The underappreciated extent ofcannibalism and ophiophagy inAfrican cobras

It is incredibly challenging to study the natural feedinghabits of snakes. Snakes are themselves secretive and rarelyobserved in the wild, and they feed relatively infrequently,making observations of feeding scarce. As a result, scientificunderstanding of snake dietary ecology (beyond simple preylists) remains remarkably superficial outside of a few species.In this context, our own research sets out to understandhow feeding and competition for spatially and temporallypatchy resources drives interspecific competition betweentwo species of African snakes and how those processesmight change in the future.During fieldwork in January 2018, we found ourselves at

our study site in South Africa’s Kalahari Desert in search ofcape cobras (Naja nivea) and boomslang (Dispholidus typus)to surgically implant with radio-transmitters. While search-ing for snakes one morning, we were alerted by a tour guidevia radio to the presence of “two large yellow snakes fight-ing”. Given that cape cobras fit the not widely applicabledescription of “large yellow snakes,” we raced to the scenehoping to capture both animals for implantation. However,arriving some 15 min later, we were greeted not by twomales in ritualized combat as initially expected, but ratherby a large (total length 1,672 mm; 840 g) male cape cobra inthe process of swallowing a smaller (estimated total length

1,300 mm) male conspecific (Fig 1). Instead of capturingtwo potential study animals, we found one well-fed studyanimal, now known as NN011, or more casually, Hannibal.Cannibalism, the consumption of conspecifics, is wide-

spread in animals (Polis 1981). Its occurrence in populationsis thought to be the result of the trade-off between the costs(including risk of injury, reduction of inclusive fitness, or riskof parasite infection) and benefits (including energetic gains,reduction in competitors for food or mates, or potentially thereduction of parasite infection risk) of the behavior (reviewedin Van Allen et al. 2017). Ecologically, cannibalism can haveimportant size-dependent impacts on the demography ofpopulations (Claessen et al. 2004), can influence aspects ofintersexual and intrasexual competition (Polis et al. 1989),and can mediate intraguild predation dynamics (by impactingeither the intraguild prey or the resource species; Rudolf2007a). Among snakes however, cannibalism is treated largelyas rare opportunism associated with the consumption of adiverse range of prey (Polis and Myers 1985) and frequentlydismissed as such. Yet snakes offer a fascinating context inwhich to examine cannibalism, and ophiophagy (the con-sumption of snakes) in general, because their gross morphol-ogy and gape-limitation mean that snakes and other elongateprey offer the greatest value in terms of prey mass relative tocross-sectional size (Cundall and Greene 2000). This meansthat snakes (including conspecifics) represent relatively enor-mous meals with important energy balance and fitness impli-cations for snakes that are consuming them. Perhapsunsurprisingly, ophiophagy has evolved multiple times withinsnakes, and several species (e.g., king cobras [Ophiophagushannah], kraits [Bungarus spp.], king snakes [Lampropeltisspp.], and several genera of African file snakes [Lamprophi-idae]) are thought to consume snakes regularly.Cape cobras are large-bodied (occasionally exceeding 2 m

in length) dangerously venomous elapid snakes that inhabit

FIG. 1. In situ cannibalism in the cape cobra (Naja nivea) from Tswalu Kalahari Reserve, Northern Cape Province, South Africa.

1

The Scientific NaturalistEcology, 0(0), 2018, pp. 1–3© 2018 by the Ecological Society of America

the arid western half of southern Africa. They consume aremarkable range of ectothermic and endothermic tetra-pods, including other snakes; a recent review (Layloo et al.2017) suggests that other snakes make up as much as one-third of their diet. A close examination of the literaturereveals that, while cannibalism by cape cobras has beenreported (Broadley 1983, Phelps 2007, Layloo et al. 2017)all of those instances had been treated as aberrant. However,if cannibalism occurs more frequently in cape cobras thancurrently perceived, it has important implications for ourlong-term field study. What if cannibalism is more commonin the wild than currently thought?Inspired by our field observation, we set out to establish

how widespread cannibalism and ophiophagy are in cobras(Naja spp.), a clade of approximately 30 species that occupiesmost of Africa as well as south and east Asia. We performedan extensive review of cobra diets, mining information frompublished scientific reports, minor and difficult to access pub-lications (e.g., herpetological and ornithological newsletters;museum bulletins), as well as a social media group that wehave created to collate trophic information for reptiles andamphibians in southern Africa (available online).4 Werestricted our analysis (see Appendix S1) to species that occurin southern Africa because of their relevance to our own workand because of our access to relatively obscure literaturesources for this region. The six species (N. anchietae, N. an-nulifera, N. nivea, N. mossambica, N. nigricincta, and N. sub-fulva) collectively have a cosmopolitan distribution across theregion, occupying habitats as different as deserts and forests(Broadley 1983). Our analysis excludes Naja nigricollisbecause the only population in southern Africa is restrictedto the extreme north of the region and our only diet recordsfor the species come from outside of this region (Appendix S1:Table S1).Our results highlight that ophiophagy is commonplace

