Ducking is another option for sure, and Beeps' points are good (esp. the pumping concern). I tend not to use this often because a generic compressor is "frequency dumb" so dives the entire signal.
For separation, frequency (EQ), depth (volume (fader/ducking), reverb (more reverb is further away)), and panning are the three biggies. Typically panning is used more for higher frequencies and not for lowest (for a few reasons). Because of this, kick/bass are also "special" in that they are both commonly center-panned, which removes the "panning option" out of the picture as a solution, but Beeps hit upon the volume aspect.
SPAN will allow you to overlay up to 4 signals which is really nice but not intuitive to set up. Frequency masking is most prominent when signals are not separated by a couple dB or so, with the caveat (very important) of the most important component frequency getting priority. Using an audio microscope is helpful for this to find the material you want to keep so you have a mental record of the key frequencies to work with, then an overlaid frequency spectrum will reveal where they are ("frequency collisions").
I apologize for the basic presentation, but after commenting about keywords not getting dropped and making posts clearer for future folks who may find it I have been trying to adjust my own explanations accordingly.