I am trying to figure out the sex-determination of a very foreign form of reproduction.

A species has three sexes X, Y, and Z. X produces X gametes, Y produces Y gametes, Z produces Z gametes. They do not correspond to male or female, as any may impregnate or be impregnated. Only two parents of different sexes are needed to reproduce, because the only possible combinations of gamete fusion are X and Y, or X and Z, or Y and Z. The offspring may be any sex, including the one that neither of their parents are. Sex is determined genetically, not by environmental factors.

How does the zygote determine whether to be X, Y or Z?

(Based on this: https://paperiapina.deviantart.com/gallery/31840163/Triaformica)

  • $\begingroup$ There are many terrestrial mating systems. Some sexual some not. There are fungus species with thousands of sexes. $\endgroup$
    – sphennings
    Nov 13, 2017 at 21:16
  • 1
    $\begingroup$ @Orphevs I will leave the thousands example to Sphennings as I am not familiar with it, but I believe that some members of this group en.wikipedia.org/wiki/Basidiomycota have 3 isogamous types $\endgroup$
    – Slarty
    Nov 13, 2017 at 22:41
  • 2
    $\begingroup$ @Mazura: Gender is a social construct, sex is a biological feature. Males produce sperm, females produce ovum, hermaphrodites produce both. The species in question is none of these: they produce a single type of gamete that is neither sperm or ovum. In real biology, these are known as mating types. $\endgroup$
    – Anonymous
    Nov 17, 2017 at 16:31
  • 2
    $\begingroup$ Might be worth investigating Tetrahymena, which has seven sexes and each one decides on its own which one it will be. en.wikipedia.org/wiki/Tetrahymena $\endgroup$
    – Dubukay
    Dec 18, 2017 at 11:20
  • 1
    $\begingroup$ @Dubukay: The biological sciences never cease to amaze. $\endgroup$
    – Anonymous
    Dec 18, 2017 at 16:37

13 Answers 13


Your 3 sex system lends itself perfectly to a "scissors-rock-paper" system of determining which sex is expressed using regular old dominance / recessiveness. Rock crushes scissors, scissors cuts paper, paper covers rock. So too as regards gene expression: X is dominant over Y, Y is dominant over Z and Z is dominant over X.

Thus XY = X sex. XZ = Z sex. YZ = Y sex. Because XX is nonviable in your system, parents of XY (X type) and XZ (Z type) could produce XZ, YZ, or XY offspring, the YZ offspring being Y sex and so unlike either parent.

The nonviable XX, YY and ZZ possibilities would mean 1 in 4 conceptions would end in miscarriage as is the case with humans when the 2 parents are heterozygotes for a recessive gene which is lethal as the homozygote.

  • 1
    $\begingroup$ +1 this is the way i was thinking of it as well when i read the question. $\endgroup$
    – Stephan
    Dec 5, 2017 at 19:43
  • $\begingroup$ You misunderstand the explanation. Gametes of the same type cannot fuse to begin with and they're determined by the sex of the parent. Same-sex couples can't reproduce because the gametes cannot fuse. $\endgroup$
    – Anonymous
    Dec 5, 2017 at 20:08
  • $\begingroup$ I see you have special properties of gamete fusion in the OP. Regular genetics can do what you want with no special gamete properties. If you stick to mammal-type genetics with a diploid organism and haploid gametes, an organism can make 2 types of gametes. A human male can make X or Y sperm. If you let your organisms be diploid you can have your 3 types of organism and all types of progeny. Homosexual parents would have a 50% miscarriage rate instead of the 25% for heterosexual because you have 2 chances for homozygous (XX, YY, XY, YX) instead of just 1. $\endgroup$
    – Willk
    Dec 5, 2017 at 20:23
  • $\begingroup$ @Will: Same-sex couples cannot reproduce. I too came to the circular dominance conclusion in the past, but I discarded it because it did not fit the observations. I am sorry, but I am trying to make the genetics fit the observations, not the other way around. $\endgroup$
    – Anonymous
    Dec 5, 2017 at 20:32
  • 1
    $\begingroup$ Did some more research, and I think I am on to something. If the organism is diploid and produces gametes through mitosis, then the compatibility mechanisms are preserved without introducing any excess complexity. The tetraploid zygote can then produce diploid zygotes through meiosis, and homozygous segregation would be prevented by a mechanism similar to whatever male platypuses use to ensure their ten allosomes segregate correctly (that is poorly understood currently). Does that make sense? $\endgroup$
    – Anonymous
    Dec 6, 2017 at 14:12

