De-nitrification, you remember, is the other "arc" that completes the nitrogen cycle, turning nitrates eventually to nitrous oxide or dissolved dinitrogen that is released to the atmosphere. Rapid rates of de-nitrification in bacterial metabolisms keep the nitrate levels in unpolluted natural waters vanishingly low. This is the regime we hope to imitate in our aquaria. I try to maintain nitrate at less than 20 ppm. Sometimes only a series of 50% water changes will bring down stubborn nitrate levels, but if nitrate builds up in a matter of a few weeks, it's a sign that your system is out of balance and you want to find out why.
Encouraging de-nitrification in the aquarium is part of the long-term answer. All surfaces in the aquarium offer a potential home to the community of aerobic bacteria that metabolize ammonia finally to nitrate. Researchers have determined that the mature biofilm on stem and leaf surfaces of water plants, which are substrates for bacterial nitrification, are also locally coupled with de-nitrification. The nitrifying process is stimulated in daylight by the oxygen diffusing from photosynthesing surfaces, in a gentle diurnal pulse that only lab experiments can detect. The uppermost surfaces of the substrate are also prime locations for these biofilm populations, as you know.
The nitrification process demands a lot of oxygen, more even than familiar cellular respiration. Only a few millimeters below the substrate's surface, or in the interstitial water of a mature biofilm, the diffusion of oxygen can't keep up with demand. As localized oxygen levels drop in a gradient, facultative anaerobic bacteria find their niche. "Facultative" in this sense merely means "opportunistic." Many ordinary bacteria are facultative anaerobes: when oxygen is in short supply, these kinds of bacteria are able to switch to a metabolism that doesn't require oxygen. Instead, some use nitrate. The familiar nitrating bacteria provide the nitrate, and their high oxygen demands also tend to exhaust the limited supply. So besides providing the nitrate, a thriving microzone of aerobic nitrifiers provides low-oxygen conditions too. You can visualize a mutually beneficial exchange between the two types of bacteria across a fluctuating boundary lying not far beneath surfaces. If there were no other reason not to disturb the substrate in an aquarium, this would be enough for me.
If plants are well-rooted in your substrate, their rootzone (rhizosphere) provides further edge microzones for de-nitrification. Oxygen transported through the vascular system of stems and roots diffuses into the spaces adjacent to each rootlet, supporting surrounding microzones of nitrification. Just beyond this aerobic network, in hypoxic porewater, bacterial de-nitrification occurs. The resultant nitrogen loss is a problem for rice-farmers but a boon to us with planted aquaria.
Chemical pathways to de-nitrification. The tribes of bacteria responsible for de-nitrification are diverse in their taxonomic groups and in their biochemistry too. In fact there are two different chemical pathways that provide energy for de-nitrifying bacteria. In one, some of them metabolize nitrate to nitrous oxide (NO), harmless, colorless non-reactive "laughing gas." It dissolves in water and is also carried to leaf surfaces of submerse vascular plants, and one way or the other it finds its way back to the atmosphere, completing the nitrogen cycle.
In the other chemical pathway, members of this anaerobic community metabolize some nitrate back to nitrite or ammonia. And if the nitrogen in those plant-available forms still hasn't been scavenged by the root hairs of plants, eventually yet other members of the anaerobe community produce molecules of gaseous di-nitrogen (N2). Chemically inert and harmless, dissolved in water and diffused back into the water column, the molecules of nitrogen gas also eventually escape into the atmosphere. You see why this "other arc" of the nitrogen cycle is called "de-nitrification." Diana Walstad says (in The Ecology of the Planted Aquarium, 1999, p. 65), "for aquarium hobbyists, de-nitrification is a harmless bacterial process that helps prevent nitrate accumulation."
A freshwater plenum. The idea of encouraging this bacterial community and applying to planted freshwater aquaria the reefkeepers' "Jaubert" system, with a low-oxygen area in the lower substrate (the "plenum") where de-nitrification proceeds, seems to have been kicked around first by Australian David Aiken, posting to the Aquatic-Plants Digest in the winter of 1997. His post is worth searching out if you have planted tanks and dare to reduce your sponge filters and bio-wheels, and let the plants handle the ammonia instead. A further less optimistic 1999 post from Roger Miller is also worth reading. To supplement my few thoughts in the filtration pages about a freshwater "plenum" try googling "freshwater-plenum".