Etter 1
Abstract
Two groups of Octopus rubescens were maintained and observed for
space usage in a 3.6 m2 tank for one month. Composite regional foraging
patterns developed. A pattern of social interactions and body postures
reflecting territorial dominance was also observed. Two octopuses were seen
to be dominant within identifiable home ranges. These data suggest that
aggressive, dominant O. rubescens may defend a territory.
Etter 2
Introduction
Access by octopuses to dens and shelters constitutes one of the most
important resources that determines the viability of an individual. Octopuses
must rely on body patterning and dens as their sole defensive mechanisms, as
they have no hard shell (Barnett 1962). It is likely, based on the important
nature of dens for octopuses, that they might defend such a valuable asset. In
turn, the defense of a territory around the den would allow the octopus a free
range within which it may forage for food.
Regardless of function, territoriality is a powerful driving instinct in
many species. Territoriality has been shown in the field voles Microtus
pennsylvanicus and M. ochrogaster (Myers and Krebs 1971) and the chickadee
Parus atricapillus (Desrochers 1989), among other species. It is an important
part of many species' ethologies (Manning 1967).
In her studies of O. joubini, Mather (1980) introcduced the two main
contrasting ideas of spatial use in octopuses. Regional occupancy could
consist of either a territorial system, where each member defends a specific
area, or a generalized social dominance system, in which space has less to do
with interactions than relative social rank.
Previous studies of various Octopus species have come to several
different conclusions regarding the use of space in octopuses. Early field
studies by Woods (1965) suggested that octopuses might be territorial. While
little field work has been done with octopuses since that date (generally due to
procedural difficulty), laboratory work performed by Ambrose (1982) and
Mather (1980) on O. bimaculatus and O. joubini have suggested that these
species are not territorial. Rather, these species appear to follow a strict social
dominance hierarchy and individuals appear to move in a common space.
Etter 3
Experience, however, has instructed caution over generalizing
behavior beyond specific species (Wilson 1975). Although one species of
octopus does not exhibit territorial behavior, that does not rule out the
behavior for the genus as a whole. There are several differences that have
been noted in the ethology of Octopus rubescens (Stricker 1987) such that the
possibility exists that O. rubescens may defend a territory. Although her
findings were inconclusive, regional defense was shown in one individual in
the study.
Territoriality, as asserted by Mather (1980, 1991), could be identified by
two criteria. First, time spent occupying an area could indicate usage of a
given home range. Second, the win/loss record of octopuses at particular
dens might be utilized for an analysis of territory as well.
The difficulty of conducting field observations of wild octopus
populations indicates that study in a semi-natural environment would be
most effective in quantifying data. While such an environment presents
such problems as crowding and unnatural disturbance, it is to date one of the
most effective means of observation.
Materials and Methods
Members of the species Octopus rubescens were collected using SCUBA
from in the Monterey Bay, California. Subjects were gathered in 6-8L plastic
bags from a depth of 11-15 meters for transport to the Hopkins Marine
Station.
The octopuses were placed in a tank with dimensions 3.3 x 1.1 x 0.6 m
and a water depth of 0.5 m. The upper edge of the tank was surrounded with
a band of artificial grass to inhibit octopus escape. The tank was maintaned by
Etter 4
an open-seawater system, with the water temperature varying from 10-12
degrees C.
The tank was kept under a tarp-covered enclosure to limit sunlight
exposure and to mimic light conditions at 40 feet. As O. rubescens is known
to be nocturnal (Strickler 1987), all observations were made between 8 pm and
3 am local time.
A dim (60W) red light was used to illuminate the tank. The lamp was
placed at the end of the tank near the water inlet. For the first five days, the
light was turned on immediately preceding observation. As the octopuses
began to recognize this as a cue for feeding and observation, the light was
then left on continuosly. The light emitted was sufficient to allow
observation after dark-adapting for several minutes, but not so bright as to
inhibit octopus activity in any obvious way. Installationof a blind or screen
would have limited observational area, thus octopuses were observed from a
stool placed near the tank in various positions. The octopuses appeared to
habituate rapidly to this situation, and would only startle in response to
sharp, quick movements.