among wild cobras (Fig. 2A; Data S1). All six species haverecords of snakes in their diet confirming that ophiophagy isphylogenetically widespread in the group. We found thatsnakes account for 13–43% of all prey species detected in thediets of wild cobras. This result was somewhat surprisinggiven that Shine et al. (2007), in an excellent empirical reviewof southern African cobra diets, found no evidence of ophio-phagy in N. nivea or N. subfulva whereas our assessment sug-gests snakes account for about 32% and 24% of their diets,respectively. In contrast to that landmark paper, our reviewrevealed a significant increase in the average proportion ofsnake prey items (paired t test, t17 = 5.26; P < 0.001; averageincrease = 24.5%; Fig. 2A) in the diet of southern Africancobras. Puff adders (Bitis arietans), a species of pan-Africanviper, are particularly obvious in the diet of all cobras, espe-cially N. nivea, N. nigricinta, and N. annulifera (making up33%, 42%, and 43% of all snake prey items, respectively).Collectively, puff adders accounted for 27% all snake preyitems consumed by wild southern African cobras.Our survey revealed cannibalistic events in the wild for

five of the six species assessed (sans N. nigricincta) whereasShine et al. (2007) found no evidence of cannibalism at all.

Our review suggests that conspecifics represent at least 4%of all prey items consumed by cape cobras, and 12.5% of allsnake prey items (N = 148). Remarkably, conspecifics werethe second most abundant snake prey species (out of 15) forcape cobras (only puff adders were more abundant). Dis-crepancies between our findings and those of Shine et al.(2007) likely arise as a result of differences in sampling effortand the number of methods employed: method-specificbiases in the detection of different prey items in animal dietscan severely limit studies of prey diversity when only onemethod is used (Glaudas et al. 2017). Regardless, our find-ings suggest that southern African cobras display frequentophiophagous behavior which predisposes them to cannibal-ism. Given that diet appears to be conserved within cobras(Fig. 2B; Spawls et al. 2018, Whitaker and Captain 2004),this syndrome likely characterizes cobras more broadly.Contrary to widely held views, snakes appear to offer an

underappreciated array of cannibalistic scenarios. Theseencompass (1) rare, isolated events seemingly without mecha-nistic explanation (Shine 1977, Greene 1984), (2) relativelycommonplace events as seen in cape cobras, or even (3) appar-ent adaptive avoidance of cannibalism, e.g., K. Wiseman andH. Greene (unpublished data) didn’t detect cannibalism in asample of over 400 prey items for the highly ophiophagousCalifornia kingsnake (Lampropeltis californiae). This range of

FIG. 2. (A) Relative frequency of ophiophagy and cannibalismin the diet of six southern African cobra species as historicallyreported (gray) and following our review (yellow), and (B) the fre-quency with which six species of southern African cobras consumedifferent prey types. Sample size is provided above each bar.

4 www.facebook.com/groups/888525291183325

2 THE SCIENTIFIC NATURALIST Ecology, Vol. xx, No. xx

potential manifestations of cannibalism in snakes offers anopportunity to examine the possible drivers of cannibalism,especially when viewed in the context of ophiophagy ingeneral, and the potential massive energetic value of snake-shaped meals.Closer to home, we are just beginning to understand the

nature of inter- and intraspecific interactions in cape cobras.Recently, “Hannibal” has been observed in the presence of anon-telemetered cape cobra, during which time he tongue-flicked and pursued the other snake for a short distancebefore moving off in another direction. To date, the winnersand the losers of recorded cannibalism events in cape cobrashave been males, hinting that cannibalism might impact intra-sexual competition, but our current sample size for this isproblematically small. Additionally, this male bias in ourobservations raises the question of whether cannibalismevolved from a male–male combat behavioral precursor, espe-cially given that male–male combat in cobras is known toinclude biting (Shine et al. 2007). Luckily, frequent cannibal-ism raises important, testable hypotheses regarding age-struc-tured utilization of point resources which might providefruitful avenues of enquiry. In our study system, for instance,visiting sociable weaver (large communal nesting passerinesthat are heavily preyed on by cobras) colonies may be toocostly for juvenile cobras given the inherent risk of intraspeci-fic predation, forcing them to avoid this resource and result-ing in size-structured feeding niche and habitat-usedifferentiation within the cobra population. Overall, increas-ing our knowledge of how snakes behave and interact withconspecifics and resources in the landscape will ultimatelyallow us to test more complex hypotheses such as whether ornot cannibalism occurs more frequently during bouts of foodscarcity (Polis 1981), or whether size-structured cannibalismcan drive indirect positive interactions between smallercobras and their heterospecific prey (Rudolf 2007b).

