Water quality guide
Most keepers test ammonia, nitrite, nitrate and pH on a schedule, and almost nobody watches dissolved oxygen at all. It is the parameter that moves fastest, bottoms out while you are asleep, and is behind a good share of the losses people find at first light and never fully explain.
Dissolved oxygen, usually written DO and read in ppm or mg/L (the two are interchangeable in fresh water), is what the fish are actually breathing. Ammonia poisons them slowly over days. Oxygen can take a pond from normal to a crisis in a single warm, still night.
The reason it goes unwatched is practical. There is no cheap, reliable drop test for it the way there is for ammonia or nitrite. Reading DO properly means an electronic meter, and a good handheld one costs more than most keepers expect to spend on testing. So the parameter with the shortest fuse ends up being the one nobody has a number for.
This is physics, not biology, and there is nothing you can do about it. The warmer the water, the less oxygen it is capable of holding even when it is fully saturated. At the same time, a warm koi is more metabolically active and needs more oxygen, and so does the bacteria bed in your filter. Demand rises exactly as supply falls.
| Water temp | Oxygen ceiling at 100% saturation | What that means in the pond |
|---|---|---|
| 50 °F (10 °C) | about 11.3 mg/L | Comfortable headroom; fish demand is low |
| 59 °F (15 °C) | about 10.1 mg/L | Still generous for a normally stocked pond |
| 68 °F (20 °C) | about 9.1 mg/L | Ceiling now close to the comfortable range |
| 77 °F (25 °C) | about 8.3 mg/L | A heavy load can pull the pond below target overnight |
| 86 °F (30 °C) | about 7.6 mg/L | The ceiling itself is barely above the floor you want |
Read that table as a ceiling, not a forecast. It is the best a clean, still, fully saturated pond at sea level could do at that temperature. A stocked, fed, planted pond in August sits below it, and the gap between the ceiling and where you actually are is the whole story.
Plants, string algae and the green water bloom all photosynthesize in daylight and put oxygen into the water. After sunset they stop producing and keep respiring, consuming oxygen alongside the fish, the filter bacteria and everything breaking down in the bottom muck. The pond draws down all night with nothing adding back.
So a midday reading, taken when the sun has been on the water for hours and the waterfall has been running, is the daily maximum. It can read a comfortable 9 mg/L and tell you nothing at all about the 5 mg/L the same pond touched at 5:00 in the morning. A pea-soup green pond has the widest daily swing of all, because the bloom that oxygenates it by day is the biggest consumer at night.
| Reading | Interpretation | What to do |
|---|---|---|
| 8-10 mg/L | Comfortable | Nothing. This is where a healthy pond should sit, including at dawn |
| 7 mg/L | Sensible floor | Acceptable, but treat it as the bottom of the range rather than a target |
| 6-7 mg/L | Margin getting thin | Add aeration now, before the next hot night takes it lower |
| Below 6 mg/L | Stress | Fish are working to breathe; growth, appetite and immune response suffer |
| Below 4 mg/L | Acutely dangerous | Emergency. Maximum aeration, stop feeding, hold off on any treatment |
Judge a pond by its minimum, not its average. A pond that averages 8 mg/L across the day and dips to 5 mg/L every night at 4:00 a.m. is a stressed pond, and the daily average will never tell you that.
Two situations that catch people out. Several common pond treatments consume oxygen or interfere with the fish's ability to use it, and formalin-based products in particular can drop DO sharply within an hour of dosing. Aerate hard before, during and after any treatment, and never treat a pond that is already gasping.
The second is weather. A summer thunderstorm brings a rapid barometric drop, heavy cloud that shuts down photosynthesis, and often a slug of warm runoff and organics into the pond. Fish are noticeably more likely to be at the surface the morning after a storm.
The classic signs are koi hanging at the surface with their mouths at the air, called gasping or piping, and fish crowding the waterfall, the return jets or the aerated end of the pond. It is worst at dawn and often eases by mid-morning as the sun comes up, which is exactly why it gets dismissed as the fish being hungry.
The problem with behavioral symptoms is that they are the last stage, not the first. By the time koi are visibly piping, the pond is already down around 3-4 mg/L and the margin is gone. Fish that were merely stressed at 5 mg/L for six weeks showed you nothing at all, and that quiet stretch is where the real damage to growth, color and disease resistance happens.
| Method | Strength | Watch for |
|---|---|---|
| Air pump and diffusers | The most reliable option. Bottom diffusers lift the whole water column and break up stratification | Size the pump to the pond, not to the box it came in; check the diaphragms annually |
| Venturi on the return | Good transfer for no extra power draw, and easy to add to an existing return line | Loses flow if the jet is set too deep; clogs with algae |
| Waterfall or spillway | Genuine gas exchange plus surface agitation, and it runs whenever the pump runs | Rarely enough on its own for a heavily stocked pond in summer |
| Fountain or spray bar | Useful surface agitation | Works the top of the pond only; does little for a deep, stratified bottom |
Three rules cover most of it. Run aeration through the night, because that is when it is needed and switching it off at dusk removes it precisely at the worst hour. Increase it in hot weather rather than leaving one setting year-round. And put aeration on the same circuit you would protect first in an outage, since a pond with no air movement on a July night has a shorter margin than most people assume.
