A fish with genitals on its head

August 28, 2012 • 4:26 am

The temptation to call this a dickheaded fish is overwhelming, but the females also have genitals in the same place.  And it’s an evolutionary mystery.

I saw this on National Geographic Daily News (h/t: M. Cobb), which gave a picture of the new species, Phallostethus cuulong, and noted that it’s the 22nd member of its family (Phallostethidae [“penis chest”]) to be described. All of these species have genitals on the ventral (lower) side of their heads.  They’re also called “priapiumfish,” another name based on their genitals (biologists are just as salacious as everyone else). 

Here are the male (top) and female (bottom) from the Nat Geo article; you can see the male’s genitals on the head:

(From the paper):Freshly-collected specimens of Phallostethus cuulong. A) CTU-P 2494, male, 22.5 mm SL, Tra Vinh Province,Vietnam; B) CTU-P 2327, female, 23.7 mm SL, Sóc Trăng Province, Vietnam. Photographed and retouched by L.X. Tran and K. Shibukawa, respectively


The species, captured in the Mekong in Vietnam is described in a new paper by Kochi Shibukawa et al. (reference and link below) in Zootaxa.

The fish have internal fertilization, so the male has an intromittent organ on the head which he inserts into the female’s urogenital opening. Here are better pictures of the male and female genitalia, also from the paper:

Head and anterior part of the body of male

The males have a clasping organ, the ctenactinium, which is used to hold onto the female when mating; here’s a cleared and stained speciment of a male clearly showing the genital apparatus:

Lateral view of head and anterior part of body of cleared and stained specimen of Phallostethus cuulong. CTU-P 5020, male, 23.5 mm SL. Photographed by L.X. Tran.

Sadly, the paper doesn’t describe the behavior (it’s a straight morphological and taxonomic description), but the Nat Geo article helps answer at least one of the two obvious questions.

1. How do they mate?  This gives new meaning to the phrase “dancing cheek to cheek,” since they obviously have to put their genitals next to each other; one of the fish must therefore be upside down. Nat Geo describes the act, obviously hard to see since it requires keeping them alive in captivity and being there at the right time:

As with all Phallostethus—“penis chest” in Greek—species, the male uses its bony “priapium” to clasp a female while he inserts sperm into her urogenital opening, also located on the head, said Lynne Parenti, curator of fishes at the National Museum of Natural History in Washington, D.C.

Parenti remembers seeing another species of priapiumfish mate at a lab in Singapore. Attached at the head and together forming a v, the fish “looked like a little pair of scissors, darting around the tank together,” she said.

For many fish, such as guppies, mating is almost instantaneous, but priapiumfish “actually couple, staying together for a remarkable period of time,” she noted.

2. How did it evolve? To me, this is the really interesting question. But the Nat Geo article doesn’t offer much help:

Partly due to these overlooked and possibly unappealing study sites, the fish “tend to be ignored by a lot of biologists,” she said.

That may explain why the fish’s front-loaded genitalia remains an evolutionary mystery, she added.

There are some clues, though: For one thing, the Phallostethidae family is part of a larger group that includes many species that fertilize their eggs internally. (The vast majority of fish species fertilize their eggs outside the body.)

Many of the males in the family have physical modifications that allow them to internally fertilize females, so it makes sense that priapiumfish would also have evolved an adaptation.

For another thing, head-to-head mating is apparently “a very efficient way to do it,” Parenti added. While examining preserved female priapiumfish, she has found oviducts filled with sperm, meaning almost all the eggs had been fertilized.

Well, presumably the fish’s relatives, whose genitals are in the normal place further back on the body, can also mate efficiently, so that’s not much of a hypothesis. And other fish also have similar types of genital modifications. Perhaps something about the fish’s environment—shallow rivers, perhaps in areas with low visibility—explains it, but who knows?

And I’m curious about how genitals on the head evolved from genitals on the rear: what were the intermediate stages of evolution, and how were they adaptive? During that evolutionary process, the male and female genitals would have to be placed in a way that made copulation easy, so perhaps the change in position evolved concurrently in both sexes.

I have no answers here, but maybe some readers with an ichthyological bent can offer solutions.


Shibukawa, K., D. D. Tran, and L X. Tran. 2012. Phallostethus cuulong, a new species of priapiumfish (Actinopterygii: Atheriniformed: Phallotstehidae) from the Vietnamese Mekong. Zootaxa 3363:45-51.

50 thoughts on “A fish with genitals on its head

  1. Wanting to know “what’s a toxactinium?”, I ask Google which only tells me the obvious: “A long curved projection from the anterior end of the priapium in some Phallostethidae.”

    I suppose a better question is “What’s a toxactinium for?” Is it just for display?

