05 Food
Feeding goldfish through the seasons
Digestion slows as the water cools and goldfish eat less of their own accord. The seasonal rule worth following is to read the thermometer rather than the fish, and many keepers stop feeding pond fish altogether at about 8 to 10 °C.
A goldfish has no internal thermostat. Its body temperature is the water’s temperature, and with it goes the speed of the chemical processes in the animal — including digestion. That single fact produces a seasonal rule that runs against instinct, and that catches keepers out every year.
The rule is that you feed by water temperature, not by appetite, not by air temperature, and not by the calendar.
Why cold slows digestion
The anatomy page covers ectothermy in full; the relevant part here is narrow. As a rough rule, biological processes roughly halve in rate for each 10 °C drop. Apply that to a gut which — as the nutrition page explains — has no stomach, and the consequence is a steep slowing rather than a stop: in one marine fish, the kingfish, food took about three times as long to clear the gut in the cold, and goldfish eat progressively less of their own accord as the water falls from 28 °C to 15 °C.
Several things slow at once: the movement of food along the tract, the digestive enzymes (in one carp species at 10 °C amylase, protease, lipase and trypsin were all reduced), and the fish’s whole metabolism, so there is little demand pulling the process along.
The reasons for feeding little or nothing in the cold are plain enough: the fish eat less of their own accord, digestion is slow, and what is not eaten fouls the water. Many keepers stop feeding pond fish at about 8 to 10 °C as a precaution, and cut the ration well before that.
The bands
| Water temperature | Feed | How often |
|---|---|---|
| Above 20 °C | Normal ration, quality staple | 1–3 times a day |
| 15–20 °C | Normal food, easing the ration as the water falls | Once a day |
| 10–15 °C | Small amounts; wheatgerm-based if you use one | Every other day or less |
| 8–10 °C | Nothing, or a trace on a warm day | Rarely |
| Below 8 °C | Nothing, the usual practice | — |
Two notes on reading that table. The bands are keepers’ conventions rather than measured limits, and they are about the water, measured with a thermometer in it, not about the weather. And the transitions belong to the direction you are moving: coming down in autumn you taper through them, and coming up in spring you climb them cautiously, for reasons the next sections cover.
Goldfish temperature ranges
- Cold below about 8 °C Pond fish slow right down; stop feeding (a pond convention, about 8–10 °C).
- Cool about 8–18 °C Fine for single-tails; fancies eat and move less.
- Comfortable about 18–24 °C The range most keepers aim for.
- Warm about 24–27 °C Metabolism up, oxygen down. Watch aeration.
- Hot above about 27 °C Oxygen runs short and parasites cycle fast. Goldfish survive it, but aerate and cool.
Measuring the temperature that matters
The instruction “measure the water” hides a step that decides whether the measurement is any use, and it is a step many people skip.
Measure at the depth the fish are using. In winter that is the bottom. Water is at its densest at about 4 °C, so as a pond cools the coldest water rises and the 4 °C water sinks — which is why a pond freezes from the top down, and why the bottom of a deep one holds steady at around 4 °C while the surface is at zero. A floating thermometer reads the surface film and can be several degrees out from where the fish actually are. Use a probe on a lead, or lower a thermometer and leave it a few minutes before reading it.
Take the reading at the same time each day. A shallow pond or an unheated indoor tank swings between afternoon and dawn, and a reading taken on a sunny afternoon flatters the average considerably. If you have to pick one time, early morning is the reliable one.
Watch the trend, not the number. A single reading tells you where you are; three or four across a week tell you where you are going, which is what the feeding decision actually depends on. The testing guide makes the same argument about the chemical parameters, and it applies with even more force here.
Autumn: the season that sets up the winter
Autumn is easy to overlook, and it affects how a fish comes through the winter.
A goldfish going into winter lives partly on what it built in autumn. Through September and October the water is still warm enough for proper digestion while the fish’s appetite is strong, and that window is when condition is laid down — body reserves that will carry it through several months of eating little or nothing.
So autumn is a season for feeding well: a full ration of a good staple, with the protein and fat the nutrition page describes, while the fish can still use it. Skimping here produces a fish that enters winter thin, and a thin fish is more vulnerable in spring.
