AGGRESSIVE BEHAVIOR AMONG MALES
OF THE PARADISE FISH,
MACROPODUS .OPERCULARIS

Thais Ior Tho Dog”. of M. S.

MICHIGAN STATE UNIVERSITY

Howard G. MacMIIIan, Jr.
1958

MGANICHI

3 1293 01591H40

LIBRARY

Michigan State
University

 

.’ fNPI"~r-.'" ‘I rx 1‘. 7:?“7'. T T’\{ ' 'N“’\","I 7.7, T"! "
nuunpublju gaudy/hula hinvl‘du 11111.1...)

(‘1 7')

--.-. *r.-'" T' Z“: .,W,r.;.-’~.-.--.nv_:
Vl‘ Till; I’hLUXDIULL I‘lbf., uu'svuvLUl/Uu

 

OPLRC UMR’ IS

 

By

Howard G. Machillan, Jr.

AN ABSTRACT
Submitted to the College of bcience
and Arts, Michigan btate University,
in partial fulfillment of the
requirements for the degree of
MAJTER OF 03

Department of Ecology

1958

Approved gum G. W

Howard G. MacMillan, Jr.

OF THE PARADISE FISH, MACROFODUS

. r .

OFiRCULARIS

 

ABSTRACT
It is generally known that the male Paradise fish,

Macroppdus opercularis, exhibits aggressive behavior. This

 

study was intended as an analysis of the activities involved.
It was discovered that certain sets of pair-mates fought,
while others established a dominance-subordination relation-
ship without fighting, and a comparison was made between
them .

The fish studied in these observations were obtained
from a dealer. All laboratory work was done in the
Natural Science Building of Michigan State University
during the spring and summer of 1957. Fifty-nine sets
of observations were recorded and fighting was observed
in twenty-five. Thirty-three fish were used.

In those observations in which fighting occurred
three distinct stages were noted. Aggressive behavior
was first noted prior to fighting and was characterized
by activities involving the pair-mates. and reactions of
the fish to their surroundings. In those observations

wnere no fighting occurred, this was replaced by a period

prior to the establishment of dominance. The period was
terminated by the first retreat by one of the pair-mates.

Periods of actual fighting involved the greates
amount of activity. Mutual exchanges of bites and look-
ing of jaws were characteristic only of this stage, and
actual physical contact between the pair-mates occurred
with greatly increased frequency as compared with the
preceding period. Fighting ended with a retreat by one
of the pair-mates.

Following surrender or the establishment of dominance,
similar activities were noted,and it was assumed that these
stages were equivalent to each other. Certain obvious
differences were observed, however, when winners were
compared with losers. The winners and dominant fish
were more aggressive while the losers and subordinates
remained motionless, reacting only when approached by an
aggressor.’

A brief‘and generalized comparison was madexwith

the behavior of the Siamese fighting fish, Betta spjendens

 

 

(Braddock and Braddock, 1955). It was found that the be-
havior of the Bettas resembled that of the Paradise fish

in many of its details

LITLHATURE {31"}; ‘

Braddock, J. C. and Zora I. 1955. Aggressive behavior
among females of the Siamese fighting fish, Betta
splendens. Physiol. 2001., 28: 152-72,

AGGRL;SIV£ BLdAVIOfi AMQLG haLnd

Oi" Th2} PAI‘LADIQLS rise, QAGRJPC‘LUS

 

7—.

OfnhCULAnIS

 

By

Howard G. MacMillan, Jr.

A‘Hflfiis
Submitted to the College of Science and
Arts, Michigan State University, in

partial fulfillment of the requirements
for the degree of

hthLR OF SCIENCE

Department of Zoology

1958

Approved

' —r-.V:r.-:‘ j" ”'77 \‘
ACAA uumiJCnLnumeu

The author wishes to express his sincere
appreciation to Dr. James C. Braddock, of the
Department of Zoology, for his criticism and guid-
ance, and his encouragement and interest during the
course of this investigation.

Special thanks is also due Dr. John R. Shaver
and Dr. T. tayne Porter of the Department of Z0010gy,
and Mrs. Jane Smith of the Department of Geology for
their help and advice.

Appreciation is also due Mrs. B. R. henderson,
secretary of the Department of ZOOIOgy, for her en-
couragement, and to the author's wife, Cheryl, for

her patience and support.

‘i‘nBiaL or Cin'i'Isléio‘

Acknowledgements

Table of Contents

List of Tables

I. introduction

II. methods and Materials

III. Results

7.

Activities Prior to the Lstablishment
of Dominance without Fighting

activities Prior to righting
Activities During the Period of Fighting

Activities Following the Establishment
of Lominance without Fighting

activities Following Surrender Sub-
sequent to Fighting

Intensities of Color Saturation

Comparison w th remale Betta splendens

 

IV. Discussion

V. Summary

VI. BibliOgraphy

rage

ii
iii

12
16

21

22

26
31
to
1+5

II:

III

IS

SS

63

Ia '
:41
2+1

I‘)
o

to

41¢

\1-‘1
o

O\

10.

Durations of Periods Prior to the
Lstablishment Oi Dominance without
figttin”

Time elapsed before rirst Social
Reaction Prior to Lstablishment of
Dominance

frequencies of ”arious nCtiVitiGS
Prior to the Ls ablisnmsnt of
Lominar ce witho t '

Iiae Llapsed before xirst so ial
hcaction Prior to iightirg
.requencias of Various 5 ctivities
During the xeriois rrior to fighti.g

g c . iious activities
D‘rin“ the Leriods of righting

Activities ro ll owing the Lstablishment
[12.

1198 n IthCUt .I.‘ iéiitirlg

1 ”7'
“t $51.6

i5“

\0

Ho

I. INTRODUCTION

It has been known for some time that the Paradise
fish, Macropgdus opercularis exhibits aggressive behavior,
but this is the first formal study to be undertaken.
Several papers have been publishedczoncerning the
aggressive behavior of other species of fish. Braddock
and Braddock (1955) reported that the female Battg
splendens displays elaborate fighting behavior. Another
study (Braddock 19h5) described the dominance-subordination
relationship in homosexual and heterosexual groups cf both
females and males of Platypoecilus maculatus. Hess (1952)
in a slightly different type of study found that by using
models representing male Bettas and by varying the tem-
perature of the water, he could cause the duration of the
challenge response in male Bettas to vary. In his study
of the Green Sunfish, Greenberg (lth) discovered that in
these fishes, heirarchy represents distinct levels of
aggressiveness while territoriality results from a balance
between almost equally aggressive individuals. Newman (1956)
found that a relatively complicated dominance-subordination
relationship occurred in twesspecies of trout. The
dominant individuals of both species were light in color
while the subordinates were darker. Also size was an

important factor in determining dominance, but sex was

not significant.

The purpose of this study was to discover and
analyze, as far as possible, the various aggressive
behavioral traits characteristic of the Paradise fish.

A comparison was then made between those fish that did
fight and those that established a dominance-subordination
relationship without fighting.

It was also hoped that certain consistent differ-
ences in detail would be found to occur between the

aggressive patterns of M. opegcularis and B. splendegg

 

 

which could be used as criteria in future experiments

involving these two species.

II. METHODS AND MATERIALS

All observations reported in this paper were made
in a laboratory located on the second floor of the
Natural Science Building at Michigan State University.
Three large windows facing south admitted insufficient
light. Therefore four desk lamps with 100 watt bulbs
were used to compensate for the lack of light from the
windows.

The laboratory was heated by thermostatically
controlled steam radiators located undert;he windows
and well away from the aquaria. "Observations were con-
ducted during the spring and summer when there is normally
a wide temperature range. Therefore, the range of 78° -
8&0 F in water temperature was unavoidable.

Tap water from the building was used after it had
been aerated for several days.

The fish were housed in clear-glass, gallon jugs
that had been cut off near the top to provide a wider
opening. Approximately 5 inches of water were kept in
these containers at all times. No gravel or plants were
present. Waste products were siphoned out every seven
days. During observation periods the fish were placed in
an aquarium whose dimensions were as follows:

30.5 x 26.5 x 26.5 cm. This was filled to a depth of
10 inches. No snails, plants or gravel were put in

this aquarium.

Small pieces of finely ground, uncooked shrimp
were given to each fish every morning. This diet was

supplemented with live vestigial Drosophilg once a week.

 

The fish observed were all adult males obtained
from a dealer. Their exact ages were not known. When
individuals were paired, their relative sizes were noted
prior to each observation and an attempt was made to
match them according to size. Thirty-three individuals
were used in this study,£nd fifty-nine pairings were
made.

