ANIMAL BEHAVIOUR, 2000, 60, 27–33

doi:10.1006/anbe.2000.1447, available online at http://www.idealibrary.com on

Sex identification and mating in the blue-ringed octopus,

Hapalochlaena lunulata

MARY W. CHENG & ROY L. CALDWELL

Department of Integrative Biology, University of California, Berkeley

(Received 28 August 1996; initial acceptance 3 November 1996;

final acceptance 10 January 2000; MS. number A7691)

We studied the reproductive behaviour of the blue-ringed octopus, Hapalochlaena lunulata, in the
laboratory by examining 15 male–male and nine male–female interactions. The initiation of physical
contact was independent of sex, size or residency status, and there were no noticeable changes in
behaviour such as sexual displays associated with courtship or aggression prior to contact. Males did not
distinguish between females or other males and copulated (defined as the insertion of the hectocotylus
into the mantle cavity of another octopus) readily with both. Spermatophores were released in all
copulations with females but not with males. The duration of copulation was significantly longer in
male–female interactions (median 160.5 min) than in male–male interactions (median 30 s). Although
male–male copulations ended passively with the withdrawal of the hectocotylus by the initiating animal,
male–female copulations were always terminated by the females following an intense struggle. These
studies suggest the inability of male H. lunulata to determine the sexual identity of potential mates prior
to the insertion of the hectocotylus and demonstrate the active role of the female during copulation.
 2000 The Association for the Study of Animal Behaviour

Although mating has been observed in a few select species
of octopuses (see Hanlon & Messenger 1996 for an over-
view of octopus mating systems), few studies have been
designed specifically to examine precopulatory behaviour
and the differences between male–male and male–female
encounters. Courtship behaviour in the form of sexual
displays has been described in only three, or possibly
four, species of octopuses. A striped body pattern
display in male Octopus cyanea was first reported by Van
Heukelem (1966) along with a similar yet more limited
display occasionally seen in females. Wells & Wells (1972)
reported that this display was given by male O. cyanea at
the approach of either sex, but that it tended to be more
pronounced in male–male encounters and those involv-
ing resistant females. However, the display has also been
observed during fights between foraging animals and in
the absence of other octopuses (Forsythe & Hanlon 1997;
Hanlon et al. 1999). Because this display occurs both
intra- and interspecifically, it is more likely a threat
than a sexual display (Hanlon & Messenger 1996). A
similar striped display by a male O. horridus (probably
O. abaculus, C. L. Huffard, personal communication)
mating with a female was figured by Young (1962).
Voight (1991b) reported a sexual display of the ligula,
the intromittent organ at the tip of the hectocotylus
Correspondence: R. L. Caldwell, Department of Integrative Biology,
University of California, Berkeley, CA 94720-3140, U.S.A. (email:
4roy@socrates.berkeley.edu).
0003–3472/00/070027+07 $35.00/0
(modified third right arm of males), in O. digueti upon
approaches to females, but did not mention whether this
display was given during male–male encounters.
In some species of octopuses, male sexual dimorphism
includes the presence of enlarged suckers on the second
and third pairs of arms. Based on encounters between
small males and larger females, Packard (1961) proposed
the display of these enlarged suckers as a means of mate
recognition in O. vulgaris. Wells & Wells (1972) were not
able to substantiate these findings. Voight (1991a) further
proposed that, because some octopuses copulate out of
visual range of one another (Wells & Wells 1972; Forsythe
& Hanlon 1988), enlarged suckers may function primarily
as chemoreceptors rather than visual cues in sexual
recognition. Therefore, the function of the enlarged
suckers as a courtship display in O. vulgaris or other
sexually dimorphic species is unclear.
That there are so few observations of sexual displays
and courtship behaviour in octopuses is surprising, given
their highly developed sensory systems (Wells 1978),
visual acuity (Messenger 1981) and potential for a variety
of body patterns and textures (Roper & Hochberg 1988).
