (fish A ; P < 0 . 0 5 , two tailed r a n d o m i z a t i o n test for matched pairs) or exclusively during stimulation with attack pattern B (animals C, E, F, and H, Table 2).
An analysis of the attack totals of all fish also yielded a significant difference for rest and attack pattern stimulation (P < 0.05; two-tailed r a n d o m i z a t i o n test for m a t c h e d pairs). The reason no attack was observed in m a n y paired experiments was that the fish did not leave their hiding-places (these experi- ments are m a r k e d by an asterisk).
EOD Responses to the Stimulus Patterns
The previous section on m o t o r behaviour described attack responses of the fish towards the dipole model.
Associated with these ' a t t a c k s ' , high discharge rates were displayed (Fig. 7). This was expected since in earlier work ( K r a m e r and Bauer, 1976) a discharge rate increase was found to a c c o m p a n y every attack on other fish.
A n o t h e r type of E O D response displayed during
social behaviour was also shown towards the active
dipole model.
G. petersii had been found statisticallyto prefer a discharge delay o f approx. 9-13 ms latency
to foreign pulses (Preferred Latency Response: Bauer
and K r a m e r , 1974; K r a m e r , 1974; Russell et al.,
1974). Very strong Preferred Latency Responses (of up
to 36% E O D s within the above latency range) were
also found in the present experiments with b o t h stim-
ulus patterns. N o systematic trend for one o f the
two stimulation patterns to elicit m o r e Preferred
Latency Responses than the other could be detected.
B. Kramer : Responses of Gnathonemus to Play-back of Social Signals 73 Table 1. Startle response rates displayed by seven G. petersii inside their porous-pot hiding-places during stimulation with rest pattern A and attack pattern B. Asterisks: no startle responses were observed since the fish immediately left its hiding-place at stimulation onset. The bar graphs show the means together with their _+ 1 standard error ranges separately for each fish. Note that during stimulation with attack pattern B, each fish displayed a higher startle response rate compared with stimulation with rest pattern A. The data were tested for significance of differences by the randomization test for matched pairs. Pooling of data yielded P < 0.001
Startle responses/s
Fish A C D E F G H
Stimulation pattern B A B A B A B A B A B A B A
Day 1 0.48 0.37 0.42 0.03 0.60 0.34 0.43 0.25 0.30 0.16 0.19 0.07 0.07 0.05
Day 2 0.39 0.20 0.20 0.08 0.20 0.17 0.37 0.14 0.25 0.18 0.22 0.07 0.12 0.05
Day 3 * * 0.18 0.04 0.33 0.14 0.19 0.05 0.23 0.18 0.13 0.09 0.06 0.03
Day 4 * * 0.18 0.07 0,28 0.25 0.19 0.09 0.31 0.25 0.12 0.05 0.11 0.04
Day 5 * * 0.15 0.08 - - 0.20 0.08 0.29 0.20 0.14 0.06 0.13 0.04
t q
~ 0 . 4 - c 0 0 . 3 -
0.2 -
,m -
0 . ~ -
U o 0 -
P<0.031 P<0.062 P<0.031 P<0.031 P<0.031 P<0.031
A C D E F G H
F i g u r e s 8 a n d 9 s h o w t h e d i s c h a r g e r a t e r e s p o n s e s o f f i s h A t o t h e o n s e t o f s t i m u l a t i o n w i t h t h e r e s t p a t t e r n a n d t h e a t t a c k p a t t e r n . I n b o t h r e c o r d s , a n E O D r a t e i n c r e a s e w a s o b s e r v e d a t t h e o n s e t o f s t i m u - l a t i o n w i t h t h e d i p o l e ( b r o k e n lines). W h i l e in F i g . 9 t h e r e w e r e 15 c l e a r - c u t E O D r a t e i n c r e a s e s ( > 60 H z ) at d i p o l e p u l s e r a t e i n c r e a s e s w i t h i n a p p r o x . 180 s s h o w n , t h e r e w e r e o n l y t w o s u c h w e a k r e s p o n s e s ( > 4 0 H z ) in F i g . 8. I n F i g . 9, t h r e e o f t h e E O D r a t e i n c r e a s e s w e r e f o l l o w e d b y s h a r p d i s c h a r g e r a t e d e c r e a s e s o r d i s c h a r g e b r e a k s (at 30, 50, a n d 116 s).