Before theifish were placed together for observation,
the jars containing them were placed near the observation
aquarium, and the water temperatures were equalized by
the room temperature. The fish were then gently netted
one after the other, as quickly as possible, and placed
in the aquarium. This never required more than twenty
seconds. Observations lasted a minimum of thirty minutes
and, if longer, were terminated only when mutual behavior
ceased.

A record of all observed activities was kept for
each observation period. This was accomplished by the
use of symbols representing different activities.

At the end of each observation period the fish were
placed in their respective containers. A small amount of
salt was added to the water in cases where a fish was

injured. ho individual was paired again before its wounds

were completely healed.

~~ T. -T 3 A \ , t. A " “ . " - J-I. 4 " - ‘i' r I“
1."qu Calljbfl¢vl Du .LIJ. fur/Drtmiic, vtiu va‘LHK/AICG UL DVJLIU»)
1 ‘45" r . . \ Q.-‘\ a . r. . r ‘- L‘r . 1 . -. fi ' : \/ ‘fi
in e on pairing “a3 divided into tie lollOMing sectiois:
9. 91...? a, .02 j n 9 . u _ ,- °-‘
10 aniJibiuo JJI’IOI‘ DO ILL-lIIULIIL) -.L on" botabllULl.‘

 

\

v “ .n w . v..~-9v,r‘—.'.‘~ '1'. ‘ ’~~ 0—. .-- 9-, >- -. -. - 3' "’ .' A
iflbnt U1 (104.1,;in U u ltLIUDLt x .LCLMLLQLLIQO Lilia pdl’iad ILL"

 

..,«. ~.. ., 2.1-9 t- 9 - ~. was - r-. .. ' - . ~ ~ .—
Cluubd all activities that occuricd irom the time Lo n

t":
[—‘o
L.
r
p.

n were placed in the aquarium until actual fi

-subordination relation—

(0

-_ . .A y‘
Umliiani‘.

(I)
C;

init

to

began or until a de

ship was established in its absence. In those instances

(2.:

.- V ”x, . - ,, .-.-, - ., , ~ " ,. A. 1,‘ .0, H ,. - . ,- r,‘ g . y
in union LO actual iiQu toox place, this period L s Judge

inite retreat by one of the

s “or od bezan

L,

w-
P-

2. Activities during fighting. Th
with the first exchange of bites before dominance was
established, and terminated with the first definite retreat

by one of the pair-mates.

U1

 

'flL: y: .'\ ~ "A ‘- '.~I~-:~~. A +- ’F‘fi .f‘1r

.3. .nccivities subsequent 13 Il .i.n.- or the establisn

W ' I . 1'": ‘-"~ :. 4‘1". x " ' n .‘ . rw r J e. W . :1 A —.. v -. INT-x
melt of dominance without fighting. ihis period began with

 

the first definite retreat by one of the pair-mates whether

or not a fight had taken place and ended when mutual

aw ressive activities c ased.

(D

III . mes ULTS

1. Activities prior to the establishment of

 

0 1

dominance without fighting. A dominance-subordination

 

relationship is not established immediately after two
male Paradise fish are placed together in an aquarium.
first the fish react in various ways to each other and
to their surroundings. These social reactions result in
fighting or the establishment of a dominance-subordination
relationship without fighting. The following concerns-the
nature of this phase in those instances where no fighting
eventuated. The period varied in duration (Table 1).
TABLE
Durations of Periods Prior to the
Lstablishment of Dominance

Rithout Fighting
I .
(34 iairings)

Time HO. Of
(nin.) Pairings
0-!2 0.0.0.0......0...0000 l3

")

LT:?.) 0.000000000.000000000 (3)
0-8 000000....00000..0.00 2
8-10 .000...0.......0.00.. 0
10-12 .000.0.0.00.00..00... 2
12-1! .0..0..00000.0......0 2
“1b 000.000.00.00000..000 1
1\-ld 0....0...0...0....... 1
10-20 .00....00..000...0000 0
20-22 0.....0...000000.0.00 1

Among the pairings the longest such period lasted 21

minutes, the shortest 0.25 minutes, and the mean was

5.2 minutes. In 22 of the BA observations the duration
I

was 4 minutes or less. On the basis of the frequencv

classes employed in Table l the mode was 0-2 minutes.

In 29 of the 3h observations, the first social
reaction occurred within 0.5 minutes, and in 19, one
or both fish reacted socially as soon as the second
fish was placed in the aquarium. In the other 15
observations the maximum time before the.flrst reaction
was h.2 minutes, the minimum 0.25 minutes, and the mean,
0.9 minutes (Table 2).4 0n the basis of the frequency
classes employed in Table 2 the mode was 0.0 - 0.25
minutes. The maximum time of h.2 minutes was exhibited
by one extremely passive pair. Not including those fish
that showed immediate reactions, the greatest number of
pairs (8) showed initial reactions at 0.25 - 0.50 minutes.
The other 5 pair-mates explored or remained motionless
prior to their first social reactions.

TABLE 2
Time Elapsed Before first Social Reaction.

Prior to the Lstablishment of Dominance
(3h Pairings)

Time . No. of
(Min.) Pairings
0.0-0.2; ................... 21
0.25-0.50 ................... 8
0.50-0.75 ................... O
0.75-1.00 ................... 2
1000‘1025 0.0000000000000000. 0
1025‘1050 000000000000.000000 O
1050‘1075 000.00.000.00...... 1
1.75'2000 000.000.00.00000000 1
.00-t.25 ................... l

I

In every case the first reaction observed between
pair-mates took the form of challenging. In 33 oft:he
3h flpairings she first challenge was individual, while
in one instance it was mutual. Individual cha lenging

approaches the other,

F)

is a stereotyped reaction. One fisl

'0

holds its body in a rigid position,£1nd spreads its cauda.
fin. The gill covers may or may not be erected. The color
of the fins and body usually becomes more fully saturated.
Individual challenges may be of either short or long duration,
and may or may not occur in a rapid series. The total number
of individual challenges for the 3h observations in which no
fight occurred was 750, an average of 12 per individual.
Both pair mates exhibited individual challenging in 29
pairings, and one only in 5. The maximum number per fish
during this period was 55, the minimum 0,21nd the mean ll.
for contact pairs these figures were 10h, 1, and 22.re-
spectively (Table 3).

Mutual challenging occurs when two fish challenge at
the same time. The two fish approach each other and stop
in a head to tail position. Their bodies become rigid, and
their tail fins are spread. The gill covers may or may not
be erected. rrequently the bodies of both fish vibrate
rapidly, and they sink slowly to the bottom of the aquarium.
Mutual challenges varied slightly in duration, lasting from
approximately 3 seconds to 10 seconds, but unlike individual
challenges, they did not occur in rapid succession. ire-
quendy an individual challenge became a mutual challenge if
the original challenger maintained his pbsition.

Mutual challenging occurred throughout the period
prior to the establish‘ent of dominance without fighting.
The total number of mutual challenges was 100, and they
occurred in 22 pairings. Considering only those pairings

in which this activity was observed, the maximum number per

contact pair was 16, the minimum 1, and the mean 5 (Table 3).
then mutual challenges are considered as individual

challenges of each pair-mate, the maximum number of all

challenges per individual was 69, the minimum 0, and the

mean 1h. for contact pairs these figures were 136, 3, and

28 respectively. some sort of challenging occurred in all

experiments with a total of 950 challenges. Only 3 of the

6d pair-mates did not challenge at all.

TABLE 3

Frequencies of Various activities Prior to the
Lstablisnment of Lominance without righting
(3h Pairings)

. Lance

Indiv. Mut. All Chas- Bib- y for

Chal. Chalfi Cnal. ingfi Ret.% ingfi Exphnfi Pos.*
Totals for _
all Pairings 50 100 950 15 11 6 57 100
No.Pair-mates
Involved 63 hh b; 2 2 6 20 Ah
N0.Pairings
Involved 3h 22 3h 2 2 5 12 22
hax./Indiv. 55 16 69 11 18
Min./Indiv. 0 1 c 1 l
V/Indiv. ll 5 1h 3 5
max./Pairing 10h 10 136 13 18
Min./Pairing 1 1 3 1 1
V/Pairing 22 S 28 5 5

*Maximum, minimum, and mean figures apply only
to those pairings where activitity occurred.

