What the shop will not tell you · Stone and filter

Aquarium water myths: rocks, hardness and filters

Two groups of corrections, both about what is happening in the water rather than what is growing in it. The stone myths, where the arithmetic is far less frightening than the forum, and the filter myths, where the organisms named in every beginner guide are not the organisms doing the work.

Iwagumi rock detail
Iwagumi rock detailMoritz Holzinger · CC BY-SA 3.0 · Resized; uncropped.

Rock and wood

The most feared beginner mistake is mostly not a disaster.

Putting the wrong rock in the tank is the thing beginners are warned about most and it has a hard ceiling that almost nobody quotes. Calcite does not dissolve forever. It stops at saturation, and saturation depends almost entirely on how much carbon dioxide is in the water. The whole calculation, with its constants and its assumptions, is on the water page.

Correction 04 · the reassuring one

A limestone rock in an uninjected tank cannot push you past about 3 dGH1. Turn the carbon dioxide on and it can reach ten.

Commonly said
Seiryu, or any rock that fizzes, will wreck a soft water tank. Take it out. Some say never use carbonate rock at all.
What the evidence shows
Solve the standard calcite-carbonate equilibrium system at 25 degrees for water sitting on calcite, using log K1 of -6.35, log K2 of -10.33 and log Ksp for calcite of -8.4823, and the answer is bounded. At air equilibrium with today's atmosphere at 420 parts per million, which is 0.63 mg/L of dissolved carbon dioxide, the water reaches about 19.4 mg/L of calcium, pH 8.2 and 2.7 dGH, and then stops1. That is inside the range most plants and most soft water fish tolerate. Under carbon dioxide injection the same solver gives a ceiling around 10 dGH and 10 dKH1, which genuinely does matter for a Caridina shrimp tank or a wild caught blackwater fish. Hardness rises as the cube root of carbon dioxide, so 20 percent more carbon dioxide buys 6 percent more hardness1, which is why the number moves so little and why the headline is stated as "under 3" rather than to two digits. The calculation is validated against the experimental figure of about 20 mg/L of calcium in contact with air, and 16 mg/L from thermodynamic calculation2.
What to do instead
Make the advice conditional, because that is what the chemistry is. No injection and community fish: use the rock. Injection, or Caridina, or a wild caught soft water species: do not. If you want certainty about a particular piece, soak it for fourteen days in reverse osmosis or distilled water and test GH and KH at the start and the end. Tap water may already be near saturation and will hide the effect. Two weeks, no cost, and it settles the argument for your rock rather than for rocks in general.
Why the common version exists
It is correct advice with its condition dropped. The people who wrote it were mostly running injected high-tech tanks or shrimp tanks, where the ceiling really is around ten and the warning is real. Passed along without the condition it becomes an absolute, and an absolute is easier to repeat at a counter than a curve.

Correction 05

"Boil the rock and it will stop leaching."

Commonly said
Boil or bake a new rock and it will be safe.
What the evidence shows
Boiling kills organisms and removes loose dust. It does not change the mineral. Calcite is still calcite after boiling, and its solubility product is a property of the mineral and the water it sits in2. Nothing available in a kitchen alters it.
What to do instead
Boil for the organisms, by all means. For the chemistry, do the fourteen day soak in correction 04. It is the only kitchen procedure that answers the question you are actually asking.
Why the common version exists
Boiling genuinely does solve the neighbouring problem, which is hitchhikers and surface dirt, and it visibly changes the water in the pot. Two true things about heat treatment get merged into one false one.

Correction 06

"Driftwood softens your water." No. Softening means removing calcium and magnesium.

Commonly said
Add wood and the water softens and the pH drops. Sold as the low cost route to blackwater.
What the evidence shows
Wood does not remove calcium or magnesium, so general hardness is untouched. What it adds is weak organic acids, which consume a little alkalinity and pull KH down slightly. And the pH effect is conditional on your KH: in water at 5 dKH or more a normal amount of wood will tint the water and barely move the pH, because that is what a carbonate buffer does2.
What to do instead
If you want soft water there are three routes and wood is not one of them: reverse osmosis, rain water, or ion exchange. Add wood for tannins, structure, biofilm and the animals that graze it, all of which are real reasons.
Why the common version exists
It is true in soft water. Most of the English language hobby literature was written in places with low KH tap water, where adding wood does visibly drop the pH, and the observation was accurate where it was made. Yerevan tap runs pH 6.0 to 7.3 across districts4 with a real carbonate reserve, and the same wood will do very little.

