Corydoras care for beginners

This is first because it is the only item on this page that a care authority ties directly to injury, and because it is the decision that is hardest to reverse once there are fish in the tank.

A corydoras feeds by pushing its barbels, which are soft sensory whiskers, through the substrate. On sharp or coarse grains those barbels wear down and can become infected, and a cory that has lost its barbels has lost the organ it finds food with. It is a slow failure that looks like nothing for months.

Fine sand is the straightforward answer. It is also the substrate this fish sifts most convincingly, which is most of the reason to keep them.

The honest costs. Sand compacts and needs occasional stirring, and it is vacuumed differently from gravel: hover above it rather than digging in, or the sand leaves with the waste. Fine sand can be drawn into a filter intake sitting too low. Twenty pounds is more than a 20 gallon tank needs for a shallow bed, so you will have some left over. And the live bacteria packet is a head start, not a substitute for cycling the tank.

The second half of the problem. Corydoras are not scavengers living on what the other fish drop, however often they are sold that way, and a community tank of enthusiastic surface feeders is a tank where nothing edible reaches the floor.

A sinking wafer is the simplest fix because it is countable. You put one in, you watch it get found, and you know the fish ate.

Hikari's claim that it is built to avoid dissolving is the useful part: a wafer that holds together is still there when the corys get busy after lights-out, which is when they prefer to work.

The honest costs. A wafer is a lump of food sitting on the substrate, so an already-overfed tank gets worse fast — it disperses into nothing and nothing in the upper water column will take it. Hikari's own instruction to remove what is uneaten after 20 minutes is the part everyone skips. Ash is listed at up to 16%, higher than the Fluval granule below. And one whole wafer is too much for two or three fish, so break it.

The alternative shape, and with a group of five or six it is often the better one. A wafer is a single prize that the boldest fish sits on; granules scatter, which means the whole group feeds at once.

The grain size is the reason it works here. A corydoras has a small downturned mouth, and 1.4 to 2.0 mm is about the largest thing it can take comfortably.

**One honesty note, because it matters.** Fluval aims this formula at plecos and ancistrus. It does not name corydoras anywhere we can find. The grain size and the sinking behavior are what make it suitable, and that is our read of the specification rather than a claim by the manufacturer.

The honest costs. Beyond the above: Fluval's own website does not respond to us, so the analysis here comes from a retailer listing rather than the maker. Granules scatter, which is good for a group and bad for seeing how much you actually added — the commonest way a cory tank gets overfed. It smells strongly of insect meal. And 1.60 ounces goes stale long before a small group finishes it.

Plenty of people have a gravel tank already and are not going to tear it down, and the honest reading of OATA's wording is that they do not have to: the sheet says sand **or fine gravel**.

Smooth rounded river stone is the version that is defensible. What is not defensible is the sharp crushed gravel and the coated novelty gravel that most starter kits photograph, because the injury OATA is describing comes from edges.

It is also easier to maintain than sand. You push the siphon into it and let the grains tumble back; nothing compacts and nothing gets sucked into the filter.

The honest costs. This is the compromise pick and the compromise is real. At 5 mm the grains are coarse enough that food falls between them and out of reach, which is precisely the problem a bottom feeder has — so a gravel tank needs a more deliberate feeding routine, not a more relaxed one. Corys also sift gravel far less convincingly than sand, so you lose some of the behavior you bought them for. If you are setting up a new tank for corydoras, buy the sand.

OATA's range is 22 to 27 C, which is ordinary tropical and comfortably met by any heater sized to the tank. The reason this is on the page at all is that corydoras get sold as a cleanup crew rather than as fish, and cleanup crews get put in unheated tanks.

A hundred watts is the right figure for the 20 gallon tank this page recommends.

An external controller also means you set the temperature by a number rather than a dial mark, which is the part people actually want.

The honest costs. The plus or minus 0.5 C accuracy is the listing's own claim, not a manufacturer document we can check, so treat it as a claim. The controller is another component to site and keep dry, and its readout is not the same thing as a thermometer in the water — buy one of those separately. No thermometer is included.

Six small catfish working the substrate all day produce more than six small catfish look like they should, and all of it ends up on the floor rather than in the water column.

The AquaClear's advantage here is the media basket. You choose what goes in it, which matters in a tank where you may want to swap the fine mechanical media more often than the biological.

The honest costs. Two of them are specific to this page. **Fine sand and a low intake do not get along** — raise the intake, or fit the pre-filter sponge from the first day, or you will hear the impeller grinding. And at full rated output 150 GPH is a lot of flow for a 20 gallon tank, so expect to use the flow control rather than treating it as a bonus. Beyond that: no heater, no test kit, and the media is a recurring cost.

Most care sheets say keep nitrate low. OATA's catfish sheet gives a number — not more than 20 mg per liter above normal tap water levels — and that is a figure you can only act on if you are measuring both the tank and the tap.

The pH test earns its place differently here. OATA's range for catfish is **6.0 to 8.0**, which is unusually wide, and the useful outcome of testing is usually finding out you can stop worrying about it.

The honest costs. It does not test hardness, though with a 4 to 18 dH range that matters less for corydoras than for most fish. It is slower and fiddlier than strips, the nitrate test needs the second bottle shaken hard and then a wait, and the color card is genuinely hard to read between adjacent shades under warm light. Reagents expire.

A bottom-dwelling fish lives in the layer where waste settles, so this is a more direct piece of welfare equipment on this page than it is on most.

On sand the technique is different and the difference is the whole trick: hover the tube an inch above the surface and let the lighter waste lift off, rather than pushing in as you would with gravel. We wrote the method up separately on our **can you use a gravel vacuum on sand** page.

The honest costs. You carry the water, which is fine at 20 gallons and the reason people quietly stop doing water changes once the tank is bigger. Starting the siphon takes a knack. And on sand it is easy to remove several pounds of substrate before you work out the hover height — practice in a bucket.

The skip entry, and it is not a bad product — it is our pick for a community tank elsewhere on this site.

The problem is geometry. Flake is eaten at and just below the surface by fish that are faster and bolder than a catfish, and in a stocked tank almost none of it reaches the floor. The corydoras is then expected to live on scraps that never arrive.

This is the actual mechanism behind corys that look fine for six months and then quietly thin out and die. Nobody watched them not eat.

The honest costs of the skip. If your tank is lightly stocked, flake does land and corys will take it, so this is not a rule. But you cannot know that without watching after lights-out, which is why a sinking food is the safer default. Keep the flake for the fish it is for, and add a sinking food for the ones on the bottom.