Chasing is the pursuit, either rapid or slew, of one

pair-mate by the other. This occurred in only 2 of the 3h

10.

pairings, . and only 2 individuals exhibited this behavior
giving a total of 15 instances for all pairings. -
Retreating occurred only as a response to chasing.
It took the form of either rapid or slow movement away
from the other fish. Eleven retreats were recorded.
Biting was noted in 5 of the 3h pairings. In one
pairing both fish bit, while in four, only one of the
pair-mates exhibited this activity. The total number of
bites recorded was 6, and all individuals showing this
activity bit only once.
Exploring was observed in 12 pairings (Table 3)
and involved swimming into various regions of the aquarium,
usually picking up bits of algae and other small particles
in what appeared to be an attempt to feed. The total
number of instances for all pairings was 57 with 20 pair-
mates involved. Considering only those individuals that
actually explored, the maximum number of instances per
individual was 11, the minimum 1, and the mean 3. Cor-
responding figures for contact pairs were 13, l, and 5.
Dancing for position is a behavior trait involving
both fishtand was observed in every period in most of the
pairings. The fish approach each other and begin to swim
in a tight circle usually maintaining a very close distance
between them. Individual and mutual challenging may both
occur. Dancing for position was observed 100 timeszand

was noted in 22 of the 3h pairings. . Based only on those

pairings in which this activity took place, the maximum
number per contact pair was 18, the minimum 1, and the
mean 5 (Table 3).

As shown in Table 3, challenging was the most common
activity during the period prior to the estab1131ment of
dominance without fighting. Dancing for position was
second. Exploring, chasing, retreating,zand biting followed
in order. Actually, biting was noted in more pairings
and more pair-mates were involved than was the case for

chasing and retreating.

l2.

2. gctivitie§_pgior £2 fighting. In those pairings where
fighting occurred the activities that preceded it re—
sembled those noted in the observations where dominance
was established without fighting. Also the range of
variation in dura ion of the initial period was similar.
The maximum was 16 minutes, the minimum 1.5 minutes, and
the mean 6.3 minutes (Table A). In 19 of the 25 pairings
the duration was 6 minutes or less. 0n the basis ofthe
frequency classes employed in Table A the mode was 6-8_

minutes.

TABLE 1+

Durations of Periods Prior to Fighting
(25 Pairings)

Time No. of
(Min. ‘ Pairings
C”? 0.00.00.00.0000000000... Z

2‘ .00000000000000000000000

4-t 00.000000000000000000000 u
0-8 0.0.00000000000000000... 7
0-10 0000.00.00.000000000000. 2
10’1? 00.00.00.000000000000000 1
12‘1‘ 000.000.00.0000w00000000 1
,-lg .00000000000000000000000 l
lb‘ld ‘ 0.00.0..0000000000000000 1.

In 19 of the 25 pairings one or both pair-mates
reacted to the other as soon as the second fishvvas placed
in the aquarium. In the other 6 the maximum time before
the first reaction was 2 minutes, the minimum 0.33 minutes,
and the mean 0.8 minutes. 0n the basis of the frequency

classes employed in Table 5 the mode was 0.0-0.25 minutes.

F]
b)
O

T ,~ .« 4‘ ‘ P) y- ( : 1 ~. vr‘ 1 -~ -' 3 ‘- . r ‘ -\ r- v.-
In 2+ oi the 25 fulllueb the first Pedutlon between
the pair-mates took the form of challeng ng, while the

other initial reaction was dancing for position.

Tine Elapsed before rirst oocial Reaction
Prior to Fighting
(25 Pairings)

Time No. of
(hin.) Pai-ring s
000 “0.25 00.000.00.00.00.....00 19
0.25-0. :0 ...................... 3
00/0‘0075 00000000000000.0000... l
0.15-1.0‘ 00.00.0000.000.00.0.0. 1
1.0 0‘1025 00.0000000000.0.0.00.. O
1.2)“1050 .......000.......0.... O
1.30-1.75 ...................... O
1.75—2.00 ...................... 1

Seven of the initial challenges were mutual,21nd 17 were

The total number of individual challenges for all
23 pairinvs was 103h. Both pair-mates exhibited individual
challenging in all pairings, and the maxitzum number per

al was 02, the minimum 2, and the mean 21. for

g-

iiuii.v

contact pairs these fQ ures were 118, 6, and hi respectively

)\

(Table
mutual challenges were observ ed in all of the pairings
and totalled 263 (Table 6). The maximum number per contac
pair was 23, the minimum L, and the mean 11. nine in,
idual challenges, mut“al challenges were more numerous
prior to fi‘hting than in the corresponding period where

no fighting occurred.

TABLE. 6

Frequencies of Various Activities during the
Periodsirior to Eighting
(2S Pairings)

Dance
indiv. Mut. All for
hal. Chal. Chal. Bitingfi hxolor.* Position
Totals for

all Pairings 103M 263 1550 1A 8 207
No.Pair-nates

involved 50 SO SO 12 _ 7 #8
Ho. Pairings

involved 25 25 2S 7 5 2h
max./lndiv. 62 23 94 ' 2h
nin./lndiv. 2 h C
V/lndiv. 21 11 '31 8
max./Pairing 118 23 1L2 ‘24
:~.i:1./Pairing 6 ii 111 O
V/rairing hl 11 62 8 ,

nMaximum, minimum, and mean figures apply only to
those pairings where activity occurred.

If mutual and individual challenges are combined and
treated simply51s general challenges the maximum per in-
dividual was 9h, the minimum 7, and the mean 31. Corres-
ponding figures for contact pairs were th, 1h, and 62.
The total;for all challenges for the 25 pairings was 1560.

Biting occurred in 7 of the 25 pairings and involved

12 air-mates. Only 1h bites were recorded (Table 6).

’U

Cons1dering only those individuals that bit, the maximum

per fish was 3, the minimum 1, and the mean 1. It was

also noted that biting occurred more frequently during this

period than in the corresponding period of the non-fighters.
The mean per pairing was 2.

exploring was considerably less prevalent in the period
prior to fighting than was he case when dominance was

established without fighting. In the former it was 0 served

0"

in only 5 of the 25 pairings. Only 7 pair-mates were in-
volved and the total number of instances was 8 (Table 6).
Considering only those individuals that actufl.ly explored,
the maximum per individual was 2, the minimum 1, and the
mean 1. The mean per pairing was 2.

Dancing for position was observed in 2h of the 25
pairings and a total 207 examples was recorded. The maximum
number of instances per contact pair was 2h, the minimum 0,
and the mean 8 (Table 6). This activity also occurred more-
often than in the corresponding period where no fighting

followed.

5.3

able 6 indicates that challenging wastshe most
frequent activity during the period prior to i
Lancing for position had the second greatest frequency and
was iollowed bv biting and exploring in that order.

— 3 .-‘ '\ 1 e ( \a .w- 1
g did not occur dor1ig the

I.)

.11

:9
cf

Chasing and retre

ds prior to actual fighting.

H0
(J

observations of the per

.J

‘
0" n
LJV

t

‘5 I 1

11¢

(1‘

1)

H-

r. o -, a
riod oi f1pnt

C .

p I1 (.3.- 0

+1-

C‘ U. 1”

L0

:3
.Lb

F10

tiv

O'

3. n

p

This period began with th first exchange of bites by the

r

pair-mates. In the 25 pairings in which fighting took place,
individual challenges immediately preceded theITights in 15,
mutual challenges in 7, and dancing for position in 3.

This period varied in duration (Table 7). The maximum
was 25.66 minutes, the minimum 0.05 minutes, and the mean
8.6 minutes. In the 22 pairings where duration was recorded,
10 such periods lasted 2 minutes or less. 0n the basis of
the frequency classes employed in Table 7 the mode was
0-1 minute. In 3 pairings fighting had not ended at the
termination of the obServation periods, and the actual
durations were not recorded. For consistency with the other
observations a limit of 30 minutes was given to these 3
instances when making computations concerning the various

activities.

TABLE 7

Lurations of flights
(22 Pairings)
Time ho. of
(221.111.) Fights

wwwoooowwoowmwmw

Certain activities characteristic of the preceding
period were also noted during fighting. iThese were indiv~
idual and mutual challenging, biting by one pair-mate, and
dancing for position. Exploring was not observed. In con-
trast, multiple exchanges of bites and locking of jaws oc-
curred only during f1: J1ting. All activities during t
figh ing period seemed to be accelerated, and both fish
generally moved very rapidly at all times.

Individual challen nging occurred in 23 of the 25 pairines,
and a total of 858 such challenges were recorded. 1he max—
imum number per individual was 121, the minimum 0, and the
mean 17. For contact pairs these figures were 230, 0, and
3h respectively (Table 3).

mutual challenges totalled 333 and were observed 1h
20 pairings. The maximum per contact pair was 65, the
minimum 0, and the mean 13 (Table 8).

