A lot depends on the depth and the strength of the roots.  Some roots cannot take much before they break. Some sedges and rushes can have very deep and dense root systems.   If you can find these grasses along streamsides, gullies road cuts, cliffs, or other surface elevation drops you might find partially exposed root systems and work to wash off the remainder with a power washer on one of the low spray intensities.  If you don't care about root damage, perhaps you can just excavate an area, estimate the number of rhizomes, or stems from areas that just contain one functional group.  You don't see many studies on this as it can be a lot of work.  If you find areas such as streams or rivers with exposed roots to the bottom, you may not know if the water table limited their depth.  If you try some of this is gullies, the paper I have on researchgate that discusses sediment from a small ephemeral gully has a filter fabric fence described and powerpoint that could catch sediment downstream from a small ephemeral gully or gully channel area so you could work and catch the sediment downstream before it travels too far off site.

Thank you for the suggestions. 

I had a co-worker at University if Missouri, David McNabb and he took an undisturbed sample of soil and broke it up enough to get the larger roots out, and then used soil sieves to separate soil from roots, much like an archeologist would do in separating out artifacts.  He picked out the fine roots with tweezers as best as he could.  I think he did this for tree roots.  I dont know if he measured the roots and classified their sizes, but sure he weighed them.  I don't remember if he washed the roots off after separating them from the soils or not.  

Sorry, I had another thought that relates to a large suction dredge mining to close along a stream bank that liquefied the bank with vibrations.  If you can get soils saturated and add a lot of vibration, they may liquefy, somewhat like puddling and perhaps allow the full plant to be pulled out, roots and all if they are not too fragile.  Sort of like a ultrasonic ring or instrument cleaner, the vibrations in the liquid break the particles of soil up.  I have seen wetland soils puddled from too much disturbance, it becomes like a slurry or soup instead of soil. 

Once again thank you very much for your suggestions