The aquarium nitrogen cycle, sourced

What the cycle actually is

The aquarium nitrogen cycle is biology, not magic. Fish excrete ammonia (NH₃ / NH₄⁺) across their gills as the main nitrogenous waste; uneaten food and decomposing organic matter add more. Ammonia is toxic above ~0.05 ppm NH₃ (the un-ionised form, which dominates at higher pH). Two functional groups of chemoautotrophic bacteria convert ammonia into nitrate, which is dramatically less toxic and which plants, macroalgae, and corals can use as a nitrogen source:

1. Ammonia oxidisers (AOB / AOA) convert NH₃ → NO₂⁻. Energy yield is low; growth is slow.
2. Nitrite oxidisers (NOB) convert NO₂⁻ → NO₃⁻. Energy yield is even lower; growth is slower.

Both groups need oxygen, a surface to colonise (rock, sand, sponge, ceramic media), and trace nutrients — phosphate, in particular, is required. The biofilm grows until it can clear the daily ammonia load to near zero within 24 hours. At that point the tank is “cycled.”

The hobby's longstanding shorthand was “Nitrosomonas handles ammonia, Nitrobacter handles nitrite.” That is half right. The half that's wrong matters when you pick a bottled bacteria product.

Stage durations

The fishless ammonium chloride cycle without a bottled inoculant runs roughly:

• Days 1–7 — lag phase. Ammonia rises; nitrite stays near zero. The starter community is establishing.
• Days 7–14 — ammonia → nitrite. AOB populations explode; ammonia begins falling and nitrite climbs.
• Days 14–28 — nitrite → nitrate. NOB catch up. Nitrite peaks and crashes.
• Days 21–42 — cycle complete. A 2 ppm ammonia dose clears with no nitrite peak inside 24 hours.

Numbers from DrTim's Aquatics fishless cycling guide. Saltwater systems generally run on the slower end of this range — 4 to 8 weeks — because the higher pH means more of the ammonia is in the toxic NH₃ form and the bacteria growth is suppressed by the higher concentrations of unionised ammonia at the start. Freshwater commonly finishes in 3 to 4 weeks per Fritz Aquatics.

Nitrospira, not Nitrobacter — the correction the hobby took 30 years to absorb

In 1996 and again definitively in 1998, Tim Hovanec and colleagues at the DeLong lab applied molecular biology — 16S rRNA gene probes — to actual aquarium biofilms instead of the laboratory cultures the hobby had been extrapolating from. The result is the most important paper most reef-keepers have never read: Hovanec, Taylor, Blakis & DeLong (1998), “Nitrospira-Like Bacteria Associated with Nitrite Oxidation in Freshwater Aquaria,” Applied and Environmental Microbiology.

The Nitrobacter-targeted DNA probe gave zero positive signals in any aquarium tested. The Nitrospira-targeted probe lit up freshwater and saltwater aquaria alike, and the appearance of Nitrospira ribosomal DNA correlated temporally with the disappearance of nitrite. The dominant nitrite-oxidising bacterium in aquariums is not Nitrobacter. It is Nitrospira.

This matters for bottled-bacteria products. Older formulations from various manufacturers claimed Nitrobacter content, which is fine for sewage treatment plants but does not establish a working aquarium biofilter. The products that actually work — Dr. Tim Hovanec's own One & Only, Fritz Aquatics' TurboStart 900 — contain aquarium-derived Nitrospira isolated from real working tanks. The marketing claim “contains Nitrobacter” is, by the lab evidence, marketing for the wrong organism.

A 2015 discovery added a footnote: certain Nitrospira strains — “comammox” complete-ammonia-oxidiser strains — can do both steps of nitrification by themselves, going NH₃ → NO₃⁻ in a single cell. A 2024 study in Applied and Environmental Microbiology found comammox Nitrospira was the most abundant ammonia oxidiser in freshwater RAS at low ammonia loads. It has not been detected in marine RAS as of that work. The take-home: if your freshwater tank shows ammonia falling and nitrite never rising, comammox is a real explanation, not a malfunctioning test kit.

