The Moon Snails

Four Names, Four Ideas About What Matters

English named this family after the moon. Other languages looked at the same animal and named something else. Norwegian calls them boresnegler and Swedish borrsnaeckor, both built on the verb to bore. They are drill snails. German settled on Nabelschnecken, navel snails, after the umbilicus, the pit at the center of the underside. English carries a second name in the register too, necklace shells, which is not about the shell at all. It is about the egg case the animal leaves in the sand.

Four names against one record, and four different decisions about what the defining feature is: how it looks, what it does, one structure on its underside, and the object it leaves behind.

The record is Naticidae Guilding, 1834, AphiaID 145 in the World Register of Marine Species, order Littorinimorpha. Its superfamily, Naticoidea, holds exactly one accepted family, so the two are the same tent. Kabat's 1996 estimate puts the living family near 260 to 270 species, richest in the tropics but running from the intertidal down to several thousand meters.

A Foot Three Times the Length of Its Shell

The first thing wrong with a live moon snail is the proportions. There is far more animal than shell, and the animal is on the outside of it.

The foot comes in two parts. The front section is the propodium, a broad plow running ahead of the shell. Behind it is the mesopodium, the main sole, which wraps up and over the shell as the animal moves. Huelsken and colleagues described living animals of eight species in a 2008 study and put measurements on it: an extended mesopodium twice the length of the shell, a propodium about as long as the shell again. A crawling moon snail is roughly three shell-lengths of animal wrapped around one shell-length of shell. In baby's ear, Sinum perspectivum, the disproportion goes all the way and the animal cannot withdraw into its shell at all.

E. R. Trueman filmed and instrumented the burrowing in 1968 and found a repeated four-stage cycle, the propodium swelling and relaxing in alternation with another region of the foot, driven by blood moving within the pedal sinus. The mechanism runs backward from what it looks like. The shell is not pushed down into the sand. The propodium dilates below, anchors, and the shell is pulled down to it. A moon snail descends the way a climber ascends: set the anchor, haul the body to the anchor.

Disturb one and it reverses in seconds. The foot deflates, fluid runs out of it, the animal withdraws, and the operculum seals the aperture. Re-inflation takes much longer. That asymmetry is what anyone handling one actually sees.

What the fluid is remains argued. Nearly every popular account says the foot fills with seawater, and there is real literature behind that: Schiemenz reported pores opening into an aquiferous system separate from the blood sinuses in 1884, and Bernard described an aquiferous system for podial expansion in 1968. But Schiemenz's pores were never corroborated by anyone else, and Trueman's mechanical work puts the whole thing down to blood movement in the pedal sinus with no seawater required. Both positions are in print and neither has been retired.

A Shell Built to Be Pulled Through Sand

Everything about a naticid shell is subtractive. Globose to ear-shaped, four or five whorls, a low spire, a body whorl that is most of the shell, and essentially no sculpture: no spines, no varices, no long canal at the front. One animal is armored and works sitting still; this one is streamlined and moves through the bottom rather than on it. The material itself is the same everywhere, and that is on how a shell is made.

What diagnostic detail the family has sits around the umbilicus. The umbilical callus is a deposit of shell material filling part or all of that pit, and it explains something people ask about: the dark spiral button on a shark eye is not a marking or a stain. It is a structure. The eye is architecture. The funicle, a cord-like ridge spiraling into the umbilicus, is used as a character of the subfamily Naticinae.

The other landmark is the operculum, calcareous in Naticinae and horny in Polinicinae and Sininae. Hold that loosely as a classification tool. The traditional arrangement of this family was built on what the door is made of, and molecular work has since argued that character cannot carry a split that deep.

The Drill Is on Its Face

Two groups of snails on this coast open other shells by drilling through them: the moon snails and the murexes. Both use the same two-tool method, a rasping radula alternating with a gland that chemically softens shell. Both leave a round hole.

They are not related in this behavior. They invented it separately, and you can show that from where the gland is bolted on.

In a murex the accessory boring organ sits on the mid-ventral sole of the foot. In a moon snail it sits on the ventral side of the tip of the proboscis, just behind the mouth, developing embryologically as a patch of enlarged epidermal cells. Same job, near-identical function, two different body parts. A murex drills with its foot. A moon snail drills with its face.

That is convergent evolution with an anatomical receipt, and it carries a practical warning: nothing about one family's drilling can be safely assumed of the other.

The beach-facing version of the difference, the countersunk moon snail hole against the straighter-sided murex one, is already worked through on our Florida shell identification guide and on the murexes and drills. Read either for the tell. The rest of this page is what those two do not have room for.

