Do variable aperture values represent the way in which depth of field is rendered by a lens or only to the transmission of light (i.e the brightness)?
For example, if i wish to calculate depth of field values for an 18-50 f2.8-4 lens when used at 50mm and with the aperture set to the largest it can be (which would be f4), should i use a value of f2.8 or f4?
update: background
I had always assumed that using a 18-50 f2.8-f4 at 50mm and f4 would mean a brightness of f4 and a depth of field equivalent to using any other 50mm lens at f4, however I started to think about this after seeing two other related SE questions and answers
This question discusses how to achieve maximum DOF with the sort of lens i’m talking about, and there is a statement that the variable aperture values are actually only apparent aperture - this is from a comment by @Matt Grum
What I mean is the solid round hole in the lens through which light passes doesn't grow or shrink when you zoom, hence you're not gaining anything by zooming out and using f/3.5 The relationship between the apparent aperture and focal length (the f-stop) does change. In fact the apparent size of the aperture increases when you zoom in, meaning using the 18mm end is the worst thing you could do for bokeh
(NB I don’t feel my question is a duplicate of the question i’ve linked to because that is a discussion considering multiple variable relating to depth of field, and also the quoted text above is in a comment, not an answer)
Secondly, i know from this question that with macro lenses when focussed close the ‘effective aperture’ falls (ie they get dimmer) but depth of field doesn’t change. So consequently, i wonder if the same could be true for variable aperture zoom lenses.
If possible, I’m after an answer that explains the optics of variable aperture lenses and why they behave in the way they do