I n t h r e e i n s t a n c e s , a v e r y l o w d i s c h a r g e r a t e o r a m o -
m e n t a r y d i s c h a r g e b r e a k w a s d i s p l a y e d b e f o r e a n E O D r a t e i n c r e a s e (at 81, 116, a n d 168 s). A t 51 s, a s h o r t d i s c h a r g e b r e a k i n s t e a d o f a n E O D r a t e in- c r e a s e w a s d i s p l a y e d o n a d i p o l e p u l s e r a t e i n c r e a s e . I n o r d e r t o r e d u c e t h e a m o u n t o f v e r b a l d e s c r i p - t i o n n e c e s s a r y t o c h a r a c t e r i z e t h e E O D r e s p o n s e s o f t h e fish t o s t i m u l a t i o n , t h e a u t o c o r r e l a t i o n a n d s p e c - t r a l a n a l y s i s m e t h o d s o u t l i n e d in M a t e r i a l s a n d M e t h o d s w e r e a p p l i e d t o t h e d a t a . I n a d d i t i o n , c r o s s - c o r r e l a t i o n f r o m t h e E O D r e s p o n s e s t o t h e s t i m u l u s p a t t e r n s w a s p e r f o r m e d . I n e a c h a n a l y s i s , t h e first 10 s o f E O D a c t i v i t y f o l l o w i n g o n s e t o f s t i m u l a t i o n w e r e o m i t t e d since v i s u a l i n s p e c t i o n o f t h e d a t a s u g g e s t e d t h a t t h e f i s h d i d n o t r e s p o n d v e r y d i f f e r e n t l y t o t h e t w o k i n d s o f s t i m u l a t i o n p a t t e r n s d u r i n g t h a t t i m e .
A u t o c o r r e l a t i o n a n a l y s i s o f t h e E O D r a t e s d i s p l a y e d b y t h e f i s h d u r i n g s t i m u l a t i o n i n v a r i a b l y y i e l d e d p o s i t i v e c o r r e l a t i o n c o e f f i c i e n t s f o r a d j a c e n t d a t a p o i n t s , w i t h a d e c l i n e t o z e r o as m o r e d a t a p o i n t s s e p a r a t e d t h e t w o p o i n t s c o n s i d e r e d ( F i g . 10). H o w -
Table 2. Attack responses of seven G. petersii on the dipole model during 3 min of stimulation with rest pattern A and attack pattern B. Asterisks: the fish did not leave their porous-pot hiding-places during the whole stimulation period. Attack rates were significantly greater during attack pattern stimulation individually for fish A, and for the paired attack totals of all fish (P < 0.05, randomization test for matched pairs)
Attack responses on the dipole model
Fish A C D E F G H
Stimulation pattern B A B A B A B A B A B A B A
Day 1 10 2 8 0* 0 0 5 0 2 0 0* 0* 0* 0*
D a y 2 27 0 1 0* 0 0 0 0* 1 0 0* 0* 0* 0*
Day 3 24 9 0* 0* 0 0 0 0 0 0 0* 0* 0* 0*
Day 4 21 5 0 0* 0* 0* 2 0 0 0 0* 0* 0* 0*
Day 5 23 13 7 0 -- -- 3 0 0 0 0* 0* 4 0*
74 B. Kramer: Responses of Gnathonemus to Play-back o f Social Signals g
d
g
o 0"3 ~
O Z u ~ _ L j ~ Z
o
uJ~
g
[12 p - - o
rl 0
t
TIME(S)
I1 2 I1 4 I1 G I1 8 ~ 0
Fig. 7. E O D activity of G. petersii during stilzaulation with social signals (attack pattern B). The abscissa of each point is the time of occurrence (s) of each discharge, and the ordinate is the interval (ms) from the previous discharge. High discharge rates (underlined by horizontal bars) and attack responses were elicited by high pulse rates of the stimulating dipole model
9O 8O 70 O 60 Z 5O 0 4O i i i 3O u3 20 10 14.1 D 0..
50 U') 50 I,.U U') 4O ...a 30 :~ 2O II,,. tO 0
o n s e t of s t i m u l a t i o n with p a t t e r n A
!i i~ ,~ ..
- % %
- - I 'q i:
'ioo 11~o 1~o 1~o 1~,o 1~o 115o 1~o 1~o lbo
T I M E ( S )
Fig. 8. Concurrent pulse rates o f G. petersii and o f the dipole model driven with rest pattern A. Abscissa: time (s); ordinate: pulse rates measured during successive 0.2-s time intervals (expressed in Hz). The onset o f stimulation (broken line) at 9 s is indicated by an arrow. The whole sequence is cut into two parts, the lower part continues the upper. White arrows indicate two instances where a stimulus pulse rate increase coincided with an EOD rate increase. Note that these instances were rare compared with Fig. 9
ever, several parameters of the EOD activities of the experimental subjects significantly depended on the type of stimulation pattern used. The correlation be- tween adjacent data points was stronger when the fish were stimulated with rest pattern A (average cor- relation of 0.67) than when they were stimulated with attack pattern B (average correlation of 0.61;
P < 0.02, paired t-test, Wilcoxon matched-pairs signed- ranks test, Fig. 11). Also the number of correla-
tion coefficients significant at P < 0.001 was greater during rest pattern stimulation (on the average 9.7) than during attack pattern stimulation (on the average 6.2; P<0.01, Wilcoxon matched-pairs signed-ranks test, Fig. 11).
Other parameters of the discharge activities also
depended on the kind of stimulus pattern used. The
peaks of the pulse rate histograms were on signifi-
cantly different positions of the Hz scale during rest
B. K r a m e r : Responses of Gnathonemus to Play-back of Social Signals 75
9O 80 70 80 0 50 Z 4 0 0 3O (--)20
o Ill i40 l Q- 130
120 ca') I 1 0 LIJ q3 100 - - s
?o 5o 4o 30 20 lO o
onset of s t i m u l a t i o n with p a t t e r n B
%1
,. !i
!i iii
ii ::
%
11o 10
II o o i11 o
V ~" l
~o !t
%ko
%~o ~o ~o
1
a
~o ~o
i1 i
Ao l~o ~o i~o 13oI
T I M E ( S )
Fig. 9. Shows the same as Fig. 8 but for stimulus pattern B. Note that the instances o f E n D rate increases elicited by stimulus pulse rate increases (white arrows) were observed m u c h m o r e often t h a n during stimulation with rest pattern A (Fig. 8)
a
~- < o.oo,
6 7
g#-
o 8 16 24 3 2 40 4B 56 64
gBDER 0F AUTOCORRELATTDN
o | \ P<O,O01
~o/ -~.~_P<0.ol
~ 0 8 i 6 24 32 4D 48 56 64
o ORDER OF RUTOCORRELRI]ON
Fig. lOa and b. Autocorrelation analysis of the E n D rate responses o f G. petersii to (a) stimulation with rest pattern A a n d (b) stimula- tion with attack pattern B. Note that the autocorrelation function o f the E n D responses start with a higher coefficient in a t h a n in b. Discharge rate correlation 'died o u t ' at a greater lag or order of autocorrelation in a (1.8 s; k = 9 ) t h a n in b (1.2 s; k = 6 ) . D a t a were taken from one fish a n d the same day
E.
