An empty filter does not cycle

The first tank I “cycled” sat humming in a corner for a week. The filter ran, the heater glowed, and the water became impressively clear. A shop employee had told me that seven days was enough. On the eighth day I added fish. A few days later they were hanging just below the surface, and the ammonia test had turned a green I had not yet learned to fear.

Nothing useful had happened during that quiet week. I had circulated dechlorinated water through clean media, but I had supplied almost no nitrogen for a nitrifying community to eat. The tank was a week older and no more ready than the day I filled it.

“Cycling” is hobby shorthand for establishing reliable biological nitrogen processing. Fish excrete ammonia through their gills, and uneaten food, dead leaves, and waste release more as they decompose. In water, un-ionized ammonia (NH₃) exists in equilibrium with ammonium (NH₄⁺). The balance depends strongly on pH and temperature; the NH₃ fraction rises as either rises. Aquarium tests often report the two together as total ammonia nitrogen, which is why a result needs context rather than a color alone.

A mature biofilter prevents that waste from accumulating. A new one cannot. Running the pump provides oxygenated surfaces and flow, both valuable later, but bacteria cannot multiply into a working population without an energy source. Time is involved, but time is not the mechanism.

The familiar diagram is right, but incomplete

The useful classroom version has two steps. Ammonia-oxidizing microorganisms turn ammonia into nitrite (NO₂⁻); nitrite oxidizers turn nitrite into nitrate (NO₃⁻). Both steps consume oxygen. Nitrate is generally tolerated at much higher concentrations than ammonia or nitrite and can be exported through plant growth and water changes.

Older aquarium explanations assign the first step to Nitrosomonas and the second to Nitrobacter. The names make a tidy diagram, but studies of freshwater aquarium biofilters found Nitrospira-like organisms, not Nitrobacter, associated with nitrite oxidation. More recent work complicates the story again: ammonia-oxidizing archaea, plus “comammox” Nitrospira that can carry ammonia all the way to nitrate on their own, can be important members of aquarium biofilters. Nitrosomonas/Nitrospira-type is therefore a useful shorthand, not a complete census.

A cycle is not something the water possesses. It is work done by a community living mainly on wet surfaces.

That distinction changes how I handle a tank. Nitrifiers grow in biofilms on filter sponge, ceramic media, substrate, wood, glass, and plant surfaces. Relatively few are drifting usefully in the water column. Replacing all filter media does not “freshen” a mature aquarium; it can discard much of its biological capacity. Conversely, moving seasoned sponge or ceramic media into a new filter can transfer something real. A bag of old tank water transfers much less.

These organisms are slow growers compared with the bacteria behind the white haze in a new tank. That cloudy bloom is usually a population of fast-growing heterotrophs responding to dissolved organic material. It can coincide with cycling, but it is neither proof of nitrification nor a finish line. Only the tank’s handling of an ammonia input tells us that.

Why I no longer ask fish to test a filter

A fish-in cycle uses living animals as both the ammonia source and the alarm. When a test shows ammonia or nitrite, the keeper dilutes it with a water change, feeds sparingly, adds conditioner as appropriate, and repeats. It can be managed, especially when mature media is available, but its margin for error is paid for by the fish. “Hardy” does not mean unharmed.

A fishless cycle separates biofilter preparation from animal welfare. Add a measured source of pure ammonia or ammonium chloride—never a household cleaner containing fragrance, surfactant, or soap—then follow ammonia and nitrite with tests. Hobby protocols commonly begin around 2 ppm total ammonia, enough to establish useful capacity without creating the very high nitrite load that can prolong the exercise. The exact dose is not sacred; what matters is knowing what went in and observing what the system does with it.

At first, ammonia remains detectable. As ammonia oxidizers establish, that reading falls and nitrite appears. Later, nitrite oxidizers catch up, nitrite falls, and nitrate usually accumulates. The practical completion test is not “nitrate exists.” Tap water, fertilizers, and aquasoil can all supply nitrate. Instead, redose the known ammonia amount and confirm that both ammonia and nitrite return to zero, or the lowest reliable reading of the kit, within about 24 hours. Then make a large water change to reduce accumulated nitrate before adding animals.

