I think gazoo may be on to the right answer. I have been reading recently, on another site, that many troubles could be traced to actually doing a thorough gravel vac on anything but a tank that is running a UGF. I'm not sure I agree with their conclusions but it had some good arguments to it. The concept was that the sand or gravel gets a somewhat layered effect going where the beneficial bacteria in the substrate were concentrated in the top layer of gravel where there was more oxygen present and that when thoroughly vacuumed, the layer was no longer a layer but was mixed into the gravel. Like I said, it has just enough common sense to it to be believeable but I'm not sure I buy it. The idea of only doing part of the tank each time would ceratinly avoid the issue though.