These organisms should be diploid (2 sets of chromosomes like humans). Whenever 2 gametes approach each other to fuse there simply needs to be some element that prohibits the formation of a zygote if both gametes are of the same type.

This might be achieved if there were 3 different possible protein structures at one key point on each gamete. The properties of the proteins would have to be such that the x protein was attracted to a y or z protein but repelled by another x protein. The y protein was attracted to the x or z proteins but repelled by another y protein and the z protein was attracted to the x or y proteins but repelled by another z protein.

In this way x or y could fuse with z, y or z could fuse with x and x or z could fuse with y. But x could not fuse with x, y could not fuse with y and z could not fuse with z.

Edit some more examples to clarify
Proteins come in an almost infinite variety of configurations and shapes so imagine 6 proteins forming these shapes

1 Sphere
2 Cup
3 Cube
4 Square hole
5 Tetrahedron
6 Triangular indent

The X gamete has proteins 1 and 6
The Y gamete has proteins 2 and 3
The Z gamete has proteins 4 and 5

X-Y Sphere fits into cup, Y-Z cube fits into square hole, Z-X tetrahedron fits triangular indent no other combinations fit.

The system could easily be extended to four gametes by the addition of another six protein pairs. As an example

7 single prong
8 single prong indentation
9 double prong
10 double prong indentation
11 triple prong
12 triple prong indentation

Each protein would provide compatibility with one other gamete type so X is compatible with W,Y and Z etc

The X gamete has proteins 6, 7 and 10
The Y gamete has proteins 2, 8 and 11
The Z gamete has proteins 1, 4 and 9
The W gamete has proteins 3, 5 and 12

  • $\begingroup$ I think you mean diploid if they’re like humans and have 2 of each chromosome $\endgroup$
    – Dubukay
    Nov 14, 2017 at 4:03
  • $\begingroup$ @Dubukay Yes! indeed I did - thanks - now corrected $\endgroup$
    – Slarty
    Nov 14, 2017 at 9:18
  • 1
    $\begingroup$ To allow for W, W can fuse with W: it is the only allowed homozygote. $\endgroup$
    – Willk
    Nov 14, 2017 at 16:09
  • $\begingroup$ That's how mating type systems work. My problem is that tetrapolar mating type systems, which I presume this is a variant of, have two such genes on different chromosomes. If the parents match in either, they are incompatible. So even if there are four mating types (2x2 genes), they are only compatible 25% of the time, not 66%. Thus, couplings will always produce spawn containing all mating types. How does this system get around that limitation, I wonder? $\endgroup$
    – Anonymous
    Nov 15, 2017 at 12:38
  • $\begingroup$ @Anonymous I might have missed something in your argument, but why must there be two genes on different chromosomes? Especially if that produces an effect that you don’t want and you are designing from scratch? The system that I described I think qualifies as tripolar, but could easily be extended to four. I will update my answer to add this and give a better example. $\endgroup$
    – Slarty
    Nov 15, 2017 at 13:45

Here's another option:

There are three genes: The X gene, the Y gene, and the Z gene. And there are three sex chromosomes: The XY chromosome, the XZ chromosome and the YZ chromosome. Each sex chromosome has exactly the two genes found in its name. Just as with humans, each individual has two sex chromosomes, but the gametes have only one.

Now the following rules apply:

  • Each individual must have all three of the genes.

  • The one gene that appears twice determines the sex. For example, an individual that has an XY and an XZ chromosome is of sex X.

  • Gametes of the same sex simply don't fuse, due to receptors on their gametes; that is, while 50% of same-sex fusions would give a working gene combination, this is irrelevant because the fusion simply doesn't take place.

