Lets look at some more of the details of this argument: The existence of recessives&such is quite bit more dramatic a bit of evidence than most people at first realize. It is not just that a line of linear descent can have variations with each 4th offspring. Because large populations mix, the net heritage of a bisexual population doesn't have to be reflected in every member. There can be recessives&such that are totally unavailable in any particular mating, that could still eventually be recaptured by one's distant offspring, through judicious subsequent matings.
To appreciate more fully what is going on here, you have to divorce yourself from the linear tree of descent creationists are so fond of arguing about. A modern bisexual species is a gene corporation, no one member of which is a fully empowered representative of, with a broader survival mandate than any one member represents.
The existence of dog breeds is a sort of butterfly-pinned-to-the glass picture of this fluidity in bisexuals, made obvious by the application for a few thousand years by humans of various diverse pressures nature only generally applies one at a time.
<sarcasm>How small is the gene pool of single celled organisms, and are they endangered?</sarcasm>