title for learn-about section on whelks & relatives in A SNAIL'S ODYSSEY
  Foods, feeding, & growth

Topics in this section includeCalifornia cone shells: cusps used as harpoons, considered here, and
, and
HATCHLINGS AS PREDATORS considered in other sections.

  black dot

California cone shells: cusps used as harpoons


schematic showing arrangement of venom gland, harpoon sac, and proboscis in a cone shell Conus This unique strategy has evolved in cone shells Conus, of which there are some 500-600 world species. Cone shells are carnivores that employ modified radular cusps or teeth as harpoons (see photograph below and in situ on Right) to stun polychaete, snail and, in some cases, fish prey, before feeding on them. Cone shells lack a "traditional" scraping radula. Prey if small are swallowed whole by drawing them via the "mouth" into an expansible sheath that surrounds the proboscis, and then ingested by the proboscis (see drawings on Right and photos below). Ingested particles are moved directly into the esophagus by muscular contractions, and thence to the stomach perhaps by cilia. If too large to be taken in to the proboscis sheath in this way, the predator scavenges on the soft parts with its extended proboscis. It is not uncommon to see several west-coast cone shells Conus califrornicus jointly scavenging on a dead, decaying prey (see photograph in Research Study 3 below).

dissection of cone shell to show venom apparatus, radula gland, and proboscisThe teeth or harpoons are constructed of a flat sheet of chitin and protein rolled up to enclose a spiral channel or duct that may be only partially closed off from the outside. The teeth are secreted in the long end of a Y- or L-shaped radula sac and stored in the short end, oriented there much arrows in a quiver mostly with their pointy ends facing towards the opening of the storage sac into the pharynx (see drawing on Right). The teeth are hardened by quinone-tanning and by inclusion of calcium. They are moved individually as needed to the end of the proboscis where they are firmly held by a special sphincter muscle. While no-one seems to know exactly how they get from the storage sac to the tip of the proboscis, one idea is that the proboscis inverts on itself, and in this inside-out form reaches back to envelope a tooth completely. Then, gripping the basal end of tooth, the proboscis everts with the harpoon now pointing in the correct direction from the tip of the proboscis. Simpler still might be for the tooth to be just moved into position through muscular paristalses of the proboscis from the storage sac into the pharynx and then into the proboscis (see drawing on Right).

Venom is produced in the venom ducts and stored there. When needed it is pulsed along the duct into the pharynx, and harpoon of cone shell Conus californicusdown the proboscis, possibly moving from pressure generated by the venom bulb or, alternatively, from the proboscis. It is unclear whether the venom is ultimately squirted through the harpoon or whether is just becomes filled with or perhaps coated with venom as the harpoon moves through the proboscis. With the violence of the strike and accompanying pressures exerted, it may be that venom is delivered by any or all of these means. The proboscis contains large hemocoelic spaces and can be extended and retracted by muscular action on the hemolymph contained within the spaces in the manner of an hydrostatic skeleton. The proboscis is highly extensible and can be moved quite quickly to make the strike or, in some cases, to wrap around a prey preparatory to swallowing (see Research Study 6 below). Several strikes may be needed to subdue a prey. When the cone senses that the prey is quiescent enough to be eaten it expands its mouth and the sheath surrounding the proboscis and, if the prey is small enough, engulfs it. Photographs of some of these actions are shown below.

The venom, or conotoxin, is manfactured in the coiled venom ducts (coloured blue in the above drawing). It is a potent neurotoxin consisting of a variety of different peptides, many specialised for a certain neurophysiological role, such as affecting membrane permeability or neurotransmitter function, or otherwise specialised to match the physiology of a certain prey type, or both. Schematic of venom gland, harpoon sac, and sheath/proboscis courtesy Gladys Archerd Shell Collection, Washington State University Tri-Cities Natural History Museum, Richland, Washington Archerd Shell Collection. Photograph of proboscis/harpoon top Right courtesy Clay Brice, In, Australian marine shells, Odyssey Publ., San Diego, CA. Photograph of harpoon of C. californicus courtesy Alan Kohn, University of Washington, Seattle (Kohn et al. 1999 J Moll Stud 65: 461). Drawing of dissected Conus courtesy A. Traica.

