In article ,
Stewart Robert Hinsley wrote:

The obvious explanation is that the parent plant is heterozyous for stem
anthocyanin (or other pigment, but google provides a citation for the
presence of anthocyanins in Amsonia) production, and a proportion of the
seedlings are homozygotes for the recessive allele. Perhaps you'll later
find a pleiotropic effect on flower colour.

For every complex problem there is an answer that is clear, simple, and
wrong :-) That may be the explanation, but it may well be mistaken.
If the flower colours are different according to the stem colour,
I agree that would make this fairly certain.

The expected mean ratio of offspring from a self-pollinated diploid
dominant/recessive heterozygote is 3:1. 50:10 isn't that far off 45:15 -
you'd need someone with more current statistical expertise to say
whether the difference is statistically significant for a sample of that
size.

Current, I can't do, but I don't need it. That is the mean ratio
ONLY if it is a single, Mendelian gene AND the parent reproduces
purely by haploid/diploid self-pollination. In my previous post,
I missed that it was a single parent, so said something irrelevant.
I agree that is the simplest plausible explanation. Anyway, as I
said, 50:10 isn't significantly different from 45:15.


Regards,
Nick Maclaren.