(o
z
W
o (.9
0 . 6 -
0.4-
O.2-
0 . 9 .
A B P < 0.02
m ,,z, :EN
_.g =<.=.-
U - - 0
A B P < 0.01 P < 0,001
~u - ~
~ ~ ~o-
=
~o- 4-Zl ,,. ~ ~o :
O ~ -
,s-I-11z ,~
_111 ,,, ,s 3o ~ /
,o LII
s - I l l ~
lo-|
P < 0,001 P < 0.05 Fig. 11. Statistical c o m p a r i s o n of autocorrelation and of pulse rate parameters o f the E n D activities displayed by seven G. petersii during rest stimulation with pattern A (left columns) a n d pattern B (right columns). The m e a n s ( n = 2 3 ) are s h o w n together with their • 1 standard error ranges. The levels of significance of differ- ences between paired series o f experiments are indicated below the base line. N o t e that the correlation between adjacent data points was stronger and that the n u m b e r o f significant coefficients (P < 0.001) was greater during rest pattern stimulation c o m p a r e d with attack pattern stimulation. The m o d e s o f the pulse rate distri- butions were on different positions of the Hz scale, a n d also the m e a n E n D rates differed significantly. The ranges of pulse rates observed (i.e., the span from the lowest to the highest rates) were greater during attack pattern stimulation t h a n during rest pattern stimulation. The d a t a were analyzed by the paired t-test except the second pair of c o l u m n s (from the left) which was analyzed by the Wilcoxon matched-pairs signed-ranks test
76 B. Kramer: Responses of Gnathonemus to Play-back of Social Signals
o d ,
I 0 -z
13.
'~ 4 ~ ~ ' ~ ' 1 0 _ 1 { ~ 4 ~ ~'~9'i0 o
F R E Q U E N C Y ( HZ )
b
~ I IC
F R E Q U E N C Y I HZ l
Fig. 12a and h. Amplitude spectrum analysis of the discharge rate fluctuations of G. petersii during stimulation with rest pattern A (a) and attack pattern B (b). Note that during stimulation with the attack pattern G. petersii displayed considerably higher fre- quencies of discharge rate changes than during stimulation with the rest pattern. Data were taken from one fish and the same day
pattern stimulation (average 12.3 Hz) and during at- tack pattern stimulation (average 16.2 Hz; P < 0.001, paired t-test, Fig. 11). Similar differences (18.2Hz and 21.3 Hz, respectively) were obtained for the mean discharge rates (P < 0.001, paired t-test, Fig. 11). The number of non-empty bins or the spans of the pulse rate histograms comprised an average 47 Hz during rest pattern stimulation, and an average 56 Hz during attack pattern stimulation (P < 0.05, paired t-test, Fig. 11).
The finding that the number of positive correlation coefficients was greater during rest pattern stimula- tion than during attack pattern stimulation (Fig. 11) shows that the periods of time for which similar dis- charge rates were maintained were longer during rest
N
>..
z LLI
t~
LI.
Lkl Q_
0 . 0 9 -
0.08 - 0 . 0 7 -
0.06 -
O.O5 - 0 . 0 4 -
0.03 -
0.02 - 0.01 - 0
A B
P < 0.01
AMPLITUDE SPECTRUM ANALYSIS 0 . 6 -
B ~05-
=o ~04 203
~ ~ OI 2
0 A B
P < 0.01
Fig. 13. Comparison of two amplitude-spectrum parameters of the E O D rate fluctuations of seven G. petersii during stimulation with rest pattern A (left columns) and attack pattern B (right columns).
