Report
Signal Use by Octopuses in Agonistic Interactions Graphical Abstract
Authors David Scheel, Peter Godfrey-Smith, Matthew Lawrence
Correspondence
[email protected]
In Brief Cephalopods show complex behaviors, including color changes. Scheel et al. assess the signaling role of body pattern displays by octopuses. Some displays are signals that mediate agonistic interactions. Comparisons of signal behavior among octopus species may provide a model for understanding the ecological factors shaping signal evolution.
Highlights d
Octopus tetricus use a range of displays during agonistic interactions
d
Displays correlated with the outcomes of interactions
d
Closely matched displays were more likely to result in escalation
Scheel et al., 2016, Current Biology 26, 377–382 February 8, 2016 ª2016 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.cub.2015.12.033
Current Biology
Report Signal Use by Octopuses in Agonistic Interactions David Scheel,1,* Peter Godfrey-Smith,2,3 and Matthew Lawrence4 1Department
of Environmental Science, Alaska Pacific University, University Drive, Anchorage, AK 99508, USA Program, The Graduate Center, City University of New York, Fifth Avenue, New York, NY 10016, USA 3Unit for History and Philosophy of Science, University of Sydney, Sydney, NSW 2006, Australia 4PO Box 3, Huskisson, NSW 2540, Australia *Correspondence:
[email protected] http://dx.doi.org/10.1016/j.cub.2015.12.033 2Philosophy
SUMMARY
Cephalopods show behavioral parallels to birds and mammals despite considerable evolutionary distance [1, 2]. Many cephalopods produce complex body patterns and visual signals, documented especially in cuttlefish and squid, where they are used both in camouflage and a range of interspecific interactions [1, 3–5]. Octopuses, in contrast, are usually seen as solitary and asocial [6, 7]; their body patterns and color changes have primarily been interpreted as camouflage and anti-predator tactics [8–12], though the familiar view of the solitary octopus faces a growing list of exceptions. Here, we show by field observation that in a shallow-water octopus, Octopus tetricus, a range of visible displays are produced during agonistic interactions, and these displays correlate with the outcome of those interactions. Interactions in which dark body color by an approaching octopus was matched by similar color in the reacting octopus were more likely to escalate to grappling. Darkness in an approaching octopus met by paler color in the reacting octopus accompanied retreat of the paler octopus. Octopuses also displayed on high ground and stood with spread web and elevated mantle, often producing these behaviors in combinations. This study is the first to document the systematic use of signals during agonistic interactions among octopuses. We show prima facie conformity of our results to an influential model of agonistic signaling [13]. These results suggest that interactions have a greater influence on octopus evolution than has been recognized and show the importance of convergent evolution in behavioral traits. RESULTS We recorded interactions between octopuses from 52.8 hr of video during a long-term study of octopus behavior in Jervis Bay, Australia (see the Supplemental Experimental Procedures). From n = 186 interactions, we recorded n = 345 body patterns and darkness of interacting individuals, along with n = 512 actions such as reaches within interactions. The number of octopuses present on any given day ranged from three to ten
octopuses counted during censuses at the study site on the day of the video. The total duration of interactions over all samples was 7.3 hr (time during which two octopuses were interacting) and comprised 14% of the sample time. Thus, interactions occupy a substantial portion of octopus time during daylight hours. Mating attempts (Figure 1A) comprised 11% (n = 22) of 186 interactions, and interactions involving a mating attempt lasted on average 5.4 min (range 3 s to 31 min). From these attempts, we identified the sex of n = 14 of 40 (35%) of the octopuses as presumed male (n = 7) or presumed female (n = 7). Re-identification of octopuses poses difficulties (see the Supplemental Experimental Procedures), but we estimate that individuals of known sex were involved in n = 80 of 186 (43%) of all interactions. Thus, mating activity by these metrics did not comprise a majority of the activity in our sample. Mating attempts occurred in 5 of 9 months in which octopuses were observed (January, March, July, August, and November, but not in February, April, June, or October). Thus, mating attempts were not limited to a particular season of the year. The most common action during interactions was reaching (Figure 1B; 72% of n = 512 recorded actions), in which one octopus extended an arm toward another without contact. Reaching was commonly initiated from within a den (Figure S1). Both touching and grappling were comparatively rare. Grappling occurred rarely during interactions in which octopuses differed more in darkness (e.g., Figure 1B) but was significantly more common among interactions with little difference in darkness (Figure 2A). Thus, interactions were overwhelmingly without contact, and smaller differences in darkness were associated with a higher likelihood of escalation to grappling. Octopuses sometimes became dark, raised the head on extended arms (stand tall), spread the arms and web, raised their mantle (all as in Figure 1C), or occupied high ground. The relative darkness of interacting octopuses correlated with their choice to withdraw or stand their ground during interactions (Figure 2B). When one octopus (initiator) approached another (reactor), in cases when the reactor stood its ground (either stand tall or reach), the initiator was lighter and the reactor darker. The relative darkness of participants was not significantly different between these two types of interactions (Figure 2B). Each of these types of interactions, however, was different from interactions in which the reactor withdrew—in these cases, the initiator was darker and the reactor lighter (Figure 2B; Movie S1). Thus, the relative darkness of the two interacting octopuses indicated whether the reactor would hold its ground or withdraw from the initiator.
