Long time lurker, first time poster.

I'm a reasonably experienced aquarist, now thinking about keeping discus for the first time.

In the course of my research, I came across David Bogert's site, where he talks at length about the role of biofilm filtration in reducing dissolved organic compounds and the number of microbes in the water column. I was initially sceptical, but I've spent an unhealthy amount of time reviewing the literature and have concluded that his theories on filtration seem sound.

Theory

A biofilm is a symbiotic colony of microorganisms that feeds on dissolved organic compounds, ammonia and nitrite, and microorganisms captured from the water column. It's the slippery film on the stones in a river, the gunk in the sediment in a wetland, and the sludge you clean out of your canister filter if you leave it long enough. Biofilms have "emergent" properties, in that the composition and relationship between organisms develops to meet whatever food sources are available - in one experiment, even pharmaceutical waste.

As to their practical applications: biofilms explain how slow sand filtration works (the oldest sewerage treatment method known the man), as well as the moving bed biofilm reactors (MBBRs) that have become widespread in water treatment, aquaculture, hydroponics and koi keeping in the last decade. They're reasonably well studied, albeit not in hobby aquaria (to my knowledge).

Why would you want a biofilm filter in your aquarium? Because microbes in the water column are bad for the health of certain fish, especially discus. An effective biofilm filter removes microbes from the water column by direct predation and resource competition (by consuming dissolved organic compounds released from fish waste and dead plant matter). It performs the same function as a UV filter, but far more effectively.

System design

There are several engineering problems with biofilm filtration in aquaria. First, the filter must not be cleaned. Second, it must not clog. Third, it must have a massive surface area (compared to what's required for nitrification alone). There are three aquarium-sized solutions I'm aware of; moving bed filters, Mattenfilters, and undergravel filters. My idea is to combine all three.

The main filter will be moving bed biofilm reactor using K1 micro in a sump (usually incorrectly referred to as a "fluidized bed" by monster fish and koi keepers) and an in-line UV filter. The sump chambers will be separated by Aquamesh.

The entire substrate will be a sheet of black Poret foam, on top of a single layer of very loosely packed sintered media to create a separate chamber, with glass airlifts - in other words, a cross between a Mattenfilter and an undergravel filter.

I haven't done the maths properly yet, but the combined surface area of the Mattenfilter substrate, the sintered glass media underneath it, the K1 micro in the sump, and the Aquamesh baffles should be far, far greater than is possible with canister filters.

Maintenance regime

Uneaten food will be left on the substrate for the biofilm to eat, maybe with a "turbo tank forcing filter" in any dead spots while the filter gets established.

Nitrate and phosphates will be assimilated by emergent plants growing from the sump to the nearby window (something very large and vigorous, like monstera).

The chamber containing the plant roots will serve as a refugium and be stocked with detritovores, like shrimp, to break down any larger particles that survive the K1 media chamber to and keep the roots clean.

Water changes will be fortnightly at most and involve changing only about 30-40%, by evacuating the sump using a Ultrazero utility pump and refilling from a Brute bin. (My water is high quality and quite soft, with only a little chlorine and no chloramine.)

The filters will never be cleaned, except for trimming roots and cleaning pipes and impellers occasionally.

Can you please share any practical or theoretical objections to this approach? (I'll experiment on a "pilot" scale with other blackwater fish first.)