Furthermore, incidental reports of hectocotylus insertion
between males in O. cyanea (Van Heukelem 1966; Wells
& Wells 1972), Hapalochlaena maculosa (Tranter &
Augustine 1973; this species is most likely H. fasciata),
and even interspecifically (Lutz & Voight 1994) cast
doubt on the ability of male octopuses to distinguish
27  2000 The Association for the Study of Animal Behaviour
28 ANIMAL BEHAVIOUR, 60, 1
between the sexes prior to copulation. In studies on
another cephalopod, a cuttlefish, Boal (1996) showed
that animals were not capable of individual recognition.
Octopuses of the genus Hapalochlaena are characterized
by the presence of iridescent blue rings and markings over
the dorsal body surface. This aposematic coloration is
thought to be associated with an extremely potent toxin,
tetrodotoxin produced by symbiotic bacteria found in
Hapalochlaena salivary glands. Although lethal to most
organisms and responsible for several human fatalities,
salivary gland extracts and injections of tetrodotoxin
have no ill effects on the blue-ringed octopuses them-
selves (Flachsenberger & Kerr 1985). The aposematic col-
oration in conjunction with extensive chromatophore
use make this group of octopuses a candidate for studies
on visual displays and signalling. A variety of colour and
body patterns in this genus has been reported by Roper &
Hochberg (1988) and aspects of postembryonic develop-
ment including specific body patterns have been
described (Stranks & Lu 1991). The pygmy nature of these
species facilitates behavioural studies in captivity because
small octopuses can be maintained in relatively large
aquaria under seminatural conditions.
We designed this study to test whether precopulatory
courtship behaviour occurs in the blue-ringed octopus,
and whether males of this species are able to determine
the sexual identity of novel individuals prior to physical
contact. We also describe the mating behaviour of
H. lunulata before, during and after copulation and com-
pare male–male and male–female interactions.
METHODS
We obtained 12 male and four female H. lunulata from
the Philippines via a commercial supplier. Additional
males and females were obtained from Indonesia after
this study was completed and qualitative observations on
their mating will also be included. Male mantle length
averaged 21 mm (range 11–31 mm); female mantle
lengths averaged 29 mm (range 15–38 mm). We housed
animals individually in closed artificial sea water aquaria
(503025 cm) at 26C under a 12:12 h light:dark
cycle. We fed the octopuses shore crabs (Hemigrapsus
and Pachygrapsus) every other day and allowed them to
acclimate to their new environment for 2 weeks prior to
the onset of behavioural experiments. Hapalochlaena
lunulata are diurnally active and show slight crepuscular
peaks (M. W. Cheng, personal observation). We con-
ducted studies during the daylight hours, using animals
that had been fed the previous day. We never tested
octopuses on consecutive days or with individuals they
had previously encountered. We excluded one female
from experiments because she spawned and began
brooding.
We studied 15 male–male and nine male–female inter-
actions by introducing an octopus into the home tank of
another animal. We randomly assigned the roles of intro-
duced versus home-tank animals by sex and we made no
attempt to size-match individuals. During an interaction,
we continuously videotaped activity using a colour cam-
corder with an observer stationed approximately 1 m
from the tank during the first hour. After 1 h, we removed
the introduced octopus and returned it to its own tank
unless the two were in physical contact. If physical
interaction was still occurring at the end of the hour, we
allowed the animals to continue their activities and
videotaped them until they separated and remained apart
for at least 10 min. In one instance when physical contact
exceeded 5 h, we reprogrammed the video recorder to
monitor 5 s of activity at 3-min intervals throughout
the day and night (under a 25-W red light during the
dark phase). We later analysed video recordings in
conjunction with observer notes to verify details of
interaction and behaviour.
Precopulatory Behaviour
We observed precopulatory behaviour for signs of
aggression or courtship. We measured ‘latency until con-
tact’ as the time period from the introduction of an
octopus into another’s tank until the first physical
encounter between them. Sex and the residency status of
the animal initiating contact were noted. We recognized
three general categories of contact based on the actions of
the initiating animal. ‘Arm touch’ was scored if the initial
contact was made by the distal half of one or more arms.
