When calculating by hand a simple example of He I realised how an interpretation made from the data could be be entirely wrong if I just visualized He as representing number of different alleles without too much weight on frequency.

Not being a mathematician – I was amazed to see that the squaring component in the He calculation reduced my precious private and rare alleles to almost nothing…and that if the frequency of one allele was dominant the presence of low frequency alleles did not really increase diversity – so there was greater diversity when two alleles were present in equal frequency than when multiple alleles were present but when one allele dominated.

I know this is basic knowledge – this is stuff I should know – but I think you can easily not know the inns and outs when you cross discipline / especially when there is such a body of info and proper understanding relies on substantial background knowledge.

So – anyway. If I am looking at a species which may have been fragmented for only two or three generations – and I am looking at the genetics of these populations to see if there is any evidence of within population mating – I might find some private alleles but not at a level detectable to He or any follow on statistic.

The other problem is that the maximum He is given by 1 divided by the number of alleles. Again -basic knowledge that I didn’t realise the significance of until I did a few experiments by hand. This makes it totally meaningless to compare He across different markers -

When you calculate a Hs from the average He what happens to the maximum He that is embedded in each individual He? I suppose it makes sense if you interpret is as expected heterozygotes rather than gene diversity. And I suppose the maximum is reduced initially because of the lack of diversity at that loci (so there is that component) and then additionally within in each subpopulation in relation to how much of the possible diversity that subpopulation exhibits. But surely it would be important to know what the maximum He for a particular marker at a particular locus before interpreting subpopulation He.