The "intelligent" octopus
  black dot
  Personality, self-awareness, & the need for "environmental enrichment"
  This section is divided into personality, self-awareness, & the need for "environmental enrichment", considered here, and BRAIN, BEHAVIOUR, & LEARNING considered elsewhere.
Research study 1

photograph of stranded octopus in a tidepool Octopus rubescensDo octopuses have personalities? Researchers at the Seattle Aquarium address this quetion using freshly collected Octopus rubescens from Elliott Bay, Washington.  Each of 44 octopuses is subjected to 3 tests, repeated 7 times in a 2-wk period, with behavioural responses being recorded after each test.  The first test is alerting where the experimenter presents himself/herself in close view of the octopus, which is in its den; the second is threat, where the octopus is touched with a brush; and the third is feeding, where the octopus is presented with a live crab (Hemigrapsus spp.).  Nineteen behaviours1, divided 7:7:5, respectively among the three tests, are scored and then averaged for the 7 replicates.  Here, the means of occurrences of the different behaviours are unimportant; rather, what is important for statistical analysis is the amount of variance2 accounted for by certain behaviours. The overall result is separation of the 19 variables into 3 principal components of behaviour variability exhibited by the 44 octopuses.  The 3 components, taken over the 3 tests, are activity (e.g., staying in den during the alerting test, coming out and grasping the brush during the threat test), reactivity (e.g., squirting, shrinking, swimming, and crawling during threat test), and avoidance (e.g., staying in den during threat and feeding tests). 

In their summary of the results the authors imply that some octopus individuals tend to be more active, some tend to be more reactive, while others tend to be more avoiding; in other words, they may be exhibiting different temperaments or personalities.  Is this the case, though?  The analyses are done on group variances, not on individual variances, so no single octopus is identified on the basis of its individual traits.  There is no doubt, however, that octopuses do vary in their individual responses to certain situations and that these behaviours are consistent within an individual3.  Octopuses learn from experience and this must contribute greatly to the variability seen in adults.  In summarising their work the authors are careful not to anthropomorphise, yet state that octopuses may have temperaments that are shaped into adult personality by experience.  The research is interesting and provocative, and follow-up studies have been done by the same research group.  Mather & Anderson 1993 J Comp Psychol 107: 336.

NOTE1 behaviours for alerting include eye and colour responses, body movements, and locational changes; for threat include siphon jets and ink release, movement, locational changes, and grasping with arms; and, for feeding, include skin-papillae responses, head movements, locational changes, and attack and capture techniques. 

NOTE2  a sequence of factor anlyses followed by a principal-components analysis separate the 19 variables into 3 major components, which together account for 45% of the variance

NOTE3  the authors remark that Aquarium volunteers often give names to octopuses and ascribe personality types to them

  black dot
Research study 2

photograph of a Caribbean octopus with shiny metal cans around its denOctopuses are curious about “found” objects and many species routinely bring empty cans and other shiny objects back to their dens.  A study on exploration and “play” in Enteroctopus dofleini in Seattle Aquarium, Washington suggests that octopuses may sometimes engage in activites that appear to have no immediate benefit – termed exploratory play by the researchers.  When presented with a floating pill bottle, for example, the octopuses will mostly touch and palpate it (presumbly testing for edibility) and then discard it.  However, 2 octopuses (out of 8 tested) use their jets to move the bottle away in a kind of “play”.  On return of the bottle in the aquarium intake current, the octopuses jet it away.  After several such events the octopuses habituate to their “toy” or perhaps become physically tired of it.  If this behaviour is indeed play and not just attempts to clear their immediate environment of unwanted clutter, then cephalopods can be added to the only other animal groups, mammals and birds (and possibly turtles) that exhibit it.  Mather & Anderson 1999 J Com Psychol 113: 333.



An Octopus (Caribbean species) has gathered some aluminum drink
cans as decoration for its den. The octopus is looking on from the upper R

Research study 3

photograph of O. bimaculoides courtesy Roger Hanlon, Woods Hole Marine Biological Laboratory, Woods Hole, MasachusettsIn a study in Portland, Oregon, Octopus bimaculoides raised from eggs1 collected at Long Beach, California are tested for temperamental2 traits at 3wk of age and then at 2 equally spaced intervals from 3-9wk of age3.  This species differs from other Octopus spp. in that after hatching the young immediately adopt a benthic life style.  Most or all other octopuses hatch from comparatively small-sized eggs and spend weeks or months in the plankton before settling to the sea bottom.  Similar methods are used as in Research Study 1 above and these lead to the description of 4 components or “temperament dimensions” (active engagement, arousal/readiness, aggression, and avoidance/disinterest) that explain 53% of the variance in frequencies of occurrence of 15 behaviours4 monitored.  Once again, the analyses are done on the behaviour of many (n = 73), not single, individuals.  In the words of the authors the study shows that octopuses can be described as displaying temperamental traits just as described from studies of other taxa, including primates.  What this means is not exactly clear, other than reinforcing our beliefs that octopuses are behaviourally complex invertebrates.  The study is interesting, but whether further comparative studies of this nature will be “useful in...understanding...the evolution of individuality itself”, as suggested by the authors, remains to be seen.  What would be of more interest would be to compare the behaviour of these young O. bimaculoides with that of another species, such as O. dolfeini, that spends the early part of its life, not behaving as an adult on the sea bottom, but floating about in the plankton.  Sinn et al. 2001 J Comp Psychol 115: 351. Photograph of O. bimaculoides courtesy Roger Hanlon, Woods Hole Marine Biological Laboratory, Woods Hole, Masachusetts.