NOTE the opening to the proboscis sheath is sometimes termed a "mouth", but of course the real mouth or opening to the digestive tract is at the end of the proboscis. The proboscis sheath is a blind cavity into which a stunned prey, if too large to be taken in via the proboscis, can be stored while its soft parts are being nibbled on by the proboscis. This cavity is sometimes referred to as a rhynchocoel (lit. "snout" "cavity" G.), analogous to the space surrounding the inverted proboscis in a nemertean worm

Cusp, tooth, or harpoon of cone shell Conus californicus. Note the barbs and
the basal end shaped like a baseball bat (for better grip?). This harpoon is about
1mm in length, but in other species may be 7mm or more (see photograph above)

Photographs of a Conus geographus attacking and eating a fish (source: unknown video on YouTube):

photograph from a video of Conus geographus hunting a fish Number 1 in a series of 4 photograph from a video of Conus geographus hunting a fish Number 2 in a series of 4 photograph from a video of Conus geographus hunting a fish Number 3 in a series of 4 photograph from a video of Conus geographus hunting a fish Number 4 in a series of 4
Conus geographus emerges from sand. The proboscis is probing out from the sheath. The tube above is the siphon In this front view the proboscis is just visible within the sheath, while the siphon is on the left side Post-strike on a fish leaves the proboscis without a harpoon. The sheath is expanding. Tentacles with eyes are visible The sheath expands greatly to take in the now stunned fish. The fish's head is at the left. A black spot marks its doral fin
Research study 1

drawing of harpoon tooth of cone shell Conus californicusAn early investigation of feeding in Conus californicus includes the description of a “radula” consisting of a mass of clear, needle-like hollow teeth, each tooth with a spirally twisted channel running its length with openings at either end.  The sample tooth figured in the publication and shown here is identical with the one shown in the introductory section above and in Research Study 4 below.   The authors posit that these teeth “may function in the injection of poison into an attacked animal although there is no known evidence of such action by this species”.  Hanna & Strong 1949 Proc Calif Acad Sci 26 (9): 247.

Research study 2

As noted in the introductory part above, conotoxins are peptides of small molecular-mass. The configuration of amino-acids, especially cysteine, is highly variable thus producing many different toxins, some specific to disabling a certain neural function; others, specific to certain prey taxa.  Summaries of a few studies are given here to show the diverse effects of Conus californicus conotoxins on nerve and nerve/muscle preparations in different animals:

Venom of Conus californicus purified and lethal component identified through disc electrophoresis as a protein or bound to a protein.  Whysner & Saunders 1966 Toxicon 4: 177.

5-hydroxytryptamine-like effects leading to.relaxation of byssus-retractor muscle in mussels Mytilus sp. caused by bathing with seawater extracts of C. californicus venom gland.  Cottrell & Twarog 1972 British J Pharmacol 44 (3): 365P.

Cholinergic receptors activated in sea hares Aplysia californica Elliott & Kehoe 1978 Brain Res 156: 387.

Cholinomimetic effects on isolated Mercenaria mercenaria (clam) heart: interferes with nerve-impulse transmission and inhibits its beating.  Elliott 1979 J Comp Physiol 129: 61.

Cholinomimetic effects on Mercenaria (clam) heart and Aplysia (sea hare) central neurons; interferes with nerve-impulse transmission.  Elliott & Raftery 1979 Toxicon 17 (3): 259.

Sodium-channels targeted by a novel conotoxin that is specific for squids and does not work in sea hares Aplysia californicaBingham et al. 2000 J Gen Physiol 116 (1): 12A.

Calcium-channel blocking  in rat neurons.  Bernaldez et al. 2011 Toxicon 57: 60.

Sodium-channel blocking: a number of unique variations of toxins are identified showing specificity for similar molecular sodium-channel targets in taxa that may be preyed upon by C. californicus. As an example of specificity, certain peptides that block sodium channels in cephalopods appear to have no activity in gastropods.  Gilly et al. 2011 J Exper Biol 214: 147.