So if e.g. an X (gene combination XY + XZ) and an Y (gene combination XY + YZ) get offspring, there are four possibilities:

  1. XY + XY: The fertilized egg won't develop, as Z is missing. Since this happens before the egg even gets to the placenta, this will be indistinguishable from no fertilization happening in the first place.

  2. XY + YZ: The child will have sex Y.

  3. XZ + XY: The child will have sex X.

  4. XZ + YZ: The child will have sex Z.

So independent of the sexes of the parents (as long as they are different), the children will have all three sexes with equal probability.

Note that while this has some similarity with Kyyshak's answer, it's not the same as it only needs standard chromosome pairs and does away with the dominant/recessive complexity by simply having the same-sex fertilization prevented by a different mechanism.


Some organisms use an XX, X0 mating system in which sex is determined by the presence or absence of a second sex chromosome. This system can be extended to provide a three sex system.

Suppose that (1) there are two sex chromosomes: X & Y; (2) non-disjunction regularly produces loss of one sex chromosome (as happens in, for example, C. elegans); and (3) the total absence of sex chromosomes (00) is non-viable.

Then we have five genotypes: XX, X0, YY, Y0 and XY. Let us further assume that XX and X0 produce the same phenotype, as do YY and Y0; whilst XY produces a co-dominant effect of a third phenotype. We than have three sexes: XX/X0 (X), YY/Y0 (Y), and XY (XY).

XX/X0 can produce X or 0 gametes, and mate with Y or XY. This can produce all three possibilities with genotypes XX, X0, Y0, or XY.

YY/Y0 can produce Y or 0 gametes, and mate with X or XY. This can produce all three possibilities with genotypes YY, Y0, X0, or XY.

XY can produce X, Y or 0 gametes, and mate with X or Y. This can produce all three possibilities with genotypes XX, X0, YY, Y0, or XY.

The probability of getting each of the sexes will vary with the exact genotype of the parents and the possibility of disjunction (producing a missing chromosome and thus an 0 from an X or Y) but all combinations can produce all three sexes.


Instead of assuming a Haploid/Diploid set, what if we instead had Diploid/Quadraploid? Thus an adult would have four gametes of three types (X, Y, Z) to make up the following sets:

XXyz = X Gender
xYYz = Y Gender
xyZZ = Z Gender

These quadraploid sets are actually a loop, with the last gamete also being connected to the first. They must be in alphabetical sequence to be viable, however a quadraploid of YZZX is the same as XYZZ, you've just started reading at a different point.

Each parent would contribute one of their dominant gametes (the one they have two of) in addition to one other. These Haploids are transferred as a unit and bonds with the Haploid from the other parent to create the full Quadraploid.

Using an XY mating pair, we thus get the following possibilities:

Xy + xY = XyxY (non-viable)
zX + xY = zXxY = XXyz (viable)
Xy + Yz = XyYz = xYYz (viable)
zX + Yz = zXYz = xyZZ (viable)

Thus the child will have equal chance of being any of the three genders, with a 1/4 chance of immediate miscarriage (likely undetectable at this early stage). In addition to requiring all three gametes, the dominant gametes must be different to maintain viability, thus eliminating all possibilities with a same-sex mating pair. Without this dominance requirement, a same-sex pair will have a 50% miscarriage chance, and will only be able to produce children of their own sex.

This is admittedly the biggest stretch in this answer, but as I have absolutely no biology knowledge, I'd find it very easy to Suspend Disbelief in this case.


If they reproduce sexually and their reproductive cells are haploids but they are isogamous (Isogamy is a form of sexual reproduction that involves gametes of similar morphology (similar shape and size), differing in general only in allele expression in one or more mating-type regions.)





All combinations are valid.


In re-reading the question, and some of the answers, it does not exclude the possibility that all THREE sex gamete individuals are required in reproduction.

