Aggressive Behaviour of Male Siamese Fighting Fish Betta splendens Regan
in Different Colour Environment
Bambang Irawan1) & Norma Afiati2)
1)Dept. of Biology, Faculty of Mathematics and Natural Sciences,
C-Campus Airlangga University, Jl. Mulyorejo, Surabaya 60115
2)Dept. of Fisheries, Faculty of Fisheries and Marine Sciences,
Diponegoro University, Tembalang, Semarang 50275
Penelitian ini adalah tentang perilaku agresif ikan cupang (Betta splendens Regan) jantan yang
memperlihatkan empat jenis perilaku agonistik terhadap rangsang visual, yaitu menyerang,
mengintimidasi, menghindar, dan kalah/pecundang. Cermin normal digunakan terhadap ikan cupang
jantan berwarna merah dan hitam untuk mengkaji masing-masing dua komponen perilaku agresif
yaitu menyerang (bergerak maju dengan cepat ke depan/fFD, ke samping/fLD) dan mengancam
(frekuensi/fGCE dan lama waktu/tGCE mengembangnya tutup insang). Data menguatkan bukti
kemampuan ikan membedakan warna-warna utama (merah, kuning, biru) terhadap lingkungan
transparan dan menyanggah anggapan bahwa semakin gelap warna B. splendens semakin agresif. Uji
statistik Wald-Wolfowitz setiap komponen perilaku kedua kelompok di tiap lingkungan warna
menunjukkan tidak adanya perbedaan pada fFD, fLD, fGCE maupun tGCE pada jenjang statistik
0,05. Pengamatan terhadap empat komponen perilaku yang biasa ditunjukkan B. splendens pada saat
berlaga memperlihatkan bahwa warna dapat mempengaruhi perilaku agresifnya, meskipun pengaruh
warna terhadap masing-masing komponen perilaku tersebut tidak sama. Percobaan ini menyimpulkan
tidak ada perbedaan tingkat keagresifan antara B. splendens merah dan B. splendens hitam pada
lingkungan warna yang berbeda.
The present study looks at the aggressive behaviour of male Siamese fighting fish (Betta splendens
Regan), which performs four principal elements of agonistic behaviour to visual stimuli, i.e. attack,
threat, flight, and submissive. Two elements of this threat behaviour on black and red fish i.e. attack
(frontal display/fFD and lateral display/fLD) and threat (frequency/fGCE and duration/tGCE of fin
spread along with gill cover erection) were examined in response to a normal mirror image. Data
confirmed fish’s ability in colour vision (red, yellow, blue) against transparency and disagreement to
the assumption that the lighter the fish colour the less the aggressiveness. Wald-Wolfowitz test over
both fish groups toward each of behavioural element in each colour environment indicates that
frequencies of frontal and lateral displays, frequency and duration of gill cover erection were not
significant (P≥0.05). Observation upon four elements of aggressive behaviour normally displayed
when B. splendens fight suggested that colour may have an effect on fish’s aggressiveness, although
the shown effect differed between elements. Result of this experiment thus indicated no significant
differences in the aggressive behaviour level between red and black B. splendens in the different
Key words: Betta splendens, colour perception, aggressive behaviour.
*) Correspondence address: firstname.lastname@example.org
Aggressive behaviour is a term of a group of behaviour. Aggressive behaviour of male
Siamese fighting fish Beta splendens Regan is stereotyped. The sequence was so called fixed
action patterns (FAPs) because they performed almost identically each time they occur;
whereas stimuli that trigger an FAP is called sign stimuli. The capacity to turn aggressive
behaviour on and off in this way, thus without putting the subject at the risk of physical
damage inherent in staging an actual aggressive conflict, and with little cost as the
information be gathered at low risk, made Betta a popular subject of study by ethologists and
comparative psychologists interested in studying aggression.