NOTE  includes 2 abstracts

Research study 3

photograph of Conus californicus feeding on a snail Caesia fossatusThere is only one west-coast species of cone shell, Conus californicus, distributed from the Farallon Islands off San Francisco south to Baja California.  Early observations by researchers from the University of Southern California, Los Angeles show that in laboratory settings the preferred prey are snails Nassarius (now Caesia) spp. (see photograph on Left), polychaete worms Glycera dibranchiata, and Olivella (now Callianax) biplicata.  Other prey include bivalves and sometimes fishes.  Dead animals such as fishes and cephalopods are commonly scavenged. In the absence of prey the cone shells generally remain buried in sand, with their siphons maintaining contact with the surface for gas exchange.  When a prey is sensed, a Conus immediately emerges from the sand, extends its proboscis to nearly shell length and waves it around searching.  Contact of the proboscis tip with the prey leads to immediate injection of a radula tooth.  The tooth is barbed and may remain imbedded for a few moments during which the prey may shake free.  This induces further attacks from the same or other Conus attracted to the scent of food.  Within a few moments the prey becomes moribund from the toxin injected via the radula tooth, and feeding commences.  A snail may be removed from its shell at this time, and consumed completely within 30min.  The authors are unclear how a Conus can feed on the muscular tissues of a snail prey, but they do note in at least one instance that a thin strand of tissue of "thread" could be withdrawn from the photograph of cone shell Conus californicus attacking a sandworm, courtesy Kevin Lee, Fullerton, Californiacone on its removal from the prey, so perhaps there is some preliminary in situ digestion via enzymes. This might be something to investigate. The authors describe several instances of group feeding on snails Caesia sp. in the laboratory, and on worms Arenicola cristata and fishes in the field. Saunders 1959 The Veliger 1 (3): 13; Saunders & Wolfson 1961 The Veliger 3 (3): 73. Photograph above courtesy the authors; photo on Right courtesy Kevin Lee, Fullerton, California diverkevin.

NOTE the authors do describe the presence of radular teeth scattered about in one of their dissections. They add, however, that their preparation was rather degraded and, by their description, it seems likely that what they found were the remains of the harpoon sac

Conus californicus attacks a nereid polychaete 0.5X

Research study 4

cusp or harpoon of cone shell Conus californicusCone-shell teeth (cusps, harpoons) can be divided into 4 morphologically different groups, and these groups roughly correspond with 3 major feeding types of molluscivores, piscivores, and vermivores (worm-eaters), with the 4th group being unique.  The vermivore type is most numerous among Conus species and also the most diverse in terms of tooth morphology.  Since there is only one west-coast Conus species the details of these groupings are outside the interest of the Odyssey; suffice it to say that Conus californicus is placed in the “unique” group, and its broad dietary preferences encompass all 3 major food types.  Nybakken 1990 32 (1): 35.

Research study 5

schematic of venomation system in a cone shell Conus californicusA different interpretation of how venom is delivered to a prey during the stabbing action of the harpoon in Conus californicus is provided by researchers working primarily at Hopkins Marine Station, Pacific Grove, California.  By use mainly of electon-micrographs they determine that the part of the venom duct nearest the venom gland or bulb is structured more for active transport than secretion, while the part nearest the pharynx is structured for secretion.  Active transport of amino acids and fatty acids for energy would be prerequisite to production of venom, but with most of its production occurring in the duct-lining proximal to the pharynx.  The opening of the venom duct into the pharynx consists of many extremely small and narrow channels, suggesting that venom injection into the pharynx is both slow and in small volume.  Granules containing putative peptides are present within the secretory cells in the proximal duct, in the lumen of the duct, and also in dense numbers within harpoon of cone shell Conus californicus showing putative peptide granules withinharpoons contained in the radular sac (see drawing).  It appears then that venom is back-washed in some way into the radular sac and that the teeth within are “pre-loaded” with toxins.  The authors are careful to note that this observation does not preclude additional venom being released along with a tooth at the time of stabbing.  Marshall et al. 2002 Biol Bull 203: 27.