NOTE1  because the eggs are harvested from the wild with no knowledge of mating, the authors caution that their experimental cohort may have involved more than 2 parental genotypes (other studies show that female O. bimaculoides in laboratory conditions may mate with more than one male)

NOTE2  while acknowledging that confusion exists over the difference between temperament and personality, even in describing human behaviour, here the authors define temperament as early-life behaviour, and personality as adult behaviour – the latter being modified through learning

NOTE3  the size of a 3-wk old animal is not given, but at 2wk of age they are about 6mm in length and 1g in live mass

NOTE4  included in the 15 behaviours are such things as pupil and postural changes, jetting and crawling, head and arm movements, and papillae and colour changes

Research study 4

Can an octopus that is held in captivity for a relatively long period be successfully released to the field?  Many indiduals of Enteroctopus dofleini have been held for a time and released, but only one, from the Seattle Aquarium, has been held, released, and monitored for sufficient time to photograph of octopus Enteroctopus dolfleini being held by a diverensure that the release has been “successful”.  The animal in question is a female and grew in the Aquarium from 50g to 18kg live mass over a 22-mo period.  SCUBA-divers monitoring the release followed her to 15m depth where she rested. Eventually, the divers had to leave.  Two weeks later a diver found a similarly large female octopus in a feshly dug den that had a midden of fresh crab pieces as evidence of recent feeding.  The released female octopus was seen periodically over the next 4mo, along with 2 different males, one of which appeared about 5wk after the female’s release.  Assuming that the octopus being monitored was the one released, and there is no way to be certain about this, the release shows that an octopus can survive in the wild after being held captive for a considerable portion of its life.  It represents the first attempted documentation of such an event.  Anderson 2000 Drum & Croaker 31: 7. 

NOTE  this is almost 2yr out of a total life span of 3-4yr

A small octopus, possibly O. dofleini being
held by a diver and preparing to jet away

Research study 5

An invertebrate that can learn to distinguish between objects of different sizes, shapes, textures, and colours, and that can navigate in laboratory mazes and in the field, and that is held usually for life-duration in captivity in aquarium tanks for public display, may be a candidate for “environmental enrichment”.  Up to the time of this Research Study, only 2 other papers have been specifically written on enrichment for invertebrates, both on octopuses.  These papers (not cited here) recommend that octopuses be held in a suitable space with “toys” and complex food and feeding strategies, and with proper den space.  Large-sized giant octopuses Enteroctopus dofleini currently held in many west-coast commercial aquariums may have extra demands.  Researchers at the Seattle Aquarium and University of Texas recommend the following: 1) provision of live and photograph of octopus Enteroctopus dolfleini in an aquarium tank“challenging” food, such as large crabs or other crustaceans that require capturing, subduing, killing, dismembering, and tidying up by the octopus afterwards, 2) inclusion of small motile prey that are able to hide away; 3) provision of food in openable containers to add challenge; and 4) inclusion of co-inhabitants such as fishes and other invertebrates.  Do such “enrichments” work for giant octopuses E. dofleini?  The authors admit that they don’t know, nor do they know if the species even needs such enrichment.  They do conclude that they are intelligent, though, and on this basis alone should be candidates for enrichment, at least until more is learned about their behavioural needs.  Anderson & Wood 2001 J Applied Anim Welfare Sci 4: 157.

NOTE  defined as animal husbandry that seeks to enhance the quality of captive life by including environmental stimuli that optimise psychological and physiological well-being

Apart from simulated rock surrounds, there is little of "challenge" in this
Enteroctopus dolfleini
's tank in a small Oregon commercial aquarium

Research study 6

cartoon showing octopus with mirrorPersonality aside, does an octopus have an awareness of “self”?  This concept is considered as part of a thoughtful discourse by research staff and visiting associates at the Seattle Aquarium, Washington on the ethics of maintaining cephalopods in captivity.  Self-awareness, that is, awareness of one’s own body and thus of “self”, is commonly assessed with a “mirror test”.  The animal is marked on its face without its knowing, and allowed to look in a mirror.  If it responds by touching or otherwise manipulating the mark, it is considered to have self-recognition.  Thus far, only primates definitively pass the test.  On viewing themselves in a mirror, octopuses may increase activity and change colour, but otherwise give no irrefutable evidence of being aware of self (the authors freely admit to not knowing how an octopus would actually behave if it were to recognise itself).  This paper should be mandatory reading for everyone involved in capture, holding, transport, and display of not just octopuses, but any invertebrate.  Mather & Anderson 2007 Dis Aquat Org 75: 119.