Fishless cycling protocols

Dr. Tim's ammonium chloride + One & Only

Pure ammonium chloride solution gives a clean, dose-able ammonia source with no organic load to confuse the test results. The current Dr. Tim's formula (post-2016) is calibrated to 4 drops per US gallon → 2 ppm NH₃-N (ammonia-nitrogen, not total ammonia). Pre-2016 bottles used 1 drop per gallon — verify which formula you have. Source: DrTim's Aquatics.

Protocol:

1. Fill tank with mixed saltwater (or dechlorinated freshwater) at target salinity and temperature.
2. Dose 4 drops of Dr. Tim's ammonium chloride per gallon. Confirm 2 ppm NH₃-N with a test kit.
3. Dose one bottle of One & Only per recommended tank size. Skimmer off, UV off for 24 h.
4. Test daily. Re-dose ammonia to 2 ppm only when both ammonia and nitrite read zero.
5. Do not let ammonia or nitrite exceed 5 ppm — inhibits the nitrifiers you just paid for.
6. Cycle complete: 2 ppm dose → both NH₃ and NO₂ ≤ 0.2 ppm inside 24 h.

Fritz TurboStart 900 (saltwater)

Fritz's TurboStart 900 Saltwater contains live Nitrosomonas-class AOB and Nitrospira-class NOB cultured for saltwater conditions. The Fritz product page specifies:

• Dose: 1 oz (29 mL) per 25 US gal (95 L) of system water. Up to 5× this dose is documented safe.
• Claimed cycle time: 5 days or less when ammonia source is added simultaneously.
• Turn off skimmer, UV sterilizer, and fleece rollers for at least 5 days post-dose.
• Ideal conditions: 77–86 °F, pH 7.3–8.0, salinity 15–40 ppt, KH ≥ 4.5 dKH, trace phosphate must be present.

The phosphate requirement is the one most people miss. RO/DI-only saltwater with no fish and no fish food will read 0.00 ppm PO₄, and the nitrifiers cannot build cell structures without phosphorus. A pinch of fish food after dosing the ammonium chloride solves this.

Seachem Stability and others

Various general-purpose live nitrifying blends exist. Seachem Stability instructions call for 1 cap (5 mL) per 40 US gallons on day 1, then 1 cap per 80 gallons for 7 days; we couldn't verify this from the Seachem product page directly during this round of research and flag it as such. The category leaders for verifiable performance remain Dr. Tim's and Fritz.

Saltwater: dry rock vs live rock cycle timing

How fast a saltwater tank cycles depends almost entirely on what came on the rock.

Live / uncured rock arrives carrying nitrifiers, coralline, sponges, micro-fauna — and a lot of dead organics from the trip. Curing means letting that die-off run in a bare vessel with strong flow and frequent water changes (2–4 weeks) before the rock goes into the display. During curing the rock cycles itself; once it's in the display the cycle is effectively continuous. The risk: imported pests (Aiptasia, vermetids, predatory flatworms) come with the biodiversity.

Dry rock (Marco, Real Reef, BRS pukani, CaribSea LifeRock) arrives sterile. No die-off, no ammonia spike on day one, no pests. You run a full fishless cycle — 3 to 8 weeks — but you start with a clean slate. BRS's 52 Weeks of Reefing Week 10 episode states the case plainly: “pest free dry rock is probably the best.” That has been the consensus position in the hobby since roughly 2014.

The hybrid play — a 90 % dry rock build with one 5 lb piece of cured live rock to seed it — gives you most of the biodiversity benefit and most of the pest-avoidance benefit. Worth it.

Cycle-complete criteria

A tank is cycled when:

• You dose ammonia to a known concentration (Dr. Tim's protocol: 2 ppm NH₃-N).
• Twenty-four hours later, both NH₃ and NO₂ ≤ 0.2 ppm.
• Nitrate is detectable and rising over the cycle — this confirms the full pathway, not just ammonia disappearing into the void.

The 0.2 ppm threshold is the limit of detection on most hobby test kits. Source: DrTim's fishless cycling protocol.

If you want a numerical first-fish stocking density once the cycle is complete, the stocking calculator works against your verified tank volume. Don't overstock day one; the biofilter is sized for the load it cycled against, and you just cycled against 2 ppm ammonia per day — not a tank-full of fish.