    What a weird fish.

  2. “And I’m curious about how genitals on the head evolved from genitals on the rear: what were the intermediate stages of evolution, and how were they adaptive?”

    Two words, “Intelligent Design.” What more proof do you need? Obviously god was feeling humorous one day and decided to inject a little levity into his creation for us to discover and enjoy. It’s the only explanation that makes any sense.

  3. Is it silly to ask why genitals are normally at the rear rather than on the head? It seems like it should be efficient to have the most important organs (i.e. those involved with feeding and reproduction) near the eyes and brain. Why is it normally only the feeding ones?

    1. Think of a tube with one opening where food enters, another where undigested stuff is excreted. A simple arrangement for external fertilization is to co-opt the evacuation apparatus for the release of gametes, and that is probably the ancestral condition. It later changed, teleost fish don’t have a cloaca and generally keep their openings separate, but the distal arrangement persisted, except for these guys.

    1. Yep, I noticed that too. The whole arrangement is moving as an ensemble. And for that reason I suspect that there may not have been any intermediate forms – this is all under control of some alteration in or control of a Hox gene, or something of that sort, and one small change had a major effect. (Cue for a developmental biologist to weigh in on this.)

      Maybe having reproductive organs near the head provides some advantage, but having the alimentary canal terminate there wouldn’t seem to.

      1. “this is all under control of some alteration in or control of a Hox gene, or something of that sort, and one small change had a major effect.”

        This occurred to me, too. But I’m not a developmental biologist either, so I can’t lend any authority to the idea.

      2. If not Hox then some similar developmental gene that underwent a duplication or fusion event (e.g., inbreeding) that occurred in a small population that became isolated from the main in some small tributary. Just a thought.

    2. The pelvic fins are variably located along the rostral-caudal axis in fish. Here, it’s like the anal-urogenital openings (and some organs associated) have followed the move forward. You can also notice that those fish have numerous tails segments and long anal fins on the ventral tail. It suggests that the whole normally-caudal area has migrated rostrally and attached itsaelf to the head. Embryological tinkering.

      If the genitals were on the dorsal part of the head, then we should have to conclude to the intervention of an “intelligent” designer 🙂

      Desnes Diev

  4. What an amazing species. However, it is not such a stretch to conceive (pardon the pun) of intermediate forms if this fish evolved from a smaller fish whose tail lengthened caudally while at the same time the urogenital structures moved cranially. With the urogenital structures out of the way, so to speak, the tail structures could have rapidly elongated giving the appearance of a long migration of the sexual apparatus, when in fact, they really haven’t migrated that far compared to ancestral forms.

    1. Yeah, this is my first guess as well. Take any normal fish, elongate the portion of the body behind the anus/genitals 4-5 times the original body length, and enlarge the head proportionally, and you get dickhead fish.

  5. LOL! What a dickhead that fish is.

    “this is all under control of some alteration in or control of a Hox gene, or something of that sort, and one small change had a major effect.”

    This seems to be the best explanation for such an arrangement.

    1. I see this macromutational scenario as problematic, simply because the first mutant individual wouldn’t be able to copulate (or at least copulate competently). And I’m not sure that such a large mutation wouldn’t have deleterious pleiotropic effects. I’d bet $100 that this movement involves a lot more than one gene. AND nobody has suggested why it’s adaptive yet!

      1. I agree with JAC about the macromutational scenario. The difference between this fish and typical fish is the relative length of the tail vs. the torso (albeit a very extreme difference). Those sort of size/shape differences are typically multifactorial, and respond to artificial selection quite well.

        1. Sorry, I’m not following your point about tail-v.-torso. While it may be true that these fish have long tails compared to their torsos, I don’t see how that by itself can explain how the genitals came to be attached to the head, forward of the torso, as they clearly are in the stained specimen. Surely there’s some other factor besides body proportions in play here.

          1. The anal-genital openings are at the back end of the torso. So what these fish have is a telescoping of the torso into the anterior end of the body. Posterior to these openings is the tail. Although the overall proportions of the fish look ‘normal’, the position of the anus shows that it has ‘normal’ head, very short torso, and long tail. What’s not clear to me from the figures here is whether the body cavity extends into what appears to be the tail (so that the body cavity and associated organs such as the gut) would be sort of U-shaped, bringing the anus and urogenital opening near the head via a curved route. These are really weird fish. I’d want to see a clearer anatomical preparation (and there may be such in the literature), but my first thought is that they are almost as strange as turtles. (Turtles’ shoulder and pelvic girdles are inside their ribcages. Think about that– what sort of morphological transformation would bring your shoulder blades, collar bone, and hips inside your ribs?)

