bluerosy,
So let me see if I understand this now. For this morph you see 4 phenotypes -
1. normal males
2. normal females (normal coloration at hatching)
3. light colored females
4. peanut butter males
Clearly the expression of this trait is influenced by the sex of the individual. Whether it is sex-linked or autosomal, the light colored females are not hets, they are in fact PB females.
As to the question of whether it is autosomal or sex-linked, you could answer this by mating two "heterozygotes" (I put het in quotes because if it is sex linked, a female cannot be a het).
This means you would have to mate a PB male to a normal (wild type) female and then cross two of those normal looking offspring to each other (i.e. - not using any light colored females from the first cross). This way you would be sure you are crossing two "hets".
If the trait is sex-linked, you will be crossing (assuming z is the PB allele, Z and W are the normal sex chromosomes)
Zz x ZW
and you should get
- all normal looking males (half of which would be het)
- 50% normal females and 50% light (PB) females
If the trait is autosomal, you will be crossing
Aa x Aa
and you should get
- 3/4 of the males should be normal, 1/4 should be PB
- 3/4 of the females should be normal, 1/4 should be light (PB)
So if you did this cross and got any PB males, you would know the trait is autosomal. Of course, not getting any PB males doesn't prove it is sex-linked, but that is the nature of science. That is why it would take repeated matings of this sort to support the sex-linked idea.
-----
Chris Harrison

