It isn't the origin alone of the tourists that matters (although it is of critical import that they are potential sources of non-native virulence), it is their impermanence. All of them are potential attack vectors for disease. Each of them might communicate a disease to immunocompromised herd members, instituting epidemic, and then leave, no longer contributing to the herd - except as a vector.
Since epidemic isn't simply an event that happens at an instant, it isn't the total number of tourists who are present at any one time that is the number of attack vectors weakening the herd that is relevant, but rather the total number of tourists that impact the herd that count for that purpose.
Were the tourists to stay for long periods of time, the rate of immunocompetence of the tourists would impact the development of epidemic. Since they do not, it is the increase in attack vectors alone that comprises their impact on herd immunity.
In the event of epidemic, only those tourists present at the time, a calculation impossible to make absent more information, or blind estimations, would be included in the herd.
I concede this does modify my original statement.
Thanks!