Ammonia in a cycled tank: which side of the balance moved

Ammonia turning up in a tank you thought was cycled is alarming, but it is almost never a mystery. A working cycle means your filter bacteria consume ammonia at least as fast as your livestock produce it, so the reading sits at zero. When a number suddenly appears, one side of that balance has moved: production jumped, or capacity dropped. In freshwater, anything above 0.1 mg/L total ammonia is outside the target band and above 0.25 mg/L is acutely dangerous — but the fix is almost never to dose the tank. It is to work out which side changed and undo it.

A cycled tank has spare capacity, not infinite capacity. The colony is sized for yesterday's bioload, not for a sudden doubling of it. So the whole diagnosis is one question with two branches: did something start producing more nitrogen than usual, or did something just kill part of your filter? Both end with the same reading on the kit. Telling them apart is what tells you what to do next.

The seven usual causes, most common first

1. Something died and you haven't found it

This is the single most common cause of a sudden spike, and the easiest to miss. A dead fish wedged behind the filter intake, a snail that expired in the substrate, a shrimp that moulted badly and didn't recover — a decaying body is a concentrated nitrogen source dumping ammonia far faster than the colony grew to expect. In a small, lightly stocked tank one lost fish can drive the reading past 0.25 mg/L inside a day. Count heads. Look behind hardscape, under the intake, in the plant mass. If a body has been rotting for a day, your nose will usually find it before your eyes do.

2. You overfed, or the leftovers rotted

Every flake that isn't eaten breaks down into ammonia on exactly the same clock as a corpse, just spread thinner. A holiday auto-feeder that jammed open, a well-meaning house-sitter, a switch to a richer food, or simply a heavier hand for a few days — any of these lifts the daily nitrogen load above what the filter is sized for. The tell is a small, grumbling reading rather than a sharp spike, and food debris visible in the substrate. Feed less, and feed only what's gone in a couple of minutes.

3. You cleaned the filter too well

The bacteria that run your cycle live overwhelmingly in the filter, gripping the surface area of the media. Rinsing that media under the tap is the classic own-goal: chlorine and chloramine are antibacterial by design, and one cold-tap rinse strips most of the colony in a single wash. Replacing all the media at once, or swapping to a brand-new filter, does the same thing more slowly. Always rinse media in old tank water siphoned out during a water change, and never change more than half your media in one go. If you've just done this and ammonia appeared, you have your answer — expect a short mini-cycle while the colony rebuilds.

4. Medication knocked back the biofilm

Anything antibacterial aimed at fish disease can't tell a pathogen from your nitrifiers. Antibiotics (Furan-2, Maracyn), methylene blue at treatment strength, and some broad-spectrum treatments all knock the filter colony back, and the ammonia usually shows a few days into a course or just after it ends. Copper-based invert treatments are a similar hazard. Where you can, treat sick fish in a separate hospital tank and leave the display's biology alone; if you must dose the display, test daily and expect a recovery cycle lasting a week or two.

5. Chloramine in your tap water read as ammonia after a water change

This one is often a near-false alarm. Where the water utility disinfects with chloramine (chlorine bonded to ammonia — common in parts of the US and UK, rarer elsewhere), a conditioner that breaks the bond frees that ammonia straight into your tank. A good dechlorinator immediately converts it to the far gentler ammonium form and detoxifies it, but your test kit still reports it as total ammonia for a day or two. Some kits also react with the conditioner itself and throw a false positive. If a reading appears within hours of a water change and nothing else changed, suspect this first: retest a day later, and check whether your supply is chloraminated.

6. The filter ran cold or sat off too long

Nitrifying bacteria need oxygenated flow. Cut the power — a blackout, a pump you forgot to restart after maintenance, a filter left unplugged "for a minute" that became an hour — and the colony starts suffocating in a stagnant filter within a couple of hours. When you switch it back on, you push a slug of oxygen-starved water and dying bacteria into the tank, and ammonia climbs. The same applies to a filter that's been off during a house move or a media that dried out. If flow stopped for more than an hour, rinse the media in tank water before restarting and watch the numbers for a few days.