Correction 07

The white slime on new wood is a sign the wood is fresh, not that it is bad.

Commonly said
White jelly on new driftwood means the piece is rotten, contaminated or was sold to you dirty. Take it out.
What the evidence shows
It is saprophytic growth, fungi and water moulds eating the sugars left in the outer wood, and it appears on almost every new piece in the first two to six weeks. The manufacturer whose layout wood is the reference product in this hobby warns buyers in its own product notes that Saprolegnia may appear at the initial stage5. It feeds on a limited supply and then disappears.
What to do instead
Leave it. Shrimp, otocinclus and snails eat it. If it offends you, siphon it off and it will come back a little smaller each time until the sugars are gone.
Why the common version exists
It looks exactly like something going wrong, and the hobby has no vocabulary for a normal succession stage. Nothing else in a new tank is white, slimy and growing.

Correction 08

The vinegar test is worth doing and it misses three whole classes of rock.

Commonly said
Drop vinegar on it. If it does not fizz it is inert and safe.
What the evidence shows
A fizz is a definite answer in one direction. Silence is not a definite answer in the other. The geological version of the test uses dilute hydrochloric acid at 5 to 10 percent; household vinegar is about 5 percent acetic acid, a much weaker acid, so the kitchen version is a weaker form of an already limited test, and solid dolomite barely reacts to cold acid at all6. Dolomite is calcium magnesium carbonate, so a piece that passes the test can still raise both GH and KH. Gypsum is calcium sulfate2, dissolves more readily than calcite, puts calcium into the water and contains no carbonate to make a bubble, so it is invisible to the test. And it is a spot test on the face you can reach: a calcite vein on the hidden side is untested.
What to do instead
Two minutes, then two weeks. Do the vinegar test on every piece, on a freshly broken face rather than a weathered outside, and grind a corner to powder if you suspect dolomite. When the silence matters, do the fourteen day soak. Both procedures are set out on find a hardscape.
Why the common version exists
It is a real geological field test, correctly transmitted, with its stated limitations left behind. Geologists know what it does not detect because that is taught alongside it.

Correction 09

Test the stone yourself, because the trade name tells you nothing and vinegar tells you everything.

Commonly said
This is Seiryu stone. This is Dragon Stone. This is Frodo Stone. Said as though these were rock types.
What the evidence shows
There is no mineral called Seiryu and no rock type called Dragon Stone. They are marketing labels attached to whatever a quarry is currently producing that looks right, and there is no customs code for aquascaping stone, so it travels under the broad building and decorative stone headings alongside kerbstones and worktops. Read from the wholesale side: a search of a main Chinese export directory returns 30 first page products under "Seiryu Stone", every one from a single company in Hebei province, at 350 to 600 US dollars per tonne with a one tonne minimum; the same directory returns 30 products under Dragon Stone and 32 under Ohko Stone, all Chinese; and "Frodo Stone" returns zero results7. A zero is informative. It means Frodo is a Western retail name invented downstream and applied to material bought under some other description.
What to do instead
Treat the name as decoration and the rock as unknown until you have tested it. Our hardscape tool files a trade description as unknown rather than as inert, and never treats unknown as safe. If you want a rock whose behavior you can predict, the cheapest reliable option in Armenia is the stone under your feet, tested.
Why the common version exists
Because the person selling it to you does not know either, and is not concealing anything. The name came with the pallet. Genuine Japanese quarried layout stone does exist and is sold under its own names, and it is a real but small and expensive part of the market.
Porous aquarium filter sponge
The open pores of filter foam give a biofilm somewhere to grow.Ofkun · CC BY-SA 4.0 · Resized; uncropped.

Water and the filter

Almost nothing in your filter is a nitrifier, and the textbook names are decades out of date.