What Actually Dissolves the Shell

The rasping half is ordinary equipment. Naticids are caenogastropods and carry a taenioglossate radula, seven teeth to a transverse row. We call that the standard caenogastropod condition rather than a naticid discovery, because our research pass did not land a naticid-specific primary citation for it.

The chemical half is not simply acid. The main event is enzymatic: carbonic anhydrase catalyzes the hydration of the animal's own metabolic carbon dioxide, and the hydrogen ions that produces are selectively transported into the borehole, where they release calcium out of the shell. Hydrochloric acid is in there too, with chelating agents that bind calcium and further enzymes still not fully identified.

An attribution note most sources skip: the classic ultrastructural work on boring chemistry is M. R. Carriker's, and it was done on Urosalpinx cinerea, a murex. Given the previous section, moving those findings across families without saying so is exactly the wrong move. For a treatment that covers naticids, the reference is Clelland and Webster's 2021 chapter on drilling into hard substrate by naticid and muricid gastropods.

The two tools alternate: soften the patch, rasp a pit inside it, soften again, with the chemical phase taking more of the total time than the scraping. Because the softened patch is wider than the rasped pit, the hole ends up wider outside than inside, a shallow parabola in section rather than a cylinder. The famous bevel falls out of the method rather than being aimed at.

The Arithmetic of a Meal

Popular accounts say the drill-and-feed sequence takes four to five days. We could not verify that against primary literature, so we are leaving the number out.

What is well sourced is better material anyway, because it is about decisions rather than duration. Kitchell, Boggs, Kitchell and Rice built a cost-benefit model of naticid prey selection in Paleobiology in 1981 and found preference tracks profitability: which clam, and which size of that clam, is predicted by what the animal gets back for the time it spends. Thin-shelled prey goes faster, and the time saved is reinvested in foraging. A moon snail is running an economy, and the hole is the invoice. When the wall gives, the proboscis goes in through it, digestive fluid with it, and tissue comes back out.

One correction to the picture. Moon snails are not obligate clam killers. Casey, Fall and Dietl ran stable isotopes on Neverita duplicata in 2016 and got a realized trophic position of 1.8 to 2.5 against the 3.0 expected of a straightforward predator, with roughly 60% of the carbon from littoral sources. Lab-fed controls confirmed normal nitrogen fractionation, so that is real omnivory rather than an artifact. They eat algae and carrion, and they drill each other routinely. They are also not the only snails that drill: murexes do, so do some marginellids, nassariids and helmet shells, and octopuses drill too.

Not Half of Them. Closer to One in Eight.

A line circulates about this family: that half the clam shells on the beach carry a moon snail hole. It is a good line and it does not survive contact with the counts.

The best modern number for our water is Calderaro, Harnik and Rillo, published in Paleobiology in 2023. They examined 3,568 whole bivalve shells from 15 stations in Holocene death assemblages of the northern Gulf of Mexico and recorded 445 drill holes. Roughly one shell in eight.

Drilling frequencies for the Florida Gulf are reported elsewhere in a range of about 5% to 35%, which we pass on as indicative rather than hard, because we read it in secondary reporting and did not open the primary table. Two beaches on San Salvador in the Bahamas came in near 11% and 37%: different fauna, different place, not ours to import.

So the honest version is a striking minority. Commonly around one shell in ten, at some sites as many as one in three. That is still an enormous quantity of predation to be able to hold in your hand, and the absolute count in a single handful stays high even at the low end. It just is not half.

The Bevel Says Probably. The Position Says Who.

Here is the correction that matters most on this page, and it cuts against the tidy version of the tip. Hole profile is not the reliable family-level test. Hole position is.

Paul and Herbert reported it in 2014, in a study of Plio-Pleistocene drilling predation in Florida bivalves, building on earlier work by Dietl and colleagues. Working on Chione, they found murexes drill almost exclusively over the ventral half of the valve and generally away from the direct center, while the naticid Neverita delessertiana drills almost exclusively on or near the umbo, the beak up by the hinge. Where the hole sits separates the two families more cleanly than what shape it is.

Three things go wrong with a profile-only test, and all three have been demonstrated.

• Large murexes drill countersunk holes too. Chicoreus florifer, which older Florida literature calls Chicoreus dilectus, and Phyllonotus pomum have both been shown experimentally to drill large beveled holes. Both are Southwest Florida animals.
• Beveled is not universal. In the Calderaro study, 94% of Gulf drill holes were the beveled naticid type. About 6% were not.
• Not every round hole is predatory. Tumbling experiments have produced abrasion artifacts that mimic drill holes.

