In quantitative traits we have many loci affecting a trait. If we took a bunch of parents where all mothers and fathers were of a certain value at the upper end of a trait range, and then looked at the offspring would we expect offspring from pairs of parents from more divergent populations to be even more extreme variance in their traits?
For example: if we look at 2 men from East Africa, both 190 cm tall. They each have children with women who are 180 cm, one of whom is from East Africa and the other from Canada. We could assume that the Canadian woman is probably more unrelated to her male than the East African woman. I would therefore expect that it is quite likely that the East Africans in the example are tall because of "tall-alleles" in many of the same loci, whereas there is a greater probability that the Canadian female is tall thanks to mutations at different loci. Therefore would it be reasonable to expect greater variation in the offspring of the Canadian woman?
Is there any evidence of this phenomenon?
Note: Ignore environmental and gene * environment interaction variance, just go with a simple additive model, and assume populations have had some barriers to migration for some time