Research study 6

graph showing number of stings required to subdue different-sized fishes by a cone shell Conus californicuscone shell Conus californicus courtesy Kevin LeeA later study at Hopkins Marine Station, Pacific Grove, California provides interesting detail of feeding behaviour of Conus californicus on 2 species of prickleback1  fishes in captivity. Two feeding methods are used, similar to ones employed by strictly piscivorous Conus in other parts of the world.  The first involves stinging with a toxic radular tooth2 , often done multiply by a single Conus to subdue a prey or even by several Conus on the same prey.  Like specialised piscivorous cone shells, C. californicus usually retains a grip on the radula tooth after injection and is able to pull the fish into the opening of the proboscis sheath, or rostrum3 , a type of prey capture known as “sting-and-pull” or “hook-and-line” fishing.  The prey is then engulfed whole via the highly expansible sheath.  The second method simply involves engulfing without stinging, although whether the prey is somehow anaesthetised by the cone shell to enable this is not known.  Larger fishes may require several stings before being captured (see graph).   The effect of successive stings is additive, and leads to a seriously disabled state in which the fish progressively stiffens and its respiration weakens.  Interestingly, the researchers discover a progressive habitituation occurring over the course of repeated encounters of the snails with fishes, with more rapid arousal of the snails with each new appearance of prey fishes.  Stewart & Gilly 2005 Biol Bull 209: 146. Photograph above courtesy Kevin Lee, Fullerton, California diverkevin.

NOTE1  these are Cebidichthys violaceus and Xiphister spp.  This laboratory account is the first of C. californicus eating fishes; the extent to which this occurs in the field is not known

NOTE2  the teeth are ancestral radular cusps modified into hollow hypodermic-like harpoons through which a toxin is injected under high pressure.  The toxins tend to be specific to a certain prey category being consumed, but perhaps not in the case of C. californicus whose diet is generally less specialised than other world species.  At the time of this publication the toxins of C. californicus had not been characterised

NOTE3  the sheath (rostrum) can expand greatly, and leads directly into the esophagus

The following shows a sequence taken from a video of an attack by Conus on a prickleback fish:

1 in a series of 6 photo/drawings of Conus californicus catching a prickleback fish 2 in a series of 6 photo/drawings of Conus californicus catching a prickleback fish 3 in a series of 6 photo/drawings of Conus californicus catching a prickleback fish
A sting is delivered here by proboscis, the first of 3 in total. Note the proboscis with harpoon spear at its tip The fish, 2cm in length, struggles from the effects of the toxin The fish continues to struggle
4 in a series of 6 photo/drawings of Conus californicus catching a prickleback fish 5 in a series of 6 photo/drawings of Conus californicus catching a prickleback fish 6 in a series of 6 photo/drawings of Conus californicus catching a prickleback fish
The sheath, or rostrum, opens to engulf the fish. Note that the proboscis originates from within the sheath The fish is now half-ingested, and the snail begins to pull the prey into its shell Only the tail of the fish remains exposed. An arrow points tothe searching proboscis of a second, buried snail
Research study 7