‘Web touch’ occurred if initial contact was made with a
loose draping of the interbrachial web rather than a tight
clasp. ‘Mounting’ was designated if one individual
pounced upon another, wrapping its arms and web
tightly around the mantle.
Copulatory Behaviour
The hectocotylus, the modified third right arm of male
octopuses, functions in copulation and spermatophore
transport. In females, all arms are essentially identical.
We defined ‘copulation’ as the placement of the hecto-
cotylus into the mantle cavity of another animal irrespec-
tive of spermatophore transport. In these studies,
placement of the hectocotylus into the mantle cavity
occurred immediately or within seconds following
mounting; therefore, the ‘latency until mounting’ was
measured as the ‘latency until copulation’. Although
other octopus species have been observed copulating in
both a ‘mounted’ and at a ‘distance’ position (Hanlon &
Messenger 1996), H. lunulata copulated exclusively in a
mounted position throughout these experiments. During
the copulatory episode, we described the activity of the
mounted animal as either ‘at rest’ or ‘crawling’ based on
its overall movement around the aquarium.
We determined spermatophore release based on the
‘arch and pump’ behaviour described by Wells & Wells
(1972) in which male octopuses inflate their mantle
cavities and release spermatophores in a large exhalation
accompanied by a wave of contractions along the hecto-
cotylus. Although this action may be very dramatic and
distinguishable from a deep respiratory exhalation in
larger species, it is sometimes not as obvious in pygmy
octopuses, especially when the length of the hectocotylus
is obscured from view. Consequently, we counted as a

spermatophore release only the most exaggerated mantle
exhalations followed by corresponding heavy contrac-
tions of the recipient animal’s mantle. Over 75% of these
were confirmed by the expulsion of what appeared to be
part of the spermatophoric capsule from the recipient
octopus’ mantle.
The ‘latency until initial or subsequent struggles’ was
measured as the time from onset of copulation until the
first attempts by the mounted individual to remove the
other octopus from its mantle. ‘Duration of copulation’
was the period from mounting and introduction of the
hectocotylus until its withdrawal from the mantle cavity.
If withdrawal occurred as a result of an intense struggle by
the mounted octopus to remove the other octopus from
its mantle, we considered the struggling octopus to have
initiated the termination. If copulation ended in a more
passive manner with the mounting individual simply
withdrawing its hectocotylus, we credited this individual
with terminating copulation.
Postcopulatory Behaviour
Following copulation, we observed three categories of
behaviour: ‘avoidance’, ‘remounting’ and ‘attack’. During
‘avoidance’, animals remained apart and did not attempt
further contact. If ‘remounting’ occurred, copulation was
reinitiated, although individual roles could potentially be
reversed. ‘Attack’ could also be initiated by either or both
individuals when an octopus attempted to envelop the
other octopus under its interbrachial web. Unlike mount-
ings, there were no attempts at hectocotylus placement
into another animal’s mantle cavity during attacks.
Ethical Note
Agonistic encounters leading to injury are rare in
octopuses. During preliminary observations on mating in
H. lunulata, we observed no interactions producing
injury. We therefore did not interfere when we observed
four females ‘attack’ males following copulation. In three
cases, the males escaped unscathed. However, one female
did envelop and hold her former mate in her inter-
brachial web making it impossible to see the male.
Attempts to intervene physically would have risked
injury to both the octopuses and the observer. The female
bit and killed the male. We allowed her to keep and
cannibalize the body.
RESULTS
Male–Male Interaction
Physical contact occurred in 12 of 15 (80%) male–male
interactions. In the remaining three, animals did not
react to the presence of one another and we detected no
changes in body pattern, ventilation rate or body position
aside from a tentative exploration around the tank by the
introduced animal. The octopuses did not come into
physical contact, nor were there any attempts to
approach or move away from the other individual during
the 1-h observation period.
CHENG & CALDWELL: MATING IN BLUE-RINGED OCTOPUS 29
60
50
Time until initial contact (min)
40
30
20
10
0
Male–Female Male–Male
Type of interaction
Figure 1. The latency in minutes until initial contact during
male–female (N=8) and male–male (N=12) interactions.