The nitrogen cycle diagram on the side of a filter box names two bacteria. Both halves of that picture have been rewritten twice since, once in 1998 and again by the discovery of ammonia oxidising archaea and of organisms that do the whole conversion alone. The long account is on the cycling page.

Correction 10

At least six in seven of the organisms in your filter are not ammonia oxidisers.

Commonly said
The filter is full of beneficial bacteria, meaning nitrifiers, and the media is where they live. Two genera do the work: one for ammonia, one for nitrite.
What the evidence shows
A 2025 study followed three ordinary home aquariums for twelve weeks and found ammonia oxidisers making up 1.6, 3.1 and 3.8 percent of the community on the biofilter sponges, and 0.6 to 14.1 percent on the beads8. Everything else is overwhelmingly heterotrophic, and the most common genera are waste processing organisms. The named organisms have moved too. The nitrite step in aquaria was shown in 1998 to be done by Nitrospira and not by the Nitrobacter in the books, in a paper received in September 1997 and confirmed within twelve months by three other groups working on a bioreactor, on municipal activated sludge and on a fluidised bed9. Six freshwater aquaria followed for four months had archaea outnumbering bacteria in all six, and archaea were the only detectable ammonia oxidiser in three of them10. The same paper notes the useful part: heterotrophs help nitrifiers establish by forming the initial biofilm that nitrifiers then colonise.
What to do instead
Stop thinking of the filter as a nitrifier box and start thinking of it as a waste processing community that contains a small, slow, valuable minority. It changes three decisions: do not sterilise media, do not replace all of it at once, and rinse in tank water rather than tap. It also explains why bottled products naming a single genus are working on a thin slice of the problem.
Why the common version exists
It was the state of the science, and the aquarium industry printed it faithfully. The two-genus diagram was correct in 1985. Packaging outlives papers.

Correction 11

"I overfed and the ammonia spike killed my bacteria." Almost certainly they were smothered, not killed.

Commonly said
Overfeeding produced an ammonia spike, the spike killed the filter bacteria, and the tank has to be cycled again.
What the evidence shows
Ammonia at those levels does not kill nitrifiers; it is their food. What the extra food is, from the filter's point of view, is carbon. Two nitrifying biofilm reactors given acetate lost nitrification, and imaging showed why: a heterotrophic layer formed on top of the nitrifying biofilm and cut off its oxygen supply11. An aquaculture biofilter study put numbers on it: at 2 mg/L of total ammonia nitrogen, raising the carbon to nitrogen ratio to 0.5 cut ammonia removal by 30 percent, and at higher ratios nitrification efficiency fell by 50 percent while the number of attached bacteria doubled12. Twice as many bacteria on the media, half as much nitrification. That is an overfed tank in one sentence.
What to do instead
This is good news, and it changes what you do. If they had died you would have to cycle again. Because they were buried, removing the organic load usually restores the rate within days. Vacuum the mulm, rinse the filter in tank water, and stop feeding for two days. Do not tear the tank down and do not buy a bottle.
Why the common version exists
Cause and effect are in the right order and only the mechanism is wrong, which is the hardest kind of error to notice. Feeding did cause the ammonia reading. It is just that the reading is a symptom of oxygen loss in the biofilm rather than of a death.

Correction 12

"A bacterial bloom means the tank is cycling." It means something quite different.

Commonly said
White cloudy water in a new tank is the bacteria colonising. Wait it out, it is a good sign.
What the evidence shows
A milky cloud is free living heterotrophs. Nitrifiers are attached organisms present in far smaller numbers and would never make water visibly cloudy: they are single-digit percentages of the community even on established media8. A bloom means one thing, a sudden supply of dissolved organic carbon. New soil, a dead snail, a large feed.
What to do instead
Read it as a carbon event and go looking for the carbon. It says nothing directly about whether nitrification is established, which is a question only an ammonia and nitrite test answers. Waiting it out is still usually right, but for a different reason.
Why the common version exists
The timing coincides. Blooms happen in the first weeks, cycling happens in the first weeks, and both end. Two processes on the same clock get read as one.

