With the weather warming up, I recently moved a couple gravid female Gambusia affinis out of their breeding tank into an outdoor "pond" consisting of a 4 ft diameter inflatable wading pool. The pond has been there about 10 months and has a fairly mature population of invertebrates including copepods, daphnia, helgramites, mosquitos, ramshorn and pond snails, chironomids, and a couple other things I can't identify, sustained by a thick wad of fallen leaves in the bottom. The fish rapidly dropped their young which I assume they had been holding due to the high population density of their indoor tank. Several weeks later I noticed a second, larger cohort of fry, and started netting out the older ones to reduce the pressure on the food supply.

As I've previously mentioned, I'm trying to select for the melanistic trait, and am saving only female offspring from this tank to breed back to wild-caught melanistic males. This brood only had two apparent females out of 14. Google suggests some chemicals may influence sex determination, and there are a couple papers suggesting temperature, but I don't see them concentrating on this species specifically. Would I be better off trying to raise the young at lower temperatures perhaps, or could there be something else going on?

I notice with wild caught ganbusia fry. That they all look like females for a very long time. Then the anal fins grow long.

That's true. I made a post discussing that phenomenon last year, as I initially thought I was saving only females for my breeding project, but kept finding males in the tank. I've never seen or heard of a transformation in the other direction, however.

Just a thought: stocking density may also affect sex ratios. If sex ratio production is under the control of the parents, they might opt for a male-oriented brood when population density is high (gambling on the ability of the males to spread genes most rapidly if females are plentiful) and female-oriented when density is low (as females are virtually guaranteed to breed, while most males may be excluded from mating in a colonization situation). I have no idea if this actually happens in Gambusia, but it might be worth fooling around with.

By the way, in your readings on temperature-mediated sex determination in livebearers, did you learn whether the sex determination ocurred pre- or post-fertilization, and if post, whether the sex of the embryos actually changed (vs. selective reabsorption of embryos of one sex)? The only temperature-based sex determination systems I am familiar with are those of turtles; it seems to me that a live-bearing animal would have more options. Just curious, really; I don't know that the answer would have any bearing on your problem.