Current Biology 26, 377–382, February 8, 2016 ª2016 Elsevier Ltd All rights reserved 377
Figure 1. Some Behaviors and Actions of Octopus tetricus during Interactions (A) Characteristic mating posture and body patterns. The male (left; body pattern pale with dark eyes) is reaching the right third arm toward the female (right). The male right third arm is pale, and the right second arm is draped over the right third. The female body pattern is mottle. (B) The initiator of an interaction (top), who is dark, extends a reach toward the reactor in that interaction (bottom), who is pale with dark eyes. (C) An octopus displays a dark body pattern, spread arms and web, stand tall, and elevated mantle. (D) Immediately after the display shown in (C), the displaying octopus (left) approaches a second octopus (right; deimatic on the right side of the body with intermediate color on the left side of the body while fleeing backward). See also Movie S1 and Figure S1.
Octopuses displaying a stand tall posture were dark (see above), and they also tended to elevate the mantle and to seek a higher position to stand tall (Figures 1C and 3B; Movie S1). Mantle elevation was negatively correlated with relative mean pixel intensity: that is, higher mantle elevation occurred with darker colors (Figure 3A; Movie S1). When mantle elevation was above the horizontal, octopuses were more likely to exhibit stand tall on top of the main den, the only nearby feature elevated above the sea floor (Figure 3B). Thus, stand tall with web spread, elevated mantle, and dark color were commonly displayed together (Figures 1C, 3A, and 3B; Movie S1), and more extreme expressions were amplified by exhibition from an elevated position. DISCUSSION Octopus body patterns (including color patterns, texture, and posture) have primarily been interpreted as systems to avoid or startle predators and not as signals to conspecifics (for exceptions, see Table 1). Our interpretation of a number of previously described behaviors is novel. Octopus stand tall behavior has been interpreted to improve their view of a conspecific [14] and has also been interpreted in the same work as a component of a ‘‘bounce display’’ used in male-female interactions. To our knowledge, seeking higher ground has not been discussed as a display. Spread web has been mentioned as a possible display component [21]. Although mantle elevation has been described or appears in photographs in ethograms [29] (J. Mather and J. Alupay, personal communication), it has not been examined as a display. We have documented significant correlations between these behaviors and the outcomes of agonistic interactions (withdrawal versus escalation to grappling). Such correlations might have multiple interpretations, however. Some or all might be visible but fortuitous byproducts of physiological events associated with more aggressive, or less aggressive, behavioral profiles. Then a correlation between color (for example) and subse-
quent behaviors might exist without this color functioning as a signal—without it affecting the behavior of a receiver and without it being a product of selection for such a signaling role [38]. To our knowledge, there is no physiological reason why skin darkening in octopuses would tend to be associated with aggression in the absence of a communicative role. Several ancillary sources of evidence support the hypothesis that some of these color changes and behaviors are indeed evolved displays with a communicative function. Different considerations bear on the displays correlated with more-aggressive and less-aggressive behaviors. First, the darkening associated with aggression accompanies a set of other conspicuous behaviors: stand tall, spread web, raised mantle, and seeking high ground. These display elements often appeared in potentially independent combinations (Figure 3). Stand tall (without raised mantle, [14]; without spread web and arms, J. Mather and J. Alupay, personal communication) has been interpreted as behavior to improve the view of the surroundings (e.g., [14]), and seeking higher ground could be interpreted in the same way. However, mantle elevation does not elevate the eyes and is unlikely to improve the view of the surroundings. Spread web also seems unlikely to have another role. Like both stand tall and moving higher, mantle elevation and spread web enhance apparent size and increase conspicuousness. Second, dark color is used to accompany aggressive behaviors in other cephalopods, as well as other taxa (e.g., [39, 40]), where a communicative role is plausible. In cuttlefish (Sepia officinalis), ‘‘dark face’’ is produced by males in agonistic interactions, paler males are more likely to withdraw from fights, and when both males maintain dark faces, fights ensue [5]. That is similar to the pattern here. Third, the apparent signals of aggression described here have a good prima facie fit to a standard game-theoretic model of signaling in agonistic contexts [13, 38, 41]. Enquist [13] found that honest communication of intent and fighting strength can be an evolutionarily stable strategy in situations where the cost