Recently, a large consortium of researchers from various U.S. and Mexican institutions has sequenced 12S, 16S, and COI gene segments of the mitochondrial genomes, and cDNA, of the venom gland of Conus californicus, and constructed phylogenetic trees including comparable published sequencing data for other Conus species and several outgroup species (other snails).  The resulting tree for 12S and 16S rRNA sequences from phylogenetic tree of 57 species of cone shells including Conus californicus57 Conus species and 2 outgroup snail species yields one large and one small clade with C. californicus occupying a branch well outside of and sister to the other clades (see schematic), and fully supports the notion that C. californicus is a genetic outlier. The authors note that C. californicus shares only half of the conotoxin-gene superfamilies found in the other Conus species, with the split thought to have occurred in the early Tertiary Period, some 50-60 million years ago.   aIn support of the notion that C. californicus has had a long history of genetic separation from the other species, they point to aspects of prey-capture behaviour that have diverged significantly from the other evolutionary lines of hunting behaviour in Conus species.  Additional to the much broader range of prey taxa comprising the diet of C. californicus, then, is an idea that its mode of hunting has evolutionarily diverged from that of the other phylogenies.  The authors  refer here to what they term “group hunting”, or “pack behavior” of C. californicus, where several individuals “cooperate” in attacking prey such as prawns or snails that are too large for a single cone shell to handle alone.  Now, C. californicus is both a predator and a scavenger, and it is not uncommon for several individuals to be found feeding on the decaying flesh of a dead fish or cephalopod, nor is it uncommon for several individuals to simultaneously attack a single large prey, but whether these activities can be termed “cooperative hunting” similar to the behaviour of a tropical-fish hunting assemblage or, in the words of the authors, “a marauding wolf pack”, is doubtful.  The description of C. californicus attacking single shrimps in “seemingly coordinated groups” is just as fanciful, and the comment that this is the first documentation of any Conus species preying upon a crustacean is simply incorrect, as amphipods are known to be a dietary component of at least one C. californicus. Biggs et al. 2010 Mol Phylogenetics Evol 56: 1. Photograph courtesy the authors.


photograph of several Conus californicus feeding on a goldfish in an aquarium tank

  a measure of geologic time lasting from approximately 65mya to 2.6mya

NOTE  the terminology used by the authors for such a "cooperative" attack on a large fish in an aquarium is imaginative to say the least: “the first snail..tether(s) the prey..and gives other snails a chance to approach; multiple snails..then rapidly converge..in a manner reminiscent of a marauding wolf pack; and all..feed together..”.  The idea for this "pack behavior" is claimed to have come from an earlier account by Saunders & Wolfson (1961) describing unusual feeding behaviours in C. californicus, including “organized/cooperative attacks” on other snails, but a careful re-reading of the 1961 publication reveals no such description (see Research Study 3 above)  

Group of several Conus californicus
feeding on a goldfish in an aquarium tank

Research study 8

Recent research at Hopkins Marine Station, Pacific Grove, California provides interesting insight into the evolution of cone-shell toxins, or conotoxins.  In comparison with other world species of Conus, most of which have highly specialised diets, C. californicus with its generalised diet would be expected to have greater variety of conotoxins.  That said, the researchers use a combination of DNA- and protein-based methods to identify in C. californicus close to 50 inferred peptide toxins, half of which appear to be unique to other known “superfamilies” of conotoxins.  The authors suggest that the difficulty of fitting C. californicus conotoxins into established toxin classification schemes owes to the unusually wide range of prey species consumed and to the large phylogenetic distance between this solitary west-coast species and the many Indo-Pacific species.  Elliger et al. 2011 Toxicon 57: 311.

NOTE  conotoxins are short-length peptides that are rich in cysteine amino acids.  Most are only 10-40 amino acids in length, of which 10 or so may be cysteine. The toxins are classified into 8 categories based on how the cysteines are configured

Research study 9

The probing tip of a cone-shell’s proboscis must be endowed with numerous types of chemo- and mechanosensory organs to enable it to locate and identify prey, both at a distance and close in, and to ensure that the harpoon will be inserted into a suitable area of soft tissue of the prey. A consortium of California researchers using a variety of microscopical techniques have recently described finger-shaped papillae on the proboscis tip of Conus californicus and related species. The papillae have columnar epithelial cells with dense apical borders of microvilli, and from external view have all the appearance of sensory structures. Length and form of the papillae appear to relate to preferred prey: those of worm- and mollusc-hunting species are short and conical, while those of fish-hunting species additionally have long tubular types. Long proboscises photographs of papillae on the proboscis of a cone snail Conus californicusand other extensible soft parts of invertebrates tend to be bitten off from time to time, so a good regerative ability is not unexpected. Indeed, experimental ablation of proboscis tips in C. californicus lead to quick regeneration and full functionality within 10d. James et al. 2014 Invert Biol 133 (3): 221.

NOTE information is provided for 8 other world Conus species, not included here

Left: tip of proboscis of Conus californicus showing tubercles
(bumps) covering surface: Right: close view of 3 tubercles

  black dot