When individual and mutual challenges were combined
by counting the mutual challenges as individual challenges
for each of the pa air— —mates, it was evident that some sort
of challenging took place in 2h of the 25 pairings. There
was a total of l52h challenges for all pairings. rorty-
five individuals were involved. The maximum number of all
challenges per individual was 18 , the minimum 0, and the
mean 30. Corresponding figures for contact pairs were 36 ,

0, and 61 (Table 8). In one pairing neither pair-mate
showed any type of challenging, and in 3, one fish did

not challenge.

Multiple exchanges of bites were characteristic only
of the period of actual fighting. These were most fre—
quently directed at the mouth, but also involved other

parts of the body. since this activity was very rapid,

it was inpo Mble to count the number of sin51e bites that
took place, and only the number of exchanges was recorded.

‘Ihe frequency oi these multiple exchanges varied greatly.
many times several occurred without bei 1n5 interrupted by
any other type o1 activity. lore often, however, a
multiple e} :change took place and was interrupted by some
other activity such as challenging or dancing for position
before it occurred again. Frequently multiple exchanges
of bites were terminated by locking of jaws.' ihe maximum
number of n1ultiple exchang es of bites per contact pair
was 52, he minimum 0, and the mean 13. A total of 317

Hult le exchan5e s were recorded.for the 22 pairings in
which they occurred \lable 3).

Lancing1 for position was a common activity d1r ing
fidhting. This was accompanied by other activities such
as challengin5 and biting. Dancing for position varied in
duration. Certain instances lasted only a few secor ids
while others continued 10r several minutes. It was ob-
served 2hh tines in the 19 pairings in which it occurred
at all. Con side ring only those pairings, the maximum per

contact pair was LS, the minimum 1, and the mean 13 (Table 8).

 

AB'“ 8
brequencies of Various Activities
During the Feriods of Eighting
(25 lairings)
Other
Challerges Bites Activity
Lance

_ Mult. for Lock
indiv. Hut. Tot. Indiy. hxch. Pos.% Jaws*

 

 

111 "Has 658 333 Isak 777 317 211 17“
no.‘air-nates

In‘Oqud L2 to L5 he an 38 36
1:3. Fairin‘fi

involved 23 20 2 25 22 19 1b
max/indiv. 121 55 166 73 52 us 32
min./indiv. o o o o o 1 1
T/Indiv. 17 13 30 lb 13 13 10
haz./Pairing 230 65 3co 137 52 as 32
min./£airing O O O 1 O 1 1
V/fairing 3t 13 61 31 13 13 10

fimaximum, minimum, and mean figures apply only
to those pairings where activity occurred.

‘

Individual biting was also characteristic of t1is

period but with much greater frequency than in the pre-
ceding period (Table 8). The maximum number per individual
was 73, the minimum 0, nd the mean 16. For contact pairs
these figures were 137, 1, and 31 respectively.-The total
number of bites for the 25 pairings was 777. forty-

eignt pair-mates were involved, but in two pairings one

of the pair-mates did not bite.

1'.-,~.,:1. 1,. :3". 1. _.,,. - .1 1. 1- .° 1 .
Lbbnlné of “ans Occurred only durinu the perlod of

0- .1, 1-.-

11511.1ng. frequently a multiple exchange of bites pre—

(T1

0 ded this activity. Locking of jaws 1nvolved mutual tugginc
and pulling while the pair-mates turned ov r and over as

they slowly sank to the bottom of the aquarium. lhe bodies

H.

as n challenging.

CL

of the fish were rigid and the fins sprea

ho accurate record of the duration of this activity was

made,k>ut some inst ances lasted orfl 1y a few seconds while

«5.5

l-’
O

/
1g as CO seconds. c011s1de11

(Ti
Cu

others last only

11’
C.)

, th

CL.
(D

those “airings in which lockinv of 'aws was observe
1‘) (1 C.

r was 32, the minimum 1, and the

H°

a

maximum per contact

l-(i

’ 9

mean 10. It occurred 1n 18 of the 25 pa a11inbs with a total
of 179 instances (Table 8).

The fighting period involved the greatest amount of
activity in the sequence, and the actual physical contact
between pair-mates was greatest then. In comparison to the
period prior to Ii“rt1nu, individual challenging and mutual
challenging occurred with less frequency. Individual biting

1‘

was much more 1requ ent during the1.i éhting period, while
the occurrence of two new acti “ti 5, multiple exchanges of
bites and locking of jaws, rives further indication of the
vigorous activity characteristic of this period.

Inezach pairing the intensity of color saturation of

both fish was‘greaterd curing this period than in the other

periods of the same pairi" fflra inter1s s1t y frequently va1ie

a

with re sfec ct to a particular fish, but during loching of
jaws, booth pair-mates were as dark in intensity of color

saturation as at any other time in the observation peritd.

inance without fighting. In the absence of.fi5hting, dom—
inance was considered established when one pa air-mate re-
tr ated without subsequent aggressive behavior.

in 2b of the 3h observations individual challenging
immediately preceded the first retreat, while in 0L6, mutual
M'ecede d the

challe “ll? was the preceding activity. Eitin

0
first retreat in 2 pairings and dancing for position in S.
in every case retreating immediately followed some type of
aggressive behavior by one or both fish.

at the moment of the first retreat all of the subordinate
fish quickly became less saturated*. ith regard to th color
of both body and fins. The color saturation of the dominant

I 1

indiv1ouals be cazne less much more slowly, if at l1.

m

0
..

During this period dominant individuals differed 1n

emany respects from their subordinates (Table 9). The former

challenged 1267 times in 3h pairings while the latter chal-
’nged only hag times in 26. The means were 37 and lb, re-

w
A

. 1 _
gca1 cha

q

llenges ex-

H-

V

Ho

spectively. in nearly all cases, ind
.hibited by the subordinates were defensive in nature, and
occu*red only when they were approached andcprnered by the
dominant individuals. The dominants bit the subordinates during
2h pairings and registered a total of 113 observed bites com-
pared to O for the subcrdina es. lhe dominant individuals

did not retreat at allhnt chased the subordinates lllS

times, while the latter retreated 1208 times and did not chase.
nxploring was recorded 225 times in 31 pairings for the dom-
inant fishzand only 51 times in 21 pairings for the sub-
ordinates. The means were, respectively, 7’and h. Thus the
dominant fish had a much greater freedom of movement within
the aquarium. Mutual challenging occurred kl times in 9
pairings, and dancing for position occurred to times in 7.
Mutual reactions were not as frequent as individual reactions
during this period.
TABhn 9
Activities Following the hstablishment
of Dominance Without righting
(3h rairings)
Dance
Indiv. Mut. for
Chal. Chal. Bites Retreat Chasing Explor. Pos.

Dominants 1267 A1 113 o 1115 225 to
Subordinates h09 hl O 1208 O 61 to

5. Activities following_surrender subsequent to fighting,
The periods of active fighting were followed by periods thought
to correspond to the periods following the establish ment
of dominance where no fighting occurred. They began with
the first definite retreat by one of the pair-mates. after
this the individual that retreated exhibited little or no
aggressive behavior.

In 12 of the 22 pairings, individual challenging
immediately preceded the first retreat.' Mutual challenging
preceded it in 2, as did individual biting, while locking of
jaws occurred in h pairings. Dancing for position and
multiple exchanges of bites each preceded the initial re-

treat in one pairing.

Il1e than es in the intensity of color saturation by
both the winners and losers were very simila r to tr1sse shown
in the corresponding periods where no fighting took place.
In every instance the winners were of a darker color sat-
uration than the loser

The winners challenged the losers in 20 nair1ings with
a total of L19 individual challenges. The loser challenged
only 68 times in 11 pairings.' The means were 21 and 6, re-
spectively. host of the losers challenges seemed to be de-
fensive in nature. The winners bit the losers he times in

12 nairin

4L

gs and were not bitten in return. The winners did
not retreat but chased the losers Slb imes. The losers
retreated from the winners hho times and did not chase. The
winners explored 86 times in lo pairings compared to 22 times
in 7 pairings ior the losers, indicating a freer movement on
the part of the winners. The means were 5 and 3, respectively.

The mutual activities, mutual challenging and dancing f

O

r
position, were even more noticeably absent in these periods
tnan in the correspondir g periods without fighting (Table 10).
There were 6 mutual challe enges observed in 3 pairings, and

only 2 instances of danc 11g for position, both in the same

pairing.
TABLE 10
activities rollowing surrender
Subsequent to Fighting
(22 Pairings)
Lance
Indiv. Mut. for
Chal. Chal. Bites Retreats Chasing hxplor. Fos.
Winners h19 t he 0 516 ob 2

Losers 68 o O LAB O 22 2

P1)
4;.