In the 12 cases of male–male interaction, there were no
noticeable behavioural changes prior to contact. First
physical contact was initiated by the larger animal in nine
(75%) cases. Subsequent contact was often initiated by
the other octopus later during the observation period.
Residency status of the initiating animal was divided
equally between introduced and home-tank octopuses.
Although ‘latencies until contact’ ranged from less than
1 min to 59 min (median 3.5 min), 75% of all initial
contacts occurred within the first 10 min (Fig. 1).
Approaches were sudden and devoid of any observable
behavioural changes. There were no associated changes in
body pattern, display of the ligula or body position. A
mounting pounce followed by probing and insertion of
the hectocotylus into the mantle cavity was the most
common form of initial contact, occurring in 11 of 12
cases. In the one exception, the initiating animal also
attempted to mount by pouncing, but the recipient
changed its body position and met the oncoming animal
with web spread leading to arm and web entanglement.
The only noticeable change in body pattern during
these interactions occurred early with the onset of
mounting. Upon initial contact, the individual being
mounted often darkened its body and briefly flashed its
iridescent blue rings while spreading its arms and inter-
brachial web. Similar intensifications of colour often
occur when H. lunulata are startled (R. L. Caldwell,
personal observation).
Copulation, the introduction of the hectocotylus
within the mantle cavity of the other animal, was
typically brief, ranging from 5 s to 6 min (median 30 s).
One pair remained mounted with hectocotylus inserted
for 44.5 h, however (Fig. 2). In 50% of the observations,
30 ANIMAL BEHAVIOUR, 60, 1
300
Duration of copulation (min)
200
100
* female responded with a brief flash of colour, which
0
Male–Female Male–Male
Type of interaction
Figure 2. The duration in minutes of copulation for male–
female (N=8) and male–male (N=12) interactions. *One outlying
male–male copulation lasted 2760 min and is not shown.
remounting and reinsertion occurred within 20 min of
the initial contact. These second copulatory events lasted
less than 1 min and half were initiated by the other
male. There were no simultaneous insertions of the
hectocotylus between paired males.
We found no behavioural or physical evidence of
spermatophore release in any of the male–male inter-
actions. During the brief period of copulation, the
recipient animal remained very active and crawled con-
tinuously around the tank but did not struggle or attempt
to remove the other animal from its position on the
mantle. In all cases, copulation was terminated by the
passive withdrawal of the initiating male’s hectocotylus
from the other male’s mantle cavity.
After separation, animals actively crawled around the
tank or twirled the distal portions of their arms rapidly
while vigorously rubbing their bodies. They either
avoided one another by moving away if the other
approached, or they remounted. There were no attacks by
either individual.
Male–Female Interaction
Physical contact and copulation occurred in eight of
nine male–female pairings. Although the male and
female came within proximity to one another in
the remaining case, there were no attempts at contact.
The octopuses simply passed one another without any
detectable change in behaviour.
Similarly, in the eight male–female interactions
observed, there were no behavioural displays or
changes in body pattern prior to the initiation of contact.
‘Latencies until contact’ ranged from less than 1 min to
23 min (median 4.1 min); 50% of all initial contacts
occurred within the first 10 min (Fig. 1). There was no
significant difference between ‘latencies until initial
contact’ in male–male and male–female pairs (Mann–
Whitney U test: U=48, N1 =8, N2 =12, NS). Sex of the
initiating animal was divided evenly between males and
females and also between the larger and smaller of the
paired octopuses. In six of the eight observations,
the residency status of the initiating animal was the
home-tank individual.
Of the four interactions initiated by males, three initial
contacts were in the form of ‘mounting’ and one was with
‘web spread’. Female-initiated contacts followed a similar
pattern with three in the form of an ‘arm touch’ and one
with ‘web spread’. Females never attempted to mount
males. In addition, female-initiated contact immediately
triggered male mounting behaviour and hectocotylus
insertion. The only noticeable change in coloration also
occurred during the initial mount when the recipient
included chromatophores and irridescent rings, and
extended arms and web apart. In its most extreme form,
the female elevated the tips of her arms and fully
extended the web similar to the shape of an inverted
umbrella. In this instance, the iridescent rings were filled
completely with a deep blue colour while the overall body
colour was a pale yellow.
