Could my creature's reproductive system work?

Question

I came up with anthropoidal creature that has several subspecies, let's call them red, green & blue, that can interbreed with each other. My problem in this setting is that:

1. I don't want any ugly hybrids,
2. I don't want any of the subspecies to disappear due to interbreeding

So this is what my proposal how should their reproductive system work:

Each individual contains chromosomes that are shared for all the colors e.g. 80, plus certain amount of chromosomes unique to each color e.g. 20. Males produce sperm which contains half of the genetic material, just as humans do in this case the sperm would contain 50 chromosomes of which 40 are shared & 10 are unique for each color.

When couple of same color (red & red, green & green, blue & blue) have a child it works the same way as in humans. The child could either be male or female, will have the same color as its parents, and has 50% DNA from its mother and 50% DNA from its father.

However when a couple of different colors (red & green, red & blue, green & blue) has a child, the mother reproductive system only uses the shared chromosomes of the specie from the father's sperm, while the chromosomes with different color are discarded and mother's own chromosomes are used instead. So in that case the child will be always be a female, will have same color as its mother, and will have 60% DNA from its mother and 40% DNA from its father.

Answer 1

The simplest way to achieve the desired result is through the Mitochondrial DNA. Humans and most other eukaryotes contain a smaller, secondary genome outside of their nuclei in another organelle called the Mitochondrion. This mitochondrial genome is very useful for our purposes because it is only inherited from the mother through the egg.

If you imagine three significantly different mitochondrial genomes: red, green, and blue you will arrive at precisely your desired scenario. There is only one set of mitochondrial DNA per organism so there can be no hybrids. In any cross-population procreation only the mitochondria of the mother is passed on meaning all of a female's offspring will have identical mitochondrial genomes regardless of the genotype of the father.

Edit: To address the plausibility of the mitochondrial genome influencing phenotypic traits.

It is certainly plausible that mitochondrial DNA could encode phenotypic traits like skin color, but perhaps unlikely. The mitochondrial genome in humans is ~30,000 nucleotides in length and contains 37 genes. This means compared to the nuclear genome it is ~1/100000 the size and contains ~1/5000 of the genes. The chance of any given trait being encoded by the mitochondrial genome is thus somewhat small, but those 37 genes (most of which aren't actually proteins, but rather encode RNAs) do have functions. There are many disorders caused by mutations in the mitochondrial DNA.

I don't know of any phenotypic traits encoded by the mitochondrial DNA but currently we are much better at finding disease alleles than we are at finding the loci that govern more complex traits like skin, or hair, or eye color. This means there may be phenotypic traits encoded in the mitochondrial DNA that we don't know of, or there may not be, but I would argue that either way it's certainly plausible that the genes transcribed in the mitochondrial DNA could influence such features.

Answer 2

Hybridogenesis

There is no system like this on Earth. The closest is hybridogenesis, where one parent is a sexual parasite on the other to produce hybrid offspring which only pass down the parasite chromosomes. Species that engage in hybridogenesis are dependent on it to reproduce and consist of only one sex.

Mendelian alleles

If your goal is to have subspecies with traits which cannot be watered down through interbreeding, then an alternative is that subspecies is determined by a single gene.

Let's call the resulting alleles Cᴿ, Cᴳ and Cᴮ. These genes are recessive or dominant, so there are only three possible phenotypes of red, green or blue. Let's assume the dominance pattern is a circular B>R>G>B. The possible genotypes (and phenotypes) are CᴿCᴿ (red), CᴳCᴳ (green), CᴮCᴮ (blue), CᴿCᴳ (red), CᴿCᴮ (blue), and CᴳCᴮ (green).

If the allele is located on the X- or Z-chromosome or its equivalent, then heterozygous (XY, ZW) offspring (who have only one allele) will be the same color as the parent who contributed the X- or Z-chromosome. (Males are XY or ZZ, females are XX or ZW.)

Answer 3

Apologies if this has been said... The premise is that a mixed marriage has only daughters in which color comes solely from the mother. Therefore

1. Doesn't matter where it comes from in a non-mixed.
2. There can never be mixed color females
3. If mixed matings happen at all, females will outnumber males
4. If mixed marriages are common, females will greatly outnumber males and the overall population will slowly decrease. (But it will stabilize probably at some lower limit.)
5. If one "color" is more desirable for some social reason, that color will increase in population and the others decrease because those females will have more mates.
6. If gestation is of significant length, polygamy will likely be the norm due to the females being so numerous.
7. There will be no mixed males, since they all come from non-mixed unions.

Whether or not there are mitochondria doesn't matter; there will be some biological explanation for the premise (first paragraph) but the story can probably be told without revealing how the genetics actually work.

Seems to me it would be rather easy to program a simulation, but I can't afford the time to do it. I shouldn't even be typing this, but it's an interesting problem and I'm quite tempted. :-)