That is, two different gamete sexes, say X and Y, form the zygote, but it is embedded in the third host, Z, to develop to a viable fetus. It would seem that, by the criteria, any of the X, Y, or Z could be the host, and any two of the X, Y, or Z could be the two donors of genetic material, the only requirement is that the genetic material donors be of opposite gamete sex and the host be the third. Perhaps it is an immunity thing. A Z host will reject Z gametes through an immune process. However, it seems a Z host COULD carry a Z zygote. Thus, the offspring gamete sex would not normally be determined by the host, which did not donate any genetic gamete material, but it COULD be.


The act of copulation would be even more interesting.


Do your types have to be genetic? Lots of reptiles (the crocodilians and the turtles) determine the sex of their hatchlings by the temperature the eggs are incubated at: https://en.wikipedia.org/wiki/Temperature-dependent_sex_determination

Obviously, it's easier to manipulate temperature when the eggs are outside the body. But for internal gestation, give your aliens a womb which can tolerate bigger temperature fluctations (i.e. core body temperature fluctuation) than most mammals can tolerate. That way, the 'mating type' can be seasonal. In winter more X, in spring/autumn more Y, in summer more Z.

If their planet is like Earth, then in December northern hemisphere folks are birthing X babies and southern hemisphere folks are birthing Z babies. It gets more complicated in the tropics, where temperatures are more stable. But perhaps that might be a useful cultural difference - Country AAA in the tropics has a high proportion of Ys and has encourages Z and Xs to immigrate to their land. Or they make all the 'surplus' Ys join the army or the priesthood.

Alternatively female honey bees end up with a very different body form depending on whether they were fed royal jelly or not when they were bee larvae. So maybe the diet of the parent is important - and they may not even know it. If the parent prefers tea to coffee, they get more X and fewer Y and Z kids.

Genetics can be superimposed upon this to make things more complicated - so your W could still be a recessive, but only expressed if both genes (WW) and a low temperature/royal jelly are present. Perhaps the invention of central heating has doomed the Ws! :-)

  • $\begingroup$ My question concerns mating types, which are found in real fungi and algae. $\endgroup$
    – Anonymous
    Nov 15, 2017 at 12:40

To answer the title-question :They determine the sex by it's physical characteristics attributed to the given sex, like we do.

Regarding the genetic mechanism I would think that a combination of X and Y partners would only produce an X or Y offspring and not Z, given that neither carry Z. Also having both organs necessary for impregnating and being impregnated might render pointless a partner.


As an alternative to my previous answer. The other option is the same way that yeast do it, although this requires a haploid organism or, at least, a haploid sex chromosome or for sex to be determined by an extra-chromosomal element such as mitochondria that is haploid.

In (some) yeast, mating type is determined by a switchable "cassette". All yeast carry the gene for both mating types (I'll call them + and -) but only one is in the active position. There are also flippases which during division, with low random probability, will "flip" out the current active cassette and replace it with the inactive one (and vice-versa), resulting in a change of mating type. There is no inherent reason that this process cannot have a third state resulting in three mating types (similar six way switches are known in some bacteria) or that the activity of these flippases cannot be upregulated to result in random assortment at every generation.


Here is my pick: The creatures are zwitters and the XYZ system is there to guarantee self-infertility.

The phenotype x has genotype YZ and produces only X gametes (small and big ones); the phenotype y has genotype XZ and produces only Y gametes; and the phenotype z has genotype XY and produces only Z gametes. This works, because "by default" the creatures can produce all three types of gametes, but by inheritance each type is self-suppressing.

Now the fun part of world building: How do the creatures find suitable mates?

There are several possibilities

  • They are promisk (at least as long as they have found a fertile, i.e. not incompatible, mate)
  • They can smell the phenotype of each other and select a mate according to smell
  • The different phenotypes are different visually (having different skin colours, skin patterns, or coloured signal spots, or some body modifications, e.g., only x's wear horns)
  • They may differ in their voices and the tunes they sing in mating songs

EDIT: I wrote up the answer before looking at the quoted deviant art site. From the depiction of the three types it is clear that they are indeed hermaphrodites (zwitters) and that differ in their phenotypes quite significantly. The only thing that I find implausible is having equal gametes while there is a pregnancy period: This would include the risk of getting pregnant with an embryo that has not genetic relationship to the "mother".