Duration of copulation in male–female interactions
generally exceeded 90 min, but ranged from 25 to
246 min (median 160.5 min; Fig. 2). Duration of copu-
lation between male–male and male–female pairs differed
significantly (Mann–Whitney U test; U=8, N1 =12, N2 =8,
P<0.05). Figure 3 depicts a typical copulatory posture and
clearly shows the placement of the male’s hectocotylus
into the female’s mantle cavity. Evidence of spermato-
phore release occurred in all eight male–female copu-
lations. First spermatophores appeared as early as 10 min
after introduction of the hectocotylus to as late as
129 min. The number of spermatophores released ranged
from one to four in these studies. Females remained very
active during copulation and frequently crawled around
the aquarium with the male firmly attached and with his
hectocotylus in her mantle cavity.
Females terminated copulation in all eight obser-
vations. Termination resulted from intense struggles
where females reached behind with their arms and
forcefully pulled the male from their body. Females
generally assumed an inverted posture with web spread
and suckers outward while releasing their attachment to
aquarium walls and tumbling down as they struggled
with the males. Often several attempts were necessary
before a female could dislodge the male from her
mantle. In other observations that we made keeping
H. lunulata in aquaria with several pieces of coral rubble,
females would often attempt to squeeze through small
openings in the rubble trying to scrape the mounted
male from their mantle.
Following termination, postcopulatory behaviour was
divided primarily between ‘avoidance’ and ‘attack’. In the

Figure 3. This line drawing was made from a photograph of a male H. lunulata (shaded) copulating with a female. The arrow points to the
male’s hectocotylus, which is being inserted into the female’s mantle cavity. The colour image from which this drawing was made is available
at: http://www.academicpress.com/anbehav. An image of a female displaying as she is mounted by a male is also available at this Web site.
eight male–female interactions, copulation was followed
by ‘attack’ in four cases, ‘avoidance’ in three, and
‘remounting’ in one. Attacks were initiated by the female
and never by the male. During an attack, the female
pounced on to the male while enveloping him tightly
under her interbrachial web. In most cases, males man-
aged to break free and escape, although complete avoid-
ance was limited by the confines of the aquarium. In one
instance, the female killed the male by pinning him
down against the aquarium wall for over 10 h while
cannibalizing most of his arms. This was also the
male–female pair in which duration of copulation was
noticeably shorter, lasting only 25 min.
While filming an interaction between a female and
male H. lunulata obtained from Indonesia, one of us
(R.L.C.) had the opportunity to observe another
instance of multiple mating. The resident female
approached the male within 2 min after he was intro-
duced into her aquarium. He immediately mounted her
and the copulating pair retreated into a coral cavity.
After approximately 45 min, the male emerged from the
cavity. A few minutes later, while he was moving along
the wall of the aquarium 15 cm from the female’s cav-
ity, she rapidly jetted out of the cavity, seized the male
in her interbrachial web and pulled him back to the
cavity. As she reached the narrow cavity entrance, the
male moved on to her mantle and began a second
copulation that lasted for over 2 h. Because both
copulations took place inside the cavity, it was not
possible to determine whether spermatophores were
transferred.
CHENG & CALDWELL: MATING IN BLUE-RINGED OCTOPUS 31
DISCUSSION
Lack of Courtship and Sex Recognition
Mating occurred during the day and in the open.
Hanlon & Wolterding (1989) speculated that males might
visually assess the receptivity of females and thereby
decrease the risk of cannibalism from encounters with
nonreceptive animals. Our study indicates that male
H. lunulata do not differentiate between the sexes prior
to insertion of the hectocotylus. Precontact behaviour
by males in male–male and male–female encounters is
identical and devoid of any detectable courtship or sexual
displays. There are no changes in ventilation rate, body
pattern or position, and neither the ligula nor any other
portion of the hectocotylus is displayed upon approach.