Inside the family the holes are close to mute. Dietl and Kelley asked in 2006 whether naticid predators can be identified by the holes they drill, and at species level the answer was largely no: only the mean ratio of inner to outer borehole diameter differed significantly between the two species compared. So read the bevel, then look at where on the valve it sits, which is the part the popular version leaves out.

The Sand Collar Is Not Litter

The thing that most often reaches a beach from this family is not a shell. It is a stiff curved gray ring, glossy, unnervingly like a discarded gasket or a strip of industrial rubber. People pick them up and put them in the trash bag.

It is an egg case. Huelsken and colleagues describe it plainly: the egg masses are collar-shaped and consist of sand grains cemented together by a gelatinous matrix with the eggs embedded in it. Kang, Tan and Liu measured collars of nine species in 2018 and listed the contents more bluntly still: egg capsules, sediment that is chiefly sand and mud, adhesive, and occasional fecal pellets. It is not a decorative object. It is whatever was in reach, glued.

The construction explains the shape. The female settles into the sand, secretes a sheet of mucus, works sand into it and pushes it up around herself. The sheet curls around her own shell and sets as a ring. Eggs are attached, then covered by a second layer of sand and mucus. The shell is the mold, which is why every collar is a circle with a central hole sized to the animal that made it. For scale, from the 2018 study: capsules run roughly 90 to 550 microns across depending on species, collars from a single turn to four and a half turns. The classification scheme still in use goes back to Giglioli's 1955 survey.

Leave them where they are. A fresh collar is a nursery holding thousands of capsules, and once stranded and dried it is fragile enough to come apart in your fingers. Biologists have publicly raised the problem of well-meaning beach cleanups carrying collars off with the actual trash.

One more thing, and it is the best fact here. Huelsken and colleagues found the collars easier to find than the snails: the adults burrow and are invisible, the collars sit out in the open. Their inference was that even species regarded as uncommon may be more abundant than is realized. The animal keeps no public record. Its nursery does.

A Hole Is a Behavior That Fossilizes

Almost nothing in the fossil record preserves behavior. Bones and shells record what an animal was, not what it did. A naticid borehole is one of the rare exceptions, and it is exceptional in a way that comes close to laboratory-grade.

• One hole is one predation event. Not an inference from two fossils lying near each other. A single attack, recorded.
• The outcome is recorded too. A complete hole is a kill. An incomplete hole is an attack that did not finish, meaning the prey lived long enough for its own shell to keep the evidence.
• Predator body size is estimable from the diameter of the hole, and position is measurable, which makes the predator's decision-making measurable.

From shells alone, then, paleontologists compute drilling frequency, prey effectiveness (how often holes are incomplete or shells bored more than once), stereotypy (how consistently the predator picks the same spot), and the ratio of outer to inner borehole diameter. Four quantitative measures of a predator-prey relationship, running through roughly 100 million years of rock. There are very few datasets like that anywhere in paleontology.

The holes even carry formal scientific names of their own, independent of the animal, because they fossilize without it. Bromley erected the countersunk and cylindrical forms as separate ichnotaxa in 1981, and the naming has been contested since, with an older name argued to have priority and a published defense of the original. That a hole needs its own taxonomy, then acquires its own taxonomic dispute, says something about how seriously this material gets taken.

The Result Nobody Predicted

The framework under test is Geerat Vermeij's escalation hypothesis, from his 1987 book Evolution and Escalation: predators and prey drive each other's evolution over geological time, and the evidence should show both sides getting better at their jobs.

Patricia Kelley and Thor Hansen went at that with the boreholes. Their 1993 paper in PALAIOS surveyed more than 40,000 specimens from the Upper Cretaceous through the lower Oligocene of the United States Gulf and Atlantic Coastal Plain. It is the centerpiece of this literature, and the result is not the story anyone expected.

Drilling frequency did not climb steadily. It was low in the Cretaceous, rose sharply just above the Cretaceous-Tertiary boundary, stayed high to the late Eocene, declined near the Eocene-Oligocene boundary, then recovered to a moderate level in the Oligocene. Up, down, up.

And on the escalation prediction itself, their finding was that contrary to prediction, no temporal trend of increasing stereotypy of drillhole site occurred. The predators did not get measurably better at choosing where to drill. What did improve was the defense: prey effectiveness, measured as the incidence of incomplete drillholes and multiply bored shells, rose significantly between the Cretaceous and the Oligocene, across entire faunas as well as within the prey groups drilled hardest. In this arms race, as recorded in these shells, the clams carry the clearer signal.

