In reality everything have a cause even random events .
David,
Modern Physics has described subatomic events as acausal. As such, only probabalistic treatments at that level seem to be useful. Naturally, you may assert that even those events must have a cause, but that would only be an opinion, not a proof.
As for milf's suggestions, I find them very interesting. It is true that there are over a thousand species in which some amount of homosexuality is exhibited.
I think the OP question is useful, since some people try to make the bogus claim that the uselessness of homosexuality in terms of reproduction would select it out of the species over time. The reference to non-reproducing genders in other species kills that theory right away.
I like the "hitch-hiking" metaphor of one of the posters. It is true that many physical traits can be traced to a single gene. If one of the traits of a gene is very valuable, that trait may outweigh the negative cost of one of the other traits, and cause it to be propagated down the line.
Sickle cell anemia is a good example of this. This is a birth defect that is related to a gene that profers malaria immunity on its bearer. In tropical climates, you find lots of sources of malaria and also a higher rate of sickle cell. The malaria immunity outweighs the sickle cell amemia.
This is not to imply that homosexuality is a disease. It only shows that if harmful traits "hitchhike" on valuable ones, certainly evolutionarily neutral traits like homosexuality could do so just as easily.
There are a number of theories on how homosexuality in pack animal species is prevalent. They suggest that it is a reaction to the fact that only the alpha male is allowed to reproduce in a pack.