Are there any check valves isolating these zones? If not, pressure differential from one active circulator creates some amount of reverse flow in the other zones, pulling from the return manifold toward the supply manifold. The amount of this flow is determined by the relative head of the active zones and inactive zones. If a zone has fairly low head relative to the active zone, can get enough induced flow to put some heat where it's not wanted when the overall heating load is fairly light (as it has been recently in Boston), even if it's pulling say, 110-120F water from the return manifold. Sometimes convection forces alone are enough to induce flow.
A system schematic would be useful here.
It may or may not be related, but EPDM tubing has a checkered history in the trade- when bleeding the apparently vapor-locked zone, was there any black gunk (or black water) evident?