Maybe you could call it conduction at the microscopic level, but I always thought conduction was a solid to solid thing, not between the surface of the water bottle and the individual air molecules. In any event, convection is carrying the warmed-up air molecules away from the surface of the bottle so I'm pretty sure this would be called convective cooling.

Regardless of the terminology, we know what we're talking about here and yes, humid air might have a percent or two of gaseous water vapor that dry air doesn't have with somewhat different heat transfer characteristics than the nitrogen and oxygen in the dry air. But I don't think there is any mechanism for this to have a large affect - even if the water vapor were dramatically different, the heat transfer characteristics of the mixture would be dominated by the main constituents, not trace gases.

For the OP's observation to be caused by humidity differences, you'd have to assume that humid air was perhaps 50% more efficient in air-cooling the bottle. If this were the case, I'd think we'd have a lot of literature discussing how air-cooled engines (as one example) in dry climates are much more prone to overheating than in wet climates (or vice versa - I'm still haven't seen something that says water vapor is better/worse than transferring heat compared to nitrogen/oxygen). Other than some fringe esoteric papers, I can't find anything on the net that talks about atmospheric humidity having a big role convective heat transfer. Humidity matters to creatures that sweat, but it doesn't matter a bit to things that don't (e.g., engines or closed water bottles).

All in all, the idea that a difference in humidity is the cause of the OP's observation just doesn't pass the common sense test.

- Mark