Eight dens were provided for the octopuses. These dens were
constructed of various materials. Two dens were built around small glass
bottles and covered with rocks. Four dens were created by stacking granite
rocks. On den was made from a quantity of abalone shells, and a final den
was created from a large barnacle shell (Fig. 1).
The octopuses were maintained on a diet of various shore crabs
(Petrolisthes eriomerus, Pachygrapsus crassipes, Hermigrapsus nudus, H.
oregonesis) and subtidal crabs (Putgettia gracilis, P. producta). The octopuses
were fed an average of one crab per octopus per day, but as octopus feeding
varied considerably from day to day, the number of crabs was varied such that
Etter 5
food was always available. Feedings were generally between 11 pm and 12
am. Observations were conducted both before and after feeding.
The octopuses were collected in two groups. The first three collected
(O1, 02, and Ö3) were observed during a two week period. At this point Ö3
escaped, at which point two new octopuses (04 and Ö5) were added to the
tank, and this group was observed for an additional two weeks.
Several characteristics were used to identify the octopuses on each
night of observations. Difference in size was a sufficient indicator to
distinguish between all but Ö1 and 04. For these, a combination of resting
body coloration and locomotion method allowed these two to be distingished.
Thus, tagging was not necessary.
Octopuses were weighed and their dorsal mantle length (distance from
the center of the eyes to the tip of the mantle) was measured (Table 1). Sexing
was performed by observation of the 6th pair of suckers on the ventral
tentacles as well as examination for presence of a hectocotylus on the 4th arm.
Octopuses 3 and 5 were not large enough nor fully developed to allow an
accurate determination of sex.
Results
The subject octopuses acclimated quickly to the tank, and began feeding
within one day of capture. Äfter several nights of intense exploratory activity,
a hierarchy of aggressiveness and dominance began to emerge. In group 1, O1
was the dominant individual, defeating 02 in approximately 75% of
encounters, and defeating Ö3 in 100% of all encounters. At this point, Ol also
began using its standard foraging patterns (Fig 7). In group 2, O1 and 04 were
the co-dominant individuals. 02 was subdominant, as was Ö5. It is
Etter 6
interesting to note that the hierarchy was not size specific, as 02 was the
largest organism, outweighing the dominant octopuses by as much as 40 g
The octopuses were observed under two primary conditions. The first
was general foraging patterns under semi-isolated conditions (other octopuses
hidden or motionless). The second condition was activity and interactions
when other octopuses were in motion or in close proximity.
Foraging
Äfter a period of observations, a general pattern to octopus foraging
emerged. A foraging octopus moved at a moderate pace across the tank floor,
using its tentacles to inspect unoccupied dens and around obstructions (Table
2).
A summary of foraging patterns by the dominant octopuses is shown
in Fig. 7. Octopus 1 used the indicated foraging paths (solid line) for 78% of its
total exploratory forays. Individually, path 1 was used 48% of the time, path 2
was used 32% of the time, and path 3 was used 20% of the time. Octopus 4
used the indicated paths (dotted lines) for 68% of its foraging periods. Path 1
was used 48%, path 2 was used 20%, path 3 was used 18%, and path 4 was used
14% of the time for the total amount of time foraging. These paths always
(unless interrupted by an aggressive act by another octopus) led back to the
home den. Simultaneous utilization of the areas where foraging paths
overlap was never observed.
Foraging behavior by the non-dominant octopuses was markedly
different. These octopuses generally waited in or near dens for a crab to
approach them. If mobile foraging was performed, it was usually a
laboriously slow process. In one instance, Ö3 moved from its position on den
4 over a period of 3.5 minutes to capture a small Pugettia crab. It then moved
slowly to den 3, at which point it paused to consume its prey. 04, a dominant
Etter 7
individual, performed similar movement and capture along this same route
with a total elapsed time of only 1.6 minutes.