6. IntgnsitiesAgf color saturation. No truly objective

 

 

 

 

 

system was used to classify the differences in color satur-
ation. The fish were merely identified as dark, medium, or
light. An individual identified as "dark" exhibited vivid
blue and orange alternating dorso-ventral stripes along the
body. The dorsalzand ventral fins were bright blue, while
,the caudal fin was bright orange. "Medium" indicated that
the fish clearly showed the blue and orange stripes along
the body, but not as vividly. Also the fins were less
saturated. A "light" fish showed only faint blue and orange
stripes along the body, and the tail fin was pale orange.
The dorsal and ventral fins were almost colorless.

The intensities of color saturation varied greatly
among the individuals, and the classification was not used
to describe the exact intensities of all the fish observed
at all times. It did enable the observer to make a com-
parison of the relative color saturations of each of the
pair-mates during the various phases of each observation
period. Thus, both were consistently dark or medium<iuring
the period prior to the establishment of dominance or fight-
ing as well as during the period of actual fighting. Immed—
iately after dominance was established either by fighting or
not the winner was alwqm; darker than the loser. This does
not mean that the dominant fish was always dark and the sub-
ordinate always light. Instances occurred in which the winner

D

01 one pairing was actually lighter than the loser of another.

In most pairings the winners and dominant fish retained a
darker color saturation than the losers and subordinates
until the observation periods were terminated. In some
cases both pair-mates exhibited the same intensity after

dominance had been established for a period of time.

I‘x)
(*\
O

7. Comparison with female Betta sple11dens.
Braddock and Braddock (1955) gave an extensive account
of t11e azzressive behavior patterns of the Siamese fighting
fish, Eetta splezicWe1s. The data gathered in the study of

the Paradise fish indicate a marked similarity between the

two species.

'u—‘o

iv

Ci-

hhile many of the ac ties were essentially the same

for Bettas and Paradise fish, there was one outstand in ex-
ceptiOn. Dancing for position was noted throughout the
entire observation period in this study while its rough
equivalent, joshe yi gfor position, was observed only
during the period of actual fighting among the Bettas.
Prior to actual fignting there was a period where
the fish reacted both to each other and to th1eir surround-
ings. In both studies the durations of these periods prior
to fighting were similar. Based on 35 pairings the mean
time for the Bettas was 6.6 minutes compared to 6.3 minutes
for 25 pairings of the Paradise fish.

T11e mean time that elapsed before the first social
reaction of the Bettas was 33.h seconds based on BS pair-
ings and hB seconds for the Paradise fish based on 25
pairings. Individual challenging was the predominant
first reaction in both studies. Among the Bettas this
cecurred in 22 of the 3S pairings; the other 13 were
.initiated by mutual chalglen es. In 2h of the 25 Paradise
fish pairings, social reactions began with challenging.
Seventeen of these challenges were individual and 7 were

mutual.

27.

A comparison of the relative frequencies of aggressive
activities throughout the periods prior to fighting
indicates that individual challenging was the most common
reaction in both studies. This activity occurred in
nearly all of the Betta pairings ard was universal among
those of the Paradise fish.

Mutual challenges occurred in all pairings in both
studies and was the second most frequent activity.

Instances of chasing and retreating were recorded
in the Betta observations and were absent in this study
during the period prior to fighting. Biting was much
more frequent in the Betta study. Exploring was noted
in both but instances were not frequent.

In both species some pair-mates fought while others
did not. The period of actual fighting among the Bettas
was of longer duration and seemed to involve greater
activity than the corresponding period among the Paradise
fish. The mean-time for the Betta pairings was 51.5 minutes
for 35 pairings compared to 6.6 minutes for 25 Paradise
fish observation periods.

Similar activities were observed in both species,
anj although they were observed less often in the Paradise
fish study, this was probably because the mean time for
the period of fighting was approximately 6 times less

than in the Betta observations.

The types of bites were analyzed to a greater re ree
in the Betta work than was done in this study. however,
single and multiple exchanges of bites were frequent in both.

Locking of jaws occurred only during the fighting
periods in both studies but was not characteristic of very
short fights (less than 15 minutes) among the Bettas.

Since only 5 pairings of Paradise fish lasted over 15 min-
utes, it is significant that for them this activity was
relatively common.

Like jaw—locking, jockeying for position was observed
only during the fighting period among the Bettas, but its
rough counter-part, dancing for position, was recorded in
all periods of this study.

Injuries were extensive among the Bettas due to their
savage biting while very few were noted among the Paradise
fish.

Definite dominance-subordination relationships were

estab i

r—J

hed in both studies with or w’thout fighting. In

(0

In both, the dominant individuals were more aggressive and
had a much greater freedom of mO'ement within the aquarium.
The subordinates tended to retreat or to make way for the
dominant fish.

No indication of a revolt on the part of a subordinate

Ft)
F 0
U1
p.-.a
5..

was not d in either study.

nelative size of the pair-mates was a fairly reliable
basis upon which to predict the winner of a fight, but there
were exceptions in both studies. based on 22 Paradise fish

pairings the larger fish won 13 times and the smaller 3.

- / o a '1 o
in o pairings tne pair-mates were 01 equal Size.

In this study, as in the Betta observations, a rouei
comparison was made between the relative frequencies of the
various types of challenging both during the periods prior
to fighting and the periods of actual fightinr. This was
done by dividing the total number of instances of each activity
by the total duration in minutes of the appropriate period.

The sum of the durations of all periods prior to
fighting was 159 minutes, while the corresponding.figure for
the fights was 216 minutes. The average number of individual
challenges per minute for the periods prior to fighting for
all 25 Paradise fish pairings was 6.5. The equivalent
figure for the fights was h.O.

This situation was reversed in the Betta study, the
corresponding numbers being 0.9 and l.h for 21 pairings.

Among the Paradise fish mutual challenges averaged
1.7 and 1.5 per minute respectively for the periods prior
to fighting and of actual fighting while for Bettas these
figures were 0.9 and O.h.

It is doubtful that dancing for position amxngPara-
dise fish is the samet;ype of activity as the jockeying for
position of the Bettas in that it represents a tendency to
replace mutual challenging as fatigue increases during the
course of a fight. This was suggested for jockeying in the
Betta study. however, if it is assumed that such a relation-

ship does exist, one may add the totals of dancing for

and otainai verage

"U
(I)
*3
i
H
'"J.
C,
C
CD
CL
H4

mg the perioo prior to fightinr of 3.0. for
i

ctual f

Cr
:74
(D
*U
(D
*3
[—10
O
Q.
0
*5
m

Betta study the correspo.ding figure involves the fighting
period only since jockeying did not occur at any other time,
and the average was 1.9. On this basis the bettas showed a
marked increase of the frequency of mutual challenging

during fights while the Paradise fish did not.

These data indicate no increase in the.frequency of

3

‘11.

H-

mutual challenging du ring the fights of Paradise f

O

Jockeying for pos iti n among the Bettas and danchqgfor
position among the Paradise fish are probe ably not comparable,
and this may he one of the causes of he difference.
1he Bettas had a complex sys stem of color changes as
compared to Paradise fish, and a comparison of the changes
in relation to various activities would be extremely
difficult. The most noticeable color changes in both species

occurred at the end of a fight when a dominance-subordination

relatior ship was established.

31.

IV. risotssiou

As was stated previously, the primary purpose of this

ous activities involved in the

H

work was to discover the var
establishment of dominance-subordination relationships among

Paradise fish, Kacropodus gpercularis. also, a com risen

F“)
m

 

F")

was made between the activities of this fish and those 0
Bettas (Braddock and Braddock, 1955), with the hope of
finding consistent differences that could be used as criteria
in future studies involving these two species.

The periods prior to fighting resembled the periods
prior to the establishment of dominance without fighting,

The maximum and minimum durations were similar (Tables 1, h)
as were the mean durations which were 6.3 and 5.2 minutes
respectively.

”he t’mes that elapsed before the first social reactions
were similar for both periods (Tables 2, S}. The mean for
the periods prior to fighting was 0.3 minutes while 0.9
minutes was the mean for the periods prior to the establish-
ment of dominance without fighting. There was no significant
difference here. Therefore, it was impossible to predict the
probability of a fight on this basis.

