Starting the Cycle

"Cycling" a tank means growing the bacteria that turn ammonia into nitrate. Until those bacteria are established, everything a fish excretes accumulates as ammonia and then as nitrite, both of which are toxic. The biology behind that conversion is covered in Nitrification; this page is about the practical question that comes first. You have a new tank and a bucket of water. What can you do on day one to make the wait shorter?

Grown from nothing, a nitrifying community takes roughly a month to establish. Almost every keeper tries to shorten that, and the methods on offer differ enormously in how well they work. The ordering below is what the evidence supports, and it is also what the simulator encodes.

What a head start actually moves

Nitrifying bacteria are not free-floating cells. They live in biofilm, anchored to glass, sand, plant leaves, and filter media by a matrix of sticky polymers they build themselves. The reasons are covered under why nitrifiers are biofilm dwellers: a surface offers shade from light, shelter from grazers, and a richer supply of ammonia than the open water. Del'Duca et al. (2019) followed a tilapia recirculating system by fluorescence microscopy and found biofilm bacterial abundance climbing continuously while water-column abundance did not change significantly over the same period.

That single fact ranks every method on this page. A method is worth what it carries in attached bacteria, so moving surfaces works, moving water does not, and a bottle sits somewhere in between depending on what actually survived in it.

Media or substrate from a running tank

Taking filter media, gravel, or a handful of substrate from an established aquarium moves the biofilm itself, complete with the ammonia oxidisers, the nitrite oxidisers, and the comammox Nitrospira that come to dominate mature filters. Nothing else on this list transfers the whole community.

The hobby rule of thumb is that about a third of a donor filter's media carries about a third of that filter's load, and it holds up as a way to think about the transfer. The transferred fraction can handle immediately whatever it was already handling, so an instant cycle is real only when the new tank's bioload sits below that. A cupful of substrate or a single sponge squeeze is a smaller dose and behaves like one.

Two caveats belong with the method. The first is that this is a transfer, not a gift: strip a third of a running filter and the donor tank can go through a small cycle of its own for a few days. That is a second aquarium, so the simulator does not model it, but it is worth planning around. The second is that the bacteria need feeding. A transferred biofilm starves at the same rate it would in the donor tank if it arrives to an empty aquarium, so the head start decays unless ammonia is arriving from somewhere (see feeding a new cycle below).

Substrate, gravel, and pond mud sit just below filter media and above everything else. In the aquariumscience.org bucket comparison, brown gunk squeezed from an established sponge, composted cow manure, and garden soil all cycled a 5-gallon bucket in single-digit-to-low-teens days while controls and eleven bottled products took several times longer. Those absolute figures come from someone else's buckets and should be read as an ordering rather than as a schedule, but the ordering matches every peer-reviewed source on this page.

Live plants from an established tank

Plants grown in an established aquarium arrive coated in the same biofilm that covers everything else in that tank, and that biofilm nitrifies. Eriksson and Weisner (1999) and Körner (1999) both measured active nitrification on submerged macrophyte surfaces. A bag of stem plants is therefore a genuine, if partial, inoculum, and it is one most planted-tank keepers get without asking for it.

The 2025 ISME Communications survey of three home aquarium biofilters is the strongest recent support. The two tanks that were not dosed with any commercial supplement, and that did have live plants, cleared ammonia by week 2 and nitrite by week 3. Comammox Nitrospira had come to dominate the ammonia oxidisers in all three tanks by weeks 6 to 12, which is the succession the simulator reproduces from competition rather than by script.

The simulator treats a planted start as a light seed on the nitrogen-cycle guilds only, roughly five times the near-sterile tap floor. Heterotrophs, algae, and microfauna recover quickly from any start, so they are left at the floor; the nitrifiers are what changes the shape of the cycle. The result is a nitrite spike that arrives earlier and peaks lower than a bare tank's.

Bottled nitrifying bacteria

This is the most popular method and the one with the weakest evidence behind it.

Scagnelli et al. (2022) is the only peer-reviewed test. Six naïve 10-gallon aquaria were dosed to 1.0 ppm ammonium hydroxide, five quick-start products plus one untreated control, with total ammonia, nitrite, and nitrate measured every 48 hours for 14 days. Four of the five products produced no statistically significant decrease in total ammonia. The fifth, Tetra SafeStart Plus, did, and took the full fortnight to move 1.0 ppm down to 0.29 ppm. The authors concluded that these products should not be recommended as a sole method of establishing a biofilter. In the ISME Communications survey, the one tank of three that received a commercial supplement was the slowest, still above 750 µg/L ammonium at week 7.

