• Part 1
• Part 2
• Part 4

Biofiltration or biological filtration is the part of the total filtration process causing the most heartache for newcomers to fish keeping, and which is too often ignored or forgotten by more experienced hobbyists - until something goes wrong. Biofiltration is specifically concerned with the oxidation (by at least two types of bacteria) of the toxic nitrogen-containing (nitrogenous) waste material produced by the normal metabolism of the resident fish. Fish release ammonia (NH3, dissolved ammonia gas or ammonium ion, NH4+) into the water primarily through their gills. This ammonia is one of the most water-soluble materials in the tank. Active transport of the material across the gill epithelium is not required. Simple diffusion from the area of higher concentration, the fish's blood, to the area of lower concentration, the tank water, is quite effective. Active transport of the ion is also available. Terrestrial (land-dwelling) animals must use other, more energy-requiring, mechanisms to rid themselves of this waste material. It is a given that any biologic creature's waste is to some (usually large) degree toxic to the producer. Otherwise it would not be waste, but an intermediate metabolite, that is, it would be used in further processes in the body. Other tank inhabitants contribute to nitrogenous wastes in the tank - bacteria, fungi, infusoria, acting on uneaten or partially to fully digested food, and decaying plants - all may add to the waste load, but we expect the majority to be from the fish. A possible exception to that could be in very heavily overfed tanks. Just remember that everything alive in the tank, fish, snails, shrimp, infusoria, and bacteria are all part of the bioload. Even if they are consuming wastes, they do contribute their own wastes to the total load.

We commonly refer to the development of the needed populations of the two bacterial types required to oxidize ammonia to nitrite (NO2-) and in turn nitrite to nitrate (NO3-) as "the nitrogen cycle". This is a bit of a misnomer. We are establishing only a small section of the nitrogen cycle in our tanks. In nature, nitrogen gas (N2) is the largest component of the air we breathe. It is effectively inert. Only highly specialized life forms can "fix" atmospheric nitrogen gas, that is, convert it from the gas (or dissolved gas) to a dissolved ion or incorporate it into a combined form, from which the rest of the life forms of our planet can use it. Generally these small creatures are referred to as bacteria, but this is a level where the taxonomists have loads of fun. Defining the classification of these little creatures can be a real trick. But there are lots of nitrogen-fixers in the world, both in terrestrial soils and in fresh and salt waters. They bring atmospheric nitrogen gas into biochemistry for their own use. Higher plants may get nitrogen from these "bacteria" directly, even cultivating them in specialized structures in their roots (legumes - the group including peas and beans), or they may get the nitrogen from the decay of the original "fixers". Animals could get their nitrogen from the original fixers directly by consuming them, but are more likely to get nitrogen from consuming higher plants, or by consuming other animals, which have consumed plants (grazers are always more numerous that predators). At each stage, the contained nitrogen may be recycled by other organisms to continue up or back down the food chain and the nitrogen cycle. These are not one-way paths. At each step, the nitrogen contained in the organism could be returned to the atmosphere if the organism decays under anaerobic conditions, so the path is not only not one-way, it branches and potentially short-circuits back to the atmosphere at any point.

One member of this group of nitrogen-fixers with which we may be all too familiar is what we call BGA, Blue-Green Algae. BGA is not really an alga, may or may not be a bacterium as conventionally defined, most likely is a direct descendant of one of the very early if not original life forms on our planet, and may have a "kingdom" (in taxonomic terminology) unto itself. In any case, it seems to be able to fix nitrogen gas directly. Unfortunately, I know of nothing that eats BGA to continue the nitrogen cycle from this source in our tanks, and most tank BGA does not appear to have the structures required to fix atmospheric or dissolved nitrogen.

A reduced (not oxidized) combining form of nitrogen is the nitrogen found in ammonia. When our fish release their metabolic ammonia in the water in our tanks, it is already several steps and organisms away from atmospheric nitrogen, so is far from the "beginning" of the nitrogen cycle. When the resident bacteria have oxidized that ammonia to nitrate, gaining energy for their own use in the process, we are still not at the "end" of the nitrogen cycle. The nitrogen could be recycled in our tanks if heavily planted and not too heavily stocked, or elsewhere if removed by water changes, plant trimming, or algae removal. To "complete" the nitrogen cycle, we would have to be using plenums or deep sand beds, certain sintered glass structures (including Ehfisubstrat or Cell-Pore) or other bacterial denitrification devices. These techniques provide an environment sufficiently low in oxygen to support those specialized denitrifying bacteria that use reduction (effectively the opposite of oxidation as used by our good guy filter bacteria) for an energy source. These creatures reduce nitrogen completely to its atomic rather than ionic form, thus produce dissolved or gaseous nitrogen and release the nitrogen back to the atmosphere. Having a complete nitrogen cycle in the broadest sense (from air, into bacteria, to higher plants, to animals, and by bacteria again, back into the air) in our tanks is highly unlikely. We are content to have oxidized the ammonia to nitrate. In tank terms, our "cycle" is complete when we have the right bacteria and enough of them to convert highly toxic ammonia through equally toxic nitrite to less toxic nitrate.