The means (n=23) are shown together with their + 1 standard error ranges. The levels of significance of differences between paired series of experiments are indicated below the baseline (Wilcoxon matched-pairs signed-ranks test). Note that the mean peak fre- quency and the mean of the highest i/3 peak frequency were signifi- cantly lower during rest pattern stimulation compared with attack pattern stimulation
pattern stimulation than during attack pattern stimu- lation. Spectral analysis confirmed that the discharge rates fluctuated at different frequencies. Two exam- ples from the same fish on the same day are shown in Fig. 12. While during rest pattern stimulation very low frequencies prevailed (the highest frequency of 1/3 peak amplitude was 0.082 Hz), during attack pat- tern stimulation a 1/3 peak amplitude was still found at 0.63 Hz. The means of the peak frequencies were near 0.02 Hz during rest pattern stimulation and ap- prox. 0.07 Hz during attack pattern stimulation (P < 0.01, Wilcoxon matched-pairs signed-ranks test, Fig. 13). The highest frequencies of 1/3 peak ampli- tude were on the average 0.16 Hz during rest pattern stimulation and 0.48 Hz during attack pattern stimula- tion (P < 0.01, Wilcoxon matched-pairs signed-ranks test, Fig. 13). No significant differences were found when the lowest frequencies of 1/3 peak amplitude were compared for rest pattern and attack pattern stimulation (Wilcoxon matched-pairs signed-ranks test; P>>0.20, paired t-test).
Several significant differences in the responses of the fish to the two stimulation patterns were also found by cross-correlation of the EOD activities with the stimulus patterns.
Examples of the results are shown in Fig. 14. The
cross-correlations from the EOD responses of the fish
to the stimulus patterns were weaker during rest pat-
tern stimulation (average maximal cross-correlation
of 0.2) than during attack pattern stimulation (aver-
age maximal cross-correlation of 0.33; P < 0.001,
paired t-test, Fig. 15). However, the number of signifi-
B. Kramer : Responses of Gnathonemus to Play-back of Social Signals 77
z tlA
w o
z 2
d k u l o a z ~ a = ~ _
~ 3
g
?
i _ _ o
h _ ..
L u o
O g O
g-
a = ~ . f_)
11
P < 0.001
8 16 24 32 40 48
ORDER OF CROSS-CORRELATION
8 16 24 32 40 48
ORDER OF CROSS-CORRELATION
Fig. 14a and b. Cross-correlation analysis of the EOD rate responses of G. petersii to (a) stimulation with rest pattern A and (b) stimula- tion with attack pattern B. Note that the strongest cross-correlation in a was weaker and occurred at a greater lag or order of cross- correlation (0.6 s; k = 3) than in b (0.2 s; k = 1). However, signifi- cant cross-correlation ( P < 0.001) was maintained for a longer period of time in a (1.4s; k = 7 ) than in b (0.8 s; k = 4 ) . Data were taken from the same fish and day
cant (P < 0.001) cross-correlation coefficients of ei- ther sign was greater during rest pattern stimulation (average 8.9) than during atack pattern stimulation (average 4.6; P<0.01), Wilcoxon matched-pairs signed-ranks test, Fig. 15). A similar result was found when only the positive cross-correlation coefficients were considered (rest pattern stimulation: an average number of 6.8 coefficients were significant; attack pattern stimulation: an average number of 3.9 coeffi- cients were significant; difference significant at P < 0.01, Wilcoxon matched-pairs signed-ranks test, Fig. 15). No significant differences between the numbers of negative cross-correlation coefficients (significant at P < 0.01) were found for the two stimu- lation patterns.
Cross-correlation was not only weaker during rest pattern stimulation compared with attack pattern stimulation, but the lag of the maximal cross-correla-
,< (.,)
~,'T - ~ gu_
< u.I 0 . 3 -
= 8
u_
o z 0 . 2 - _a ..a
_a g l
o u l 0
03 o
113 112
< o P < 0.001
m
z
< , o _ ~ < ~ _ - o
#~ ~
- - z
~ ~ ~, _ _
P<0.01 ~ o ~ P < 0 . m <, o ~ P < 0 . m .J
Fig. 15. Statistical comparison of various parameters of the cross- correlations from the rest pattern A (left columns) and the attack pattern B (right columns) to the EOD rate responses of seven G. petersii. The means (n=23) are shown together with their + 1 standard error ranges. The levels of significance of differences be- tween paired series of experiments are shown below the base line.
