I agree with Bob on the importance of testing but I'm going to diverge a bit from there. First do Bob's tests, which are critical, then consider the following as building on that:
Most (but not all the time) in doing landscapes one wants max DOF - as much as possible of the scene in focus. If I have content in the foreground, midground, and background, I usually want all of it to be in sharp focus. If I have to make a sacrifice, I will usually accept a bit of de-focus to the background, since this seems natural to the eye, but not to the foreground.
Simplistically, I could simply use the min aperture setting on my lens and work from there. But the problem is the trade-off between DOF blur and diffraction blur. The larger the aperture the greater the DOF blur and the smaller the diffraction blur and vice versa. There is a cross-over point at which using a smaller aperture gains you nothing, because diffraction blur overrides the improvement in DOF. This will vary by focal length, because the actual size of the aperture at a given f/stop varies according to focal length, and quite possibly from lens to lens. I work with the same 35-70 zoom most of the time and found that f/11 gives me the most sharpness from foreground to background; i.e. represents the best compromise between DOF and diffraction.
Everyone needs to do his/her own tests on this over all the different focal lengths and lenses he/she employs. My own tests were to set up my camera on a tripod in front of a static scene with objects at every distance from a few feet to the horizon then take shots at all focal length and aperture combinations. Be sure to read Michael's article on MLU and cable releases first. Such a test will be a great time to practice minimizing all vibrations.