Kelley and Hansen later expanded to more than 150,000 specimens and characterized escalation as episodic rather than steady, possibly tied to climate, while flagging themselves that samples from different intervals came from different latitudes. Their modern calibration in 2007 refused to behave either: beach samples from Maine to the Florida Keys put peak successful drilling at mid-latitudes, falling off both north and south, rather than intensifying toward the tropics the way intuition runs.

Then the one clean signal got complicated. Prey effectiveness depends on reading an incomplete hole as a prey animal that survived. In 2013 Hutchings and Herbert published work titled "No honor among snails," showing that competition between naticids also produces incomplete holes: one snail interrupts another's meal, and the abandoned hole records a robbery rather than an escape. Since prey effectiveness was the single metric showing a clean escalation signal in 1993, that reaches the headline rather than a footnote. This is what a live scientific question looks like from the inside, and the hole in the clam in your hand is the same instrument.

Two Readings of the Same Holes

When drilling started is a second open question, with the same shape: two groups reading the same evidence differently.

Sohl placed the origin of predatory boring in the Polinicinae in the Upper Cretaceous, around 100 million years ago. Then in 1984 Fursich and Jablonski published naticid-type drillholes in Late Triassic bivalves from northern Italy in Science, roughly 120 million years earlier than anything else known. Their title states the conclusion: carnivorous gastropods gained a major adaptation but failed to radiate. The behavior appeared, was lost, and evolved again.

In 2003 Kase and Ishikawa took the opposite route in Geology. They observed the sole living representative of the group those early snails are assigned to, Cernina fluctuata, and found it is an algal grazer that drills nothing. If the pre-Cretaceous animals were that kind of snail, the early holes are not theirs and drilling has a single Cretaceous origin. A comment and reply followed in the same journal. The question is open.

Which is why we will not print a single confident age for this family either. Late Triassic and Early Jurassic are both cited as live alternatives, and one post-2003 revision argues naticids were absent altogether before the Cretaceous.

Six Moon Snails, Checked Against the Register

This roster came out of two independent sources that agree: an occurrence query against OBIS for the whole family across a Southwest Florida polygon, and the Bailey-Matthews National Shell Museum's own family list for the region. Every name carries an AphiaID, the number the World Register of Marine Species files a record under, so you can check any of them yourself.

Two notes on those names. "Gaudy natica" is a field guide name rather than a register entry; the register lists colorful Atlantic moonsnail. And several of these animals have no English vernacular in the register at all, so any common name attached to them came from a book rather than an authority.

Now the part that says something about the family. Across its four subfamilies the register carries 67 genus-level names for moon snails, and only 35 are accepted. Roughly half the genus names ever coined here have been sunk as synonyms, and the reason is on the shell: globose, smooth, nearly featureless burrowers give a taxonomist almost nothing to work with.

The clearest case is recent. Until 2006 the shark eye was one species with two shell forms. Hulsken, Clemmensen and Hollmann ran 16S and 18S rRNA, a calmodulin intron and COI, and showed the two forms are two species. In regional occurrence data roughly a third of the shark eyes belong to the one that had no working name until then. A shell this large, this common and this collected went two centuries misread, because there was not enough on it to argue with. The higher structure is unsettled too: name Naticinae, Polinicinae and Sininae and you are safe, but ask how many subfamilies there are in total and three major databases will currently give three different answers.

What We Ask Aboard the Boat

Nearly everything on this page concerns an animal you are unlikely to see. Moon snails hunt below the surface of the sand, and published intertidal densities rarely exceed one or two individuals per square meter. What you meet is the evidence: a drilled clam, and a ring of glued sand.

The two want opposite handling. A drilled valve is a record. Note where and when you picked it up, and note which part of the valve the hole sits on, because you now know that position is the part carrying information. A shell without a label is a decoration. A shell with one is data. A sand collar is the other case: leave it alone.

Our house rule aboard the boat is empty shells only, and it is our own rule rather than a statute. The law underneath it is on what the law says about taking shells. Several of the islands we run to sit inside Rookery Bay National Estuarine Research Reserve, whose collecting rule appears stricter than ours. We have asked the Reserve in writing and have no answer yet, so we will not tell you where you may fill a bag.

Trips run three times a day out of Goodland. You can reserve a seat, or work through the rest of the shell families from here.