Correction 13

"Deep sand makes hydrogen sulfide that will kill everything." True in the sea, and mostly not in a river.

Commonly said
Never let sand go deeper than about 2 cm, and stir it weekly, or a gas pocket will release and wipe out the tank.
What the evidence shows
Real in principle, overstated in fresh water, and for a specific chemical reason: sulfide production needs sulfate to reduce. Seawater carries about 2,700 mg/L of sulfate. World average river water carries about 8 to 11 mg/L2, roughly 250 to 350 times less1. Hydrogen sulfide also oxidises quickly when it meets oxygenated water. The black patches under sand that people photograph in alarm are normal and are mostly iron sulfide, which is why they are black rather than absent.
What to do instead
Stop stirring the substrate on a schedule, which does more harm than the thing it prevents. Plant it instead: roots carry oxygen down. If you want to check, push a finger in and smell your finger. Rotten eggs is unambiguous and nothing else smells like it.
Why the common version exists
It came from marine keeping, where it is correct, and crossed over intact. The deep sand bed debate is a saltwater debate and the sulfate figures above are why it does not transfer.

Correction 14

"Just top the tank up." The careful keeper is the one this happens to.

Commonly said
Keep the water level right. Top up what evaporates. Old tank syndrome happens to neglectful people.
What the evidence shows
Evaporation removes water and leaves every dissolved solid behind, so topping up with tap water adds a further load each time. Concentration after n top-ups follows C(n) = C_tap multiplied by (1 + n times the evaporated fraction), and a repeated partial change sets a ceiling rather than removing the accumulation13. Topping up promptly and keeping the level exactly right is precisely the behavior that drives the line. Someone who lets the tank drop and then refills with a large change is accidentally doing the right thing. One honest limit: sodium and chloride have no sink at all so the line is exact for them, while calcium bends below it because snail shells, shrimp moults, pruned plant tissue and scale on the heater all remove some. A hardness reading that has stopped rising has not stopped accumulating. It has run out of the one thing you were measuring.
What to do instead
Top up with distilled or reverse osmosis water. It carries nothing, so evaporation plus top-off becomes a null operation and the line goes flat. This is the cheapest intervention on the water change page and almost nobody does it, because the advice is never phrased in a way that makes you ask what the top-up water is.
Why the common version exists
"Top up the tank" is a complete sentence and correct as far as it goes. The missing half is three words long and nobody says it.

Correction 15

"Never do a big water change on a neglected tank." The rescue is an order of magnitude safer than the story.

Commonly said
A large change on an old tank shocks the fish and the pH jump converts ammonium to toxic ammonia. Do 10 percent a day for a fortnight.
What the evidence shows
Model it properly, with the pH moving in proportion to the fraction changed rather than jumping straight to the tap value, and two things fall out. The worst case is in the middle, at about 76 percent, not at either end13, so "small changes are worse" is wrong as stated and so is "bigger is always safer". And the size is the real finding: the peak un-ionised ammonia in the realistic model is 0.0155 mg N/L, which is 0.019 mg/L expressed as NH313, against tissue damage in fish beginning at about 0.05 mg/L as NH314. Total ammonia has to be above roughly 13 mg N/L before the peak crosses the threshold13, and a tank reading that is a catastrophe already in progress whether or not anyone touches the water.
What to do instead
The ammonia route is not the reason to be careful. The remaining reason is ionic and pH shock, which is standard physiology, is the universal hobby explanation, and which we could find no controlled aquarium study of at all. It is now the larger of the two routes and it is the unmeasured one, so the mechanism is rated Medium rather than Strong on the water change page. Practically: match temperature, dechlorinate, and go steadily rather than fearfully.
Why the common version exists
The chemistry it invokes is real, the direction is right, and the arithmetic was never done. We know, because we got it wrong too and in the opposite direction. That entry is in what we got wrong.

Correction 16

"A hardy fish will cope with untreated tap water." The chlorine is not aimed at your fish.