Then taper as the water drops. Through the mid-teens, reduce the amount; through the low teens, many keepers switch to a wheatgerm food; at about 8 to 10 °C, stop. Do not stop abruptly on a date — the fish should arrive at the bottom of the range already eating very little.
Autumn is also the right time for the year’s last real maintenance: a thorough water change, leaves and debris out of a pond, the filter serviced while there is still time for the bacteria to settle. Whatever is left undone in October stays undone until March.
Winter: nothing
Below about 8 °C a goldfish eats little or nothing of its own accord, and most keepers feed nothing at all. In a pond it will settle near the bottom, become almost motionless, and stay there. This is normal, it is what the species does, and it is not distress — and it is worth knowing the difference between that and the sitting on the bottom that means something is wrong, which is a fish at summer temperatures doing the same thing.
Three things matter more than food during this period.
Gas exchange. A pond that ices over completely traps the products of decomposition beneath it. Keep a hole open, and melt it with a pan of hot water rather than smashing the ice.
Not stirring the deep water. A pump running all winter at the bottom of a pond mixes the warmer deep layer up to a freezing surface and destroys the one stable place the fish have. Either lift the pump to a shallower position so it moves only the upper water, or shut it down for the season. The same logic argues against large winter water changes: a big volume of near-freezing tap water does more harm than the nitrate it removes.
Leaving them alone. A fish in winter torpor should not be disturbed, netted or moved unless something is genuinely wrong, because its energy reserves are finite.
Spring: the dangerous one
Spring is a high-risk period for pond goldfish.
A fish emerging from winter has spent months eating little, and its energy reserves are at their lowest of the year. Disease in carp-family fish often appears as the water warms: the World Organisation for Animal Health describes spring viraemia of carp as often observed in spring, and says that poor condition in over-wintered fish may be a contributory factor.
Spring is when fungus, ulcers and white spot often turn up in ponds, and the fish that struggles is often the one that went into winter in poor condition — which returns the argument to autumn. The diseases overview sets out why a weakened animal against an ordinary background of organisms is a common recipe, rather than anything arriving from outside.
Feeding through the transition needs the same caution in reverse:
- Wait for a settled rise above about 10 °C, measured at depth, for several days.
- Start small, with wheatgerm if you use it, every second or third day.
- Build over two to three weeks as the water climbs through the teens, rather than resuming a summer ration the moment they look interested.
- Restart maintenance gently too — small water changes first, and the filter back in service before the feeding increases rather than after, since the bacterial colony has to catch up with the load.
- Watch them closely in this window. It is the time of year when a daily look is most likely to catch something early, as the healthy baseline page describes.
Why wheatgerm exists
It is a traditional food rather than a winter one: a shoulder-season food for the cool weeks at either end.
Wheatgerm foods are marketed as easier to digest in cool water, and that is a manufacturer claim; there is no controlled evidence for it. One major brand’s own advice is simply to reduce the amount fed below about 10 °C, and its wheatgerm food has the same minimum protein as its staple. Keepers reach for it in the cool weeks because a rich pellet fed at 12 °C to fish that are eating little leaves food uneaten.
So the traditional sequence through autumn is staple → wheatgerm → nothing, and through spring it is nothing → wheatgerm → staple. The wheatgerm weeks are the transitions at either end, not the middle.
The tub that sat in the shed all winter
A detail specific to seasonal feeding, and one few people think about: the food you stop using in October is the food you reach for in March.
Fish food degrades once it is opened. Fats oxidise, and the vitamins — vitamin C in particular — break down steadily as air and moisture get in, faster in a warm or damp place. A tub that has been open since the previous spring and spent the winter in a shed is not the same food that went into it, and the fish most in need of a complete diet is the one coming out of winter.
So the sensible practice is to buy the autumn food in a quantity you will actually finish — the Merck Veterinary Manual says a tub kept at room temperature should be discarded two months after opening — and to start spring on a fresh tub rather than the remains of last year’s. Keep it cool, dry, dark and sealed, and throw out anything that smells rancid or has clumped. The nutrition page covers what a complete food has to contain; none of it survives indefinite storage.