            1. Thanks for the explanation. I take it then that head, torso, tail, etc. are defined not by skeletal markers (e.g. by counting vertebrae) but by soft tissues and the layout of internal organs. So by this definition a fish can have its entire torso anterior to its first vertebra, and its entire spine in (what’s defined to be) its tail. Do I have that right?

              1. Fish don’t have necks, so the torso/head boundary is not as distinct as in tetrapods. In fact, the shoulder girdle of fish (part of the torso or body in tetrapods) is often attached to the skull. Generally, we think of the torso ending where the anus or cloaca is. Behind this is the tail, which has muscles and skeleton, but no organ-containing body cavity. In advanced teleost fishes (e.g. perch), the pelvic girdle (which in fish is generally not attached to the vertebral column) is moved far forward on the body, sometimes lying below or even anterior to the pectoral girdle. The anus and urogenital openings are closely associated with the pelvic girdle, and when it moves forward, so do they. In some advanced teleosts (e.g. whiting), the gut is sort of U-shaped passing back from the mouth in the body cavity, but then looping forward to the ventral surface to the anteriorly placed anus (where it is because it has followed the pelvic girdle). In many fish (see the ratfish post linked to above), the pelvic fins are modified into intromittent organs (i.e. male genitalia). The priapium of the fish under consideration here is, in fact, composed mostly of the highly modified pelvic girdle. These fish are pretty weird, but not so weird when you think of the far forward position of the pelvic fin in many advanced fish (which I well knew) and the U-shaped gut in some of them (which I hadn’t quite realized till I began looking into it, perhaps because I’ve dissected so many sharks, which have straight guts). I’m still not sure about these fish, but I stongly suspect that, like whiting, the gut (and body cavity) extend posterior to the anus, and is U-shaped. The standard work on the morphology of the priapium is by Lynn Parenti of the Smithsonian: Parenti, L.R. 1989. A phylogenetic revision of the phallostethid fishes (Atherinomorpha, Phallostethidae). Proceedings of the California Academy of Sciences 46: 243-277.

      2. This had to emerge as a response to the difficulties in fertilization in a fast-moving, riverine environmnet. The move towards internal fertilization would be more successful than external in a fast moving stream. In such an environment having the genitialia located further away from the back and forth movement of the tail might allow copulation to occur more easily and not allow the copulating fish to get swept away downstream while doing the deed, so to speak.

        1. Attached at the head and together forming a v, the fish “looked like a little pair of scissors, darting around the tank together,” she said.

          For many fish, such as guppies, mating is almost instantaneous, but priapiumfish “actually couple, staying together for a remarkable period of time,” she noted.

          If their genitalia were further back on the tail, they wouldn’t be able to dart around efficiently (without uncoupling) and would be vulnerable to predation during the prolonged mating.

      3. It could have happened to a cluster of eggs right after fertilization or even to the sperm cells that latter fertilized those eggs.

      4. Like the paper said, many males in the family have strange modification that allow them to copulate internally. How do we know that hybridization is impossible?

  6. Aha! Try to explain THIS one, evolutionists! Proof positive that all of Evolutionary theory is wrongedy-wrong, wrong! The answer is that God was in a particularly playful mood when He instantly created these creatures, 6,000 years ago! Or- maybe He just wasn’t paying attention after having made so many other animals….

  7. I believe the most obvious answer to the question why would this fish have their genitals in the head would have to be related to develpment. It may be a wild guess, but at least I think it is an informed wild guess, that when they study this, it will turn out to be the product of a positioning gene for the genitalia turning on, on commando of some head inducing factor instead of their usual cue, or something like that.

    I love this piece, as most of what you post. Thanks a lot for the information and your clear presentation.

  8. Jerry, Hi. I loved your post. I believe the most probable reason for this would turn out to be a developmental one. My guess would have to be that the development of genitals would have changed (by a mutation, of course) their signal cue to one that is being fired in the head (or the other way around, the signal that used to trigger genital formation in the rear, ceases to appear there to do so in the front, especifically in the head).

    Again thanks, it was a nice post.

  9. Not knowing exactly how many fish this species usually gives birth to in a lifetime, but guessing it probably gives birth to a lot of them, a lot of the progenie of a fish bearing this kind of mutation I just propose may be an answer. This kind of mutation would produce the same effect in males and in females. Endogamy may be the reason, albeit not a perfect one, their genetic diversity will be extremely limited, nonetheless I see it posible. Surely much more reasearch has to be done.

    The gradual, step by step alternative sounds, to me at least, a little problematic too. Since I know too little about this species (and I may not be the only one here), I find it difficult to imagine an scenario in with gradually moving the reproductive organs was step by step an improvement. Of course my ignorance does not disprove this to have been the case, I just find it a little easier to think it could have happened at once, and luckily, the new arrangement worked (considering that most don’t).