There is a second, more specific problem. Hovanec and DeLong (1996) and Hovanec et al. (1998) established that the nitrite oxidiser in freshwater aquaria is Nitrospira, not the Nitrobacter of older textbooks. Nitrobacter is nonetheless what much of the industry cultures and names on the label, because it is the organism that grows in a fermenter, and Nitrospira is harder to keep viable on a warehouse shelf. So a bottle is more reliable at delivering the ammonia oxidiser than the nitrite oxidiser. Comammox are worse still: they are slow-growing specialists that colonise biofilms over weeks, and no product delivers them.

The simulator encodes that asymmetry directly, seeding ammonia oxidisers 30 times above the tap-water floor and nitrite oxidisers three times below the ammonia oxidisers, with comammox left at the floor. What emerges is a change in timing rather than in magnitude. The ammonia oxidiser head start pulls the whole cycle forward, so ammonia clears early and the nitrite phase begins about ten days sooner, but the nitrite peak is no smaller and the nitrite phase lasts longer, because the nitrite oxidisers are still behind. The tank finishes a few days ahead of one that was left alone.

That is the most common complaint levelled at these products, and it is worth noticing that the simulator was not told to produce it. The seed densities come from the literature above and the behaviour follows from the bacteria competing.

"But my bottle worked"

This is a fair objection and it deserves a direct answer rather than a defensive one, because for a large number of keepers the bottle really did coincide with a fast, uneventful cycle. Several things explain that without either side being wrong.

• The bottle was rarely the only variable. A tank that got a bottled dose usually also got plants, a handful of substrate, or media from a friend, and any of those carries far more biofilm than the bottle does.
• Products differ, and one of the five tested did work. Scagnelli et al. found a real effect for Tetra SafeStart Plus. The finding is that most products in the category do nothing measurable, not that the category is uniformly inert.
• The carrier may be doing some of the work. Hovanec has argued that the modest effect some products show comes from the liquid medium feeding the bacteria already resident in the tank, rather than from the bacteria in the bottle.
• An uneventful cycle is not the same as a fast one. Ammonia and nitrite both spike and clear on their own in most lightly stocked new tanks, and a keeper testing every few days can easily miss a nitrite peak that lasted a week.
• The simulator agrees the bottle helps. It shortens the cycle in the model. The disagreement is over how much, and over what happens to nitrite in the meantime.

If you want to see the argument rather than read it, run the same tank twice, once started with a bottled dose and once with fresh tap, and compare the two runs side by side.

Water carried over from another tank

Carrying water over from an established aquarium is one of the most widely held beliefs in the hobby, and for the nitrogen cycle it does essentially nothing. This follows directly from the biofilm result at the top of the page. The bacteria are on the surfaces, and a bucket of donor water leaves the surfaces behind.

The simulator makes that claim explicitly rather than by omission: the ammonia-oxidiser, nitrite-oxidiser, and comammox seeds for carried-over water are set to exactly the same values as for fresh tap water, and the two runs produce the same cycle length and the same nitrite peak to three significant figures. If a future change to the model ever gave this method a head start, the test suite fails rather than quietly letting the claim slide.

What donor water genuinely does carry is the plankton. Heterotrophic bacteria, green algae, rotifers, ciliates, and nanoflagellates all live suspended and travel with the water, so those guilds start several times higher than they would from tap. Some suspended mulm settles and seeds the floor. Run the two starts side by side and the difference is visible in the water column and absent from the nitrite curve, which is a more useful thing to see than either run alone.

Fresh tap water, nothing added

Municipal water arrives with a chlorine or chloramine residual that kills most of what it touches, so a tank filled from the tap and dechlorinated starts close to sterile. Nitrifiers then arrive slowly, from the air, from your hands, from anything you put in the tank, and the community builds from a very small base. This is the full ammonia and nitrite cycle that a new keeper expects to watch, and it is the simulator's default because it is the honest starting assumption for a tank built from scratch.

Chlorine does not remove the two guilds evenly. Nitrite oxidisers are more chlorine-sensitive than ammonia oxidisers, by a factor the literature puts somewhere between 3 and 10 (Holler and Bachofen 1969; Regan et al. 2003; Wahman et al. 2009). The simulator uses the conservative end of that range. It is one of two reasons nitrite lags ammonia in a new tank, the other being that nitrite oxidisers cannot start growing until the ammonia oxidisers have made them some nitrite.

Feeding a new cycle

A seed is only half the problem. Nitrifiers grow on ammonia, so a tank with a strong inoculum and no ammonia supply does not finish cycling, it just holds a small starving population. There are two ways to supply it.

Fishless cycling doses ammonia directly, and the simulator supports it through the scenario's nitrogen dosing settings. It is the cleaner experiment: nothing is at risk, and the dose is a number you control.

Fish-in cycling uses the fish as the ammonia source, which is what a stocked tank does anyway. Every nitrogen atom in the food you add ends up as ammonia, so the load tracks the feeding rather than the fish count (see Fish and Feeding). It works, and it is how most tanks are actually started, but the fish are living inside the experiment and both ammonia and nitrite are toxic to them.