ACKNOWLEDGMENTS

We thank the National Research Foundation of South Africa(Grant no. 99186 to B. Maritz) and the University of the Witwater-srand University Research Council Fund (to G. J. Alexander). Wealso thank the Tswalu Foundation and the staff at Tswalu KalahariReserve. Finally, we thank H. Greene and two anonymous reviewersfor their comments on an earlier version of this manuscript.

LITERATURE CITED

Broadley, D. G. 1983. Fitzsimons’ snakes of southern Africa. DeltaBooks, Pretoria, South Africa.

Claessen, D., A. de Roos, and L. Persson. 2004. Population dynamictheory of size-dependent cannibalism. Proceedings Royal SocietyB 271:333–340.

Cundall, D., and H. W. Greene. 2000. Feeding in snakes. Pages 293–333 in K. Schwenk, editor. Feeding: form, function, and evolutionin tetrapod vertebrates. Academic Press, San Diego, California,USA.

Glaudas, X., T. C. Kearney, and G. J. Alexander. 2017. Museumspecimens bias measures of snake diet: a case study using the

ambush-foraging puff adder (Bitis arietans). Herpetologica73:121–128.

Greene, H. W. 1984. Feeding behavior and diet of the eastern coralsnake, Micrurus fulvius. Pages 147–162 in H. S. Fitch and R. ASeigel, editors. Vertebrate ecology and systematics: a tribute toHenry S. Fitch. University of Kansas Museum of Natural His-tory, Special Publication 10. University of Kansas, Lawrence,Kansas, USA.

Layloo, I., C. Smith, and B. Maritz. 2017. Diet and feeding in theCape Cobra, Naja nivea. African Journal of Herpetology 66:147–153.

Phelps, T. 2007. Observations of the Cape cobra, Naja nivea (Ser-pentes: Elapidae) in the De Hoop Nature Reserve, Western CapeProvince, South Africa. Herpetological Bulletin 99:29–35.

Polis, G. 1981. The evolution and dynamics of intraspecific preda-tion. Annual Review of Ecological Systematics 12:225–251.

Polis, G. A., and C. A. Myers. 1985. A survey of intraspecific preda-tion among reptiles and amphibians. Journal of Herpetology19:99–107.

Polis, G. A., C. A. Myers, and R. D. Holt. 1989. The ecology andevolution of intraguild predation: potential competitors that eateach other. Annual Review of Ecology and Systematics 20:297–330.

Rudolf, V. H. 2007a. The interaction of cannibalism and omnivory:consequences for community dynamics. Ecology 88:2697–2705.

Rudolf, V. H. 2007b. Consequences of stage-structured predators:cannibalism, behavioral effects, and trophic cascades. Ecology88:2991–3003.

Shine, R. 1977. Habitats, diets, and sympatry in snakes: a studyfrom Australia. Canadian Journal of Zoology 55:1118–1128.

Shine, R., W. R. Branch, J. K. Webb, P. S. Harlow, T. Shine, and J. S.Keogh. 2007. Ecology of cobras from southern Africa. Journal ofZoology 272:183–193.

Spawls, S., K. Howell, H. Hinkel, and M. Menegon. 2018. Fieldguide to East African reptiles. Bloomsbury Natural History, Lon-don, UK.

Van Allen, B. G., F. P. Dillemuth, A. J. Flick, M. J. Faldyn, D. R.Clark, V. H. Rudolf, and B. D. Elderd. 2017. Cannibalism andinfectious disease: friends or foes? American Naturalist 190:299–312.

Whitaker, R., and A. Captain. 2004. Snakes of India: the field guide.Draco Books, Chennai, India.

BRYAN MARITZ,1,3 GRAHAM J. ALEXANDER2, AND

ROBIN A. MARITZ1

Manuscript received 26 June 2018; revised 21 August 2018;accepted 30 August 2018. Corresponding Editor: John Pastor.

1Department of Biodiversity and Conservation Biology,University of the Western Cape, Private Bag X17, Bellville 7535South Africa.

2School of Animal, Plant and Environmental Sciences, Universityof the Witwatersrand, Johannesburg, P.O. Wits, 2050 South Africa.

3 E-mail: bmaritz@uwc.ac.za

Additional supporting information may be found in the online ver-sion of this article at http://onlinelibrary.wiley.com/doi/10.1002/ecy.2522/suppinfo

Xxxxx 2018 THE SCIENTIFIC NATURALIST 3