A gynandromorph moth comes to the light – and tells a story about science

September 2, 2015 • 11:00 am

by Matthew Cobb

This tw**t popped up in my feed the other night, from “wildlife illustrator and invertebrate enthusiast” Richard Lewington [Richard has a website showing his art here]. Richard was running a moth trap in the night when he found this beauty:

If you look carefully, you can just see the male’s feathery antenna on the left; the female side presumably had a straighter antenna (these different shapes relate to the different functions – males have to detect female pheromones from far away; females primarily need to be able to detect food plants on which to lay their eggs). You can see this clearly in another example Richard tw**ted:

Gynandromorphs are mixtures of male and female, often occurring because of a developmental problem – we highlighted the potentially gynandromorph cardinal bird here three years ago. There is a link between birds and moths, in that both groups have an unusual form of sex determination. In mammals, females have identical sex chromosomes (XX) while males have one X and one Y chromosome – they can produce two kinds of gametes (X and Y sperm) and so are called the heterogametic sex. For reasons that are unclear, in birds and lepidoptera (moths and butterflies),  females are the heterogametic sex (to avoid confusion, their sex chromosomes are called Z and W; males in both groups are ZZ).

It seems probable that these moths are gynandromorphs because, at a very early stage of development – probably when a fertilised female ZW egg divided into two cells – one of the daughter cells ‘lost’ the W chromosome because of some glitch. The tissues that were produced by that cell were therefore ‘ZO’ – you need the W chromosome to be female, so the tissues became male. The sharp dividing line down the middle of the moths, and the ‘mirroring’ of sexually dimorphic external structures on either side reinforces this intepretration.

There are many examples of gynandromorph lepidoptera on Google, which is probably a combination of people’s interest in these insects and the striking sexual dimorphism that exists in many species, making it easier to spot:

Image taken from here.

Here’s a photo of a gynandromorph gypsy moth, clearly showing the different shaped antennae (the male side is on the right):

Image taken from Jerry’s colleague Greg Dwyer.

As Jerry pointed out in his original cardinal post, those of us who work on the fly Drosophila (which, like us, has XX females and XY males) would occasionally see gynandromorphs in our stocks, although unless you are doing some funky genetics with sex-linked eye- or body-colour, male and female flies are not as different as the examples of the moths seen above. However, I do recall finding an apparently female fly with a male foreleg (male forelegs have ‘sex combs’ that are involved in sexual behaviour). Jerry’s explanation bears repeating:

In flies the sex is determined by the ratio of X chromosome to autosomes.  Flies, like all diploid species, have two copies of every autosome. If you also have two X chromosomes, you’re a female because the ratio of autosomes to Xs is 1:1. If you have one X chromosome and one Y chromosome, your ratio is 2:1 and you’re male.  The Y doesn’t matter here: if you lose a Y chromosome, and hence are XO, you still look like a male, although you’re sterile (the Y carries genes for making sperm).

So to get gynandromorphs in flies, all that has to happen is that one X chromosome gets lost in one cell when the initial cell in a female (XX) zygotes divides in two.  One half of the fly then becomes XX, the other XO, and the fly is split neatly down the middle, looking like the one below.  But gynandromorphs don’t have to be “half and halfs”.  X chromosomes can get lost at almost any stage at development, so flies can be a quarter male, have irregular patches of maleness, have just a few male cells, or even a male patch as small as a single bristle.

Way back in the day (i.e., 1970s), making mosaic flies in which different patches of tissue are either male or female was the only tool we had for identifying which tissues were involved in controlling various behaviours. This was fastidious work pioneered by one of the greats of post-war science, the physicist-turned-molecular-geneticist-turned-behaviour-geneticist, Seymour Benzer. [JAC: see my mini-post at bottom in which I used these methods for another purpose.]

Along with Yoshiki Hotta, Benzer was able not only to show tissue-level genetic control of behaviour, but also to show where in the embryo those tissues were determined, thereby constructing what he called a fate map of the action of a particular mutation. They adapted this technique from one of the founders of genetics, Arthur Sturtevant, who originally proposed it in 1929.