  • $\begingroup$ Please define 'fertile' in this context. 'Fertile' usually denotes a female with a functioning and viable egg or a male with functioning and viable sperm. $\endgroup$ Dec 8, 2017 at 15:10
  • $\begingroup$ Ah, fertile means: A partner which whom offspring can be produced successfully and no incompatiblities exist. $\endgroup$ Dec 8, 2017 at 15:56
  • $\begingroup$ Does that include in vitro fertilization? This is a very complex OP question. Too many assumptions have to be made. The means and methods of reproduction are not specified, and open to speculation. These really have to be known before any credible answer can be given, unless 'sex' and 'reproduction' are not equated. $\endgroup$ Dec 8, 2017 at 16:14
  • $\begingroup$ Your edited comment about the possibility that the zygote, in whatever form, is not a product of the host, but that the host is merely a surrogate. Much like in vitro fertilization. Interesting. $\endgroup$ Dec 8, 2017 at 16:17
  • $\begingroup$ @JustinThyme: In the given setting this is indeed possible (e.g., in the case of a menage à trois) that two gametes fuse and implant, but none of the two gametes is from the "mother". $\endgroup$ Dec 8, 2017 at 21:54

This question is basically a riddle/logic trap

"There are 3 sexes: X Y & Z"

"There are only 3 viable offspring combinations: XY, XZ, YZ"

The fundamental question is: how do you determine the sex of those pairings?

Even though the OP has invalidated my first answer by completely redesigning this question, the question is still mechanically flawed and or a fabricated riddle/logic problem:

In human reproduction gender in genetics is represented as: XX (female) and XY (male). With the male being able to contribute either an X or Y chromosome ultimately determining the offspring's gender. The problem with this scenario is it's starting off with 3 genders ambiguously defined as X Y and Z with the potentially misleading implication that the starting parents are XX YY and ZZ which by the OP's own rules is impossible as the only offspring combinations are XY XZ and YZ. The parents would have to have the same genetic logic as their children.

This means that the 3 sexes aren't really X, Y & Z but are really XY, YZ, and ZX. Such that: a XY parent and a XY parent can only have XY children or technically none by the OP's desire, a XY parent and a XZ parent can have all 3, XY and YZ parents can have all 3.

So just like in human reproduction the X chromosome doesn't alone mean female or male, its the combination of both chromosomes that ultimately determine gender.

If however that is unsatisfactory;

If X Y Z individually denote gender and there can only by hybrid representation of these combinations then that would mean gender must ultimately be determined by another method. Here are a few options:

1) another regulatory gene handles the determination acting as a proverbial coin flip between the two possible options making one chromosome's representation more pronounced. This isn't unheard of in genetics, it generally doesn't happen on this level or with gender but why not.

2) Pseudo environmental- RNA inherited by the hosting mother parent makes the determination. Would be an odd occurrence but also not impossible.

NOTE: A sex is nothing more than an organisms distinct template discrepancies between members, broadly covering physical and behavioral traits. This does not constrain itself exclusively to genetics with humans concocting "3rd genders". Even genetically there are cases where offspring aren't either XX or XY but are XXX. So in relation to this argument, definition of the pronunciation of a "sex/gender" is fairly subjective.

  • $\begingroup$ I swear this has to be a deliberate logic problem $\endgroup$
    – anon
    Dec 5, 2017 at 21:28
  • $\begingroup$ I feel the exact same way. It is very frustrating. $\endgroup$
    – Anonymous
    Dec 6, 2017 at 14:20
  • $\begingroup$ LOL ill take that as confirmation $\endgroup$
    – anon
    Dec 6, 2017 at 14:41

You are starting with a false assumption. You state that sex is determined genetically. It is not, not even in humans. The default sex in humans is female. To get a male, the default structures have to be modified and transformed. A long process of links and dependencies have to occur before the fetus becomes completely male. In the absence of any of these steps, the process reverts back to the default female.

