In the open air (RH 15%), very little water is pulled out of the air after two days.  No water was dripping into the planter.  You can see below there is only a few dribbles of water.

 

After 24 hours

After closing the tank and adding two inches of water to the bottom, there is over 200ml of water pulled from the air.  It is important for me to remember to run the experiments in the environment they are designed for.  It's a hard lesson to remember that the outside biosphere (namely the earth) is different than an enclosed system.  Here in CA the air temperature is always the same, but there are parts of the world where the RH is higher and more similar to what is in the biosphere I'm making.

Water can be observed to move horizontally with the sifted soil.  The water reached the sand and is stopped.  When the plants take more water, it can be seen soil drying out near the sand layer.  This soil layer is acting like a water reservoir.  

In the past with random soil particle sizes, water would rush through the bottom and everything else becomes dry.  Very few plants could survive unless it was right on the dehumidifiers.  In that case, too much water would rot the plant.   This method seems to be promising.  When mixed with some compost or other aggregate, it can be a good water storage vessel.

Perennial Veggies

Below is sorrel.  This is being planted in the beaker as part of the test.  It is a perennial vegetable.  That means it does not die back after the summer is over.  It grows all year around.  A lot of Asian vegetable plants are good for growing in a farming tank.