Here are some figures from Hotta and Benzer’s 1972 paper in Nature: ‘Mapping of behavior in Drosophila mosaics’. The first shows the range of mosaics that they produced – they were much more varied than the naturally occurring gynandromorphs because of the way they manipulated a special kind of X-chromosome in these flies, called a ring-X chromosome (known as X-R). This X-R chromosome could be lost at varying times in development, changing tissues from female (XX-R) to male (XO). The later the chromosome was lost, the more specific the tissues that would be male. By using a body-colour mutation on the X-chromosome, Hotta and Benzer could track from the outside of the fly which tissues were male and female, because they had different colours.

The top left fly in the figure apparently lost its X-R chromosome at the earliest stage of development, hence the straight line. As you can see, the effect doesn’t need to be symmetrical – if the chromosome is lost at a later stage, then a very specific part of the fly could be affected, such as the right wing in the top right fly (the left wing is still female).

The second figure shows how they interpreted which parts of the fly embryo were involved in determining the behaviour of a mutation called hyperkinetic in which the fly shakes its legs when anaesthetised (this rather odd behaviour turned out to be of major importance, as it is produced by changes to the activity of ion channels in the fly’s neurons). Unsurprisingly, it appears that the hyperkinetic gene was exerting its influence in three separate regions (one for each of the fly’s pairs of legs), all of which are involved in producing the part of the fly’s nervous system that controls movement.

240527a0The arduous nature of the technique – it was not possible to predict which tissues would lose their X-R chromosome, and often no detectable change occurred – and the problems of identifying which tissues underneath the cuticle had changed sex, meant that it was not not widely adopted. By the late 1980s this method was  overtaken by direct manipulation of genes and the tissues they are expressed in, but for many years it was cutting edge science, available in only a few leading laboratories.


Jerry’s addendum: I used gynandromorphs, and Benzer and Hotta’s ring-X stock, to determine where in the fly the females’s sex pheromone (a waxy substance on her cuticule that incites the males to court her and mate with her) resided.  As Matthew noted, that stock of flies, which still exists, is prone to losing X chromosomes when they’re contributed by a male parent. The male’s XX (female) zygotes often lose the X at different stages of development, producing patches of tissue that are XO and therefore male. You can tell which patches are male because the female’s X carries a recessive gene causing yellow body color, so male bits (XO) are yellow and female bits (XX, with one gene for normal coloration) are normally pigmented.

XX females have very different sex pheromones from XY and XO males, so by correlating which bits of a gynandromorph fly were male vs. female, and then extracting each fly’s sex phreromones with hexane and testing the chemicals’ identities on a gas chromatograph, Ryan Oyama (an undergraduate student) and I were able to determine where in the fly’s body the sex pheromones were produced and/or sequestered. It turned out that this was in the cuticle of the abdomen only: flies with female heads, legs, or thoraxes but male abdomens produced only male pheromones. The amount of female pheromone was proportional to the amount of female tissue in the abdomen, at least as seen in the visible cuticle.

This correlated with behavioral observations, too, for when gynandromorphs were tested with normal males (always horny), those males courted gynandromorphs most vigorously when their abdomens were female.  (This could, of course, have been associated with behavior or morphology of those gynandromorphs rather than pheromones, so we needed to do the pheromone tests as well.) Later workers actually localized the pheromone-producing cells to a layer right below the abdominal cuticle, confirming our results.

We published our results in the Proceedings of the National Academy of Sciences (reference and free download below), and I thought it was a very clever way to use old genetic technology to study behavior and biochemistry. Sadly, the paper didn’t get much notice!


Coyne, J. A. and R. Oyama. 1995. Localization of pheromonal sexual dimorphism in Drosophila melanogaster and its effect on sexual isolation. Proc Nat. Acad. Sci. USA 92:9505-9509.

52 thoughts on “A gynandromorph moth comes to the light – and tells a story about science

  1. It is important to mention that unlike in Drosophila, gynandromorphism is not based on sex-chromosome loss.

    There is multiple evidence of this, some of it buried in old publications (refs below), some of it simply summarized by Nipam Patel here: http://nautil.us/issue/13/symmetry/half-male-half-female-total-animal

    In summary, lepidopteran gynandromorphs are mosaics resulting from double-fertilization of a binucleated oocyte: a pronucleus and a polar body. The trick is that, due to the way meiosis works in Lepidoptera, the polar body of a ZW female necessarily inherits the alternative sex chromosome: if the pronucleus is Z, the polar body is W. If there is double fertilization, the result is a mosaic individual that is necessarily half-female, half-male, but that is also mosaic for its entire genome. Imagine two dizygotic twins, one boy and one girl, fused together in the midline: this would produce a genetically equivalent outcome – with a mosaic of not two but four haploid genomes.