Bacterial and algal spores, resting, or vegetative forms are everywhere; they may even be carried by air currents. Unfortunately, the two species we want for oxidation of ammonia to nitrate have no spore forms. By the way, they are unlikely to be Nitrosomas species and Nitrobacter sp. as the standard references cite, but that discussion is out of scope here. For our purposes, the best source for these bacteria is an existing, mature ("cycled") tank. For the beginner, this is a big hurdle. They do not have an established tank yet. LFS (local fish stores) are singularly unhelpful on this. They may recommend commercial products such as prepared shelf-stored bacterial suspensions [entirely worthless IMHO (In My Humble Opinion)], sacrificial hardy fish (goldfish, especially disease-ridden feeders, or Danios), or even ignore the issue entirely. The novice, by definition having no standards by which to judge the purportedly "professional" advice they are given, and no experience by which to judge the health of the fish they are offered, is in deep water from the start. If they are computer literate at all, they may seek out or stumble across the Internet discussion boards and hear of "fishless cycling". This technique is (again IMHO) great for beginners, and was formulated into its present state and popularized by Nomad, Dr. Chris Cow: Considerations in Fishless Cycling

The process of fishless cycling not only saves fish, it teaches the newcomer multiple lessons, including: 1) the basic idea that what you cannot see can still hurt you. 2) How to do simple water tests on their tank and to use those data as the basis for tank handling. 3) A bit about concentrations and dilutions of toxins or other dissolved materials in the tank water. 4) And, as a very worthwhile sidelight, a bit about patience.

So, in some manner we have assembled a tank, with substrate (gravel or sand), a heater, a filter, and a cover, and probably a light. We have treated the water with a chemical to inactivate the chlorine or chloramines added by our local water system (and hopefully that chemical does not add lots of "extras" such as Aloe vera gel and other "slime coat enhancers" or what have you - personal opinion again). The temperature has been checked for stability, the filter for operation (positive water flow). We have added clear ammonia (if fishless cycling) or a small number of hopefully healthy fish (if conventionally cycling). We are going to feed very little (if we are using fish), and we are going to monitor the ammonia levels regularly. This means we also had to buy test kits for ammonia, nitrite, and nitrate. If we are using fish and as a general rule, we should have a pH test kit as well. Freshly drawn tap water may not give a stable or accurate reading for pH. If your source water is from wells (directly or via a municipal water system) or is highly pressurized, there is likely to be dissolved gas in the water at levels beyond air equilibrium or even saturation at room temperature. Tap water samples should be tested after standing overnight in a shallow open container, or after aeration for a shorter time. This treatment will allow excess gases to equalize with the room air. Excess CO2 dissolved in the tap water will lower the pH, as in solution carbonic acid is formed(H2O + CO2 H2CO3). After excess CO2 has vented or off-gassed, we can read the stable pH determined by the non-volatile mineral content of the water, especially by the carbonates and bicarbonates naturally present. The reason for this test is that ammonia itself is quite toxic, but its ionic form ammonium ion (NH4+) is relatively non-toxic. The form in which ammonia is found in the tank will depend upon the pH and temperature of the water. In acid conditions (pH <7.0), the majority of the ammonia present will be ionized as ammonium. At higher pH, a somewhat higher percentage of the ammonia present will be just that, ammonia, with greater toxicity. This is not a hard break. There is an equilibrium (between dissolved ammonia and ammonium ion), which in acid conditions favors ammonium ion, and as the pH increases on the alkaline side, favors higher and higher percentages of ammonia. We will of course carefully read the instructions accompanying these test kits. Those chemicals currently best and most often used to inactivate chloramines are incompatible with Nessler's reagent based kits. Salicylate based kits will be better. If we are using fish, we will not try to modify the pH of the water to make it more acid and "safer", but we will monitor and partial water change the tank to maintain the ammonia concentration in non-injurious ranges, or better, in undetectable ranges. Do remember that not detectable by hobby test kits does not mean there is none present, only that it is below detectable levels. This will avoid the long term to permanent gill damage to the fish from ammonia burns, or even the death of the fish (definitely permanent). In tanks with mature filters, any ammonia or nitrite reading at all by hobby test kits is not normal and should not occur. We expect to see positives during "cycling" however. The highest level seen - the ammonia or nitrite "spike" - is not an absolute term, but is dependent on the bioload (number and mass of fish per unit volume of the water, along with the amount of food added). To me, any positive ammonia reading can be considered a spike, as in a normal cycled tank there should be no positive ammonia or nitrite reading at all. In fishless cycling there is no concern with toxicity, as there are no fish present, and higher concentrations of ammonia in the tank can and will allow development of larger populations of ammonia-eating bacteria.

This ammonia provided by the fish, or from a bottle by the fishkeeper, is going to be "food" for the first good guy bacteria. These bacteria get energy for their own metabolism by oxidizing (adding oxygen to) NH3/NH4+, resulting in nitrite ion, NO2-. The nitrite is released into the water - it is a waste product for these bacteria. These bacteria also need other materials, which they can extract from the water, and they need oxygen. Their energy metabolism is aerobic (oxygen requiring), just like ours. The denitrification bacteria mentioned earlier are anaerobic (without air, literally), in fact being injured by oxygen at the levels the aerobic types require. The nitrifying bacteria also require a substrate, something to which they can firmly attach. Because of this requirement for attachment, they may also be referred to as lithotrophic (rock loving) bacteria, or lithotrophs. Where will they establish? Anywhere with a solid substrate and water flow or diffusion to bring food and oxygen to them. A bacterium's idea of solid is not quite the same as yours or mine. Gravel, sand, rocks, plastics, ceramics, the tank glass itself, etc, all obviously qualify. But so do sponges, floss, and quite possibly the surface of many things in the water. So, where will they live long term? Any, or all, of that list. Where will the greatest numbers of them live? They will live where the substrate is acceptable, and the water with food and oxygen is available most consistently. In my experience, those conditions are best met in filters with relatively stable water flow bringing constant exposure to any "food" and oxygen available in the tank. This is the case whether they are simple sponge filters, HOB/HOT (Hang On Back/Hang On Tank) units, canisters, FBFs (Fluidized Bed Filters), W/Ds (Wet/Drys), or UG/RFUGs (Under Gravel/Reverse Flow Under Gravel filters). So long as the filters are well maintained, such that the bacteria are not smothered by debris buildup captured from the water column, or by their own growth, or by secondary channeling (bypass) or flow reduction from such buildup, and the media in the filter is not totally replaced routinely (disposable cartridges), the filters should provide the best conditions in the tank. If the filters are overly dirty, the flow a small fraction of the clean flow rate, and the bacterial biofilm itself severely overgrown, then this may not be the most hospitable environment. And obviously, if you routinely discard the filter media, the bacteria are discarded with it. If you replace only one-quarter to one-third of the media per change, placing the new medium downstream of the old, you probably have enough reserve capacity to show no effect on filter functionality. This is another of the great category of "it all depends", and like many of such, what it depends upon is the level and nature of the maintenance practiced on the tank. You determine where the nitrifying bacteria live in your tank by the way it is set up initially and the way you keep it up week to week. If you maintain your filter well, routine rinses or vacuuming will not disturb the biofilter (used tank water is most conservative, brief rinses in tap water of matching pH and temperature should not hurt - but do not soak in untreated tap water). If you neglect the filter for extended periods and neither rinse nor vacuum, a later thorough cleaning may reduce nitrification functionality. This is under hobbyist control, not the filter's inherent design.

Back on the ammonia processing, once ammonia eaters are in the tank and attached somewhere suitable, they will multiply if given ammonia and oxygen. As is the case with snails, or guppies, or most organisms, the population or mass of these bacteria is controlled first by the food and oxygen supply. Higher availability of food (ammonia), with sufficient oxygen, will result in larger populations. This is the principle underlying fishless cycling. In cycling with fish, the bacterial population is matched to the fish population. If you want to increase the fish population, the bacterial population to support them may or will lag behind a bit, depending on the proportional increase in the bioload. To expand on that a bit, if you have 20 one-inch Tetras in your cycled and stable tank and add 2-3 more, you are unlikely to be able to detect any re-cycling. That is because the percent increase is likely to be less than your own variation in food amounts from meal to meal. In the same tank situation, had you added 10 more fish, a reduction in feeding (every other day?) for the first few week or couple of weeks would be conservative, to allow the bacterial capacity to increase to match the load. In fishless cycling initially we use such high concentrations that the tank could not support that number of fish in normal operations, but we can stock fully in the beginning with a comfortable margin. It is safer and more conservative to have the good guy bacteria die back to operating levels than to have them be required to multiply to operating levels while exposing the fish to increased ammonia and nitrite. Even at mature or cycled levels, the bacterial population is not static. Bacteria receiving sufficient food and oxygen will not just metabolize, they will multiply. The steady state population is still a constantly changing and renewing population, not a static one, but maintaining approximately even total numbers.