Aggressive interactions between conspecific individuals contain information on attributes
such as fighting ability, physical condition and motivation, which may not be transmitted by
individual signals alone. In aggressive behaviour, agonistic displays combined
aggressiveness and fear. De Bruin (1977) stated that aggression includes such elements as
attack (butting, biting, snapping and chasing away the opponent), threat (erection of gill
covers, spreading of fins, undulating movements of the body, tail beating), flight (swimming
away from the opponent), and submissive behaviour (characterised by a decrease in body
colouration, a folding of the fins, and certain body postures mostly with the head turned
B. splendens is a native to the Mekong basin in Southeast Asia. It belongs to the gourami
family (Osphronemidae) of order Perciformes. The family consists of 17 genera (De Bruin,
1977; Charles, 2003). Of those, fourteen genera having their distribution in South East Asia,
whilst the remaining three occurring in various parts of Africa. This particular genus has been
for years domesticated by ornamental fish hobbyists in South East Asia; only second to
goldfish in length of domestication. Over many years of cultivation, two distinct breeds of B.
splendens, namely – fighting Betta and display Betta have evolved. The major external
differences between wild and domesticated forms are the extended fins and the great variety
of colour and pattern found in the latter. B. splendens specimens are also bred for their tail
shapes, i.e. ranging from the common veil-tail to the much-desired 180° half-moon tails
(Charles, 2003). The breed used in this study is the veil-tail fighter.
De Bruin (1977) has the impression that the lighter the colour the less the aggressiveness;
means the most aggressive is those of dark-blue forms. So, differences in body and fin
colouration may be accompanied by differences in behaviour, e.g. the level of aggressive
behaviour. Hess (1952) gave evidence that water temperature affect the attack response; and
that temperatures of 28°C led to consistently more responses than if a temperature of 22°C
was used. Thus, aggressive behaviour in fish has a correlation with some environmental
factors. Hess (1952) stated that fish has well developed colour vision, as indeed some fish has
various brilliant colours like B. splendens. More over, this particular fish could give response
to visual stimuli resemblance to their figure (Houpt & Wolsky, 1982). According to Hess
(1952) its aggressiveness could affected by environmental factors such as water temperature.
Ryan et al (2000) who studied the relationship between aggression and dominance ability in
the hermaphroditic fish, Rivulus marmoratus, using three standard protocols, the mirror test
(Mi), model test (Mo), and standard opponent test (So), found that general rank-order for
eliciting strength of these three different stimuli was Mi > So > Mo. This experiment,
therefore, sets to study the difference on the aggressive behaviour level between red and
black B. splendens in different colour environment by means of normal mirror image.
Material and Methods
Subject of the experiments were 20 adults male B. splendens consist of ten fish that have red
colour on trunk and fins, whilst the rest were ten fish of black trunk and dominant red fins.
Purchased from a dealer, they were not in the onset of reproductive season. Each group of
those experimentally naïve fish were divided into two batches, kept isolated in rearing jars
fortnightly (Plexiglas of 20x10x12.5cm) to acclimatised, and fed daily, ad libitum, with
Tubifex-worms pellet. Food for treated fish was supplied daily just after the treatment. A
mirror as sign stimulus was used as a tool of aggression-eliciting stimulant, along with some
red, yellow, and blue wallpaper. Following McGregor et al (2001), that none of the absolute
measures of size, size difference or aspects of display was correlated with the subjects’
response, thus no physical measurement was conducted upon the test fish. In this series of
experiments, water temperature ranged from 25°C – 27.5°C, with 26°C as the average.
Based on the position of two male B. splendens, two terms of displays were defined, i.e.
lateral and frontal display. Indeed, a variety of mutual position may occur, such as both fish
face each other frontally, one fish exhibit frontal display whilst the other in a lateral pose, or
both fish exhibit lateral display either in a head-to-tail or head-to-head positions.
For colour treatment, the base and three vertical walls of the containers were covered by
wallpaper of certain colour (red, yellow, or blue) leaving the fourth side (20x12.5cm) free for
observation. Water temperature measured daily after each treatment, which is at 11.00 am.
First series of experiment was conducted using transparent containers filled with water up to
10cm. Three days acclimatisation in the experimental containers was allowed prior to the
treatment. Mirror was inserted to the left narrow side of the container for three minutes a day
during six days of experiment. Thus, following Hess’s procedure (1952), each group of fish
has seven intermitted days before enrolling the next experiment.
Baenninger (1966; 1970), reported that fish given constant exposure to the mirror terminated
their respond after a period of time as a result of satiation or habituation to the mirror images.
Similarly, Charles (2003) noted that repeatedly eliciting aggressive responses in the fish
might be a source of stress, even if no tissue damage occurred. In 1974, Hinkel and Maier
studied the recovery of aggressive responses in fish that had been habituated to their own
images. The amount of aggressive display was greatest for animals that had not been
presented with the mirror for four days and progressively less for animals deprived of the
mirror for shorter periods. Therefore only particular elements showed significance
differences with Friedman test thus further examined (Table 2).
