The nitrogen cycle explained

Almost every "my fish keep dying and I don't know why" story is this one thing. A new tank has no bacteria in it, so the waste the fish produce has nowhere to go, and the fish sit in it. Cycling is the process of growing the bacteria first, so the tank can handle waste on the day the fish arrive rather than three weeks later.

What the cycle actually is

Fish waste, uneaten food and decaying plant matter all break down into ammonia. Ammonia burns gills and is lethal in small quantities — well under 1 ppm is enough to do lasting damage. Left alone it just accumulates.

Two different groups of bacteria deal with it, in sequence:

1. Ammonia-oxidizing bacteria consume ammonia and excrete nitrite. This solves nothing yet — nitrite is also acutely toxic, and it stops fish blood carrying oxygen properly.
2. Nitrite-oxidizing bacteria consume nitrite and excrete nitrate, which is on the order of a hundred times less toxic. Fish tolerate tens of ppm of it without obvious harm.

Nitrate is where the biological process ends in a normal tank. Nothing removes it for you at any meaningful rate, so it climbs until you take it out — which is what a water change is for. That is the whole loop, and the last step is the only part you are responsible for.

The bacteria are not in the water

This is the single most useful fact in this guide, and most of the practical advice falls straight out of it. Nitrifying bacteria form biofilms on surfaces. The overwhelming majority live in your filter media, where water is forced past them continuously; the rest are on substrate, glass, rock and decor. The population actually suspended in the water column is negligible.

So:

• Water changes do not reset a cycle. Change as much water as you like, whenever ammonia or nitrite is high enough to worry about. You are removing the toxin, not the bacteria.
• Never rinse filter media under the tap. Chlorine kills biofilm on contact and this genuinely can undo weeks of progress. Rinse it in a bucket of water you just siphoned out of the tank.
• Never replace all your media at once. Cartridge-based filters that tell you to swap the whole cartridge monthly are telling you to throw away your cycle. Replace part of it, or add new media alongside the old for a few weeks before removing anything.
• A cycled filter survives a move. Keep the media wet and reasonably cool and it will be fine for hours. This is how you move a tank without starting over.
• Seeding works, and it is the real shortcut. A used sponge, a handful of established media or some mature substrate from a healthy tank transplants a working population instead of growing one.

How long it takes

Three to six weeks is the honest answer for a fishless cycle started from nothing at normal room temperature. Four things move that number, and only these four are worth your attention:

• Temperature. These bacteria multiply far faster warm. Running the tank at 82–86°F (28–30°C) during a fishless cycle is the cheapest speed-up available, and there are no fish present to mind — worth more than three times the wait between a cold room and a warm tank.
• pH and KH. Nitrification runs well above pH 7 and slows sharply below about 6.5. It also consumes alkalinity as it goes, so a long cycle erodes its own buffer — see stalls.
• Oxygen. Both groups are aerobic. Run the filter, keep some surface movement, and do not "rest" the filter overnight.
• A steady ammonia supply. The bacteria are food-limited. No ammonia means no growth — which is why an empty tank of clean water does not cycle no matter how long you leave it running.

Bottled bacteria products are the obvious question, and the honest answer is mixed. Some contain live cultures of the right organisms and can shorten a cycle substantially; many contain the wrong species, or the right ones dead on arrival after a hot warehouse. Treat a bottle as something that might help rather than something that replaces testing, and never stock fish on the strength of the label alone.

The 24-hour test

There is one unambiguous test, and it is worth doing properly because everything else is guesswork:

1. Dose ammonia up to roughly 2 ppm — the fishless cycling calculator gives the milliliters for your volume and bottle strength.
2. Wait 24 hours. Do not change water in between.
3. Test both ammonia and nitrite.
4. If both read zero, the tank is cycled. If either shows anything at all, it is not — wait, and test again in a few days.

The reason to insist on both is that the two halves establish at different speeds. Ammonia-eaters colonize first, so it is completely normal to hit zero ammonia in a week and then watch nitrite sit stubbornly high for another fortnight. Zero ammonia on its own means you are roughly halfway.

One last step before fish: nitrate will be high after a cycle, often well over 40 ppm, because you have been manufacturing it for a month. Do a large water change to bring it down first.