    Blanchard, R., & Descimon, H. (1988). Hybridization between two species of swallowtails, meiosis mechanism, and the genesis of gynandromorphs. J. Lepid. Soc, 42, 94-102

    Ebinuma, H., Kobayashi, M., Kobayashi, J., Shimada, T., & Yoshitake, N. (1988). The detection of mosaics and polyploids in a hereditary mosaic strain of the silk moth, Bombyx mori, using egg colour mutants. Genetical research, 51(03), 223-229.

    1. Do lepidopteran embryos develop similarly to those of Drosophila, i.e. with nuclear divisions in a common cytoplasm, which subsequently becomes a cellular blastoderm? I realize that the Drosophila embryo cytoplasm is by no means homogeneous, but I’m trying to envision how the male and female nuclei of a “fused” gynandromorph embryo would respect a midline boundary at the syncytial blastoderm stage.

      1. Yes to your first question. Early insect embryos are a syncytium of nuclei at first. Toward your second question, the lineage boundaries tend to be sharply defined, as in left from right, because the nuclei do not wander from where they are even in the syncytial stage. I do not know why, but perhaps they are sort of held in place by the cytoskeleton.

        1. Interesting! My background is in vertebrate developmental biology, and I have only a smattering of invertebrate (mostly echinoderm) embryology.

          Once the left/right lineage boundaries are established early in development, it’s easy for me to accept that the structures derived from imaginal discs (wings, legs, antennae) would continue to respect the midline boundary, e.g. in the butterfly gynandromorph photos above. However, it’s more difficult for me to accept that the same midline boundary would be respected by neurons in the ventral nervous system, given that their axons may cross that boundary (sometimes more than once). I haven’t read the Hotta and Benzer paper, but Matthew’s summary of the results with the hyperkinetic mutation indicates that lineages in the nervous system control pairs of legs, which is more aligned with my recollection of anterior-posterior lineage compartments in Drosophila. Although we don’t have the appropriate markers and mutations, I’d want to know about the distribution of the male and female cells in the nervous system of a lepidopteran gynandromorph.

      2. If you look at collections of butterfly gynandromorphs, you will notice that most individuals are not bilateral. Rather than with a clean divide at the mid-line, in most cases, aberrant cell clones are visible as small patches of male/female color patterns on the wings. It thus seems that there is no strong determinism dictating that the mosaicism resulting from the double-fertilization should necessarily result in a midline boundary. In my understanding, bilateral gynandromorphs are more spectacular and probably the product of chancy division patterns in the early syncytium (?).

    1. As far as the lepidopteran bilateral gynandromorphs are concerned (see my comment above), their gametes are normal and should produce normal offspring: one side of the body makes normal sperm, the other side normal eggs. Of course, this comes with anatomical complications and the genital apparatus is also split into an impossible morphology, so they would probably not mate in the wild or be able to oviposit normally etc…

    1. there are several kinds of chimerism including karyotypes of 46 XX/46 XY or chimeras where dizygotic twin zygotes merge and the resulting person develops clones of cells from each of the different twins and ends up a mosaic of tissues from two different zygotes.

      And, of course, women are natural mosaics as each cell activates only one of the two X chromosomes, so if there are genetic differences between the X chromosomes, those will be expressed in the subsequent clones.

  2. How many different chromosomal approaches to sex determination exist in insects? I had rather assumed, apparently incorrectly, that a common mechanism was conserved throughout the whole group. For that matter what about other arthropod subphyla?

    1. Hymenopterans (ants, wasps, and bees) typically exhibit haplodiploidy: females are diploid (two sets of chromosomes); males are haplodiploidy (one set of chromosomes, they develop from unfertilized eggs and do not have a father).

        1. A cue, if one were needed, for a detailed post about the many and varied methods of sex determination in plants that produce spices.

      1. I had no idea that mammals were so unusual in their consistency of sex (or is it gender?) determination.

        As others have said, one of those posts that makes you realise how much you don’t know.

        1. I think it’s more that mammals are all rather closely related and relatively recently derived compared to insects. And mammals are the only lineage that survived past the ancient split between sauropsids and synapsids. Many branches of the sauropsid family tree have living descendents (turtles, snakes, lizards, tuataras, crocodilians and birds) whereas all the synapsid lines died out except for mammals. Humans and rodents share 80% of their DNA. I’d bet that’s not true for different orders of insects or crown sauropsids. Does anyone know?