Note that the maximal cross-correlation coefficient (taken as the absolute value) is lower but that the number of significant coeffi- cients (P < 0.001) is greater during rest pattern stimulation com- pared with attack pattern stimulation. The lags of maximal cross- correlation to the stimulus patterns were approx. 2.6 s during rest pattern stimulation, and 0.8 s during attack pattern stimulation.
The data were analyzed by the paired t-test (left pair of colmnns), and by the Wilcoxon matched-pairs signed-ranks test (all other pairs of columns)
tion was also greater. During rest pattern stimulation an average lag near 2.6 s was found; during attack pattern stimulation t]he average lag was only 0.8 s (P < 0.01, Wilcoxon matched-pairs signed-ranks test, Fig. 15).
Discussion
Time series analysis methods were extensively used in the present work in order to detect differences in response patterns to two kinds of stimulation pat- terns. It was shown that these methods were able to quantify characteristic parameters for both the stimulation and the response patterns.
The results clearly showed that
G. pertersiidis-
tinguished the two types of conspecific discharge time
patterns, resting EOD activity (pattern A) and ago-
nistic (attacking) EOD activity (pattern B), since motor
as well as electric responses were significantly differ-
ent for both stimulation patterns. It has to be kept
in mind, however, that the attack pattern used in
this study had been recorded from
a G. petersiiattack-
ing a
M o r m y r u s rume,a mormyrid fish displaying
distinct EOD patterns (Kramer, 1976a). A compari-
son with the EOD interaction observed in pairs of
G. petersii(Moller and Bauer, 1973; Bell et al., 1974)
showed that the EOD patterns of
a G. petersiiattack-
ing
a M. rumeare very similar to those, especially
the attack-associated discharge rate increases and
78 B. Kramer : Responses of Gnathonemus to Play-back of Social Signals high discharge rates (which seem to be the i m p o r t a n t
parts o f the signal; see below). T h u s it a p p e a r s justified to c o n c l u d e t h a t at least two specific time patterns o f pulses e n c o d e different ' m e s s a g e s ' in the intraspecific c o m m u n i c a t i o n system o f G. petersii. It w o u l d be intersting to k n o w whether G. p e t e r s i i w o u l d also recognize t w o o t h e r types o f E O D p a t t e r n s : s w i m m i n g ( l o c o m o t o r y ) E O D activity ( K r a m e r a n d Bauer, 1976), a n d fleeing E O D activity d u r i n g ago- nistic b e h a v i o u r ( K r a m e r , 1976 b).
T h e relationships between stimulus patterns a n d response patterns do n o t seem to be simple. While the pulse rate h i s t o g r a m d a t a (Fig. 11) m i g h t p a r t l y be explained by a very simple ' i m i t a t i o n m o d e l ' (a stimulation p a t t e r n with a b r o a d distribution o f pulse rates, s h o w n in Fig. 3b, elicited a similarly b r o a d response distribution o f pulse rates) n o such h y p o - thesis c a n be m a i n t a i n e d for p a r t o f the a u t o c o r r e l a - tion results o f the s a m e data. T h e correlations be- tween adjacent d a t a points were stronger during rest p a t t e r n stimulation t h a n during a t t a c k p a t t e r n stimu- lation (Figs. 10 a n d 11), a l t h o u g h for the t w o stimula- tion patterns the reverse results h a d been f o u n d (Fig. 4). Despite this, the periods o f time o f positive correlations o f the E O D response patterns were lon- ger d u r i n g rest p a t t e r n stimulation (2 s) t h a n d u r i n g a t t a c k p a t t e r n stimulation (1.2 s, Fig. 11), a n d thus a p p r o x i m a t e l y m a t c h e d properties o f the stimulation patterns (rest p a t t e r n : 2 s ; a t t a c k p a t t e r n : 0 . 8 s , Fig. 4). I n a similar fashion, high frequencies o f dis- charge rate changes were elicited by the a t t a c k p a t t e r n (Figs. 12 a n d 13), itself c h a r a c t e r i z e d by a high-fre- q u e n c y c o n t e n t (Fig. 5b). Surprisingly, the rapidly fluctuating a t t a c k p a t t e r n was s h o w n to exert a con- siderably stronger influence o n the response p a t t e r n t h a n did the rest pattern, as m e a s u r e d by cross-corre- lation (Figs. 14 a n d 15). I n addition, this stronger c o r r e l a t i o n o c c u r r e d at a shorter lag. This p r e s u m a b l y shows t h a t r a p i d discharge-rate changes m a y be a m o n g the i m p o r t a n t ' f e a t u r e s ' t h a t are detected b y the fish a n d t h a t influence a r o u s a l a n d m o t i v a t i o n a l state.