Commonly said
Age the water overnight, or just use a hardy species, and the chlorine will not matter.
What the evidence shows
Chlorine and chloramine are in tap water because they kill bacteria. That is their entire purpose and they are good at it. The nitrifying population is the part of the tank least able to absorb an insult: single-digit percentages of the biofilm community8, and slow, turning a gram of nitrogen into about 0.15 g of new cells15 against heterotrophs that convert most of a gram of food into cells and grow several times a day. So a dose knocks back the whole biofilm, the heterotrophs are back in a day, and the nitrifiers take weeks. The damage does not appear as a dead fish. It appears as an ammonia reading three days later in a tank that was cycled and stable.
What to do instead
Dechlorinate every time, regardless of species. If free and total chlorine read the same, ageing the water works. If total is clearly higher, that is chloramine, ageing is useless, and you want a conditioner that also handles the ammonia released when the bond is broken.
Why the common version exists
"Ageing the water" was correct advice for free chlorine, which is what municipal supplies used to run on, and it still is where they still do. Chloramine is more persistent by design, because that is the point of using it.
  1. Our own arithmetic on the carbonate and carbon dioxide systems. Not a published source. The calcite ceiling table, solved as the standard calcite-carbonate equilibrium system at 25 degrees C for a fixed dissolved CO2 with charge balance 2[Ca2+] = [HCO3-], using the constants in sources 1 and 2, activity corrections ignored. The uninjected row is solved at 420 ppm, which is 0.63 mg/L dissolved at 25 degrees, and returns 19.4 mg/L calcium, pH 8.2 and 2.7 dGH. Also the cube root sensitivity of hardness to CO2, the 250 to 350 times seawater to river sulfate ratio from source 1, and the air against water molar table computed by the ideal gas law and Henry's law from source 30 at sea level and at 887 hPa, including the 12.5 percent altitude reduction. calculated, repeatable, unverified by a third party
  2. Hem, J.D. (1985). Study and Interpretation of the Chemical Characteristics of Natural Water, 3rd edition. US Geological Survey Water-Supply Paper 2254. Public domain. Calcite solubility in contact with air, about 20 mg/L calcium from experiment and 16 mg/L from thermodynamic calculation, page 92; the hardness unit conversions; gypsum as calcium sulfate; and the seawater and river water sulfate figures. Full text PDF
  3. Morin, G. Understanding the General Chemistry of the Planted Aquarium. Seachem Laboratories. The equilibrium constants, the statement that an alkalinity measurement does not indicate the presence of any one compound, and the pH drop method. PDF
  4. Shahinyan, H.V. (2005). On the hydrogeochemical characteristics of some drinking waters supplying the city of Yerevan. Proceedings of the National Academy of Sciences of the Republic of Armenia, Earth Sciences LVIII(1): 41 to 45. Sampling 2002 to 2004. Hardness printed in mg-eq/L and converted here at 2.804 dGH per unit: 2.9 dGH in Avan, Kanaker and Nork to 10.9 dGH in Davtashen, pH 6.01 to 7.30 across all districts. The dataset is twenty years old, varies by district and rises in summer, so it is a map of the city rather than a reading of your tap.
  5. Layout Materials: Driftwood. Aqua Design Amano. The manufacturer's own Saprolegnia warning on branch wood, and its descriptions of its quarried layout stones. adana.co.jp
  6. The Acid Test for Carbonate Minerals and Carbonate Rocks. Geology.com. The dilute hydrochloric acid protocol at 5 to 10 percent, the 5 percent acetic acid comparison, and the cold dolomite limitation. geology.com/minerals/acid-test.shtml
  7. Made-in-China business to business directory, product searches. Searches for Seiryu Stone, Dragon Stone, Ohko Stone and Frodo Stone, accessed 8 September 2026. The single Hebei supplier behind the first page of Seiryu listings, the 350 to 600 US dollars per tonne price with a one tonne minimum, and the zero result for Frodo Stone. Seiryu search
