The usual (high school or intro to bio) explanation for diversification of species comes from multicellular, usually sexually reproducing organisms, and seems to be closely tied to the biological species concept. It is that once organisms are different enough, they will be unable to exchange genes and therefore gradually become more distantly related.
This is not really applicable to species of bacteria which can exchange genes through horizontal gene transfer (obviously not all can, but many can). It seems to me that if HGT rates are low relative to mutation rates, the same process of diversification could take place. However if they are high (and random), it seems like either species would not exist as discrete entities or would exist only because the environment was only favorable to specific collections of genes.
Is there a commonly accepted reason why species should exist among species which have horizontal gene transfers (ie is there a reason to expect that HGT rates will usually be very low), rather than bacteria existing as a continuous gradient?