Fst Differences Between Mature Trees and Seeds

Question

For a single co-dominant locus in two populations of blue oak, I am given genotype frequency data (genotypes are AA, AB, BB) for adult trees (at least 150 years old) in both populations (pop 1 and pop 2) and from sampling of seedlings taken in fragmented woodland from the same populations (pop 1 and pop2) very recently. I am asked to detect any difference in Fst among adults and seedlings, and propose a hypothesis explaining how any difference observed could have come about.

To tackle this question, I calculated the Fst value between Pop1 and Pop2 using the adult's genotypic frequencies and then Fst between Pop1 and Pop2 for seedlings using the respective genotypic frequencies given for the seedlings.

The Fst values for the adult's was 0.004 and the seedling' was 0.113.

Clearly there is genetic differentiation between population 1 and 2 in seeds, but not much in adult trees.

I'm mostly stuck on the explanation for why this would happen. My gut response was that there is some selection against certain adult genotypes in population 2 in that they are unable to produce seeds? Or maybe that adults in population 2 are mating non randomly and therefore produce seeds that do not match the genotypes of the adult population?

Answer 1

I will assume the claim Clearly there is genetic differentiation between population 1 and 2 in seeds, but not much in adult trees. is correct and that you have done your stats correctly.

A complete analysis of the situation depends of how you calculated Fst. Did you use Weir-Cockerham Fst or standard Nei's formulation? How did you average (with or without weights) over population and loci? But without going to these details, here is an extreme scenario that explain why you could observe such patterns.

Imagine, the migration rate from any population and to any population is 0.5. In such case, in absence of selection you would not expect any divergence between populations. However, if genotypes AA and AB both have seed survival probability of 0 in pop2 (but not in pop1), then pop2 will be fixed for A2A2 at the adult stage even though when looking at the seeds you cannot see any evidence of divergence.