Frequency and duration of gill cover erection (f GCE and t GCE), frequency of lateral
display (f LD) and frontal display (f FD) in three (k3) and four (k4) different colour
environments were directly encoded. By nature, tail beats and attempted bites are
characteristic of close-range aggressive interactions (McGregor et al, 2001) for which they
were not counted. No detectable injury or death occurred during the experiment.
The behavioural acts of contestants were not considered as independent data but as systems
influencing each other. All statistical tests comparing, within the same sample on behavioural
frequencies, proportions, or rates were thus always of the matched-pairs, correlated, or
repeated measurements types. A Friedman two-way analysis of variance by rank was applied
to test null hypothesis that treatment has no effect in the four (k4), i.e. transparent (Tr), red
(Re), yellow (Ye) and blue (Be); as well as three (k3) colour environments (Re, Ye, and Be).
As H0 rejected at P ≤ 0.05, a Wilcoxon matched-pairs-signed-ranks test was further applied to
obtain in which colour environment fish threat behaviour differ significantly. A Wald-
Wolfowitz runs test performed in the need to determine one element of threat behaviour of
On presentation of the mirror image, a species-typical response is characterised by erection
of the dorsal and ventral fins, a deepening of body colour, and stereotype undulating
swimming movements parallel to the ‘opponent’. Whilst it was rejected for k4 (P ≤ 0.05;
Table 1), Friedman test of null hypothesis (H0) for k3 frontal display frequency (f FD) in
black fish as well as for k4 and k3 environment of red fish was accepted. This finding
indicates that f FD of black fish in the transparent tank differed significantly from that in the
three other colours (Re, Ye, Be).
Table 1. Friedman test of colour environment examined upon male Betta splendens
B. splendens Colour
Black fish k4 f FD 150 9.00 0.05 – 0.02 *)
k3 42 4.20 0.20 – 0.10 ns
Red fish k4 f FD 127.
7.65 0.10 – 0.05 ns
k3 12.5 1.25 0.70 – 0.50 ns
Black fish k4 f LD 126 7.56 0.10 – 0.05 ns
k3 56 5.60 0.50 – 0.10 ns
Red fish k4 f LD 182 10.9
0.02 - 0.01 *)
k3 128 12.8
0.01 – 0.001 *)
Black fish k4 f GCE 147.
8.85 0.05 – 0.02 *)
k3 24.5 2.45 0.30 – 0.20 ns
Red fish k4 f GCE 208.
0.01 – 0.001 *)
k3 39.5 3.95 0.20 – 0.10 ns
Black fish k4 t GCE 144 8.64 0.05 – 0.02 *)
k3 54 5.40 0.10 – 0.05 ns
Red fish k4 t GCE 184 11.0
0.02 – 0.01 *)
k3 86 8.60 0.02 – 0.01 *)
k4 : transparent, red, yellow, blue FD : Frontal Display
k3 : red, yellow, blue LD : Lateral Display
f : Frequency GCE : Gill Cover Erection
t : total time *) : Significantly different at P≤ 0.05
ns : Not significant
Wilcoxon test thus applied to examine frontal display frequency of black fish in transparent
tank in comparison to the other three colour environments; there, only red tank showed a
significant difference to the translucent one (P ≤ 0.05; Table 2). On the contrary, lateral
Table 2. Wilcoxon test for each element of aggressive behaviour between two colour environments examined upon male B. splendens
Element of Fish Aggressive Behaviour
F FD f LD f GCE t GCE
T Z P Signif. T z P Signif. T z P Signif. T z P Signif.
Tr – Red Black 5 -2.29 0.02 *)2 -2.60 0.0094 *)23 -0.46 0.65 ns
Red 27 -0.05 0.96 ns 5 -2.29 0.02 *)9 -1.89 0.06 ns
Tr- Yellow Black 10 -1.78 0.08 Ns 9 -1.89 0.06 ns 6 -2.19 0.03 *)
Red 22 0.56 0.58 ns 5 -2.29 0.02 *)14 -1.38 0.17 ns
Tr– Blue Black 15 -1.27 0.20 Ns 15 -1.27 0.20 ns 6 -2.19 0.03 *)
Red 12 -1.58 0.11 ns 1 -2.70 0.007 *)7 -2.09 0.04 *)
Red 12 -1.58 0.11 ns 23 0.46 0.65 Ns
Red 15 1.27 0.20 ns 9 -1.89 0.06 Ns
Red 0 -2.80 0.005 *)8 1.99 0.05 *)
Abbreviation: see Table 1.
