I'm surprised too, especially since -- unless my understanding is very wrong -- something close to group selection is currently the most popular explanation for the natural occurrence of homosexuality in social animals: social groups with small populations of homosexual members benefit from the work contributed by those members, without the expense of additional offspring.