Your point about the dominance dynamic partially being about coordinating the release of one ovum is an excellent one, and depending on the degree to which that is true it might unravel much of the underpinnings of my argument. Although I find it somewhat unlikely that follicular selection would take the particular shape that it has if there’s not a significant element of selection. But I don’t really have a reference frame for the possibility space of available strategies for such coordination procedures. Are there any other mammals that do not have follicular selection? Any other animals with any clues? Fish and such just outsource the whole selection shebang, but is egg selection amongst birds comparable? I have no clue.
As to the relative selection pressure, I can think of at least one way to try to do the math. I’m just going off cursory search numbers here, but apparently the peak amount of oocytes is actually prenatally at 20 weeks of gestation, where it’s about 6-8 million, this declines to 1-2 million at birth, and 300-500k by puberty. Now, there is an unknown factor here which is the degree of selection pressure before the peak at 20 weeks. Apparently pruning starts as early as week 14, but I couldn’t find any established numbers. A rough estimate was apparently that well over half of the original germ cells don’t survive to be part of the privileged 8 million oocytes. So let’s take that at face value and say that it’s somewhere between 12 and 20 million competing candidates initially. There is another unknown here which is that not all of this pruning might mean the same thing. The two main mechanisms are atresia of primordial follicles, and death during meiosis, the latter of which sounds more like the results of gene-related defect. Perhaps the first is more mitochondria related, but there is actually a great deal of cell death going on at this stage all over. For comparison, apparently very roughly 75% of male germ cells die around this period. But they might also be subject to entirely different specific selection pathways that are just unrelated to mitochondria. In conclusion I have no idea how to make sense of this mess. I think it seems reasonable enough to just to add 25% to the previous upper estimate of oocytes at birth to inflate the starting point and call it a day, bringing us up to a nice even 10 million. So using the lower bound of 300k, that gives us roughly 30 times more initial oocytes candidates at puberty then at birth.
Here there is a little snag in the logic, which is that as far as I can tell pruning from week 20 to puberty looks nothing like follicle selection. There is no “growth competition”, the cells just die off steadily by apoptosis. What does this mean exactly? No clue. If we accept Nick Lane’s thesis that women are all about mitochondria, it’s probably somewhat relevant? The connection doesn’t seem as phenotypically obvious as it does in follicular selection though, where I think the “best grower” concept intuitively seems related to energy output. Then again mitochondria are also very strongly related to the apoptotic process. So I guess the relevance to mitochondria of all that cell death that happened before follicle selection starts is somewhere between 0% and 100%. Great. Let’s cut our losses evenly in the most simplistic way imaginable and just go for a 50% penalty to the initial oocyte count to account for this uncertainty to their relevance, putting us at 5 million or 15 times the number of remaining cells at puberty.
So at this point, when follicular selection starts. How many follicles start growing? Some sources say 3-30. The narrowest I found was 5-20. I’m just going to cut it at a nice looking number once again and say 15. So from 15 candidates 1 emerges. So 15 times less again. Here there is another big uncertainty, which is that out of those 15, how many truly bad candidates are there really? I mean maybe the runt of the litter is a bit off, but the biggest and second biggest are most likely not that different mitochondria wise. So once again let’s apply a 50% penalty to the importance of follicular selection, and then 50% again to account for your argument about the fact that follicular selection might have a significant element of coordination rather than selection. Let’s round it up for a nice even number of four.
So plugging that naively into your formula for X=15 and Y=4 we get 79% of the selection pressure before follicle selection and 21% during. I’m writing this on my phone so I just had GPT 5.4 do the math, and it also suggested that using the logs of each number is mathematically more sensical for reasons I’m not sure I understand. Anyways in that case it’s a 69⁄31 split instead, so roughly two thirds before and one third during follicle selection.
These numbers mean almost nothing, and have mostly been useful for me in illustrating my profound ignorance. However, the final conclusion of it being “a third” of selection pressure seems like a prudent prior to have, so I’ll just anchor mine there for no good reason. At least I can pretend to have done my Bayesian due diligence.
Oh, and GPT-5.4 gave me some lip when I asked for a sanity check:
“Your 50% penalty for pre-pubertal attrition being only partly relevant to mitochondrial quality is fine as a rhetorical placeholder, but biologically it is doing a lot of work”.
Fair enough.
I spent perhaps a bit too much of the days bandwidth allocation on this conversation myself, but I’ll see if I can find the time during the week to summarize it in case anyone’s interested.
Also I would have loved to attend your conference, and its actually during a quiet part of my holiday, but unfortunately I think the 15 hour transatlantic flight might be a little bit too steep of a threshold 😅