378 Current Biology 26, 377–382, February 8, 2016 ª2016 Elsevier Ltd All rights reserved
A
A
B
B
Figure 2. Relative Darkness of Interacting Octopuses and Behaviors during the Interaction (A) The occurrence of grappling among octopuses similar or differing in darkness. Grapple was significantly less common when the difference in darkness between two interactors was large (mean pixel intensity difference < 60 [initiator darker] or >+40 [reactor darker]; chi-square test, degrees of freedom [df] = 1, c2 = 3.52, p = 0.037). This comparison uses all available interactions and does not account for repeated measures of the same individual. In a subset of n = 12 interaction pairs, each where interactors were different in darkness in one interaction and similar in another and where each initiator is represented only in a single interaction pair, grapple still occurred significantly less often when interactors were dissimilar than when similar in darkness (paired Wilcoxon signed-rank test, Nsr = 5, z = 1.69, p = 0.046). (B) Darkness of octopuses (mean pixel intensity; more-negative numbers are darker) during behaviors of reactors when an initiator reached toward or approached the reactor. Initiators were darker than reactors at the peak of interactions from which the reactor withdrew (Wilcoxon signed-rank test, n = 15, W = 276, p = 0.000). The reverse (reactors darker) was true during interactions in which the reactor reached in response to the approach of the initiator (n = 22, W = 263, p = 0.000) or stood tall (n = 8, W = 116, p = 0.000). The relative darkness of the two octopuses was not significantly different between the two types of interactions where the reactor did not withdraw (MannWhitney U test, reactor stand tall compared to reach, U22,8 = 104, p = 0.475), whereas the relative darkness significantly differed between interactions during which the reactor withdrew or did not withdraw (reactor withdraw compared to reach, U15,22 = 51, p = 0.000; reactor withdraw compared to stand tall, U15,8 = 99, p = 0.011). See also Figure S1.
Figure 3. Mantle Elevation Was Associated with Darkness and Display from High Ground (A) The darkness of octopuses during stand tall displays was significantly correlated with the highest elevation of their mantle (Fisher’s F statistic, F1,23 = 7.9, p = 0.010). (B) Octopuses with mantle elevations above the horizontal (90 ) were significantly more likely to climb to an elevated display position than those whose mantle was below the horizontal (Chi-square test, c2 = 7.7, df = 1, p = 0.006; the sample size of paired displays by the same individual was too small to account for individual effects). The octopus line drawings illustrate elevation measurements for two example mantle positions, along with the position of the octopus relative to high ground (the main den). See also Movie S1.
of losing a fight as a weaker individual is high in relation to the cost of losing a fight when equally matched, and also in relation to the value of the resource being fought over. The ‘‘dishonest’’ strategy of signaling high strength and willingness to fight while being a weaker individual does not invade the population, as the costs of fights with stronger individuals are too high. To fit the Enquist model, a display of aggression must be correlated with some feature that gives rise to a heightened risk or cost of injury for the weaker animal. This may be strength or size, but it may also be aggressive disposition. At present, we are not able to determine how aggressive signaling might correlate with some aspect of resource holding power, and thus how these behaviors relate to a more detailed model of behavior in animal
Current Biology 26, 377–382, February 8, 2016 ª2016 Elsevier Ltd All rights reserved 379
Table 1. Some Exceptions to the Predominant Views that Octopuses Are Solitary and Asocial and Interpreting Octopus Body Patterns as Anti-predatory Tactics Behaviors Taxa
Aggr.
