I'd like to see if I can help a bit here by giving a very basic, non-mechanistic hypothesis for the development of sexual reproduction into the two-sexes thing we know today.
Let's start with one thing here, though:
Lyston writes:
And, how are you sure that things didn't start out with sexual reproduction and asexual was a mutation?
You've used the word "mutation" like this several times, and it's not entirely accurate. The word "mutation" refers only to a change in a gene, not to something that's different. It's likely that several mutations contribute to the transition between asexual and sexual reproduction.
Now, on to the hypothetical scenario:
Some organisms are isogamous (meaning that both types of gametes are identical--this is generally considered more "primitive"). Then, there are some that are anisogamous (meaning one gamete type is bigger than the other). Then, there are oogamous (oh-AH-guh-muss) types (egg and sperm--this is generally considered more "derived").
In some simple, single-celled organisms (such as algae), two cells (constituting two organisms) could feasibly combine into one with twice the genetic material of the former two, giving it added flexibility in the face of a pathogen (i.e. if one of the two genomes had a defective gene, the other genome's copy of that gene could cover for it). You can see how this would be beneficial? Perhaps this wouldn't include two cells completely fusing, but one cell transferring a copy of genome into the other.
Later, multicellularity could come around, and certain cells (with only one genome, instead of two--this is called "haploid") specialize in breaking away from the organism and fusing with other gamete cells to produce variation. This is what would be considered "isogamous." Eventually, the multicellular organism could produce tissues, which could produce organs, which could specialize in producing haploid gametes.
Providing care for one's offspring is also beneficial, wouldn't you agree? So, if an organism in a gene pool develops a pouch in which it's gametes could be kept, and gametes from other individuals could enter and fuse there, this would also increase the offspring's chance of survival, right? So, retaining gametes would be favorable to letting them loose in the cruel world.
However, if everybody retains their gametes inside themselves, how would the gametes find each other? Under these conditions, someone whose gametes can go out and find gametes that are retained would have an advantage. Thus, swimming gametes (sperm) are formed, going after the retained gametes (eggs), and we now have the two sexes that we are currently familiar with.
This may or may not be the way it actually happened, but it's reasonably close, at any rate, and it provides a simple background for the general process. This pattern of isogamy--anisogamy--oogamy can actually be seen as you follow certain branches of the evolutionary tree of the algae. I didn't propose any mechanisms, but the mechanisms are generally mutation and natural selection.
Feel free, other evolutionists, to add to, alter or clarify anything in this little scenario. Likely, there are other, sub-cellular factors to consider before the "two cells fusing" phase I listed first (such as lateral gene transfer, as mentioned a few times in this thread), but this is generally where our current knowledge is still in its infant stages.
Edited by Bluejay, : I wrote "genders" where I should have written "sexes": I did this for the sake of molbiogirl's sanity.
There was a point to this [post], but it has temporarily escaped the chronicler's mind. -modified from
Life, the Universe and Everything, Douglas Adams