Prim wrote: ... I have a hard time envisioning the kind of sustained selective pressure that would produce that effect over time in populations wiith long generation times.
Hm ... I'm not sure I follow this. We do evolve, in spite of our relatively long lifetimes ... For the top predator it's the supply line that matters, so it seems to me that all the selective pressures we've experienced must have taken that form. Anything that kept us midway between critical mass and carrying capacity would be favored in the long run. But I probably misunderstood what you meant.
Just speculating off the top of my head here, because I don't know anything about the biology of sexual preference, but it occurs to me that this might be like the question, "what good is half an eye?" The answer depends on what that half an eye was doing.
It might be that bisexual behavior was originally favored because it reduced intragroup conflict. In all mammals except humans, the females are only available when they're in heat, but selection would favor males that were johnny-on-the-spot whenever the female was ready (that is, synchronizing two cycles would be a relative disadvantage in species with low numbers of offspring), providing that what you did with the testosterone the rest of the time was to make love and not war.
Overpopulation might damp the reproductive urge in general via hormonal changes, i.e. non-human females don't come into heat when there is insufficient food, and then male homosexuality is the fallback position. We don't know the basis for sexual attraction really ... it
could be that human females are also less attractive, for reasons not fully understood, when there is resource scarcity. And when the scarcity is severe, as you noted, women become unable to reproduce anyway.
Thing is ... any mechanism that operates primarily to remove males from the reproductive cycle is a weak inhibitor. It would be a lot more effective if female homosexuality were the primary effect ... but then, most of us would benefit from having three arms (I know I would!) and we haven't evolved one of those either.
Or, it could be that female homosexuality did appear as a primary inhibitor at one time and it was
too effective.
Sociologically, female infanticide is the common response to overpopulation. That's effective within one generation, without changing reproductive capacity in the generations beyond.
There was an interesting longitudinal study done at U. Penn on determinants of family size, but I don't have time to tell about it right now 'cos I have to get to school.
Jn