One side effect is worth knowing. Vigorous aeration strips carbon dioxide out of the water, and as CO2 leaves, pH rises. How far it rises, and how sharply it swings between dawn and afternoon, depends on your buffering capacity, so it is worth reading the pond's carbonate hardness and total alkalinity alongside oxygen rather than in isolation.
Cooler, fresh, well-oxygenated make-up water also helps during a heat wave, which is one reason we build bespoke automated water changes for koi ponds. We have never found another system that does it, so it stays a custom build rather than a product on a shelf.
A handheld DO meter gives you an honest number, and it is the right instrument for verifying anything else you rely on. It has two limitations. A good one is a real investment, and it only tells you about the moment you dipped the probe. Almost nobody is standing at the pond at 4:30 a.m., which is the one reading that matters.
A continuous sensor answers a different question. It sees the overnight trace: how deep the pre-dawn trough went, whether it is getting deeper week over week as the fish grow and the water warms, and whether the aeration you added actually lifted the minimum or just the afternoon peak.
Weekly testing works well for parameters that change slowly. Nitrate creeps. Alkalinity is consumed gradually. A Sunday-morning reading is a fair picture of the week for those.
Oxygen does not behave that way. It has a large daily cycle, the low point falls in the few hours you are least likely to be outside, and the direction of travel over a hot week is downward. A weekly spot check samples the daily maximum roughly once every 168 hours, which means it can return a comfortable number every single week while the pond quietly spends every night in the stress band. The gap is not carelessness, it is the shape of the parameter against the shape of a testing routine.
A simple habit that costs nothing: on the next warm, still night, go out at first light before the sun is on the water and just look. Where the fish are sitting at that hour tells you more than a month of afternoon observations.
What H2Koi does, and what it does not. H2Koi measures dissolved oxygen directly, in mg/L and as percent saturation. This is one of the parameters it genuinely measures rather than derives, and because it reads continuously it sees the pre-dawn minimum a daytime spot check will never catch.
It is continuous early warning from industrial-grade sensors. It is not a substitute for your own eyes on the fish. It does not detect disease or parasites. It reports carbonate hardness as a derived, computed value on a 0-20 dKH scale rather than a titration, and it does not report GH at all. What it changes is when you find out, not what happens next.
Roughly 8-10 mg/L is comfortable, and 7 mg/L is a sensible floor. Below about 6 mg/L the fish are under real stress, and below 4 mg/L the situation is acutely dangerous. The number that matters is the overnight low, not the afternoon reading.
Almost always low dissolved oxygen. Overnight the plants and algae consume oxygen instead of producing it, so the pond bottoms out just before dawn and the fish come to the surface film where oxygen is highest. It often eases once the sun is up, which is why it gets missed. Add aeration and run it overnight, and check ammonia and nitrite as well, since gill irritation produces similar behavior.
Overnight is the critical window. Aeration during the day is the easy part; a pond will usually manage in daylight because the plants and algae are producing. Switching the air off at dusk removes it exactly when demand is highest and supply is lowest. Run it continuously, and add more during hot spells.
It helps genuinely, since falling water is real gas exchange rather than just movement. But for a well-stocked, well-fed pond in mid-summer a waterfall alone often is not enough, and it only works while the main pump is running. Most keepers with a serious fish load add an air pump and bottom diffusers as an independent source.
With an electronic DO meter. There is no practical, accurate drop test for oxygen the way there is for ammonia or nitrite. A good handheld meter is expensive and gives you a spot reading at the moment you dip it, which is why the pre-dawn minimum usually goes unrecorded. A continuous sensor is what captures that low point.
In practical terms you are very unlikely to harm koi with too much air. The realistic downsides are that heavy aeration strips carbon dioxide and lets pH climb, that it can stir bottom sludge if a diffuser is badly placed, and that it will chill the water faster in cold weather. In summer, err on the side of more.
Both drive oxygen down. A storm brings a barometric drop, heavy cloud that shuts off photosynthesis, and warm organic runoff. Several pond treatments, formalin-based ones especially, consume oxygen or impair the fish's ability to take it up. Aerate hard before and throughout any treatment, and do not dose a pond that is already showing fish at the surface.
H2Koi watches dissolved oxygen continuously, including the hours before dawn when nobody is at the pond. Understated, honest instrumentation, built for koi keepers who would rather know early.
Explore H2Koi