In both periods the most common initial reaction was
challenging. In both, individual challenging predominated
over mutual challenging (periods prior to the establishment
of dominance without fighting 33 to 1, and the periods prior

to fighting 17 to 7).

SD
Cf
&
H.
I?
H.
c1—
a
H
U)
O
O
Ho
9.)
H
*3
(I)
£0
0
Cf“
f...)
U
:3
S
m
L1

probability of subsequent fighting

*4.

ever, those pairings in which f

Show a s igiii

as the first social reaction betwe

O
.‘

individual

cha

ll

enging

not a criterion of the

or its abs

en pair-mat

(D

es,aru

fl

0

8.

now-

ghting later occurred did

icantly greater proportion of mutual challen es

it is

evident that fights were much more likely to occur in those

pairinus where the first reaction was an

initial challenge .

Similar activities were observed throughout both

periods (Tables 3, 6). Individual

frecuency during the periods prior

1 j H: ~Yx.
enailengin

t0 fl. lit in

EL had
It

‘ i 'v‘t"
i Lg].
V

{EU/61'"

be

aged 6.5 per minute as compared to h. 3 individual challenges

per minute in the periods prior to the establishment of

dominance without fighting. Also, mutual challe ngin

greater frequency in the periods

Cu

compare d with 0.0 for the perios
ment of dominance withOut

Both types of challenging w
pairings in which fights followed.
occurred one and one—half times as

longing more than twice as often.

I. )
r4
m
(at-
P0
9
.3
01
I 3‘
Ho
(J

81"

prior to the

liblrlt irlg‘ 0

prior to fighting,

g had

1.7 as

establish-

” P6; more C 0111211011

11'}.

those

T -‘ : ‘Q .' ' .P -.
individual challenging

often and

mutual chal-

This clearly indicates

between theiTrequency of challenging in

general and the probability of a sub 3 Hque t fight and

SL1

1'":

may be a

he

mechanism that builds ag

mates to the point where actual fights are possible.

gressiveness

in

gests that this activity when repeated frequently

the

eiioiigii

pair-

R)

K .o
kg
0

Lancing for position occur'ed almost tWice as often
in the periods prior to fighting as in those prior to the
establishment of dominance without fighting. The fre-
quencies per minute were 1.3 and 0.6 respectively.

Taken with the relative frequencies of both types of
challenging, dancing for position may be reparded as an
indicator of fight probability. The nature of this activity
is such as to suégest a relationship with challenging and
it may also contribute to a building up of general
aggressiveness.

Exploring was observed in the periods prior to the
establishment of dominance without fighting with a fre—
quency of 0.3 per minute. It occurred so seldom in the -
periods prior to fighting that a determination of its
frequency was not recorded here (fable 6). Although this
activity was infrequent as compared with challenging and
dancing for position it involved more time per instance.
Therefore, its presence in the periods prior to the
establishment of dominance without fighting in contrast
to its general absence in the periods prior to fighting
has significance. Exploring prooably is the result of a
lack of aggressiveness between pair-mates.

Biting was observed so infrequently in either period
that an attempt to compare frequencies would have no

significanCe.

31+.

in the peri .xior to iitnnihi-. .eiploring, which may
be associated with a lack of aggressiveness was mire

characteristic of the periods prior to the establisnmert
of dominance ti hout fi¢hting. T’itin; and lo king of
ressive activities that occur only
'1

"rcssivoness has occu rice.

or position may be

‘r‘ u ‘ V‘. —1'~ -‘ ~ ~. ~. ‘ ‘fi fi ‘ -. .‘V - 3“
Peefirucd 2s Dullgcrh and mai‘tuinei Oi ae.resSWVLL o
z . u ° W. q. : ' : «
.uu‘t.is species iniile o tir%:CUMq Jun’lflmd{ are its

(1)

I. fi ’ 1‘ I’ r ‘- ’3 " I". o J .. x A R
J. Jl‘blu’jo U bd‘herbU gull o

(D
C...
‘5
3.x
C?
:7

I
minat

CU
"3

*S
l“.

iue pe ods prior to fighting were t
the first rapid exc han es oftzites by the pair-mates.
ihis action has the criterion chose: fer the beginning

. .-‘/' *‘ f‘ D' . h‘ 1 n‘, " _“ 1‘
ibhts VHi?‘ i~u¢1eiet lo

preceded by some form oi aggressive action on the part of

9
Fl
(D
O
*3
( '7‘
O
('1'
P \ J
p.)
*1
fl)
'- o
'1
I
E
(+-
(J
’2
O

C.
re
( ’3
O
’7‘
'1
} 0.
T1
5
CT'
(I
UJ
O

f.

’ .J
p.

G)
E
p
t“,

periods had a meangi‘ration
. 1 o 1 3 1 I/ O
ior the per1oas prior to iigntir~ was 0.3 minutes.

Two activities ccur‘ red only duririg ;fights : multiple

e“0hanges of bites ard locking of jaws. furthermore,

: « ~ ' ' . . 2 : V. ,. g m. ., , V. -
indiVidual bi tinb, thich .a common here occui 1e d Only
c‘~ »\ “a 3 '9 -.(‘ 1.1 1v r‘I L (git . t. . ltd r‘ .- 11:»! A'Y (3““ ’2' _- p: "1‘ tffi
u UI ulna .LJ (AU (61" inn} 0 o lIcht'aLLbJ c4“... .Luo .Libii D

.. (u p" ye ~ : V\‘ "‘1‘ ,\ n ‘,.~ .r. - .a‘ A - ‘ . 1 .... _. - .7. \‘ r. v" '.‘ .2
was 3.c pwr’rnhhihe. inese three Vanni 1&P;ULJ onlxhb:Lble

ighting were also observed ddiih fighting

cf
h
5*
(x
L)
C)
"(3
('l
’3
'1- o
(I
f .
O:
O
y.
H
H
C.
n
(“F
O
C‘?‘
ha
y}
(I;
(J
(7‘)
*1
w-
(_
I J
(l
P
t“)
H-
C»
L—
r:-
F4.
{2
(7'
f o
(—0-
U
F‘)
F1
0
I

cuercy oer minute was ‘.3. The corresp ndiné irequency
for the fights was t.c.

There w re no significant differences in the fre-

h
\

qu'nc cis of mutual challenging and dancing for position

(
(.

‘

when the two periods were compared.

in summary, when the periods prior to fidhting are
t

compared .ith peri ace of ac

that cert ain activities UCCurIed only during fiLhts.
“bgsv we * lti 1 bit — rd ‘“'-lo' "1 t ' t
J.iiC..a b rd “U. 8 1.”. 8.1 Jada UC K. .Lh - o 8 QC 1‘! l J ,

exploring, was recorded only during the periods prior to
fighting. all of the other activities, individual chal-

1__._‘O 7.. ‘. ‘r . 11—A..: ,. “1' _‘,O ‘» _ '“V‘ ‘ -,.—.
“31.51716, mutual Cilaliul'élfig, indiVidual bi ii... and

O
H)
P.
C)
(.44
r.
()4
I:
{.3
FJ
('3
’D‘
W
F"
f4
(T)
F
’—
’.
01
f
C
“'3‘
F“
CS
Cr
Cf
h
F-
C.
F.
5....
W
5‘
(.9-
Po
:3
(‘1;
*t‘
(u)
*1
f o
(I
C.
(n
o

1 o n o

drawn from the co:npa rison of tne periods prior to iightin

)

o
.‘

Cr

with those prior to the es ablishment of dominance without
fighting. Thus, the supposedly non-aggressive activity,

exploring, decreased g;a time when initial aggressiveness

\J‘L

) \

would be expected to be at a high level, while the in—
dicators of aggressiveness, mutual challenging, dancing
for position, and biting oi all types ither remai12ed t
same or increased in frequency. Ihe decline in individual
challenging probably indicates its replacement by aggressive

)

activities of a higher lave; namely jaw-lockinv and biting.

If, for purposes of comparison, the average per
minute of all aggressive activities are totaled for the

periods prior to fighting and for the periods of actual

fighting, the results are 9.b and 12.5 respectively. This

Ho

97

+-
U

V)

a.

ind c s, in a :eneral way, the presence of an actual

C

increase d‘ring fichts.
hxcept for a few instances of chasing and retreating
“b3 crved prior to the establisn1cnt of dominance without
fighting in two pairing 8 these activities were recorded
only after dominance was established either by fighting or
not.
fter dominance was established by<;ither method the

winners or dominants were observed to have complete free-

£11

on of movement within the aguarius. This is shown clearly
by compar ing the frequencies of exploriné for the dominants
and winners to those of the subordinates and losers

(lables 9, 10). The mean per pairing for the winners was
S.h and fort;he losers 3.1.

1

Corresponding figures are 7.3 for'the dominants anc

on the behavior of tne pair-mates
the frequencies of cnasin5

observed instances
but they chased frequently. Also,
were not observed to chase

Lit in5 was also displw

instances.

and retreatin5.

of winners or dominant 1

ed only by the des

U.)

can be shown by compariax

There were no

[5

retreating

lsh

losers and subordinates

but retreated frequently.

pot in

Challen5in5 was a frequent activity on the part of

winners and dominants.
were observed

when

The losers and subordinates
to challenge but less frequently and only

approached and cornered by the a55ressors.

Mutual activities were almost absent in the periods

subsequent to fi5htin5.

Six mutual challenges were

observed in 3 pairings and 2 instances of dancing for

posit ion we re recorded in one.
These figures were Sii5n1ificantly

periods followin5 the establi

fi 5htin5. Mutual challer1n5

/ . .
an avera5e of h.o per pairing.

recorded in 7 pairin5s with an

The very low frequencies of mutual activities durin5

the period subsequent to fi5htin5

result of stron5 don1inance~s ubordi

between the pair—mates.

Mutual activities

of a55r ressiveness in
decrease after dominance was tablisn

shment of

occurred in 9

hi 5h3r for the
dominance without

pairin5s

Lancin5 for posit ion was

avera5e of 5.7 per pairing.

must have been the

nation relationships

which indicate a hi5h level

both pair-mates, showed a sharp

with or without

v; i t h

\ .

I
O

hi htin5.. Individual a55re essine behavior was lar5ely
confined to the dominants and winners, as was the non-
a55ressive explorin5.

then the a55ress’ve ”ctivitics of individuals that

have won fi5hts are compared t those of fisn that havb
established dominance without fi5htin5, certain imilar1ti
and differences are apparent. The winners of fi. 511ts avera5ed
21 individual 011a lle ngjes per pairin,5 while the dominants
avera5ed 37. The winners chased 23 times while the
dominants chased 33 times. These differences may have
been due to simple fati5ue on the part of the winners or
m'y reflect a str0153r dominance-subordination relationship
established by actual fi5htin5.

O

Bitin5, which is characteristic of the periods of
actual fighting occurred with similar frequency in both
irstarces. Also, there was no si5nif'cant difference in
the frecaencies of explorin5 when the winners are cor: Lars d
With the dUlhlIla nts.

n simil ar comparison be tween losers and subordinates

10 achieved this status without fi utin:
to data come: rr1inn5 individual ch‘ lelbin;
since bitin5 did not occur amon5 them and
too infrequent to have si 5nificance.

it udeiuo si 511ifica11t that the losers
clallen5od on the avera5e Oi only 5 times
uhil the subordinates challen5cd lo times
suuo1u1nates retreated 35 tiexes compared u

must be confined

and retreat1n5

explorin5 has

. “-.1 fl: ,1 L‘
OJ. aCtLLOL+ 4.1.5.11in
O O
‘3 1 r ‘ n~. -.
pvl ka;l .Lilé

his 0 the

i‘. 'J LJ. a 1:16 {SL1

per pairing of 20 times for the '13s are. Thus the losers and
subordinates appear to reflect the behavior of winners and
domimums, and the same causes probably underlie the differ~
ence in each case.

Many of the same activities noted here were also

noted in the Betta study (Braddock and Braddock, 1955).

Both individual and mutual challenges were observed prior

4‘1

to and during i h‘ting.
The Bet tas di “fered from the Paradi fish in that
individual challenges increased from 0.9 per mir ute to
l.h per minute during fighting while among tile Paradise
fish a frequency of 6.5 per minute during the periods
prior to fighting decreased to u.0 during active fighting
Unlike the condition among the Paradise fish where
mutual challenges had a mean frequency of 1.7 during the
periods prior to figr tii.g and 1.5 during actual fights
the bettas showed a decrease in these activities from
0.9e0.h over corresponding periods.
Individual bites were more frequent among the bettas

prior to fighting but vm re not as fre«uent during fighting.

,1)

Based on 33 fights the Bettas showed 0.3 bites per minute
prior to fizhting and 1.6 per minute during fighting.
Lased on 25 fights corresponding frequencies for Paradise

fish were .09 and 3.5 respectively.

0

h A
Mali .

ing for position was observed both prior and
during figl ting in this study, while in the betta study the
rough equivalent, jockeying for position, was recorded

vit’r was

FM

only ouiing.iibnti.5. Phe fre cquency of ti is set

C.)
o

1.3 per minute prior to fightir : and 1.1 per minute durirg
fighting among the Paradise fish. Jor Bettas the_frequency
of jom - n5 vas l. 5 per n;inut e during fighting only.

in both studies locking of jams was observed only
during tre fighting period. ror the settas this activity
occurred 0.5 times per minute while amonb the iaradise
fish it occurred 0.0 times per minute.

In both studies exploring was observed prior to
fightinr but lot during fighting. for hettas this occurred
0.2 times per minute and only .05 times for Paradise fish.

By totalling the various activities per minute for
the periods prior to fi:hting and duri ng Lighting for both

vs activity

:Jo

species relative frequencies of ger e laagbm
may be compa red.For Paradise fish 9.0 activities per mir ute

per minute

H

occurred prior to lighting; 12.5 activit
occurred during fighting. Corresponding frequencies for

Bettas were 2.5 and 6.9.

LJJ

Although the duration of fi hts was cor si

J a erably
1on gar amonb the bettas a comparison bette n the frequencies
of aggressive activities indicates that the fighting period

in the Paradise fish involved more such activ1ty.

Pk
(-2]
:3
C'?‘
C?
:3
(D
(t.
v—.
p.
('1‘
f—k
w
(1)

it should be noted hat frequency 0

+/\ '

cory in such comparisons. Luration of each instance of

U)

F1

any activity should also he considered. n this case, the
bettas may merely have sustained many or all of the

ecorded individual reactions longer.

The effect f dominance-subordination relationship

L)

on the pair-mates in both c* s was very similar. The

9,
L)

winners showed coniplete freedom of movement within the

ag arium and complete dominance over 'he losers. inc losers,
in contrast, had little freedom of hovement and were

subject to attack by the winners.

The frequencies of the various activities of both
species and the types of aggressive activities displayed
indicate that both are very belligerent. There is, however,
a greater tendency for Paradise fish not to fiéht. Thirty-
four of the 59 pairs in this study did not fight. Ho other

explanation than the general nature of the species can be

“htins among Bettas causes greater damage than
occurs amo ng Paradise fish.

it must be rememoere d when consid eri n~ the frequencies
of the various activities and when determining the relative
belligerency of the twosspecics, that male raradise fish
are being compared with female Bettas. Ii
species had been compared, undoubtedly the frequencies of

the various agrres ssive activities among he be tt as would

I»

g

have been greater, since noble (1939) stated that males

are too belligerent to establish any type of dominance

(D

relationship. now ver, noble, and omith (1937), indicated
that extreme male aggressiveness and rapid activity "ould
make a detailed analysis difficult, if not impossible.
tic behavior displayed by male Paradise
fish may be related to territoriality.e The dominant fish
may establish a territory which is def nded a brainst in~

truders. if a nest is built and mating occurs, the male

- -- ‘ “- 'u‘- ,r‘: .‘ - . vv‘ “9" pl'ftL"? un-:'-‘ ’r“-
aual'db t;i¢ 11:20 b o .Li’luL-J .LQ all llquI‘ quilt .L Liv uul H 51) 1 l'ebpcct

to survival of the young.

Another poss sible advanta

. I ., .° - 0 -1._ .0. ,, v. - ,. ,. I, t. ,l , ,, k - VP 4, .0 .._, _
LOIile 1.01" UJ. tLlLD "1J8 [may GuilCcf'll ddbbdos in h. U¢le0

.. "j. ('- ,,. 3,“-.. .«L . -mfi this! :_3 on. _
nMuflg Lettas matidb Uu es huu occur unless been inwi Viuusls
.. -u ‘_£ 4. .. 3 .n .i .— ,_ ..,-. L"- .- L .— a. r ' -. ,_. I»: 1. - v° .. ..
einioit Certain oi the patterns that occui Jdrlflc iibntind

(Braddock, J. C.--personal comm nication - 1950). These

may be mutuall Ml ases for the various elenents of the
mating sequence.
Lominanc can be establisned without fighting. This

is an advantage in that fibhting individuals are subject

or-

’ioilbil’l'i‘, (“.le tilt,

cf.
0
9
Cf
(“1”
(3
:3.
Us
‘4
C)
(—r
'3'
(D
*3
F.
l-
E3
94
H
L
I.
y .
F"
H
F\

increase in color saturation during fighting may actually

attract predators. also, the ifljufii:3 due to fightihb
migtt groutlv "pan“n one or been iish hus making them
more "uspectible to prclation.