A m o n g other possibilities, this h y p o t h e s i s c a n be tested b y the experimental use o f a synthetic stimula- tion p a t t e r n c o m b i n i n g s o m e properties o f the rest p a t t e r n ( E O D interval distribution, a n d thus E O D m e a n rate) with a high f r e q u e n c y s p e c t r u m o f dis- charge rate changes, a p r o p e r t y o f the a t t a c k pattern, by a p p r o p r i a t e l y r e a r r a n g i n g the E O D interval se- quence. I n this w a y the effects o f the stimulus pulse m e a n rate (which was lower in the rest p a t t e r n t h a n in the a t t a c k pattern, Fig. 3) could p r o b a b l y be dis- tinguished f r o m the effects o f a s p e c t r u m o f h i g h frequencies o f stimulus pulse rate changes alone.
Acknowledgments. This work was supported by research grants Kr 446/4-6 from the Deutsche Forschungsgemeinschaft. I wish to thank Prof. Dr. W.T. Keeton and Prof. Dr. H. Markl for valuable suggestions and critically reading the manuscript, and H.-J. Burger for expert technical assistance.
References
Bauer, R. : High electrical discharge frequency during aggressive behaviour in a mormyrid fish, Gnathonemus petersii. Experientia 28, 669 (1972)
Bauer, R.: Electric organ discharge activity of resting and stimu- lated Gnathonemus petersii (Mormyridae). Behaviour 50, 306 323 (1974)
Bauer, R., Kramer, B.: Agonistic behaviour in mormyrid fish:
Latency-relationship between the electric discharges of Gna- thonemus petersii and Mormyrus rume. Experientia 30, 51 52 (I974)
Belbenoit, P. : Relations entre la motricit6 et la d6charge ~lectrique chez les Mormyridae (Teleostei). J. Physiol. Paris 65, 197A (1972)
Bell, C.C, Myers, J.P., Russell, C.J. : Electric organ discharge pat- terns during dominance related behavioural displays in Gna- thonemuspetersii (Mormyridae). J. Comp. Physiol. 92, 201-228 (1974)
Boudinot, M.:Contribution A l'6tude des caract6ristiques 61ec- triques de l'organe effecteur de Gnathonemus petersii. M6moire pr6sent6 au conservatoire national des Arts et m6tiers pour obtenir le dipl6me d'ing6nieur C.N.A.M., Paris 1972 Box, G.E.P., Jenkins, G.M.: Time series analysis: Forecasting
and control. San Francisco: Holden Day 1976
Harder, W., Schief, A., Uhlemann, H. : Zur Empfindlichkeit des schwachelektrischen Fisches Gnathonemus petersii (Gthr.1862) (Mormyriformes, Teleostei) gegenfiber elektrischen Feldern. Z.