display frequency (f LD) of black fish showed no significant differences either in k4 or
k3 treatment (P ≥ 0.05; Table 1). Any colour on k4 and k3 treatment for red fish was
then further examined in pair only to reveal that in yellow environment lateral display
frequency of red Betta differed significantly to that in the blue surrounding (P ≤ 0.05;
Friedman null hypothesis of black and red fish in k4 treatment for frequency of gill
cover erection (f GCE) was rejected (P ≤ 0.05; Table 1), whilst in k3 treatment it was
accepted. This finding implied that frequency of branchiostegal flare deployed in
transparent milieu differed from that in the rest of the colour setting; which for red fish
was further confirmed by means of Wilcoxon matched-pair test (P ≤ 0.05; Table 2). Yet
for black fish f GCE was only significant between transparent and red colour treatment.
Duration of gill cover erection (t GCE) of black fish in k4 was statistically different,
followed by corresponding results of red fish in both k4 and k3 surrounding (P ≤ 0.05,
Table 1). Wilcoxon matched-pair test were therefore performed only for k4 black fish,
along with k4 and k3 treatment of red fish. The result showed that transparent setting
showed no significant differences in the time period of branchiostegal membrane
spread of both black and red fish - in contrast to the blue colour setting (P ≤ 0.05; Table
2). Meanwhile, red fish put in the yellow tank showed statistically different duration of
gill cover erection compared to its behaviour in the blue environment (P ≤ 0.05; Table
After all, a Wald-Wolfowitz test for every element of tested behaviour within every
colour treatment suggests that statistically no significant differences on the level of
aggressive behaviour shown either by red or black fish specimen (P >0.05; Table 3).
Table 3. Wald-Wolfowitz test of black and red B. splendens for their f FD, f LD, f GCE
and t GCE in four-colour interior (k4)
Colour Treatment Element of
Behaviour r z P
Transparent f FD 10 0.23 0.409
Red 8 / 6 1.15 / 2.07 0.072
Yellow 14 / 12 1.15 / 0.23 0.267
Blue 7 1.61 0.054
Transparent f LD 10 0.23 0.409
Red 12 / 10 0.23 0.409
Yellow 14 1.15 0.125
Blue 13 / 11 0.69 / -0.23 0.327
Transparent f GCE 10 0.23 0.409
Red 12 / 10 0.23 0.409
Yellow 11 -0.23 0.409
Blue 7 1.608 0.054
Transparent t GCE 11 -0.23 0.409
Red 9 0.69 0.215
Yellow 10 0.69 0.409
Blue 7 1.608 0.054
Note: Same values in both specimens bring about more than one r-value. P values by two tailed test,
significantly different if P ≤ 0.05. Abbreviation: see Table 1.
Agonistic behaviour performed when a male Siamese fighting fish B. splendens, sees
another male Betta, it may become aggressive or submissive. In fact, even the sight of
its own reflection in a mirror will stimulate those behaviours, which are thought to help
the dominant organism keep greater access to food, space, or mates (Clotfelter &
Paolino, 2003). The physiological mechanisms responsible for increased aggression in
fish are not known though, but most likely involve changes in androgens such as
testosterone and 11-ketotestosterone, or amines such as serotonin (Wingfield et al.
1990; Huber et al. 1997; Oliveira et al. 2001). Ryan et al, (2000) agreed that aggression
plays an important role in animal contests, but the extent to which aggression correlates
with dominance has been a topic of much debate. Indeed, Herbs et al (2002), recorded
that female Siamese fighting fish, B. splendens, mates the winner of male-male
aggressive interactions based upon information she extracts from eavesdropping.
Eavesdropping is gathering information without being directly involved in the
communication interaction. Naïve females, those that have not witnessed the
interaction, show no consistent mating preference for either male. By courting naïve
females, male losers increase their chance of being accepted by a female, while
reducing the cost of courting females that will ultimately reject them. This suggests that
male losers would only successful in courting a naïve female. A stronger response to
winners than losers was also found in similar studies of songbirds (Naguib et al. 1999;
Peake et al. 2001). Aggressiveness, however, is especially important for male B.
splendens, since only the winner has the right to mate, and afterward he must protect
the bred eggs to the fry (Bronstein, 1981) as seen in another series of experiment
(Afiati, pers. obs.). In this period males may or may not eat any food; and as such, this
could be the most aggressive period of the male (Charles, 2003).
