> I'm confused though by his comments about "our species’ one-time amazing levels of genetic diversity." Overall, he seems to be espousing the idea that "diversity" is something we once had lots of, but have lost over time. Is this true?
"Diversity" can mean a lot of things. In this case, he's referring to the variety of individual alleles. Due to population bottlenecks, most of humanity has reduced variation: for example, at a certain spot in the genome everyone has a T, because everyone who survived the bottleneck happened to have a T there; but in the population that didn't go through the bottleneck, some people have a G instead.
On the other hand, for phenotypical diversity it's not just the set of individual alleles that matters, but their distribution and correlation. For example, let's say height is affected by thousands of genes, for each of which there is an allele that makes you slightly taller and one that makes you slightly shorter, and the combined effect of all these genes is approximately additive.
Then you may have a population A which has both variants for each gene, but they're independently distributed, so everyone is of medium height, with some variance. We also have a population B, which has only one variant left for several of these genes (they have lost diversity!). But B is split into two sub-populations, which differ in the distribution of the remaining height genes: B1 has more tall alleles, and B2 has more short alleles.
Now the average height of B1 is higher than that of A, and the average height of B2 is lower than that of A. You will find very tall people in B1 that you would never see in A, and very short people in B2 that you would also never see in A. So B has more diverse height phenotypes, even though B1 and B2 (even put together!) still have lower genetic diversity than A in terms of the available repertoire of alleles.
"Diversity" can mean a lot of things. In this case, he's referring to the variety of individual alleles. Due to population bottlenecks, most of humanity has reduced variation: for example, at a certain spot in the genome everyone has a T, because everyone who survived the bottleneck happened to have a T there; but in the population that didn't go through the bottleneck, some people have a G instead.
On the other hand, for phenotypical diversity it's not just the set of individual alleles that matters, but their distribution and correlation. For example, let's say height is affected by thousands of genes, for each of which there is an allele that makes you slightly taller and one that makes you slightly shorter, and the combined effect of all these genes is approximately additive.
Then you may have a population A which has both variants for each gene, but they're independently distributed, so everyone is of medium height, with some variance. We also have a population B, which has only one variant left for several of these genes (they have lost diversity!). But B is split into two sub-populations, which differ in the distribution of the remaining height genes: B1 has more tall alleles, and B2 has more short alleles.
Now the average height of B1 is higher than that of A, and the average height of B2 is lower than that of A. You will find very tall people in B1 that you would never see in A, and very short people in B2 that you would also never see in A. So B has more diverse height phenotypes, even though B1 and B2 (even put together!) still have lower genetic diversity than A in terms of the available repertoire of alleles.