Dechlorinate all new water because chlorine and chloramine are intended to suppress microbes. Keep the filter running and the media oxygenated. Avoid scrubbing the filter or replacing its media during the process. If pH falls sharply in very soft water, nitrification can slow because the reactions consume alkalinity; measure KH and pH before assuming the bacteria “died.” A pinch of food can provide ammonia through decay, but it is a vague dose with a delay, so I prefer a product whose concentration I can calculate.

A planted tank has another route for nitrogen

Plants do not wait politely for bacteria to make nitrate. Aquatic plants can take up ammonium directly through leaves and roots, and many use it readily when it is available. This matters because assimilation into new plant tissue removes reduced nitrogen before it travels through the conventional ammonia → nitrite → nitrate sequence. Fast-growing stems and floaters can be surprisingly effective competitors for that nitrogen.

This is the basis of what Diana Walstad calls a “silent cycle”: a tank planted heavily from the beginning may never show the textbook ammonia and nitrite spikes. That does not mean the aquarium is sterile or that nitrifiers never establish. Biofilms still develop as small amounts of nitrogen escape plant uptake. It means the plants have changed what can be seen in the water.

The phrase is easy to overextend. Three newly purchased Anubias plants are not a large nitrogen sink. Slow growers, recently transplanted plants, and emersed-grown leaves melting after submersion may contribute less uptake—or briefly add organic load. A silent start is most convincing when there is abundant, visibly growing plant mass, including fast stems or floating plants, and very modest initial stocking.

I still test such tanks. I simply interpret a flat graph differently. In an unplanted fishless setup, disappearing ammonia with a later nitrite rise points toward nitrification. In a dense, growing planted tank, disappearing ammonia without nitrite may be plant uptake, microbial oxidation, or both. The animal does not care which route removed it; the keeper should care that the capacity remains reliable when plants are trimmed, melt, or stall.

What three to six weeks can look like

Three to six weeks is a reasonable planning range for an unseeded fishless aquarium, not a promise. In a recent home-aquarium study, two tanks developed undetectable ammonia and nitrite by week three while another took until week eight. Temperature, pH, alkalinity, oxygen, starting organisms, media, ammonia loading, and the test method all affect the result. Mature filter media can shorten the process dramatically; adverse conditions can stretch it.

Days 1–7: dechlorinate, run the filter continuously, add the measured ammonia source, and record a baseline for ammonia, nitrite, nitrate, pH, and temperature. Test ammonia every day or two. It may appear that nothing is happening. Resist adding more simply because the first dose has not moved.

Weeks 2–3: ammonia often begins to fall and nitrite becomes detectable. Test both. Redose ammonia only when it is near zero; repeatedly topping up to the original concentration can create an unnecessarily large nitrite backlog. Cloudy water may come and go, but it is not the measurement that decides anything.

Weeks 3–6: nitrite eventually declines as the second population expands. Nitrate may rise, though plants can obscure that trend. When both ammonia and nitrite reach zero, add the test dose again. If both clear within roughly 24 hours, the filter has demonstrated capacity. If either remains, keep going.

Before livestock, change enough water to bring nitrate into the range appropriate for the animals and the fertilizing method, and match temperature carefully. Add the first group modestly rather than filling the final stocking plan at once. For the next week, test ammonia and nitrite daily and feed lightly. A cycle is capacity matched to a particular load, not permanent immunity from overstocking, a dead fish, a power outage, medication, or a filter washed too aggressively.

The lesson I took from that first green test tube was not that aquariums require a ritual wait. It was that clear water can hide unfinished biology. Give the tank a measurable source of nitrogen, give its wet surfaces oxygen and time, and ask it to prove what it can process before an animal has to.

Sources