The keen eyes of Siamese fighting fish can see colour and are situated on their axis
such that they can rotate outwards and provide excellent peripheral vision to the
individual. A detail study on visual sensation of B. splendens by means of multi-unit
recordings from the optic tectum of B. splendens while they viewed the agonistic
display of conspecifics, showed that movement of displaying animals evoked brisk,
arrhythmic bursting in tectal units. Particularly high frequency spike bursts coincided
with the onset of full display, a behavioural element known to be of ethological
significance (McDonald et al, 2004). The ability to see colour and peripherally gives
Siamese fighting fish a good advantage over other fish of relatively the same size in
that they can avoid predation better (Afiati, pers. obs.). White (1975) concluded that the
ability to see colour also affects the Siamese fighting fish habitat selection process. In a
particular experiment, Miley & Burack (1977) found that visually experienced Siamese
fighting fish showed longer latencies to conspecifics than did isolated Siamese fighting
fish. Therefore, understanding complexity of visual communication networks through
aggressive signalling will broaden ideas, for instance, about the mechanisms of sexual
selection, parental protection, predator avoidance, or food finding.
Here the data confirmed fish’s visual ability which results in performing three of
principal elements of threat behaviour which were exhibited when male Betta fight, i.e.
FD, LD, and GCE. Frequencies of some aggressiveness element were indeed significant
when both specimens were tested partially to its mirror image in the different colour
environment. In the transparent environment, frequencies of frontal display and
opercular spread in black fish were higher than did in the red tank; duration of
opercular flare was also significantly longer than in the yellow and blue colour interior.
Red fish elicited gill cover erection more often in transparent surrounding than in
coloured interior, as well as longer duration in see-through and yellow tanks than in
blue container. Yet lateral display performed more frequent in yellow than in blue
interior setting. Considering that elements of behaviour in each group which
demonstrated significant differences were not the same, leading to a suggestion that
each behavioural response of either black or red fish to a certain colour was also
The finding that level of aggressiveness in either group in various colour settings was
not significant (P≥0.05) differed from either De Bruin (1977) who has the impression
that the lighter the colour the less the aggressiveness, means the most aggressive is
those of dark-blue forms, nor those reported by Gallup (1968) and Bronstein (1981)
where in the case of mirror images, a perfect correlation exists between the behaviour
of the fish subject and the behaviour of the image. The degree of colour change, gill
cover, and fin erection, as well as other components of the ‘opponent’s’ aggressive
display, is a direct function of the subject’s aggressive display (Gallup, 1968). More
over, Bronstein (1981) proposed that the intensity of aggressive display of the Siamese
fighting fish is directly related to the intensity and persistence of the aggressive display
of the ’opponent’. Thus, although mirror-image stimulation indeed appears to substitute
effectively for actual confrontation between two male Siamese fighting fish (Ryan et al,
2000), results relying on a single technique might suffer from a lack of generality.
Besides colour environment, light intensity seemed to contribute to fish’s aggressive-
ness. Results of the present study often were significant when it comes to compare
colour containers to the transparent ones - though not in all behavioural elements nor in
all colour. This finding corresponds to the work of Bando (2001) who used models
composed by means of image processing techniques. Using models with the contour
shape of a side view of B. splendens in an aggressive state, the responses were vigorous
when there was a fine distribution of brightness and naturalistic colour producing
textures like a scale pattern. Reactions became weaker as the brightness, colour
distribution reverted to levels that are more homogeneous, and the scale pattern
Indeed, the present study generates more questions than answers; it nonetheless
reinforces the hypothesis that prior-experiences are not qualitative states but come in
various degrees so that it can be treated as continuous variables. As a corollary, results
obtained through the application of alternate techniques might be comparable, but
difficult to conciliate unless one understands that they cover distinct segments of the
continuum of experience. Future studies shall reveal how taxonomically widespread the
aggressive phenomena are.
Baenninger, R. 1966. Waning of aggressive motivation in Betta splendens.
Psychonomic Science, 4: 241-242
Baenninger, R. 1970. Visual reinforcement, habituation, and prior social experience of
Siamese fighting fish. J. Comparative and Physiological Psychology, 71: 1-5
Bando, T., 1991. Visual perception of texture in aggressive behavior of Betta
splendens. Journal of Comparative Physiology A: Neuroethology, Sensory,
Neural and Behavioral Physiology, vol. 169 (1): 51-58, Springer Berlin
Beaugrand, J.P., & C. Goulet. 2000. Distinguishing kinds of prior dominance and
subordination experiences in male Green swordtail fish (Xiphophorus helleri).