In addition to being slow, mobile foraging by subdominant octopuses
was not limited to any particular region of the tank. Rather, these
individuals utilized any region possible, based on whether or not it was
occupied by a dominant octopus.
Among the three subdominants (02 and Ö3 from group 1, 02 and O5
from group 2), a strong dominance hierarchy was also observed. Both Ö3 and
05 strongly avoided 02 (which was nearly double their size), never winning
any encounters with that individual.
Social Interactions
Several standard interactions were observed between octopuses (Table
2). These interactions could be divided into agressive and submissive acts.
Aggressive acts consisted of (1) Motion-To: an octopus moved in the direction
of another without any clear signs of foraging behavior, (2) Touch: one
octopus extended one or several tentacles (usually 1 and 2) and briefly
touched another, and (3) Wrestle: octopuses wrestled, using all tentacles and
occasionally web-to-web fighting. An act of submission was noted when an
octopus withdrew over approximately 0.5m and assumed a submissive
posture.
Associated with these interactions were three standard body postures.
The resting posture of octopuses was used during occupation of a home den
as well as during pauses on foraging excursions (Fig. 2). The attack motion
was used in two instances: prey capture and defense of a den or region (Fig. 3).
The submissive posture, as noted above, was used to indicate loss of an
encounter (Fig. 4).
Etter 8
All interactions were recorded and win/loss records of octopuses 1 and
4 (the two most aggressive octopuses) were calculated with respect to each den
(Figs. 7 and 8). A general illustration of a typical series of interactions is
shown in Fig. 6 (see legend for action sequence). In this series, Ol won an
encounter with 02 at den 1 by Wrestling, and 02 won an encounter with Ö3
at den 4 by Motion-Toward.
In several isolated incidents, Ö1 and 04 were also observed to inspect
dens within their territorial perimeter (dens 1,2, and 3 for O1, dens 4, 5, 6, and
7 for 04) immediately following emergence from their dens after feeding.
This action was assumed to be aggressive in nature (rather than foraging
behavior). In one instance, Ö3 was resting in den 2 and was subsequently
evicted by O1, who then immediately returned to its home den, den 1.
Discussion
In this experimental situation, two of the subject octopuses underwent
activities that suggest territoriality. The use of local foraging patterns
indicates a selective area of resource use. The observed social interactions,
including den defense and evictions, suggest the formation of a home range.
The foraging patterns exhibited by Octopus rubescens in this study were
somewhat different from those noted in other species (Yarnall 1969). Rather
than exhibiting use of a common foraging space, foraging paths of the two
most aggressive octopuses were strongly regional. This suggests that the
octopuses were utilizing specific resources from a defined area of the tank.
Use of a region alone, however, does not necessarily constitute
territoriality. Supplemental information can be derived from the win/loss
record of an individual in a given region. Wilson (1975) hypothesized that
species territoriality might be characterized by a gradient of aggressiveness and
Etter 9
dominance measured from a territorial perimeter to the den. Such a gradient
has been observed in the two dominant octopuses of this study.
While territoriality was not clearly shown for the subdominant
octopuses in this study, there are two possible explanations for their behavior.
Crowding is the most plausible cause. When describing territory usage in
several species, Wilson (1975) has compared varying territory size to an elastic
disk. This disk expands or contracts proportional to population density.
When that density becomes too high, however, the territorial system begins
to disintegrate. It is possible that under the semi-crowded conditions of this
study that only the most dominant and aggressive octopuses were able to
maintain control of their territories.
An alternate explanation may apply to the case of the two juveniles. It
may be that octopuses do not determine or begin to defend a territory until
they are more fully developed. In the wild, they may spend their time as
juveniles looking for a suitable home den, isolated from the pressures of
other octopuses. They may also simply exist in the areas between territories
until they are adequately developed to challenge the aggressive dominant
octopuses. This type of variation in territorial structure has been related to
life history stages in several other organisms, such as the black-capped
chickadees (P. atricapillus) (Wilson 1975).