7. The pH crashed and the cycle seized

Nitrifiers slow sharply below about pH 6.5 and largely stop below 6.0. When KH drains away — nitrification itself consumes it, and a low-buffer tank drifts down between water changes — the daily CO₂ swing is no longer absorbed and pH can fall off a cliff overnight. The cycle doesn't die so much as freeze, and ammonia creeps up while nothing else has obviously changed. Check pH and KH together: a low reading on both, with rising ammonia, is the signature. Restore the buffer with a water change or a little baking soda, and the colony wakes back up.

The same reading, a different danger

Here is the part that makes ammonia confusing to reason about. Your test kit reports total ammonia, but only a fraction of it is the toxic uncharged form, NH₃. The rest is the far gentler ionised form, ammonium (NH₄⁺), which fish tolerate far better. The split is set by pH and temperature, and it moves fast — every step up in pH multiplies the toxic share roughly ten-fold. At 25 °C:

pH	Share present as toxic NH₃
6.0	~0.06%
7.0	~0.6%
8.0	~5.3%

So an identical 0.25 mg/L reading is close to a non-event in a soft, acidic tank at pH 6.5 and a genuine emergency in a hard, alkaline African cichlid tank at pH 8.2. Warmer water pushes every figure up too — at 30 °C the pH-8 share climbs to roughly 7–8%. This is also why "just raise the pH to help the fish breathe" is exactly the wrong reflex during an ammonia spike: lifting pH converts gentle ammonium into toxic ammonia. Fix the ammonia first, and don't chase pH while a number is showing. The NH₄ parameter page lays out the same math at more pH values.

Your action plan

Work down this list. It's ordered so the fastest, safest levers come first.

1. Confirm it's real. Retest with fresh reagent. If you did a water change in the last day or two and your tap is chloraminated, retest tomorrow before doing anything drastic — see cause 5.
2. Do a water change now. A 25–50% change with dechlorinated, temperature-matched water is the single fastest way to drop the reading. It dilutes the ammonia directly and buys the fish time. Repeat daily until you're back to zero.
3. Find and remove the source. Hunt for a body, scoop out uneaten food, and stop feeding for a day or two. You cannot filter your way out of an ongoing input.
4. Add a detoxifying conditioner. A dechlorinator that binds ammonia converts NH₃ to the harmless ammonium form for roughly 24–48 hours — a genuine safety net for the livestock while the colony catches up. It doesn't remove the nitrogen; it parks it.
5. Increase aeration. Warm water holds less oxygen, ammonia stress raises the fish's oxygen demand, and more surface agitation also off-gasses CO₂. An air stone or a lowered outlet costs nothing and helps on every front.
6. Leave the filter alone. Do not clean it, do not replace media, do not "help" it. If the spike came from a knock to the colony, it needs the surviving bacteria intact to rebuild.
7. Check pH and KH. If both are low, the pH crash in cause 7 is your culprit, and restoring the buffer is the actual fix.
8. Test daily and hold off new stock. Expect a short mini-cycle if the colony took a hit. Add nothing until NH₄ and NO₂ both read zero for several days running.

What reefnotes does with these numbers

reefnotes treats NH₄ and NO₂ as cycle-stability indicators, not dosing targets. There is no "suggested dose" for ammonia — the tank isn't short of it, and the only correct response is dilution and removing the source, never adding a product to a target band. So the app will never nudge you to dose these two, the way it does for a nutrient a planted tank is genuinely consuming.

What it does do is log them alongside everything else and flag when a fresh reading climbs past the 0.1 mg/L warn line, so a spike shows up as a clear step in the trend chart rather than a number you have to remember to eye. During a mini-cycle, daily entries turn "is it recovering?" into a line you can actually read. If you want to see the expected shape of each curve — ammonia first, nitrite trailing it, both settling to zero — the nitrogen cycle timeline tool sketches it against a calendar, and the plain-language cycle guide covers the biology underneath. A spike in a cycled tank isn't a failure. It's the cycle telling you something changed, days before the fish would.