Behavioural Processes, 50, 131-142.
Bronstein, P.M., 1981. Commitments to aggression and nest sites in male Betta
splendens. J. Comparative and Physiological Psychology, 95: 436-449
Charles, D., 2003. The effect of agonistic encounters on aggressive response in socially
isolated Siamese fighting fish, Betta splendens. Faculty of the Graduate School of
the University of Maryland, College Park.
Clotfelter, E. D. & A. D. Paolino, 2003. Bystanders to contests between conspecifics
are primed for increased aggression in male fighting fish. Animal Behaviour, 66:
De Bruin, J.P.C., 1977. Telencephalic functions in the behaviour of the Siamese
fighting fish, Betta splendens Regan (Pisces, Anabantidae). Academisch
Gallup, G. G. Jr. 1968. Mirror-image stimulation. Psychol. Bulletin. vol. 70 (6):782–
Herb, B. M., S. A. Biron & M. R. Kidd, 2003. Courtship by subordinate male Siamese
fighting fish, Betta splendens: their response to eavesdropping and naïve females.
Behaviour 140, 71-78
Hess, E.H. 1952. Temperature as a regulator of the attack response of Betta splendens.
Z.f. Tierpsychologie. Bd. 9: 379-382
Hinkel, T.J & R. Maier. 1974. Isolation and aggression in Siamese fighting fish ( Betta
splendens). Psychological Reports 34, 1323-1326.
Huber, R., K.Smith, A. Delago, K. Isaksson, & E.A. Kravitz, 1997. Serotonin and
aggressive motivation in crustaceans: altering the decision to retreat.
Proceedings of the National Academy of Sciences, U.S.A., 90, 5939–5942.
Lattal, K.A. & B. Metzger. 1994. Response acquisition by Siamese fighting fish (Betta
splendens) with delayed visual reinforcement. J. of the Experimental Analysis of
Behaviour, vol. 61(1): 35–44.
McDonald, C.G., D. H. Paul & C. W. Hawryshyn. 2004. Visual sensation of an
ethological stimulus, the agonistic display of Betta splendens, revealed using
multi-unit recordings from optic tectum. J. Environmental Biology of Fishes, vol
70 (3): 285-291
Houpt, K.A. & T. R. Wolsky, 1982. Domestic animal behaviour for veterinairs and
animal scientists. 1st ed. The Iowa State University Press, Ames, Iowa
McGregor, P.K., T. M. Peake & H. M. Lampe, 2001.Fighting fish Betta splendens
extract relative information from apparent interactions: what happens when what
you see is not what you get. J. Animal Behaviour, vol. 62: 1059-1065
Miley, W.M. & G. Burak. 1977. Strength of aggressive display in Siamese fighting fish,
(Betta splendens) toward a conspecific, an alien species (Macropodus opercularis
L.), and a mirror image as affected prior conspecific visual experience.
Behavioral Biology 21, 267-272
Naguib, M., C. Fichtel, & D. Todt. 1999. Nightingales respond more strongly to vocal
leaders of simulated dyadic interactions. Proceedings of the Royal Society of
London, Series B, 266: 537–542
Oliveira, R. F., M. Lopes, L. A. Carneiro, & A. V. M. Cana´rio, 2001. Watching fights
raises fish hormone levels. Nature, 409: 475.
Peake, T. M., A. M. R. Terry, P. K. McGregor, &T. Dabelsteen. 2001. Male great tits
eavesdrop on simulated male-male vocal interactions. Proceedings of the Royal
Society of London, Series B, 268: 1183–1187.
Ryan, L., R.L. Earley, Y. Hsu & L. L. Wolf. 2000. The use of standard aggression
testing methods to predict combat behaviour and contest outcome in Rivulus
marmoratus Dyads (Teleostei: Cyprinodontidae). J. Ethology Vol. 106 (8): 743-
White,W., 1975. The Siamese Fighting Fish: its life cycle. Sterling Publishing Co., Inc.,
New York, pp. 11-32.
Wingfield, J. C., R. E. Hegner., A.M. Dufty, & G. F. Ball, 1990. The ‘challenge
hypothesis’: theoretical implications for patterns of testosterone secretion,
mating systems, and breeding strategies. American Naturalist, 136, 829–846.