Several previous studies of octopus behavior in other species have
indicated that those species are not territorial, but instead exhibit a size-based
dominance hierarchy (Ambrose 1982, Mather 1980, Forsythe and Hanlon
1988). There are substantial differences between Octopus rubescens and the
other species studied, however, that may lead to a contrast in the use of space
by these various species.
Etter 10
One such difference between octopus species lies in the length of time a
given shelter is utilized. The amount of time a given octopus inhabits one
specific den varies widely between species. The longest period Yarnall (1969)
observed Octopus cyanea inhabiting one specific den was 23 days. This
species was determined to be non-territorial. In contrast, Octopus
bimaculatus may occupy dens for a period of 5 months or more (Ambrose
1982). Territorial studies on this species have not been conducted. Öther
work with various Octopus species, which concluded that the species were
not territorial, often did not allow sufficient time for subjects to establish a
home territory. Mather's study of Octopus joubini (1980) concluded that this
species was not territorial. The dens in this experiment, however, were not
fixed, and were moved around the tank by both octopuses and experimenter,
thereby keeping the octopuses from establishing stable territories.
The difference in long-term den use in Octopus species may be one key
to territoriality. Resource utilization by stable octopus communities may be
quite different from that of migratory or transient populations (Ambrose
1982). Individuals that remain in a given area would likely have a greater
interest in defending exclusive rights to that area. The total amount of time
Octopus rubescens may utilize a given den is currently unknown, but the
octopuses in question used one den as their home for the entire four weeks of
the study. Further studies on stable den habitation is necessary.
Etter 11
Acknowledgments
A very special thank you goes out to Jim Watanabe, for without his
wonderful assistance, from advising when the course was unclear to leading
octopus collecting dives, this project would never have left the ground.
Thanks to Joe Wible and the staff of the Hopkins Marine Station Library, for
without their amazing skill to find even the most obscure articles, the
research for this project could never have been completed. Without the help
of everyone in the shop, these octopuses would never have had such a nice
place to spend their time. I am indebted to Jim Van Houten, both for his
willingness to go on late night dive trips to collect crabs as well as his
tolerance of my odd hours. Thanks also to Chris Morace for helping collect
shore crabs; the zoo was great while it lasted. Finally, an extra scratch behind
the ears goes to the Station cat, Vanilla, for keeping me company at all odd
hours of the night as I conducted my observations.
Etter 12
Literature Cited
Ambrose, R. 1982. Shelter Utilization by the Molluscan Cephalopod Octopus
bimaculatus. Mar. Ecol. Prog. Ser. 7:67-73.
Barnett, S.A. 1962. Brain and Behavior in Cephalopods. Stanford University
Press, Stanford.
Desrochers, A. 1989. Sex, Dominance, and Microhabitat Üse in Wintering
Black-capped Chickadees. Ecology. 70(3): 636-645.
Forsythe, J. and R. Hanlon. 1988. Behavior, Body Patterning, and
Reproductive Biology of Octopus bimaculatus from California. Malacologia.
29(1): 41-55.
Manning, A. 1967. An Introduction to Animal Behavior. Reading: Addison-
Wesley Publishing Co.
Mather, J. 1980. Social Organization and Üse of Space by Octopus joubini in a
Semi-Natural Situation. Bull. Mar. Sci. 30(4): 848-857.
Mather, J. 1991. Foraging, Feeding, and Prey Remains in Middens of Juvenile
Octopus vulgaris. J. Zool., London. 224: 27-39.
Myers, J. and Krebs, C. 1971. Genetic, Behavioral, and Reproductive
Attributes of Dispersing Field Voles Microtus pennsylvanicus and Microtus
ochrogaster. Ecological Monographs. 41(1): 53-77.
Strickler, K. 1987. Behavior and Body Patterning in Octopus rubescens.
Unpublished MS. on file at Hopkins Marine Station Library.
Wilson, E. 1975. Sociobiology: The New Synthesis. Cambridge: Balknap
Press.
Woods, J. 1965. Octopus Watching off Capri. Animals. 7:324-327.
Yarnall, J. 1969. Aspects of the Behavior of Octopus cyanea Gray. Anim.