MF Adj.
Disp.
Str.
Abdopus aculeatus
D
D
D
D
A
A
A. horridus (O. horridus) Eledone moschata
Tol.
D
Eledone moschata
A
Octopus bimaculoides
A
D
O. cyanea
D
O. cyanea
A
D
O. cyanea
A
O. cyanea
A
O. digueti
A
A
A
D
O. joubini O. joubini
D
D
Octopus bimaculoides
O. insularis
Heir.
D
D
D
O. laqueus
A
Larger Pacific striped octopus
A
Larger Pacific striped octopus
A
A
A
A
O. maya
D D
Citations [14, 15]
F
[16]
C
[17]
F
[18]
FC
[19]
C
[20]
C
[21]
F
[22, 23]
F
[24]
F
[3]
FC
[25]
F
[26]
C
[27]
F
[7]
FC
[6]
F
[28]
C
[29]
C
[30]
O. rubescens
A
F
[31]
O. rubescens
A
F
[1], p. 151
O. tetricus
D
[32] and this report
D
F
O. vulgaris
D
C
[8, 33]
O. vulgaris
D
C
[21]
O. vulgaris
D
D
*
Study Type F
A
O. vulgaris
D
D
F
[34]
C
[35, 36]
Behaviors are as follows: Aggr., aggregations are reported in the field; MF Adj., males and females occupy adjacent dens in the field; Disp., displays directed at conspecifics; Str., complex mating strategies such as sneaker males or deception; Tol., species tolerates conspecifics in the lab with limited fighting or cannibalism; Hier., dominance hierarchies form. A, anecdotal mention, without quantitative analysis or detailed qualitative description of a particular site (careful discussions of a single instance of behavior were also considered anecdotal); D, detailed qualitative description of a particular site or quantitative analyses; *, cited by others for dominance hierarchy but no dominance observations included in reference. F, field study; C, captive study.
contests [41]. We note the generality of the Enquist model, however: informative signaling can be maintained as long as some feature gives rise to a heightened risk or cost of injury for the weaker animal in an asymmetrical fight. We note also that intense fights do occur at our site and individuals have been observed with significant damage (although we do not know whether the damage was sustained in a fight with another octopus or a predator, and we have not observed a fight to the death or obviously serious injury). Data from other octopus species indicate that death and cannibalism can result from intraspecific aggression [42, 43], so the costs of entering and losing a fight as a weak individual may well be high. The use of darkening as a signal of aggressive intent may then have a good fit to the Enquist model. Further, stand tall and mantle raise are displays of size that are hard to fake. So we suggest that the combination of behaviors seen in aggressive individuals at our site—dark color, stand tall, mantle raise, spread web, and seek
high ground—may be a combination of several indicators of size that are hard for small individuals to fake, along with a more arbitrary signal of aggressive ability and intent. Somewhat different considerations may apply to the behaviors of octopuses that retreat from aggressive individuals. Paleness, first, is the natural contrast to darkness, the sign of aggression. But these paler individuals often also produce a high-contrast ‘‘deimatic’’ display, especially on the side of the body close to the attacker. The deimatic behavior has been documented in various cephalopods as a response to a threat [33]. It might thus be suggested that this display at our site is merely a pre-existing accompaniment to flight behaviors. However, among 11 deimatic displays in our sample, most were not produced during flight, but during a return to a den in the face of either an aggressively displaying individual (6 of 11) or after an eviction. Deimatic displays were produced when entering the den in a manner that also withdrew from the aggression.