Common cycling mistakes

Ghost-feeding

Throwing fish food into the tank as “food for the bacteria” was popular in the 2000s and is wrong. Fish food decomposes into ammonia, ammonia spikes above 5 ppm, nitrifiers get inhibited, the cycle stalls. Dr. Tim's warns explicitly: do not let ammonia or nitrite exceed 5 ppm. Use pure ammonium chloride if you want a clean ammonia source.

Antibiotic interference

Erythromycin, metronidazole, kanamycin, tetracycline — any antibiotic in the water column during cycling will kill the bacteria you're trying to grow. If you treated for ich or finrot before the cycle finished, the cycle restarts from zero after the antibiotic course ends. Run quarantine in a separate hospital tank, not the display.

pH crash below 7.0

Nitrification produces H⁺ — every NH₃ → NO₃⁻ reaction releases two protons. Those protons eat carbonate alkalinity. In a freshwater tank with low natural KH, or a saltwater tank that wasn't buffered up before cycling, pH can crash below 7.0 and nitrification slows to a halt. The fix is a 25 % water change with properly buffered water, which simultaneously restores pH and dilutes any accumulated nitrate / ammonia. From DrTim's: “If the pH gets too low, however, the nitrification process will stop.”

Phosphate = 0 in an RO-only fishless cycle

Nitrifying bacteria need trace phosphorus to build cells. An RO/DI saltwater tank with no fish food added will read 0.00 ppm PO₄, and the cycle simply doesn't complete. Fritz's product page calls this out directly. A pinch of flake food, or 0.02 ppm of intentionally dosed PO₄, fixes it.

Adding chlorinated water mid-cycle

Tap water with chlorine or chloramine added straight to the tank as a top-off kills nitrifiers on contact. Always dechlorinate top-off water or use RO/DI.

Trusting bottled bacteria with the wrong organisms

Per the Hovanec / DeLong work, any bottled product claiming Nitrobacter as the nitrite oxidiser is selling the wrong bug. Stick with products that specify aquarium-derived Nitrospira — Dr. Tim's One & Only, Fritz TurboStart 900. Refrigerated storage helps; both products are live and degrade above room temperature.

Skipping the test

“It smells like it's cycled” is not a cycle test. Run an API or Salifert test at 24 h after an ammonia dose, read both NH₃ and NO₂, and only add livestock when both are ≤ 0.2 ppm.

Freshwater and the Walstad approach

Diana Walstad's Ecology of the Planted Aquarium (2023 ed.) argues that heavily-planted freshwater tanks with a fertile soil substrate bypass the conventional bacterial cycle: plants take up ammonia directly, faster than nitrifying bacteria can grow, and the soil layer hosts an established mixed-microbial community from day one. Walstad's position is documented at dianawalstad.com.

This is not a free pass. The Walstad approach requires:

• Heavy planting from day one — not “a few stems.”
• A real soil substrate (Walstad uses MGOCPM, Miracle-Gro Organic Choice) capped with sand or fine gravel.
• Low to moderate stocking, slow to add fish, and ammonia testing before each new addition.

It works. It is also not what most beginners are doing when they ask “do I need to cycle?” The answer for a typical freshwater community tank without dense planting is yes, cycle it the conventional way, and use a bottled Nitrospira inoculant if you want to compress the timeline.

After the cycle

Once cycled, the tank's biofilter is sized for the load it cycled against. Add livestock in steps — one or two fish a week, not a stocking blitz — and re-test ammonia and nitrite for 48 hours after each addition. The biofilter scales up to the load over a week or two; outpacing it gives you a mini-cycle and a stressed fish.

Once steady, nitrate becomes the variable to watch. Targets are 2–10 ppm for a reef per Randy Holmes-Farley's Optimal Parameters for a Coral Reef Aquarium. For freshwater, < 20 ppm is the standard healthy band; planted tanks pull this down on their own, low-tech tanks need 25 % weekly water changes. The nitrate parameter reference covers test kits, drift symptoms, and target ranges in detail.

From here, parameter management is a steady-state problem, not a startup problem. Welcome to the boring middle, which is where every successful tank lives.