  3. Like birds, most snakes also have the opposite form of chromosomal sec determination from mammals, with heterogametic females. As a consequence of this, when individuals of normally sexually reproducing snake species reproduce by parthenogenesis, the resulting offspring are male. Usually parthenogenesis results in female offspring. Only one species of snake is known to be routinely parthenogenetic, and that appears to be caused by polyploidy resulting in a triploid karyotype.

  4. Not sure if I’m more amazed that flies can be a percentage male or that you can figure out how and why. You’ve published a few science posts that make me realize how little I know, this is one of them.

  5. At a perhaps more elementary level…why are the sex chromosomes of some species specified as “XY” and others as something else, such as “XO” in the case of these flies? If I remember right, birds are something else entirely.

    Is there something fundamentally different about the chromosomes in question? Wouldn’t it make sense to designate all female-specific chromosome halves as “X” and all male-specific chromosome halves as “Y”?


    1. XO flies are male but sterile (the O means they only have one sex chromosome – an X. There is no O chromosome.). The reasons why an XO fly is male relate to sex determination in flies, which is different from that in say humans. As the post says, in birds and Lepidoptera, females are ZW and males are ZZ. Sex determination is very varied (see link in one of the other comments), so a Z chromosome in a Lepidoptetan is not the same thing as an X or a Y in a beetle (say).

      1. Does the ZZ/ZW modus operate in reptiles, inasmuch as reptiles seem (to my layman’s eye) appear to be fairly closely related to dinosaurs, from which sprang boids?

        1. The objection to the term “reptile” is that it’s polyphyletic – it doesn’t describe a group that consists of one species of organisms and all that species’ descendants. If “reptiles” consisted of birds, the other non-bird dinosaurs, crocs, lizards, snakes and a few other bits and pieces (mammal-like reptiles, and possibly mammals too – that branch point is a bit unclear), then it’d be a valid clade. But it doesn’t, so it’s passed over.
          On the grounds of phylogenetic bracketing, I’d expect that bird sex genetics to be more similar to, say, crocodilian or turtle sex genetics than to mammalian sex determination. But since there must have been major changes on several occasions in the methods of determining sex through the animal kingdom, you might have to search deeply actually see the similarities.

      2. the O means they only have one sex chromosome – an X. There is no O chromosome

        Thanks! An important piece I was missing.

        As the post says, in birds and Lepidoptera, females are ZW and males are ZZ.

        That’s the bit I’m still confused about. Why Z and W for birds and butterflies, but X and Y for monkeys and cats?


        1. XY and ZW describe the sex determination systems of the species, not the sex of an individual of the species.

  6. it’s interesting that crocodilians, the other crown archosaurs, have a temperature dependent sex determination scheme and don’t have sex chromosomes at all. One wonders at the sex determination mechanisms of non-avian dinosaurs. I imagine that therapods had the ZW system found in modern therapods and that triassic crocodiles had the same as their modern descendants. So there was a switch somewhere along the line. There was a paper recently http://www.sciencemag.org/content/346/6215/1254449.abstract that looked at archosaur evolutionary relationships, but did not mention mechanisms of sex determination. Anyone have illumination of this issue?

  7. Thanks for the effort in producing another wonderful biology lesson. This is a very cool phenomenon I was entirely unaware of.

  8. Late to the partay.
    I love gynandromorphs and other such genetic mosaics. One lesson that had been learned from them is that during later development in insects the body becomes divided into lineage ‘compartments’, so genetically marked clones formed later in development are not only smaller (since they get fewer cell divisions) but the clone of cells can be confined to a compartment with sharp edges along the boundary to the adjacent compartment.
    For example, in the late clones above the segment boundaries are seen to be compartment boundaries.

  9. Please do not say we are uninterested in science, Professor Coyne.

    An awed silence should not be mistaken for indifference.

    We may make fewer comments on these posts but we learn so much more.

  10. One waits with almost bated breath for the Discovery Institute’s theological interpretation of gynandromorphy. I can barely contain myself.

  11. Very interesting, thanks. Life on this here planet is always more interesting and complicated than one tends to realize…

  12. Late to the party. But then science posts takes longer time to read for comprehension…

    Sex is not boring. Who would have thought?

Leave a Reply