  10. One more thing. I do not believe this arrangement necesarrily had to be advantageous, if competition in its niche was permissibly enough, it may have endured and given time to adjust other things in order to improve, now gradualy, their innovative design.

  11. The first question that occurs to me is: “Is this feature adaptive, maladaptive, or neutral?”

    The first question that occurs to Jerry Coyne is: “What’s the best adaptive explanation?”

    1. Sorry, Larry, but I think there’s more likely to be an adaptive explanation than a neutral or maladaptive one. That doesn’t mean I rule the other explanations out of court.

      And anyway, that was not the first question that occurred to me. As I said in my post, that one was, “How do they bonk?”

          1. No. Maladaptive does not mean instantly lethal. It just means that overall fitness has been lowered. Think of people who are near-sighted or the massive antlers of the extinct Irish elk.

            When you see something that looks difficult to explain by an adaptive scenario then that’s the time to start thinking that it may not be adaptive.

            1. I didn’t mean instantly lethal. I mean exactly overall fitness has been lowered. In this particular species, has it? Compared to what?

              It’s easier to see (no pun intended) that near-sightedness is maladaptive because only a minority suffers the condition. If everybody is near-sighted (which would change our defintion of normal sight), then we’d start questioning whether it is maladaptive.

              I wouldn’t call the huge antlers of the Irish Elk maladaptive, but rather “seemingly maladaptive” in the same way peacock tails are.

  12. That’s a nasty looking clasper (the ctenactinium) – but it’s not obvious to me how the fish would use it.

  13. A whole heckuva lot of changes have occurred in this critter, and asking why the genitalia have moved rostrally is adaptive immediately shoves all those other changes into the shadows. For all we know, the adaptiveness (if any) of the altered morphology may have nothing at all to do with the location of the genitalia.

    My first question(s) would be about the development of the altered morphology; first asking a question about adaptation is putting the evolutionary cart before the developmental horse.

  14. RE #15, maladaptive and neutral traits: Morphologically complex, functional traits with a multi-factorial genetic basis can originate as neutral or selected-against traits only under extremely limited circumstances, the unlikelihood of which fully justifies the search for an adaptive explanation of their origin. Consider a trait with, say, 5 alleles at different loci required for its expression. If we posit these alleles have no effect on fitness, then the fixation of any one allele is improbable, and the fixation of all 5 is some low probability to the 5th power. Now consider that the alleles have a negative effect on fitness. Now they can be fixed only against selection, via genetic drift, which could only occur in a small population (exactly how small depending on the strength of selection). Again, we have (very) low probabilities to an exponent. Neutrality of complex morphological changes (sizes and connections of bones, position of reproductive and digestive tracts), especially when, as in this case, they so directly affect reproductive success, seems very unlikely, so the neutral scenario seems very unlikely, and the alternative maladaptive scenario becomes even more so.

    Probably no complex multifactorial trait originates as maladaptive. Traits once adaptive may, however, be maladaptive in a changed environment. Curious and inquisitive wolves might once have been better at finding prey; once men with guns became widespread, shyness might prove more reproductively profitable, so that curiosity is now maladaptive.

    I didn’t know anyone thought the massive antlers of Irish elk were maladaptive since the demise of orthogenesis, and especially since Steve Gould (who was rather fond of non-adaptive explanations) studied the elk and concluded “The immense antlers of Megaloceros were advantageous in themselves.” (1974:217).

    (Neutral [and even maladaptive] traits can be fixed when there is simple genetic basis. A synonymous DNA sequence which does not alter the amino acid sequence of the resulting protein will drift relative to its synonymous alleles, and eventually one will be fixed.)

    It is sometimes supposed that a neutral or maladaptive complex morphological change can occur because of mutations of large effect at a single locus, which greatly increases its probability of fixation.(It is also sometimes supposed that adaptive changes can occur that way too.) This is the question of macromutation, and the extent to which they are incorporated in evolution. It’s a bit of a separate question; see Jerry’s thoughts on Carl Zimmer’s blog, and my further comments (#30) here.

      1. @ Gregory C. Mayer

        Isn’t paragraph one just a recycled version of the Behe argument against protein evolution? Saying something is rare is not disproving it. You are allowing that same “it’s so rare” argument could be applied for both an adaptive and maladaptive explanation, as well.

        And for the maladaptive argument, one could argue a single major change did occur, and then all other changes were compensatory. Compensatory evolution is well documented (www.ncbi.nlm.nih.gov/pubmed/20182427).

        The real point is that either position is currently unsupported by evidence. All of this is speculation, which tends to reflect the bias of the speaker as to their personal interests.

Leave a Reply