Vergl. Physiol. 54, 89-108 (1967)
Heiligenberg, W. : Electrolocation and jamming avoidance in the mormyrid fish Brienomyrus. J. Comp. Physiol. 109, 357 372 (1976)
Heiligenberg, W. : Principles of electrolocation and jamming avoid- ance in electric fish. A neuroethological approach. In: Studies of brain function, Vol. 1. Braitenberg, V., Barlow, H.B., Florey, E., Grfisser, O.-J., Loos, H. van der (eds.). Berlin, Heidelberg, New York: Springer 1977
Kramer, B. : Electric organ discharge interaction during interspe- cific agonistic behaviour in freely swimming mormyrid fish. A method to evaluate two (or more) simultaneous time series of events with a digital analyser. J. Comp. Physiol. 93, 203-235 (1974)
Kramer, B. : Electric signalling during aggressive behaviour in Mor- myrus rume (Mormyridae, Teleostei). Naturwissenschaften 63, 48 (1976a)
Kramer, B. : Flight-associated discharge pattern in a weakly electric fish, Gnathonemus petersii (Mormyridae, Teleostei). Behaviour 59, 88 95 (1976b)
Kramer, B. : The attack frequency of Gnathonemus petersii towards electrically silent (denervated) and intact conspecifics, and towards another mormyrid (Brienomyrus niger). Behav. Ecol.
Sociobiol. 1, 425-446 (1976e)
Kramer, B. : Spontaneous discharge rhythms and social signalling in the weakly electric fish, Pollimyrus isidori (Cuvier et Valen- ciennes) (Mormyridae, Teleostei). Behav. Ecol. Sociobiol. 4, 61-74 (1978)
Kramer, B., Bauer, R. : Agonistic behaviour and electric signalling
B. Kramer : Responses of Gnathonemus to Play-back of Social Signals 79 in a mormyrid fish, Gnathonemus petersii. Behav. Ecol. Socio-
biol. 1, 45-61 (1976)
M6hres, F.P. : Elektrische Entladungen im Dienste der Revierab- grenzung bei Fischen. Naturwissenschaften 44, 431~432 (1957) Moller, P. : ' C o m m u n i c a t i o n ' in weakly electric fish, Gnathonemus niger (Mormyridae). I. Variation of electric organ discharge (EOD) frequency elicited by controlled electric stimuli. Anim.
Behav. 18, 768 786 (1970)
Moller, P. : Electric signals and schooling behaviour in a weakly electric fish, Marcusenius eyprinoides L. (Mormyriformes).
Science 193, 697-699 (i976)
Moller, P., Bauer, R.: ' C o m m u n i c a t i o n ' in weakly electric fish, Gnathonemus petersii (Mormyridae). II. Interaction of electric organ discharge activities of two fish. Anita. Behav. 21, 501 512 (1973)
Russell, C.J., Myers, J.P., Bell, C.C. : The echo response in Gna-
thonemus petersii (Mormyridae). J. Comp. Physiol. 92, 181-200 (1974)
Stinger, K. : Ein Beitrag zur Kl~irung der Rolle der Entladungen des elektrischen Organs im Verhalten yon Gnathonemus petersii (Gthr 1862) (Mormyriformes, Teleostei). Sitzungsber. Oesterr.
Akad. Wiss. Math. Naturwiss. KI., Abt. I, 180, 29 48 (1972)
Serrier, J. : Modifications instan6es du rythme de l'activit6 ~lec- trique d'un Mormyre, Gnathonemus petersii, provoqu6es par la stimulation 61ectrique artificielle de ses ~lectror6cepteurs. J.
Physiol. Paris 66, 713-728 (1973)
Siegel, S. : Nonparametric statistics for the behavioral sciences.
New York: McGraw-Hill 1956
Szabo, T., cited by Lissmann, H.W. : Ecological studies on gymno- tids. In: Bioelectrogenesis. Chagas, C., Paes de Carvalho, A.
(eds.), pp. 215-226. Amsterdam: Elsevier 1961