What the numbers do, week by week

Timings vary a great deal, so treat the weeks as a rough sequence rather than a schedule. The order is reliable even when the pace is not.

Why cycles stall

A cycle that has not moved in two weeks nearly always has one of these causes, roughly in order of how often it turns out to be the answer:

• A pH crash. Far and away the most common. Nitrification consumes carbonate hardness, so weeks of it can strip a soft-water tank's KH to nothing and drop the pH below 6.5, where the bacteria nearly stop. The symptom is textbook: nitrite frozen at a high reading for weeks. Fix it with a large water change, or a small amount of baking soda to restore KH, and progress usually resumes within days.
• Chlorine or chloramine. Untreated top-off water, or media rinsed under the tap. Dechlorinate everything that goes in, every time.
• Too much ammonia. High concentrations inhibit the very bacteria you are growing. Dose to 2 ppm, not to 8 "to be safe" — more is actively slower.
• Medications. Anything antibacterial will hit your filter as hard as the pathogen you are aiming at.
• Cold water. An unheated tank in a cold room can take months rather than weeks.
• A test kit problem. Before you troubleshoot the tank, rule out the kit: reagents expire, and the second bottle of the common liquid nitrate tests needs shaking far harder and longer than people expect or it reads low.

If there are already fish in there

Plenty of people find this page after being sold a tank and fish on the same afternoon. That is a fish-in cycle, it is harder on the animals, and it is entirely recoverable. Do this:

• Test daily for ammonia and nitrite. You are managing a moving target, so yesterday's reading is not good enough.
• Change water whenever either exceeds about 0.25 ppm, and change enough to matter — a 50% change halves the concentration.
• Use a conditioner that detoxifies ammonia, not just a plain dechlorinator. It converts ammonia to a far less harmful form for a day or so, which buys time between water changes without starving the bacteria you are growing.
• Feed lightly. Every meal is more ammonia. Fish are fine being fed sparingly for a few weeks; they are not fine swimming in nitrite.
• Do not add any more fish until the 24-hour test passes.

The water conditioner dose calculator will give you the exact dose for your product and volume, including the larger dose most brands specify for an ammonia emergency.

Things you will be told that are wrong

• "Run the tank for 24 hours, then add fish." This is printed on a lot of starter kits and it is the single most destructive piece of advice in the hobby. Twenty-four hours lets the water reach temperature and the chlorine disperse. It grows no bacteria whatsoever.
• "A water change will restart your cycle." The bacteria are on surfaces, not in the water. This myth causes real harm, because it talks people out of the water change that would have saved their fish during a nitrite spike.
• "Cloudy water means the tank is cycling." Cloudiness in a new tank is a bacterial bloom — fast-growing heterotrophic bacteria feeding on dissolved organics. Different organisms, unrelated process. It clears on its own and tells you nothing about your nitrite.
• "Plants mean you don't need to cycle." This one is half true, which is why it persists. Plants take up ammonia directly and prefer it to nitrate, so a heavily planted tank stocked lightly and slowly really can be safe far sooner. A few stems in gravel cannot. If you are relying on plants, you still test — you are just likely to be pleasantly surprised.

Common questions

How long does it take to cycle an aquarium?

Three to six weeks for a fishless cycle at normal room temperature. Seeding the filter with media from an established tank can cut that to a few days, because you are transplanting the bacteria rather than growing them.

How do I know when my tank is cycled?

Dose ammonia to about 2 ppm and wait 24 hours. If both ammonia and nitrite read zero the next day, the tank is cycled. Ammonia alone reaching zero is not enough — the nitrite-eating bacteria are the slower half to establish.

Does a water change restart the nitrogen cycle?

No. The nitrifying bacteria live on surfaces — filter media, substrate, glass and decor — not suspended in the water. You can change water freely during a cycle, and you should if ammonia or nitrite is high enough to harm fish.

Why is my nitrite stuck and not dropping?

Usually a pH crash. Nitrification consumes alkalinity, so a long cycle drives KH down; below about pH 6.5 the bacteria slow dramatically and can stop. A water change restores buffering and typically restarts progress within days.