Human sexuality is determined hormonaly, after the zygote starts developing. What is determined genetically is what hormones will be produced in utero in the first few hours, but these are not absolute. The default sex of humans is female. If after those first few hours, the process of masculinizing is not started, you get a female. Any interruptions in the process, you get feminization. There are many other factors besides the and Y chromosome that come into play. The sex of humans can theoretically entirely be determined in the complete absence of the X and Y chromosome. In fact' many researchers posit that the Y chromosome in humans may eventually completely disappear. It is pretty useless. It will be replaced by a standardized fully functional X chromosome. There are mammalian species in which this has happened. The sex of the offspring is determined by other genetic factors besides the Y chromosome.

In order for a genetic system to work, there needs to be a default setting. Nature seems to prefer the offspring-producing 'egg' sex as the default. That makes sense as a fail-safe mechanism. If all else fails, what is left will still be able to produce offspring. Genetic donor sexes are really not needed. They just provide extra genetic material for the sake of diversity. And genetic material can be provided by means other than sperm.

However many 'sexes' you have, somehow there has to be a process for one form to carry the offspring. Unless, of course, the 'egg' is simply deposited and abandoned, in the hopes that it is fertilized. The 'sex act' would not be between two members of the species, but would be something done over the egg. In which case, any gender can deposit the egg, and any gender can add genetic material to fertilize it. Gender becomes something other than sexual determination.

But if you stick to an exclusively gamete-determined sexual identification, you need to determine which gamete produces the genetic contributor, which gamete produces the offspring carrier, and, well, what IS the third one? Nature has a habit of making the third one 'neutral' (that is, incapable of egg OR sperm production).

So however you decide to do it, you will not end up with three 'reproducing sexual genders', you will end up with three different'functional genders' at a minimum, and maybe up to 6 combinations of X, Y, and Z. (Even more, if you specify that THREE gametes have to be donated. That is, viable offspring need a gamete from each of three different donors. So you need, for instance, XXY or XZY or ZZZ. This gives a possible 10 unique combinations. But now you have to decide which combinations are genetic donors, and which combinations are offspring carriers. There is perhaps no neutral, if it takes three gametes to produce a viable offspring). That is, it might result in functional genders of perhaps worker, leader, builder, reproducer and so forth. (Workers, for instance, might have more hands, leaders might have a bigger brain, and builders might have more legs). The only feasible purpose of having three form identifier gametes is to produce a different species functional forms for a different function.

Nature seems to prefer that, in such cases, one or more forms are infertile and incapable of reproducing. So your third 'gender' would probably be 'neutral'. You can definitely see the complications of this being left up to genetics, and not environmental, conditions. Species reproduction is now invested in only one of three identifications. If left only to the probability of mating, either that third is very prolific, or species numbers would decline.

So your answer is that you really need to re-think what your objective is in having three gametes.

  • 1
    $\begingroup$ "The default sex in humans is female. To get a male, the default structures have to be modified and transformed." Amusingly enough, embryonic research came out in the last few weeks indicating that this is not actually true. $\endgroup$ Dec 6, 2017 at 17:14
  • $\begingroup$ @Keith Morrison References please. That goes against a tremendous volume of research, Sex Determination in Humans $\endgroup$ Dec 7, 2017 at 16:07
  • $\begingroup$ @JustinThyme I recently read something about the Foxl2 gene. I think this is what Keith was talking about? $\endgroup$ Dec 8, 2017 at 3:26
  • $\begingroup$ @James Khoury If it is the Foxl2 gene, it does not refute the idea that the female is the default sex. It explains how the default sex is, in part, 'constructed'. Proteins are the building blocks of bodily structures, and if the correct proteins are not present, then the structures are not built. I do not dispute that the default sex requires several other genes to do their job in order for all the parts to be available, but these other genes just do the building, they do not do the determination. They are 'after the fact', 'after the decision has been made'. $\endgroup$ Dec 8, 2017 at 4:46
  • $\begingroup$ @JustinThyme As I understand it that research is showing suppression of this gene makes female organs start form like males. Iirc it was the ovaries that started to form cells that could produce sperm. That would suggest that the default here are the male genitals. I'm just guessing that was what Keith was talking about. I don't even have the article/paper for this. $\endgroup$ Dec 8, 2017 at 4:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .