Pillar
The Nitrogen Cycle, and Why Half of Everything Leads Back to It
Most of the problems on this site are the nitrogen cycle wearing a different hat. Cloudy water in a new tank, nitrite that will not fall, an ammonia reading two days after a filter clean — these are all the same story told from different points in it. This page is the story. It is worth reading once properly, because after that a great many aquarium symptoms stop being mysterious and become a question of which week you are in.
What is actually being built
The nitrogen cycle in an aquarium is not a chemical process you perform. It is a population of bacteria you grow, and the tank is the habitat. Two groups matter. The first converts ammonia — the continuous output of every fish and every scrap of decaying food — into nitrite. The second converts nitrite into nitrate. Neither group is added by you in any meaningful quantity; they arrive on their own from the air, the water, a plant, a piece of gravel, and then multiply to whatever size the food supply supports.
That last clause is the whole thing. Bacterial populations size themselves to the waste stream. They do not grow in anticipation of fish you have not added yet, and they shrink when the waste stops. This explains almost everything that seems arbitrary about cycling: why adding six fish at once to a tank that comfortably held two causes an ammonia reading, why a tank left unfed for a fortnight has to re-establish itself, and why there is no way to hurry the process that does not involve importing an already-grown colony from somewhere else.
They also live almost entirely on surfaces rather than in open water. Filter media, substrate, decor and glass — anywhere with flow over it. Very little of your biological filter is in the water you pour down the sink during a water change, which is why water changes do not slow a cycle, and why what you do to the filter media matters enormously.
The three stages, and what the kit shows at each
Cycling looks like three overlapping waves rather than three separate events. Ammonia rises and then falls. Nitrite rises later and falls later. Nitrate begins accumulating once the second wave is under way and then does not stop. If you test both ammonia and nitrite in the same session, the pair of numbers places you on that curve accurately enough to know whether to wait or to act.
The timings in the table below are typical for a room-temperature freshwater tank without seeded media, and they vary a great deal. Warmer water speeds the whole thing up. Water below about 70 °F slows it noticeably. A pH that has drifted low slows it too, because nitrifying bacteria work poorly in acidic water — a tank whose buffer has run down can stall its own cycle and look like a mystery.
| Stage | Typical timing | Ammonia | Nitrite | Nitrate |
|---|---|---|---|---|
| Stage one — ammonia climbs | Roughly week 1 to week 2 | Rising, often 1 – 4 ppm | 0 ppm | 0 ppm |
| Stage two — nitrite appears | Roughly week 2 to week 4 | Falling toward 0 ppm | Rising, often 1 – 5 ppm | Just detectable |
| Stage three — nitrate only | Roughly week 4 to week 8 | 0 ppm | Falling to 0 ppm | Climbing steadily |
Stage one — ammonia climbs
Roughly week 1 to week 2
Waste is entering the water and nothing is processing it. Whatever ammonia source you are using — fish food left to decay, a dose of pure ammonia, or fish themselves — is accumulating unopposed while the first bacterial group finds the filter media and begins to multiply.
What to doKeep the ammonia source steady and keep the water moving through the filter. Resist the urge to change anything. If there are fish in the tank, change water whenever ammonia goes above 0.25 ppm and accept that this stretches the timeline.
Stage two — nitrite appears
Roughly week 2 to week 4
The first group has grown enough to convert ammonia as fast as it arrives, so the ammonia reading collapses. The second group, which eats nitrite, is slower to establish, so nitrite piles up in the gap. This is the longest and least satisfying part of the process.
What to doNothing clever. Keep feeding the tank its ammonia source and wait. This is the stage where people conclude something has gone wrong and start adding products, and almost always what was actually needed was another ten days.
Stage three — nitrate only
Roughly week 4 to week 8
The second group has caught up. Both intermediate readings now sit at zero and the nitrogen you have been adding for weeks is arriving at the end of the chain as nitrate, which nothing in an ordinary tank removes. From here on, nitrate is your job rather than the bacteria's.
What to doConfirm it before you trust it: dose ammonia to about 2 ppm and test twenty-four hours later. Zero ammonia and zero nitrite the next day is a finished cycle. Then do a large water change to bring the accumulated nitrate down, and stock gradually.
Cycling with fish in the tank, and cycling without
A fishless cycle means you supply the ammonia yourself — with a bottle of pure ammonia, or by dropping in fish food and letting it rot — and let the readings do whatever they do without anything alive being harmed by them. It is faster, because you can hold ammonia at a level that grows bacteria quickly instead of a level that fish can survive, and it is entirely uneventful. You test, you wait, you test again.
A fish-in cycle means the fish are producing the ammonia and living in it. It is slower and more work, because you now have two competing goals: keep the readings low enough not to harm the fish, and keep them high enough that the bacteria have something to eat. In practice this means testing daily and changing water whenever ammonia or nitrite goes above about 0.25 ppm. Every one of those changes extends the timeline, which is the honest trade.
If you are already in a fish-in cycle, you are not doing it wrong — most people arrive here that way, because the shop sold the tank and the fish on the same afternoon. Test daily, feed lightly, change water on the readings rather than on a schedule, and add nothing else living until both intermediate numbers hold at zero.
What bottled bacteria actually do
A bottled nitrifying bacteria product is an attempt to sell you the colony instead of making you grow it. The good ones shorten the process meaningfully; none of them make it instant, and a bottle that has been warm on a shelf for a long time may contribute very little. Treat one as a head start with an uncertain size, not as a switch.
The practical consequence is that adding a bottle changes nothing about how you should behave. You still test daily. You still wait for both ammonia and nitrite to read zero. You still stock gradually. What changes is that the wait may be two weeks rather than six — and you will only know which by testing, which you were going to do anyway.
A far more reliable shortcut, if it is available to you, is borrowing established media: a used sponge, a handful of gravel, or a filter cartridge from a healthy tank you trust, moved across and kept wet. That is a real transplanted colony rather than a hopeful one. It also carries whatever else was in that tank, so it is a favor to accept from a tank you know.
The one thing that undoes it all: rinsing media under the tap
Your municipal water is disinfected on purpose. Free chlorine or chloramine is maintained in the distribution system specifically to kill bacteria, and it does not distinguish between the bacteria the utility is worried about and the ones you spent six weeks growing. A sponge held under a running tap for thirty seconds can lose most of its colony, and the tank will read ammonia two days later with no visible cause.
The same logic applies to replacing everything at once. Cartridge-based filters encourage this — the packaging says replace monthly, and the whole cartridge goes in the bin with the colony attached. In a filter that has room for more than one piece of media, stagger the replacements so an established piece always remains. In a filter that holds only one cartridge, run a piece of foam or a second cartridge alongside it for a few weeks before you swap, so there is somewhere for the bacteria to be.
What to do instead is unglamorous and takes two minutes: during a water change, fill a bucket with the water you just siphoned out, swish the media in it until the worst of the sludge is gone, and put it straight back. It will not look clean. It is not supposed to.
What a water change does at each stage
There is a persistent belief that water changes slow a cycle. They do not remove the bacteria, because the bacteria are on surfaces rather than in the water column. What a water change does is lower the concentration of whatever is dissolved, and whether that helps or hinders depends on which stage you are in.
In stage one, a water change lowers ammonia. If there are no fish, that removes the food supply and does genuinely slow things down, so do not bother. If there are fish, do it anyway — the fish come first, and a longer cycle is a price worth paying.
In stage two, a water change lowers nitrite without touching the first bacterial group at all, since ammonia is being consumed as fast as it appears. This is the safest and most useful moment for water changes in a fish-in cycle: you are protecting the fish from the reading that hurts them most, and the colony that is still growing has plenty to eat regardless.
In stage three, once both readings hold at zero, a large water change is the correct finish. Weeks of accumulated nitrate come down in one go, and the tank starts its ordinary life at a sensible baseline rather than at the top of the sawtooth.
How long it takes, and what makes it take longer
Four to eight weeks is the ordinary range without seeded media. Two to three weeks is achievable with a good bottled product or borrowed media. Anything faster than that is either a tank that was seeded heavily or a test result being misread.
Things that genuinely stretch it: water below about 70 °F, a pH that has drifted into the low sixes, a filter that is not running continuously, an ammonia source that keeps stopping and starting, and — the most common of all — changing something every few days out of impatience. The cycle is one of the few situations in this hobby where doing less is doing more.
Things that do not stretch it, despite the folklore: water changes in stage two, a light left off, an air stone added, or a substrate vacuum. None of these remove a meaningful share of the colony.
How to know it is finished
Two readings at zero on one afternoon is not a finished cycle. It is a snapshot, and it can happen by accident after a water change. The test that actually settles it is a load test: add ammonia to roughly 2 ppm, note the time, and test both ammonia and nitrite twenty-four hours later. If both read zero, the tank processed a full day of waste in a day and it is done.
On a fish-in cycle you cannot dose ammonia, so the equivalent is three consecutive days of zero ammonia and zero nitrite with normal feeding and no water change in between. That is less rigorous but it is the honest version of the same question: can the tank keep up with what it is actually being given?
Then stock slowly. A finished cycle is sized to the waste stream that finished it. Adding a full stocking list to a tank that cycled on a pinch of fish food a day will produce an ammonia reading within a week, and the tank has to grow into it either way.
What breaks a finished cycle
An established tank is fairly robust, and the things that knock it over are a short list. Rinsing or replacing all the media at once. A filter left off for many hours, which suffocates the colony in place. A medication or treatment that is unkind to bacteria — read whatever you put in the water for that, and if a treatment is involved at all, that is a conversation for an aquatic veterinarian rather than for this page. Untreated tap water going in during a water change. And a long unfed absence, which shrinks the colony to match a waste stream of nearly nothing.
The tell is always the same and it is always the same two tests. Any ammonia or nitrite reading in a tank that has been reading zero for months means the biological filter has lost capacity, and the question is which of the items above happened in the last week. Test, change water, and work backward.
A note that belongs here rather than anywhere else: if fish are behaving abnormally — gasping at the surface, sitting on the bottom, clamped fins, visible marks — test the water immediately, because bad water is the most common cause and the one you can fix in an hour. But the water test is where this site's competence ends. Diagnosing and treating a sick animal is veterinary work, and the right next call is to an aquatic veterinarian, not to a bottle from a shelf.
Where this site stops
Everything above is about water and filtration. If an animal in the tank is unwell, test the water immediately — bad water is the most common cause and the one you can fix in an hour — and then speak to an aquatic veterinarian. AquaReefBasin does not diagnose fish disease, does not recommend treatments, and publishes no medication dosages anywhere.