Genetics/Sex question

[EternalFrustration]

So... if I'm reading everything that everyone is saying correctly, the idea of multiple sexes in complex creatures probably isn't realistically feasible... right?

[Salmoneus]

So... if I'm reading everything that everyone is saying correctly, the idea of multiple sexes in complex creatures probably isn't realistically feasible... right?

The idea of multiple sexes, where 'sex' just means a reproductive compatibility setting, in complex creatures probably isn't particularly realistically feasible, no (though I'd hesitate to call it completely impossible, just highly improbable).
Multiple sexes in the sense of genetically-determined morphs is probably less likely that two sexes, but should be feasible, with sufficient reasoning. [The most likely form of this, however, is a plain male-female-neuter three-way division.]

I'd also say that any sort of deviation from disexuality in complex creatures will have an evolutionary cost, and will probably be associated with more integrated societies (with division of sex paralleling division of labour). It's not a coincidence that the most prominent examples of three-sex species are the colonial insects.

[Miekko]

So... if I'm reading everything that everyone is saying correctly, the idea of multiple sexes in complex creatures probably isn't realistically feasible... right?

The idea of multiple sexes, where 'sex' just means a reproductive compatibility setting, in complex creatures probably isn't particularly realistically feasible, no (though I'd hesitate to call it completely impossible, just highly improbable).
Multiple sexes in the sense of genetically-determined morphs is probably less likely that two sexes, but should be feasible, with sufficient reasoning. [The most likely form of this, however, is a plain male-female-neuter three-way division.]

I'd also say that any sort of deviation from disexuality in complex creatures will have an evolutionary cost, and will probably be associated with more integrated societies (with division of sex paralleling division of labour). It's not a coincidence that the most prominent examples of three-sex species are the colonial insects.

There is a game-theory based argument as to why two genders is the most likely stable configuration, alas, I don't recall the details nor do I find the paper

[2+3 clusivity]

I guess one funny effect is if 1 or 2 sexes from a 4 sex system died off, the species would overall remain viable--with a two (or three) sex system.

[EternalFrustration]

Thanks for your insight everyone!

[Chuma]

It's true as Torco says that if A+B can only have A and B children, then it might happen that some genders die out. But it doesn't have to happen (within a reasonable timeframe). Compare with the situation if A can only have children with A; that would just be different species, and of course some of them might die out, but they don't have to. In this particular case, it might also be that each gender is invaluable in some other sense. That all depends on what kind of animal you're trying to make (or it might be a plant, for all I know).

In a more exotic conworld, there's a very easy way: They don't need to have double chromosomes. In our system, as you probably know, cells first divide so that each has single chromosomes, and then merge with the other parent's cell so you again have single chromosomes. You could do it the other way around - first the cells merge, giving double chromosomes, and then they split, giving single chromosomes. Arguably you would always get twins, always of different gender, and interestingly with no genes in common. If you don't like that, you can just let the body reject one of the cells/foetuses at some point.

If you want to do it with plain old chromosome pairs, that should surely be possible too. I've thought about similar four-gender systems, and the best I could come up with was an AB0 system: Only A0 and B0 can have children, as AB and 00 are gender-neutral and thus infertile. I imagined this with humans, and the fun thing is we can basically tell what the 00 and AB would be like, by comparing with various sex-related genetic disorders.

Your system would be a bit more tricky, but one way would be an ABCD0 system, where only A0, B0, C0 and D0 survive. Genes that are lethal when you have them double are quite common. That would mean that only half of all pairings (between different genders) would make a child, but depending on the mating system, that might not be a problem - many animals after all have thousands of offspring in one go.

[Basilius]

It's true as Torco says that if A+B can only have A and B children, then it might happen that some genders die out. But it doesn't have to happen (within a reasonable timeframe).

Note that the time frame needed here is probably not just one species' history, for (their equivalent of) sapientization must have put forward some much more urgent problems than rearrangement of sexes.

<...> it might also be that each gender is invaluable in some other sense.

However, if such "genders" are invaluable in some sense not related to fertility and reproduction, we're basically back to something like sterile castes in eusocial insects.

They don't need to have double chromosomes. <...> Arguably you would always get twins, always of different gender, and interestingly with no genes in common. If you don't like that, you can just let the body reject one of the cells/foetuses at some point.

This is not genetically more stable than with diploid organisms if each individual can mate with all sexes except one's own: extinction of sexes and degeneration to a two-sex system is still the most probable scenario (without some strong counterbalancing factor).

I've thought about similar four-gender systems, and the best I could come up with was an AB0 system: Only A0 and B0 can have children, as AB and 00 are gender-neutral and thus infertile.

Yes. Note, however, that this is not too different, again, from two fertile sexes plus some irreversibly sterile castes.

Your system would be a bit more tricky, but one way would be an ABCD0 system, where only A0, B0, C0 and D0 survive. Genes that are lethal when you have them double are quite common. That would mean that only half of all pairings (between different genders) would make a child, but depending on the mating system, that might not be a problem - many animals after all have thousands of offspring in one go.

Thought of this, too. Stability issues, again. Every mutation that breaks the system will lead to increased fertility.

[KathTheDragon]

What if you prevented such doubled chromosomes from letting the offspring develop at all? Would that not then force all births to be one of these four genders?

[Basilius]

Every mutation that breaks the system will lead to increased fertility.

[svld]

Having dikaryote, diploid, and two haploid mating types could work.
You have to first classify them as four different genders, though.

[jal]

Mmmm... getting a genetically stable 4 sexes is kinda tricky. I can think of something that could work, maybe, but I still have doubts regarding its stability.

Assume there are four sexes A-D, and four sex markers a-d that combine to form the sexes as follows:

A ab
B bc
C cd
D da

Then we have these combinations assuming mating of the same sex is not possible or just doesn't happen (it's a bit difficult to see how this should work physically or behaviourly, but we're discussing the genetics so I'll leave that):

A+B -> ab ac bb bc = A † † B S
A+C -> ac ad bc bd = † D B † R
A+D -> ad aa bd ba = D † † A S
B+C -> bc bd cc cd = B † † C S
B+D -> bd ba cd ca = † A C † R
C+D -> cd ca dd da = C † † D S

For each mating, two of the possible combinations do not result in gemination (or not in a viable embryo/foetus), the other two do. Also, the distribution is stable: if the sexes mix freely, on average the offspring will have 25% of each sex. The possible threat to stability is the fact that of the possible six combination, four are self-replicating, i.e. the offspring has the same sex as the parents (S). Only two combination yield opposite sex offspring (R). I'm not a statistician, but that seems threatening stability.

I've played around a bit with other combinations, but that doesn't seem to improve the situation much. Maybe someone else can come up with a system that has a better distribution of offspring amongst the possible combinations.


JAL

[jal]

Come to think of it, it may work if we let go of the assumption that every gender needs to be present in the same distribution. I've worked out a quick example with two sets of gender determining genes, and it seems that one will dominate, another will wither, and the other two will do fine. Someone with l33t programming skillz and a better understanding of mathematics than me should create a simulation for that. Would be interesting...


JAL

[Chuma]

How about this: There are two species, i.e. four genders, which enter a symbiotic relationship. That way you would immediately get a four-gender society with no infertile genders. If you're not happy with the lack of inter-species sex, you could say that they are closely related, close enough that they can have infertile offspring. If the infertile offspring happens to have serious advantages - which is at least not unheard of in real life - you would get a relatively stable multi-gender system. It also wouldn't need millions of years to develop, as the symbiosis could be a recent invention.