380 Current Biology 26, 377–382, February 8, 2016 ª2016 Elsevier Ltd All rights reserved
Our results add to a growing set of primary literature reports of social behaviors or displays to conspecifics among over a dozen different octopus species (Table 1). Conditions for interactions among octopuses may occur wherever there is a superabundance of prey in a habitat with limited shelter [44]. Several other species of octopus are known to occur in aggregations in the field (Table 1). Such aggregations may result from den limitation (e.g., [7, 19]), the aggregation of food supply (e.g., [31]), or attraction to conspecifics (e.g., [26]), and thus we expect interactions among octopuses to occur whenever these conditions occur. The accumulation of scientific reports that describe non-cannibalistic octopus interactions indicates that we should no longer consider octopuses as solitary and asocial or their body pattern repertoires and behaviors as having evolved solely in the context of anti-predator camouflage. If signaling among octopuses occurs, as we argue here, and is variable among species, then comparisons of signal behavior among species may provide another important model system for understanding the ecological factors shaping signal evolution in general. Communication of this kind also becomes a further example of convergent evolution in behavioral traits across cephalopods and vertebrate taxa whose evolutionary lineages have been separate since the Ediacaran [45]. The social and mating system complexities of octopus behavior deserve continued interest from behavioral ecologists. SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, one figure, and one movie and can be found with this article online at http:// dx.doi.org/10.1016/j.cub.2015.12.033. AUTHOR CONTRIBUTIONS Conceptualization, D.S., P.G.-S., and M.L.; Methodology, D.S., P.G.-S., and M.L.; Investigation, D.S., P.G.-S., and M.L.; Formal Analysis, D.S.; Writing – Original Draft, D.S.; Writing – Review & Editing, D.S. and P.G.-S.; Funding Acquisition, D.S. and P.G.-S.; Resources, D.S., P.G.-S, and M.L.
2. Kro¨ger, B., Vinther, J., and Fuchs, D. (2011). Cephalopod origin and evolution: a congruent picture emerging from fossils, development and molecules: extant cephalopods are younger than previously realised and were under major selection to become agile, shell-less predators. BioEssays 33, 602–613. 3. Mather, J.A. (2004). Cephalopod skin displays: from concealment to communication. In Evolution of Communication Systems: A Comparative Approach, D.K. Oller, and U. Griebel, eds. (MIT Press), pp. 193–214. 4. van Staaden, M.J., Searcy, W.A., and Hanlon, R.T. (2011). Signaling aggression. Adv. Genet. 75, 23–49. 5. Adamo, S.A., and Hanlon, R. (1996). Do cuttlefish (Cephalopoda) signal their intentions to conspecifics during agonistic encounters? Anim. Behav. 52, 73–81. 6. Ikeda, Y. (2009). A perspective on the study of cognition and sociality of cephalopod mollusks, a group of intelligent marine invertebrates. Jpn. Psychol. Res. 51, 146–153. 7. Mather, J.A. (1982). Factors affecting the spatial distribution of natural populations of Octopus joubini Robson. Anim. Behav. 30, 1166–1170. 8. Packard, A. (1961). Sucker display of Octopus. Nature 190, 736–737. 9. Packard, A. (1988). Visual tactics and evolutionary strategies. In Cephalopods - Present and Past, J. Wiedmann, and J. Kullmann, eds. (Schweizerbart’sche Verlagsbuchhandlung), pp. 89–103. 10. Josef, N., and Shashar, N. (2014). Camouflage in benthic cephalopods: what does it teach us? In Cephalopod Cognition, A.-S. Darmaillacq, L. Dickel, and J. Mather, eds. (Cambridge University Press), pp. 177–196. 11. Hanlon, R.T., Forsythe, J.W., and Joneschild, D.E. (1999). Crypsis, conspicuousness, mimicry and polyphenism as antipredator defences of foraging octopuses on Indo-Pacific coral reefs, with a method of quantifying crypsis from video tapes. Biol. J. Linn. Soc. Lond. 66, 1–22. 12. Mather, J.A., and Mather, D.L. (2004). Apparent movement in a visual display: the ‘passing cloud’of Octopus cyanea (Mollusca: Cephalopoda). J. Zool. (Lond.) 263, 89–94. 13. Enquist, M. (1985). Communication during aggressive interactions with particular reference to variation in choice of behaviour. Anim. Behav. 33, 1152–1161. 14. Huffard, C.L. (2007). Ethogram of Abdopus aculeatus (d’Orbigny, 1834) (Cephalopoda: Octopodidae): Can behavioural characters inform octopodid taxomony and systematics? J Molluscan Stud. 73, 185–193. 15. Huffard, C.L., Caldwell, R.L., and Boneka, F. (2010). Male-male and malefemale aggression may influence mating associations in wild octopuses (Abdopus aculeatus). J. Comp. Psychol. 124, 38–46.
ACKNOWLEDGMENTS We thank Kelley Voss for assistance tabulating data from video, Elisabeth Zieger, Christine Huffard, and Jennifer Mather for assistance with literature, Stefan Linquist for assistance with collection of images, Brian Farm for octopus line drawings, and Mick Saliwon and Lyn Cleary of the OceanTrek Diving Resort for their support in the field. The Graphical Abstract was drawn by Eliza Jewett-Hall (ª P.G.-S.). This study was unobtrusive observation only of undisturbed wild non-protected invertebrate animals that were not manipulated in any way and therefore was not required to be reviewed by the Alaska Pacific University Institutional Review Board or the CUNY Institutional Animal Care and Use Committee. Financial support for this study was provided to P.G.-S. by the City University of New York and to D.S. through Alaska Pacific University from donations by the Pollock Conservation Consortium. Findings and conclusions presented by the authors are their own and do not necessarily reflect the views or positions of the supporting organizations.
16. Young, J. (1962). Courtship and mating by a coral reef octopus (O. horridus). In Proceedings of the Zoological Society of London, Volume 138. (Wiley Online Library), pp. 157–162. 17. Mather, J.A. (1985). Behavioural interactions and activity of captive Eledone moschata: laboratory investigations of a ‘social’ octopus. Anim. Behav. 33, 1138–1144. 18. Mangold-Wirz, K. (1963). Biologie des ce´phalopodes benthiques et nectoniques de la Mer Catalane. Vie Milieu 13 (Suppl. ), 1–285. 19. Forsythe, J.W., and Hanlon, R.T. (1988). Behavior, body patterning and reproductive biology of Octopus bimaculoides from California. Malacologia 29, 41–55. 20. Cigliano, J.A. (1993). Dominance and den use in Octopus bimaculoides. Anim. Behav. 46, 677–684. 21. Wells, M.J., and Wells, J. (1972). Sexual displays and mating of Octopus vulgaris Cuvier and O. cyanea Gray and attempts to alter performance by manipulating the glandular condition of the animals. Anim. Behav. 20, 293–308.
Received: September 30, 2015 Revised: November 11, 2015 Accepted: December 10, 2015 Published: January 28, 2016
22. Yarnall, J.L. (1969). Aspects of the behaviour of Octopus cyanea gray. Anim. Behav. 17, 747–754.
REFERENCES 1. Hanlon, R.T., and Messenger, J.B. (1996). Cephalopod Behaviour (Cambridge University Press).
23. Norman, M.D. (1991). Octopus cyanea Gray, 1849 (Mollusca: Cephalopoda) in Australian waters: description, distribution and taxonomy. Bull. Mar. Sci. 49, 20–38.
Current Biology 26, 377–382, February 8, 2016 ª2016 Elsevier Ltd All rights reserved 381
24. Tsuchiya, K., and Uzu, T. (1997). Sneaker male in octopus. Venus (Jap. J. Malac.) 56, 177–181. 25. Voight, J.R. (1991). Ligula length and courtship in Octopus digueti: a potential mechanism of mate choice. Evolution 45, 1726–1730.
34. Mather, J.A., and O’Dor, R.K. (1991). Foraging strategies and predation risk shape the natural history of juvenile Octopus vulgaris. Bull. Mar. Sci. 49, 256–269. 35. Boyle, P.R. (1980). Home occupancy by male Octopus vulgaris in a large seawater tank. Anim. Behav. 28, 1123–1126.
26. Leite, T.S., Haimovici, M., Mather, J., and Oliveira, J.E.L. (2009). Habitat, distribution, and abundance of the commercial octopus (Octopus insularis) in a tropical oceanic island, Brazil: information for management of an artisanal fishery inside a marine protected area. Fish. Res. 98, 85–91.
36. Tricarico, E., Borrelli, L., Gherardi, F., and Fiorito, G. (2011). I know my neighbour: individual recognition in Octopus vulgaris. PLoS ONE 6, e18710.
27. Mather, J.A. (1980). Social organization and use of space by Octopus joubini in a semi-natural situation. Bull. Mar. Sci. 30, 848–857.
38. Searcy, W.A., and Nowicki, S. (2005). The Evolution of Animal Communication: Reliability and Deception in Signaling Systems (Princeton University Press).
28. Moynihan, M., and Rodaniche, A.F. (1982). The Behavior and Natural History of the Caribbean Reef Squid Sepioteuthis sepioidea: with a Consideration of Social, Signal, and Defensive Patterns for Difficult and Dangerous Environments (Verlag Paul Parey). 29. Caldwell, R.L., Ross, R., Rodaniche, A., and Huffard, C.L. (2015). Behavior and Body Patterns of the Larger Pacific Striped Octopus. PLoS ONE 10, e0134152. 30. Van Heukelem, W.F. (1977). Laboratory maintenance, breeding, rearing, and biomedical research potential of the Yucatan octopus (Octopus maya). Lab. Anim. Sci. 27, 852–859. 31. Laidig, T.E., Adams, P.B., Baxter, C.H., and Butler, J.L. (1995). Feeding on Euphausiids by Octopus rubescens. Cal. Fish Game Tech. Rep. 81, 77–79. 32. Godfrey-Smith, P., and Lawrence, M. (2012). Long-term high-density occupation of a site by Octopus tetricus and possible site modification due to foraging behavior. Mar. Freshwat. Behav. Physio. 45, 1–8. 33. Packard, A., and Sanders, G.D. (1971). Body patterns of Octopus vulgaris and maturation of the response to disturbance. Anim. Behav. 19, 780–790.
39. West, P.M., and Packer, C. (2002). Sexual selection, temperature, and the lion’s mane. Science 297, 1339–1343. 40. Ligon, R.A. (2014). Defeated chameleons darken dynamically during dyadic disputes to decrease danger from dominants. Behav. Ecol. Sociobiol. 68, 1007–1017. 41. Hardy, I.C., and Briffa, M. (2013). Animal Contests (Cambridge University Press). 42. Iba´n˜ez, C.M., and Keyl, F. (2010). Cannibalism in cephalopods. Rev. Fish Biol. Fish. 20, 123–136. 43. Mather, J.A., and Scheel, D. (2014). Behavior. In Cephalopod Culture, J. Iglesias, L. Fuentes, and R. Villanueva, eds. (Springer), pp. 17–39. 44. Scheel, D., Godfrey-Smith, P., and Lawrence, M. (2014). Octopus tetricus (Mollusca: Cephalopoda) as an ecosystem engineer. Sci. Mar. 78, 521–528. 45. Peterson, K.J., Cotton, J.A., Gehling, J.G., and Pisani, D. (2008). The Ediacaran emergence of bilaterians: congruence between the genetic and the geological fossil records. Philos. Trans. R. Soc. Lond. B Biol. Sci. 363, 1435–1443.
382 Current Biology 26, 377–382, February 8, 2016 ª2016 Elsevier Ltd All rights reserved
Current Biology, Volume 26
Supplemental Information
Signal Use by Octopuses in Agonistic Interactions David Scheel, Peter Godfrey-Smith, and Matthew Lawrence
300 Number of reaches Number of contacts
250
Count of behavior
Number of grapples 200
150
100
50
0 In den
Out of den
Figure S1, related to Figure 1B: Reach was the most common action during Interactions, while grapple was associated with smaller differences in darkness. Counts of actions (reach, touch, grapple) by Initiators when in and out of their dens. Reach without contact was the most common action.
Supplemental Experimental Procedures Observations were made at 17 m depth in Jervis Bay, an eastern Australian temperate marine embayment [S1, S2]. The site is a flat area on silty substrate that has apparently formed around an unidentified, partially buried artifact (of approximate size 30 cm long by 30 cm emerged height above the substrate) that provides a single hard-substrate den (sometimes occupied on two sides). Remains of scallops preyed upon by octopuses have accumulated as an extended midden, forming a shell bed of rough oval shape around 3 meters along its longest diameter, where Octopus tetricus excavate additional dens. We recorded behavior during daylight hours at irregular intervals on 20 different days during 12 of the 19 months from Jul 2011 (when sampling using time-lapse photography and video began) to January 2013 (the last sampling before we began this analysis). Sampling days were determined by local conditions and opportunity to visit the site. All images and video collected over this period were analyzed. Temporarily placed GoPro cameras were mounted about 20 cm above the substrate at the edge of the shell bed on small tripods by SCUBA divers. We also counted the total number of octopuses present on the site each sampling day. An estimated summed total of 56 presumably different octopus individuals occupied the site on sampling days across the study period. We have no basis between site visits on which to decide whether the individuals present were the same as a previous visit, and not all of these individuals were active in front of the camera. We recorded interactions from video or still image sequences taken at two to five-second intervals. To operationally define an interaction, we regarded approach of one octopus (the Initiator) towards another (the Reactor) as the start of an interaction. Interactions continued until either octopus withdrew from the other [e.g., as in S3]. When octopuses were within arms’ reach of one another, approach included reaching behavior of one or more arms (the first and second arm pairs) toward the other octopus, and withdrawal included retreat into a den. During portions of the sampling, only one octopus was visible and no interactions could occur. During interactions, we also counted the total number of times either octopus reached toward, touched, or grappled with the other octopus. Grapple was defined as a touch that visibly restrained a part of the other octopus, such as pulling on an arm. We refer to the collective counts of reaching, touching and grappling during interactions as “actions” to distinguish them from interactions. A single interaction could include multiple actions by either octopus before one octopus withdrew. The sex of interacting octopuses could only be ascertained behaviorally. Individuals reaching the right third (R3) arm were engaged in mating attempts and presumed males. We did not assess mating success, but we presumed that the targets of such interactions were female. Although male-male mating attempts have been observed in other octopus species [S4], we did not observe any individual that both initiated and was the target of mating attempts. Within a single continuous video or still sequence, octopuses that occupied the same den and were visually indistinguishable based on estimated size were assumed to be the same individual regardless of brief gaps in onscreen presence. Octopuses were not recognized as individuals except by size and den location within a single continuous sampling period. The main den artifact provided a fixed-size reference against which octopus size could be roughly estimated despite variable viewing angles and distances to the camera. We assessed octopus color in two ways: first as predefined body-patterns and second as pixel intensity. Body pattern categories included Dark (even dark color, Figure 1B center octopus and C), Mottle (low contrast complex mix of light and dark areas, Figure 1A on right); Deimatic (high contrast including white spots, transverse mantle stripes, Figure 1D right octopus on one side of body), Intermediate (even color of intermediate darkness, Figure 1D right octopus on the other side of body), or Pale with dark eyes (arms and web pale with eyes and sometimes adjacent regions of head or mantle dark, Figure 1A on left, 1B lower right octopus). Mean pixel intensity was recorded using Gimp 2.0 Histograms command for selected areas of an image. Pixel intensities ranged from 0 (black) to 255 (white). We calculated the difference between the histogram mean pixel intensity (HMPI) of the largest oval portion of each octopus (mantle if available, web, or head) and an adjacent oval reference area of the substrate or water column at the same distance as the octopus from the camera. The selected reference area was of the same pixel
dimensions as the selected mantle area. Positive values of this difference of octopus to background, the relative histogram mean pixel intensity (rHMPI), indicate the average pixel intensity of the octopus is brighter than the background; negative values that the octopus is darker than the background. Where possible, the darknesses and patterns of the two octopuses in any interaction were assessed from the same video frame at the moment of peak display (e.g. highest contrast, darkest, or moment of highest mantle elevation) to control for changes in lighting between frames. However, when required, the reference area controlled for differences in lighting conditions over brief periods of time within different frames of a video sequence. Some dyads and triads had repeated interactions within a single recording. Each interaction was counted only once because darkness was analyzed as the difference between two individuals in an interaction. Subsets of paired interactions (where each Initiator was represented only once and interactor darkness differed in one interaction and was similar in another) were analyzed to control for individual affects.
References: S1. S2. S3. S4.
Godfrey-Smith, P., and Lawrence, M. (2012). Long-term high-density occupation of a site by Octopus tetricus and possible site modification due to foraging behavior. Mar. Freshw. Behav. Physiol., 1-8. Scheel, D., Godfrey-Smith, P., and Lawrence, M. (2014). Octopus tetricus (Mollusca: Cephalopoda) as an ecosystem engineer. Sci. Mar. 78. Jachowski, R.L. (1974). Agonistic Behavior of the Blue Crab, Callinectus sapidus Rathbun. Behaviour 50, 232-253. Lutz, R.A., and Voight, J.R. (1994). Close encounter in the deep. Nature 371, 563.