Two environmental controls decide whether either version finishes at all. Nitrification runs best around pH 7.5 to 8.5 and effectively stops near pH 6, and it has a thermal optimum near 28 to 30 °C. Between them these are behind most stalled cycles, and both are covered under why cycles stall. A cycle that has stalled on pH will not be rescued by any amount of seeding.

How long will it take

Less precisely than most sources will tell you. Cycle length depends on the tank volume, the ammonia load, the pH, the temperature, the amount of surface available, and whether the tank is planted, and those factors move the answer by more than the choice of starting method does.

The numbers below are one comparison on one rig, held constant across the four arms: a bare 60 L tank with no plants and no fish, dosed to 2 mg N/L whenever ammonia cleared, at 26 °C and pH 7.8. They are useful for the ordering and the spacing between methods. They are not a schedule for your tank.

How far those numbers travel is worth being blunt about. The same simulator, given a heavily planted tank rather than a bare one, puts the nitrite peak somewhere around week 7, because a planted tank's plants, substrate, and larger surface area change the ammonia supply and the places bacteria can live. Move the pH down a unit and the whole table stretches. The ranking of the methods is robust; the calendar is not.

What the simulator does with your answer

The starting method you pick in the setup wizard scales the initial population of every microbial guild in the tank. Each method is a set of multipliers on per-species seed densities, applied once when the scenario is built, and from that point the simulation is ordinary competition between bacteria. Nothing about the cycle is scripted: the nitrite spike, its timing, and the eventual comammox takeover all emerge from the guilds' different growth rates and affinities.

Every multiplier, and the source behind it, is tabulated under Cycling stage in the Parameter Reference, including the sweep that set the bottled tier's ammonia-to-nitrite-oxidiser gap and the two mechanisms that were measured and deliberately left out.

Further reading

• Nitrogen Cycle — where nitrification sits in the full journey of nitrogen through the tank
• Nitrifying Bacteria — the three guilds, their competition, and why they live on surfaces
• Filters — what a filter contributes to cycling, and why flow matters more than media volume
• Fish and Feeding — the ammonia load a stocked tank generates
• Biofilm Maturity — how a young biofilm becomes an old one, and what changes when it does
• Starved Biofilters and Dormancy — what happens to a cycle you established and then left alone
• Parameter Reference — every seed multiplier on this page, with citations

Key references

• Del'Duca, A. et al. (2019). Variability of the nitrifying bacteria in the biofilm and water column of a recirculating aquaculture system for tilapia (Oreochromis niloticus) production. Aquaculture Research 50, 2229–2238.
• Eriksson, P.G. & Weisner, S.E.B. (1999). An experimental study on effects of submersed macrophytes on nitrification and denitrification in ammonium-rich aquatic systems. Limnology and Oceanography 44, 1993–1999.
• Holler, S. & Bachofen, R. (1969). Influence of chlorine on nitrifying bacteria. Schweizerische Zeitschrift für Hydrologie 31, 67–72.
• Hovanec, T.A. & DeLong, E.F. (1996). Comparative analysis of nitrifying bacteria associated with freshwater and marine aquaria. Applied and Environmental Microbiology 62, 2888–2896.
• Hovanec, T.A., Taylor, L.T., Blakis, A. & DeLong, E.F. (1998). Nitrospira-like bacteria associated with nitrite oxidation in freshwater aquaria. Applied and Environmental Microbiology 64, 258–264.
• Körner, S. (1999). Nitrifying and denitrifying bacteria in epiphytic communities of submerged macrophytes in a treated sewage channel. Acta Hydrochimica et Hydrobiologica 27, 27–31.
• Microbial community succession of home aquarium biofilters associated with early establishment of comammox Nitrospira. ISME Communications 5(1), ycaf212 (2025).
• Regan, J.M., Harrington, G.W., Baribeau, H. & De Leon, R. (2003). Diversity of nitrifying bacteria in full-scale chloraminated distribution systems. Water Research 37, 197–205.
• Scagnelli, A.M., Javier, D., Mitchell, M.A. & Acierno, M.J. (2022). Efficacy of quick-start nitrifying products in controlled fresh-water aquaria. Journal of Exotic Pet Medicine 43, 20–24.
• Wahman, D.G., Wulfeck-Kleier, K.A. & Pressman, J.G. (2009). Monochloramine disinfection kinetics of Nitrosomonas europaea. Applied and Environmental Microbiology 75, 5555–5562.
• Aquarium Science, 2.11 Inoculate for Cycling and 2.8.1 Bacteria in a Bottle in Depth (hobbyist bucket comparisons; cited for ordering, not for absolute cycle lengths). https://aquariumscience.org/index.php/2-11-inoculate-for-cycling/