Bill is right that plant selection is important to having a successful floating island. Wetland plants are most suitable for creating extensive root systems in water column because they are built to tolerate low oxygen conditions -- they have
aerenchyma (hollow) tissues that allow for gas exchange throughout the plant. Lettuce and other likely food plants aren�t tolerant of saturated conditions like wetland plants are and I think root dieback would be predictable without some method of aeration. System and plant choices are going to be determined by your objectives.
I like John�s foam island and think it looks pretty cool. It looks like he was able to get some petunias to do well in it and it turned into a decent turtle hang-out. That design stays moist by wicking and infiltration of water, but the roots are probably well separated from submergence. It�s a really different approach with different possibilities for plant choice, aesthetics, etc. than a typical floating wetland. Adirondack�s planter probably does have a fair amount of roots under water (haven�t seen pictures) because of plant choice and design differences. If you want to grow lettuce in something like adirondack's planter (I like the plastic fencing idea a lot!), or in what you (adam) described, aeration would probably be a big help.
I think the filtration John refers to above may have more to do with the microbiology that breaks down ammonia than with filtering particulates. I looked at the post he linked to about the lettuce growing experiment and I�m not 100% convinced of the conclusions there. (I don�t know much about aquaponics (yet!) besides the fact that they usually use circulation systems, so I had to research the filtration aspect of it.
Here's a decent article that seems to stress importance of �surface area� for microbial communities in aquaponics.)
The explanation of the problem in the lettuce experiment says that the roots weren�t getting enough oxygen because the solids (fish poo) were coating them. Solids probably do play a role in limiting oxygen to the roots, but I�d guess the dead roots in that trial just drowned as much as anything. If the water is too static, it�s not going to be as oxygenated as a re-circulating aquaponic system (RAS) would be. I wonder if aeration would be enough to overcome that root �coating� problem? I�d try experimenting by running a few feet of �� soaker hose (or and airstone) under one of those floating trays of lettuce, hook it to an (even dinky aquarium) air pump, and see if the roots and growth come out any different before I�d conclude that filtration is the main/only problem there. I could be wrong, but all the aeroponic, auquaponic, and hydroponic systems for horticultural products I�ve seen use some form of aeration and I think not having it was the biggest limiting factor for that experiment.
In any case, whether you are growing lettuce or wetland plants, floating plants and the associated biofilms on their roots are going to be more successful with aeration. Without it, you�re gonna have to adjust your plans and/or your expectations.