With ammonia at detectable levels, ammonia eaters will divide and increase their numbers until ammonia is no longer at detectable levels from fish and other bioload. From once per day or every other day additions by the tank-keeper using fishless techniques, ammonia may be detectable anything from briefly up to several hours or days initially. The ammonia is converted to nitrite, also quite toxic to fish. The increasing nitrite concentration provides energy food for the second type of good guy bacteria, which get energy by oxidizing the nitrite to nitrate. These bacteria follow exactly the same population dynamics as the first type, expanding to meet the available food supply.

A possible snag in fishless cycling is that during the high-ammonia period, if there was not a significant original population of ammonia eaters, not enough nitrite would be produced to support a functional population of nitrite eaters. This means that any original inoculation could well have died/starved by the time nitrite is present in detectable levels. I believe that if the ammonia phase prior to nitrite detection lasts more than ten days to a couple of weeks, then re-inoculation may be needed to introduce the bacteria in sufficient numbers. But I have not rigorously tested this for confirmation, so call this personal opinion and in my experience again.

Once the nitrite eaters have stabilized, nitrite as well as ammonia should be undetectable. Nitrate should be present and may be at high levels with fishless cycling, requiring a large volume water change prior to introduction of the fish. Nitrate is an order of magnitude (or more) less toxic short-term than ammonia or nitrite. This does not mean that it is not toxic long-term, it is. When the tank is "cycled", it should no longer show any intermediates (ammonia, nitrite), only a gradual increase in nitrate concentrations. Those are most often controlled by partial water changes. The nitrate level should be monitored periodically to insure that partials are maintaining suitable levels of water quality. If you are being very cautious, KH (carbonate hardness, or alkalinity) would be tested as well. Nitrification is an acid-producing process, and can reduce carbonate/bicarbonate buffering if this is not renewed regularly through partial water changes. In very soft water, there can be risk of pH "crash" (slow decline followed by rapid drop in pH when buffering/alkalinity is exhausted). Such crashes have also been reported in fishless cycling in soft to very soft water when high ammonia levels are used.

Once the filters are cycled, that is, they can convert all the ammonia produced in the tank to nitrite and on to nitrate without detectable levels of either being seen, the tank is biologically pretty safe. But it is really not yet mature. There are numerous other bacterial types, and algaes, and fungi, and infusorians, all in balance in our tanks. Collectively they eat many different dissolved and particulate wastes and or byproducts. These are discussed a bit here on this site as well: Aquarium Microbes Part 2:

Planted Tanks: If a newly set tank is heavily planted (75% to 100% of the substrate covered by plants when viewed from above) and allowed settling in time (from less than one week, up to more than three weeks IME) sufficient to see positive growth prior to introduction of fish, the scenarios discussed above may or may not apply. If the fish stock is added slowly, not too many at once, the plants may well fully short-circuit the conventional cycle. This is frequently referred to on the boards as a "silent cycle". Aquatic plants prefer their nitrogen source to be ammonia, and they can remove this from the water quite rapidly. But remember that they must be healthy, growing plants. Arrested or degenerating or dying plants cannot and will not do this job. In fact, such plants are likely to contribute more to the waste load than they remove. If a healthy heavily planted tank is lightly stocked with fish and inverts, nitrogen in the form of nitrate may have to be added to maintain the growth of the plants. This balance will have to be determined for each tank; there is no way to generalize this. Nitrate is used, as it is not as hazardous to livestock as ammonia or nitrite would be. Plants, which must expend a little energy to reduce nitrate to ammonia, can in this way utilize nitrate. They cannot mimic the denitrification bacteria, whose reactions gaining energy from a similar process are anaerobic. This is not a serious or damaging expenditure. It is just slightly less efficient.

Lower planting densities and less rapidly growing plants will have lowered ammonia absorption, and such tanks will rely upon bacterial oxidation of ammonia with normal fish loads. The exact proportions again are indeterminable other than by observation, but will be related to the rate of growth inherent to, and evidenced by, the plants. That is, plants capable of very rapid growth, when growing rapidly will take up more nutrients than slower growing plants - many stem and floating plants are far "better" (more rapid) nutrient sponges than slower growing plant such as Anubias or Java Fern would be.

Next: Part 4: Summary, Putting it all together, and keeping it there.

Robert T. Ricketts, a.k.a. RTR

Main index Aquarium Supplies Navigation Badmans Tropical Fish! Fishkeeping ABCs Favorite Fish Fish Families Detailed Information In Depth Definitions Natures Way The Equipment Whats That Mean? Want More Just For Fun Guestbook Site Search, What's New