There appear to be both ad“antaées and di sadvantaies
when the adaptive significance of ascressivo hc‘,‘.ia\.'ion

of this tvoe is considered. It ma be ’udSibld the the
at“ 1"

“ .’~ A "" ‘. 'I “\ r \~ " ‘ j , .3 i '1 ‘ x -\ -' .‘1 L‘« ". f‘ w . y ‘ 1 ' '«
CotdbllohMUuu of 00mln&ubo is the adaptive value rather

a sressive ParadiSe iish, nhich are able to ac hieve this
result freouentl" without preliminary fighting are more
advanced in th s resm; ct than the Lottas, and the latter

should then be considereeltpeha viorally the more primitive

of the two Species.

’43.

V . S U121.-IARY
1. Male Paradise fish established dominance-subordination
relationships with or without preliminary fighting when
paired in a small aquarium.
2. There was a period prior to fighting consisting of

aggressive activities both individual and mutual, e.g.
individual and mutual challenging, biting, and dancing

for position, along with certain non-social activities,

e.g. exploring and restinz.

3. There was a corresponding period prior to the establish-
ment of dominance without fighting in which occurred activ-
ities similar to those observed in the period prior to
‘fighting, but the frequencies of aggressive activities
were less and that of exploring was greater. This probably
indicated a marked difference in belligerency between such
pair—mates.

h. righting began as a response to anzggressive act, and a
high threshold of mutual aggressiveness was probably nec-
essary if it was to occur.

5. During the period of fighting all hen-social activities
ceased. Mutual aggressive activities and actual physical
contact between pair-mates increased. This was indicated
in part by two new activities, locking of jaws and multiple
bites which were observed only during this period.

6. flights ended with a definite retreat by one of the
pair—mates. Initially the defeated fish lost much of its

color saturation but in some cases regained it later.

7. after the estaoiiShment of a dominance-subordination

relationsnip with or without fighting the winner and dom-

inant ”ish freely bit and chased its subordinate. The

b—t

frequencies of mutual aggressive activities were less than
those of preceding periods.

a. ho revolts on the part of the subordinates or losers
were noted.

Color saturation reached its greatest intensity during

6
Q
0

fighting. This was also try i r Betta splendens,

 

(Braddock and Braddock, 1955).

10. Aggressive activities of the Paradise fish were similar
to those noted in the betta study. available data indicate
that in both studies, individual challenging was an important
mechanism for establishing a threshold of aggressiveness
between tne pair-mates.

11. Dancing for position among Paradise fishzand jockeying
for position among Bettas may not be completely comparable.
ihis might be a factor in determining the decrease in the
frequency of mutual challenging during the period of fighting
among Paradise fish, and the increase in this activity in

the bettas.

v

VI. Dlpulvuuni 1:3:

allee, W. C. 1931 Animd. azgregations: a study in
genera i sociology. Chicago: Lniver sity of Chica go iress.
. 193b‘- The social life of animals. new York:
'V-Iu- o ‘i'i‘ o .L‘ 01‘ t Us]. 0 -
. 1940. tenccining the or gin of sociality in animals.

a \

Scientia, 34: lSa—CO.

. l9h2. eecial don inance and subordination among
Veitebrates. A chaptei in levels of int=gration in
biological and social systems, ed. R. nedfielu.
Lancaster, Fa.: Jacques dattell Press.
T fir“. ‘. A ‘ .‘\ r‘ \L‘~{P .v-\ H '5. L. '_-‘ 4“
o 1722. Muiiilijc Iiue and IlldI‘dl’LuJ lu QUDLOUlCu OJ.
vertebrates. Poll. internat. du Centre nat. de la
fiech. sci., 34: 157-01.
glee, W. C., and Guhl, A. l9h2. Conceinihg possible survival
value of sociall‘ organized as con pared with disorganized
“ \' V V . ,— r “'.
groups uf hens. nhut. nec., ~uppi., U4. h9F-7d.
r 1 9 ’r' wfi P,‘ o , A o
allee, h. 0., and Douglis, m. o. 17%;. n durinance rdcr in
the hermit crao, Pagurus lengicarpus bay. ncology,

 

26: nil-12.

n, h. 1., and himkoff, M. F. 1953. A

1,111.88, Ya". 00,1. 6
on of the concept of instinct. fhychol.
T-

is
h3—examinat
hev., 60: as

p.» a:

7"?
7‘ o

Allee, W. C., and Lickinson, J. C. Jr. IQSh. Dominance and
subordination in the smooth degfish, hustelus canis
(hitchill). rhysiol. Zool., 27: 3'7-64.

 

1

Bennett, M. A. 1939. The social hierarchy in rin: doves.
T‘ , " r"
geology, 20: 33f-97.

Braddock, J. C. l9h5. Some aspects of the dominance-
subordination relationship in the fish, Plat"poc cilg:
maculatus. Physiol. 2001., 18: 170- -95.

 

 

. l9 9. The effect of prior residence upon dominance
in the fish, llatypoecilus maculatus. Physiol. Lool.,
22:101‘67.

 

Braddock, J. C. and Lora I. 1955. Aggressive behavior among
females of the Siamese fighting fish, Betta splendens.
Physiol. Zool., 28: 132-72.

 

Breder, C. K., Jr. 1936. The reproductive habits of he berth
American sunfishes (Family Centrarchidae).
Zoologica, 21: 1-i8

-F-—-
0\

 

Eovbjerg, E. J. 1953. pcminance order in crayiish dronectes
o o o r~ v\: ' _' 4. - .~ { 1'
Virilis nagen. rnysiol. Lool., 20: 173—70.

 

4“ 0 *-r r '.l. .1 w.2, Aw”. 3 r. A ,._ -
collias, n. L. 19nd. aggressive behavior among vertebrat
EliVSi01. LOO-lo, 17: (33-1230

 

 

{\l _ . - ,, . ,i A '

Gree org, 3. l~i7. uOLfi relations between territory, _seCial
hierarchy, and leadership in the green sunfish, (Leoomis
cyanellus). Physiol. 2001., 20: 207-93.

hess, h. 1. 1~S2. Temperature as a regu atcr oi tne attac‘-
resronse of betta splendens. Leitcchrift fur mi rpsvcholo

_Edo 9, Left 3 (1933).
hewman, h. A. 1950. Social behavior and intersp ecific com-

petition in two trout species. r‘hysiol. Lool., 29: eh-Sl.

hoble, G. K. The function of the corpus striatum in the
social behavior of f‘is hes. nnat. hec., ouppl. 3, ch.: ch
(abstract).

1930. sexual selection among fishes. biol. rev.,

i3: 133-

. 1939. The experimental animal fr m the naturalist's
. . . \ 7 F, ./
point of View. nmer. nat., {3: 113-20.

Roble, G. n., and Curtis, B. $1939. The social behavior of
the jewel fish, hemi mis bima cula tus. Bull. nmer.
Mus. ”at. n1 st., 70: -Tb.

Noble, G. x., and Borne, R. 19h0. lhe effect of sex hormones
on file social hierarchy of Xiphophorus helleri. Anat.
Rec., u;pl., 70: 147 (abstract t}.

\‘ I l‘ — e" :‘a - -
. 19Hl. ine eiiect of for brain lesions on the sexual

and fi ghting behaVior of Betta Splendens and other fis me .
flrlato lLeCo, 79, .LVO. 3, supple 2, p. 4}.

 

"h 1 . v v ‘ m1 o 1 . 1 _ -
:araer, G. n. 1914. the Lirective iniluence of tne sense 01
smell in the dogfish. Bull. U.L. bureau Fisheries, 33:

63-66 0

 

Schlos be erg, H., Duncan, h., and Laitch, h. l9h9. hating
behavior of two live-bearing fish Xiphophorus helleri
and Platypoecilus maculatus. n siol. Lool., 22: lib-01.

 

Smith, H. M. 1937. The figyiting fish of Siam. Nat. nist.,
39: 205-710 .

G. 0.1957. homing behavior of Caliiornia rocky
re iishes. Univ. balif. Iub. Lool., S9: ho. 7, 2h9- -b4.

0).:
S‘U)
Ob

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1
1

 

l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Denice-293

 

"11111111“1111111111“