  8. McKnight, M.M., Szabolcs, N., Graham, A. and Neufeld, J.D. (2025). Microbial community succession of home aquarium biofilters associated with early establishment of comammox Nitrospira. ISME Communications 5(1): ycaf212. Open access. The twelve week succession study in three ordinary home aquariums, the 1.6, 3.1 and 3.8 percent sponge figures and the 0.6 to 14.1 percent bead range. DOI 10.1093/ismeco/ycaf212 open access
  9. Hovanec, T.A., Taylor, L.T., Blakis, A. and DeLong, E.F. (1998). Nitrospira-like bacteria associated with nitrite oxidation in freshwater aquaria. Applied and Environmental Microbiology 64(1): 258 to 264. Free on PubMed Central, PMC124703. Received 4 September 1997. Confirmed within twelve months by three independent groups: Burrell, Keller and Blackall on a nitrite oxidising bioreactor, May 1998, DOI 10.1128/aem.64.5.1878-1883.1998; Juretschko and others on municipal activated sludge, August 1998, DOI 10.1128/aem.64.8.3042-3051.1998; and Schramm and others on a fluidised bed reactor, September 1998, DOI 10.1128/aem.64.9.3480-3485.1998. DOI 10.1128/aem.64.1.258-264.1998 open access
  10. Bagchi, S., Vlaeminck, S.E., Sauder, L.A., Mosquera, M., Neufeld, J.D. and Boon, N. (2014). Temporal and spatial stability of ammonia-oxidizing archaea and bacteria in aquarium biofilters. PLoS ONE 9(12): e113515. Open access. Six freshwater tanks over four months, archaea outnumbering bacteria in all six and the only detectable ammonia oxidiser in three. DOI 10.1371/journal.pone.0113515 open access
  11. Nogueira, R., Melo, L.F., Purkhold, U., Wuertz, S. and Wagner, M. (2002). Nitrifying and heterotrophic population dynamics in biofilm reactors. Water Research 36(2): 469 to 481. Paywalled; abstract free. The heterotrophic layer forming on top of the nitrifying biofilm and cutting off its oxygen supply. DOI 10.1016/s0043-1354(01)00229-9
  12. Michaud, L., Blancheton, J.P., Bruni, V. and Piedrahita, R. (2006). Effect of particulate organic carbon on heterotrophic bacterial populations and nitrification efficiency in biological filters. Aquacultural Engineering 34(3): 224 to 233. Author version free at Ifremer. The 30 percent and 50 percent nitrification losses, and the doubling of attached bacteria. DOI 10.1016/j.aquaeng.2005.07.005
  13. Our own arithmetic on water changes and ammonia speciation. Not a published source. Two models. The evaporation and top-off ceiling: C(n) = C_tap (1 + n e) for topping up only, and C* = C_tap (1 + e (1 - f) / f) for the ceiling under a repeated change. The rescue water change: the fraction as NH3 = 1 / (1 + 10^(pKa - pH)) with pKa 9.25 at 25 degrees, checked against the published fractions in source 23, for a tank at pH 6.0 holding 5 mg N/L of total ammonia and tap at pH 7.8, with pH after a change moving in proportion to the fraction replaced. Returns a peak at about 76 percent, 0.0155 mg N/L un-ionised, and a crossing of the damage threshold only above roughly 13 mg N/L total. calculated, repeatable, unverified by a third party
  14. Francis-Floyd, R., Watson, C., Petty, D. and Pouder, D.B. Ammonia in Aquatic Systems. University of Florida IFAS Extension, FA031. The hundredfold toxicity difference between the two ammonia forms and the 0.05 mg/L tissue damage threshold. ask.ifas.ufl.edu
  15. Blackburne, R., Vadivelu, V.M., Yuan, Z. and Keller, J. (2007). Kinetic characterisation of an enriched Nitrospira culture with comparison to Nitrobacter. Water Research 41(14): 3033 to 3042. Paywalled; abstract free. The nitrifier yield of about 0.15 g of cells per g of nitrogen oxidised, and the oxygen half saturation constant of 0.54 mg/L. DOI 10.1016/j.watres.2007.01.043

Continue

Where to go next.

Materials and water

The carbonate ceiling in full, which is the arithmetic behind the stone myths.

Find a hardscape

Material by material, what each rock and each piece of wood actually does to the water.

Cycling and the bacteria

The cycling account with the right organisms in it, and the paper that identified them.

Plants, the room and the pond

The other half of the corrections: light, the room, and the outdoor pond advice written for the wrong country.