Behav., 17: 747-754.
Octopus +
Sex
Table 1. Octopus Statistics
Wet Weight
Dorsal Mantle
Date in
Length (cm)
4/22/94
149.70
7.00
118.70
4/22/94
6.00
4/22/94
4.50
87.90
110.20
6.00
5/4/94
4.50
5/4/94
69.80
Etter 13
Date Out
5/20/94
5/20/94
Esc. 5/3/94
5/20/
5/20/94
Foraging
Motion-To
Touch
Wrestle
Submission
Table 2. Octopus Activities
Description
Octopus moves slowly across tank bottom or side, using
tentacles to search within dens and behind obstructions
One octopus moves in the general direction of another,
without any clear foraging signs
One octopus extends one or several tentacles (usually 1
and 2) and briefly touches another
Octopuses wrestle, using all tentacles and web-to-web
fighting
Octopus retreats » 1.5 feet, assumes submissive posture
Etter 14
Etter 15
Figure Legend
Figure 1. Tank Diagram. The tank used measured 3.3 x 1.1 x 0.6 m and the
water level was maintained at 0.5 m. Dens and den numbers are shown, as is
the location of the dim red lamp used for illumination. A 3 cm layer of sand
covered the bottom of the tank.
Figure 2. Resting Position of Octopus rubescens. Observation of size and
coloration from this position allowed for a discrimination of individuals.
Note the raised head and exploratory tentacles.
Figure 3. Aggressive Motion of O. rubescens. This type of motion was used
in two situations: 1. Prey capture and 2. Attack in defense of a territory. Water
jets were used to propel the organism.
Figure 4. Submissive Posture of O. rubescens. This posture was assumed by
an octopus after it lost an aggressive encounter. Note the head pulled close to
the wall, as well as the tentacles curled around the mantle in a protective
position.
Figure 5. Octopus 1 Win / Loss Record. The total number of wins and losses
by this individual were calculated for each den. Wins were recorded over 02,
Ö3, 04, and Ö5. Losses were recorded to 02 and 04.
Figure 6. Octopus 4 Win / Loss Record. The total number of wins and losses
for this individual were calculated for each den. Wins were recorded over
O1,02, and Ö5. Losses were recorded to Ol and 02.
Figure 7. Typical Foraging Patterns of Ol and 04. These foraging paths
comprised 78% and 68% (respectively) of total foragin excursions by each
octopus. Foraging trails of Ol are represented by solid lines; those of 04 are
dotted. Individually, Ol used path 1 was 48% of the time, path 2 was used
32% of the time, and path 3 was used 20% of the time. 04 used path 1 48%,
path 2 was used 20%, path 3 was used 18%, and path 4 was used 14% of the
time for the total amount of time foraging.
Figure 8. Dominance and Territorial Defense. Represented here is a diagram
of a typical series of interactions. Event sequence: (1) 02 moved from its
position on the right hand side of the tank to rest on den 1. (2) Ol dropped
from its position directly above den 1 upon 02 and octopuses began wrestling
(3) O2 moved away from Ol to den 3 where it exhibited the ’submissive
posture, and Ol returned to its previous position. (4) After several moments,
02 moved from den 3 to den 4, at which point (5) Ö3 moved from its position
on den 4 to den 7. (6) 02 then continued on to rest at den 5.
Den 6
W
Den 8
Den 7
Lamp (60W Red)
A
Water Inlet
Water Outlet
Den 4
Den 2


Den 5
Den 3
Den 1

Etter 16
l
2





2
9
Etter 17
(











—
-
—
Etter 18
E
6
2933
Etter 19
12:
—
1 2 3
—
ataa.
Den Number
Etter 20
—0— Wins
--4-: Losses
10-
—
—0— Wins
--+-: Losses
1 2
Etter 21
4—4

3 4 5 6 7 8 9
Den Number
Octopus 4
A
Octopus1
Water Outlet
Etter 22
Octopus 3

Octopus 2

Octopus

Water Outlet