Although courtship may serve a multitude of functions
including species identification, sexual recognition,
reproductive receptivity, intrasexual competition and a
mechanism for mate choice (Krebs & Davies 1987), the
lack of courtship behaviour in itself is not an indication
of the inability to determine the sex of another octopus
prior to contact. However, the fact that male H. lunulata
copulate readily with both sexes does support this
argument that they do not distinguish between the sexes
prior to physical contact.
In both male–male and male–female encounters, initial
contact by either sex resulted in the immediate place-
ment of the hectocotylus into the mantle cavity of the
other animal by the male. Latencies until physical con-
tact and subsequent mounting and copulation did not
32 ANIMAL BEHAVIOUR, 60, 1
differ in these studies. Male–male copulation was clearly
evident and could not be confused with aggressive behav-
iour because visual confirmation of the introduced
hectocotylus was made in all cases. Van Heukelem (1966)
considered his single observation of male–male copu-
lation in O. cyanea to be an artefact of confinement based
on the smaller animal’s inability to escape and the ‘hyper-
sexuality’ of the larger male because it had been isolated
for 6 months. This explanation is not applicable to the
present study, however, for several reasons: (1) males did
not attempt to avoid one another prior to initial contact,
(2) the recipient male was passive during the entire
copulatory episode and did not attempt to resist or
remove the mounted individual, (3) although the larger
male initiated copulation in the majority of these trials,
copulatory roles were often reversed within the same
hour, (4) male–male interaction studies were intiated
within the first 3 weeks of an animal’s arrival, and (5)
males tested in male–male interactions also copulated
with females during the study period.
Although sexual recognition may be an age-dependent
phenomenon with male–male copulations limited to
juveniles and subadults, the fully developed hectocotylus
and the ability to transfer spermatophores during copu-
lation with females confirmed the reproductive maturity
of males in this study. Similarly, female maturity was
confirmed by later spawning events.
Differences between male–male and male–female inter-
actions did not occur until after placement of the hecto-
cotylus within the mantle cavity, suggesting that males
did not distinguish sex until after copulation was initi-
ated. Stimulation received through the tip of the hecto-
cotylus may be essential for spermatophore release to be
initiated (Wells & Wells 1972). The proper stimulus may
be present within a female’s mantle cavity and absent in
a male’s, thereby explaining the difference in copulatory
behaviour in these interactions. Male–male copulations
were brief, ended with the passive withdrawal of the
initiating animal’s hectocotylus, and transfer of sper-
matophores never occurred. Male–female copulations,
however, lasted over 2 h on average and involved sper-
matophore transfer in all cases. These interactions were
always terminated by the female. It is unclear why
copulation lasted for 44 h in one male–male interaction.
However, this may be associated with the lack of active
termination, which is usually initiated by the female.
Although Cigliano (1995) found that prior mating history
may affect mating duration, possibly through male
assessment of sperm presence in females, this would
not account for differences in duration within same-sex
copulations.
Duration of Copulation
The duration of male–female copulations is quite vari-
able in octopuses, 1–2 min in O. chierchiae (Rodaniche
1984) and Bathypolypus arcticus (Wood et al. 1998), and
2–4 h in O. dofleini (Hartwick 1983). During the long
mating periods of O. dofleini, males remain in the copu-
latory position with the hectocotylus inserted to hold the
spermatophore in place within the opening of the
female’s oviduct until sperm are released after 1–2 h
(Mann et al. 1970). In octopuses with relatively short
mating durations like O. joubini, which averages 5 min,
spermatophores may be held within the oviduct after
copulaiton has ended by a mechanical process such as the
swelling of the spermatophore case (Mather 1978). In
addition to ensuring transfer of sperm, prolonged copu-
lation may also be involved in sperm competition and
mate guarding. Sperm storage occurs within the sper-
matheca of the female’s oviducal glands (Froesch &
Marthy 1975). Cigliano (1995) reported that male pygmy
octopuses of an undescribed species increased mating
duration in copulations with recently mated females and
suggested that they might be removing or displacing a
previous male’s sperm.