Summer, and the other constraint
Warm water does not simply mean feed more. It means the fish can digest more, and it also means the water holds less oxygen while the fish needs more of it — the combination the aeration guide describes and the heaters and cooling page covers from the temperature side.
So the summer adjustment is not only about quantity but about timing and margin:
- Feed lightly, and not late in the day, in hot spells: a meal raises a goldfish’s oxygen use for about seven hours, and pond oxygen is lowest around dawn and highest in late afternoon.
- Reduce the ration during a heatwave, because digestion itself consumes oxygen and a fasted fish has more margin.
- Watch for gasping at the surface, which in hot weather is an oxygen problem rather than a feeding one, and which a reduced ration helps with.
- Never compensate afterwards. A hot week with light feeding is not a debt to repay.
In very hot weather, feeding lightly or skipping a few days is a reasonable response for a healthy adult, which is the same principle as winter arriving from the other direction.
The exceptions
The bands assume a healthy adult fish. Three cases need handling differently.
Young fish. A first-year goldfish has little stored reserve, so the bands above do not suit it. It also has growing to do, and months without food at that stage may leave it smaller than well-fed fish of the same age, though many catch up partly or fully when feeding improves. Fry and small juveniles are often overwintered indoors in a heated tank, where they keep feeding and keep growing, and go out the following year at a size that can sit out a winter. The growth page covers how the early years shape size.
Fish that are unwell going into autumn. A goldfish with an ulcer, a fungal patch or an unexplained loss of condition is unlikely to improve in cold water, and treating it there is harder. Deal with it in autumn while the water is still warm, or bring the fish indoors and treat it there. Hoping it will hold until spring is a decision that can turn a treatable problem into a dead fish.
Very old fish. An elderly goldfish may come through winter with less margin than one in its prime, and its autumn condition is worth more attention. Feed it well while the water is still warm, and if it goes into the cold visibly thin, consider overwintering it somewhere warmer.
Indoor tanks
None of this applies to a tank in a heated room at a steady 20 to 22 °C, which is most indoor aquariums. The routine does not change across the year, and the how often to feed guide covers it in full.
It does apply wherever the water genuinely goes cold: a conservatory, a porch, an unheated hallway, a room whose heating is off while the house is empty. A tank in any of those can spend weeks in the low teens, and the same bands apply on a smaller scale.
The practical difference indoors is that the change can be fast — a house left unheated for a fortnight cools faster than a pond does — and fast change is the thing goldfish handle worst. If an indoor tank routinely drops into that range, the heater question is worth revisiting, since a heater set as a floor removes the whole problem. The same thought applies before a winter trip, which the leaving them page goes into.
Read the fish as well as the thermometer
The bands are a framework, not a rule with legal force, and two observations refine them.
Is food being eaten promptly? Food still sitting on the bottom several minutes after it went in is food the fish are not interested in, and at cool temperatures that usually means you are past the point of feeding rather than that they are being fussy. A genuine loss of appetite in warm water is a different matter entirely.
What do the faeces look like? Long, pale, trailing strings in cool weather can mean the ration is more than the gut is handling — the picture the constipation page describes — and keepers treat it as a prompt to cut back.
Both of those are more informative than how enthusiastically the fish come to the surface, which they will do at almost any temperature.
When it has gone wrong
| What you are seeing | What it usually means | What to do |
|---|---|---|
| Food uneaten on the bottom in cool water | Appetite has dropped with the water temperature | Stop feeding; siphon out what is left |
| Long pale strings trailing from a fish in autumn | The ration may be more than the gut is handling | Cut back a band; a lighter food such as wheatgerm |
| A fish bloated and listless after a winter feed | Constipation or illness; the cause is not certain | Stop feeding; test the water; do not reach for a medication first |
| Fish at the surface under ice | Oxygen, not hunger | Open a hole; still do not feed |
| Fish thin and ragged in March | Probably went into winter underconditioned | Restart very gently; plan next autumn differently |
| Fungus or sores appearing as the water warms | A common spring pattern in carp-family fish | Treat promptly; warming water helps rather than hinders |
The year in one table
| Season | Water, typically | Feeding |
|---|---|---|
| