Yes but the two are very related, a phd student will almost always publish papers with their advisor.
Exactly - the smallest ones often go through teacher/pupil, and since almost all PhD mathematicians studied under a PhD it makes for a relatively small group.