In H. lunulata, termination of copulation is primarily
initiated by the female. If duration of copulation is
related to sperm transfer or associated with sperm com-
petition and mate guarding, female choice may play a
greater role in octopus reproductive behaviour than pre-
viously thought. These studies reveal the active role of
female H. lunulata throughout the copulatory process
from the initiation of physical contact in half of the
male–female interactions, to the termination of copu-
lation in all cases. In addition, both the mobile nature of
females during copulation and the eventual struggle to
terminate mating by dislodging the male may have
driven male H. lunulata to copulate exclusively in the
mounted position, rather than at a distance. Following
copulation, octopuses generally avoided one another or,
in two cases, remounted. In both remounts, the female
vigorously initiated the second copulations, which fol-
lowed relatively short initial mounts. Also, attacks were
only initiated by the female.
Male–Male Copulation
The degree to which male–male copulations occur
among the different species of octopuses is unknown. In
one of the few studies reporting male–male interaction,
Wells & Wells (1972) observed one incidence of hecto-
cotylus insertion out of six O. cyanea interactions and
none out of four O. vulgaris interactions. Within the
blue-ringed octopus complex, Tranter & Augustine (1973)
mention males frequently initiating copulation with
one another, but soon withdrawing. Even between
species, Lutz & Voight (1994) have reported male–male
copulation of deep water octopuses.
Whether the frequency of male–male copulation as
shown by H. lunulata is a universal phenomenon among
the different octopuses, or relatively unique to this group
remains to be seen. However, the potential for a more
common occurrence of this behaviour may be attributed
to general characteristics of many cephalopod life his-
tories, such as a relatively short life span (Mangold 1987),
which could be a driving force for octopuses to mate at
every encounter. In addition, if population densities are
relatively low (which is not known for the majority of
octopuses, including H. lunulata), encounters between
individuals may be infrequent and reflect low mate
availability. Furthermore, octopuses are solitary and

encounters among conspecifics are infrequent compared
with those of social cephalopods such as shoaling squids
(Hanlon & Messenger 1996). Seasonal breeding could
also restrict mating opportunities. Although population
densities and seasonal reproduction have not been
studied in H. lunulata, male–male interactions within this
species are nonaggressive, relatively brief, and do not
include the transfer of spermatophores. Initiating copu-
lation with any individual encountered, male or female,
may be a good strategy, or at least a neutral one, because
the costs of the interactions are negligible while the
benefits may result in an increase in fitness.
Acknowledgments
We are grateful to R. Hanlon, P. Jutte, D. Lindberg and
an anonymous referee for helpful comments on the
manuscript. M. Norman and F. Hochberg verified the
species identification. C. Huffard prepared the drawing of
copulation from a photograph, helped with subsequent
observations on H. lunulata, and provided valuable
comments on the manuscript.
References
Boal, J. G. 1996. Absence of social recognition in laboratory-reared
cuttlefish, Sepia officinalis L. (Mollusca: Cephalopoda). Animal
Behaviour, 52, 529–537.
Cigliano, J. A. 1995. Assessment of the mating history of female
pygmy octopuses and a possible sperm competition mechanism.
Animal Behaviour, 49, 849–851.
Flachsenberger, W. A. & Kerr, D. I. B. 1985. Lack of effect of
tetrodotoxin and of an extract from the posterior salivary gland of
the blue-ringed octopus following injection into the octopus and
following application to its brachial nerve. Toxicon, 23, 997–999.
Forsythe, J. W. & Hanlon, R. T. 1988. Behavior, body patterning
and reproductive biology of Octopus bimaculiodes from California.
Malacologia, 29, 41–55.
Forsythe, J. W. & Hanlon, R. T. 1997. Foraging and associated
behavior by Octopus cyanea Gray, 1948 on a coral atoll, French
Polynesia. Journal of Experimental Marine Biology and Ecology, 209,
15–31.
Froesch, D. & Marthy, H-J. 1975. The structure and function of the
oviducal gland in octopods (Cephalopoda). Proceedings of the
Royal Society of London, Series B, 188, 95–101.
Hanlon, R. T. & Messenger, J. B. 1996. Cephalopod Behaviour.
Cambridge, Massachusetts: Cambridge University Press.
