I'm not sure how it's being done, but Rockhounding Surface proves that this behavior is possible. The red liquid is wildberry juice, the other is wood aging bath.
I think that the Rockhounding fluids might be set to spread slower than the water can replenish itself. Thus preventing dehydration.

It doesn't actually seem to be an issue of speed, as my vertical honey test comparison just displayed.
I should mention that the Rockhounding fluids do in fact spread into the water. They get replaced by water and spread back again in a loop.

This behavior is actually intentional; I won't be changing it, sorry.