For an ideal gas (and for our purposes, air is close enough), PV=nRT (where P is pressure, V is volume, n is the number of moles of gas, R is a constant, and T is temperature in degrees Kelvin). If the mast is sealed airtight, then n will not change. If we assume that the mast does not balloon up (a good assumption, I think), then V does not change either. As T rises, P rises proportionately. That's why there's more pressure when it's hot.

Let's look at it differently, though. Instead of holding n constant, we hold T constant. That is, we put a valve on the mast and pressurize it. Remembering that PV=nRT, as we add molecules of air, the pressure goes up proportionately. The volume doesn't change. Because we've added molecules (which have mass) without increasing the volume, the density of the mast has increased. Denser means less bouyant.

So technically, if you seal your mast on a hot day, then you'll wind up with fewer moles of gas inside than if you did on a cold day. That's because in this scenerio, P and V are constant (14psi at sea level and the volume of the mast). As T varies, so does n in inverse proportion. Fewer moles of gas means less mass and therefore less dense and more bouyant. I doubt you could measure the difference in any meaningful way however.

As for nitrogen making a difference, well I don't believe that either. If the mast is sealed, then the first scenerio above holds: as temperature rises, pressure does too. It doesn't matter what the gas is.

When it comes right down to it, your mast doesn't need to be airtight, just watertight.

Regards,
Eric