Hanlon, R. T., Forsythe, J. W. & Joneschild, D. E. 1999. Crypsis,
conspicuousness, mimicry and polyphenism as antipredator
defenses of foraging octopuses on Indo-Pacific coral reefs, with a
method of quantifying crypsis form video tapes. Journal of the
Linnean Society, 66, 1–22.
CHENG & CALDWELL: MATING IN BLUE-RINGED OCTOPUS 33
Hartwick, B. 1983. Octopus dofleini. In: Cephalopod Life Cycles. Vol. I
(Ed. by P. R. Boyle), pp. 157–200. London: Academic Press.
Krebs, J. R. & Davies, N. B. 1987. An Introduction to Behavioral
Ecology. Oxford: Blackwell Scientific.
Lutz, R. A. & Voight, J. R. 1994. Close encounter in the deep.
Nature, 371, 563.
Mather, J. A. 1978. Mating behavior of Octopus joubini Robson.
Veliger, 21, 265–267.
Mangold, K. 1987. Reproduction. In: Cephalopod Life Cycles. Vol. II
(Ed. by P. R. Boyle), pp. 277–291. London: Academic Press.
Mann, T., Martin, A. W., Thiersch, J. & Thiersch, J. B. 1970. Male
reproductive tract, spermatophores and spermatophoric reaction
in the giant octopus of the North Pacific, Octopus dofleini martini.
Proceedings of the Royal Society of London, Series B, 175, 31–61.
Messenger, J. B. 1981. Comparative physiology of vision in
molluscs. In: Handbook of Sensory Physiology Vol. VII/6C: Com-
parative Physiology and Evolution of Vision in Invertebrates
C: Invertebrate Visual Center and Behavior II (Ed. by H. Atrum),
pp. 94–200. New York: Springer-Verlag.
Packard, A. 1961. Sucker display of octopus. Nature, 190, 736–737.
Rodaniche, A. F. 1984. Iteroparity in the lesser Pacific striped
octopus Octopus chierchiae (Jatta, 1989). Bulletin of Marine Science,
35, 99–104.
Roper, C. F. E. & Hochberg, F. G. 1988. Behaviour and systematics
of cephalopods from Lizard Island, Australia, based on color and
body patterns. Malacologia, 29, 153–193.
Stranks, T. N. & Lu, C. C. 1991. Post-embryonic development of the
blue-ringed octopus Hapalochlaena maculosa. In: The Marine Flora
and Fauna of Albany, Western Australia. Vol. II (Ed. by F. E. Wells,
D. I. Walker, H. Kirkman & R. Lethbridge), pp. 713–722. Perth:
Western Australian Museum.
Tranter, D. J. & Augustine, O. 1973. Observations of the
blue-ringed octopus Hapalochlaena maculosa. Marine Biology, 18,
115–128.
Van Heukelem, W. F. 1966. Some aspects of the ecology and
ethology of Octopus cyanea Gray. M.S. thesis. University of Hawaii.
Voight, J. R. 1991a. Enlarged suckers as an indicator of male
maturity in Octopus. Bulletin of Marine Science, 49, 98–106.
Voight, J. R. 1991b. Ligula length and courtship in Octopus digueti:
a potential mechanism for mate choice. Evolution, 45, 1726–1730.
Wells, M. J. 1978. Octopus: Physiology and Behavior of an Advanced
Invertebrate. New York: J. Wiley.
Wells, M. J. & Wells, J. 1972. Sexual displays and mating of Octopus
vulgaris Cuvier and O. cyanea Gray and attempts to alter perform-
ance by manipulation of the glandular condition of the animals.
Animal Behaviour, 20, 293–303.
Wood, J. B., Kenchington, E. & O’Dor, R. K. 1998. Reproduction
and embryonic development time of Bathypolypus arcticus, a
deep-sea octopod (Cephalopoda: Octopoda). Malacologia, 39,
11–19.
Young, J. Z. 1962. Courtship and mating by a coral reef octopus
(O. horridus). Proceedings of the Zoological Society of London, Series
B, 138, 157–162.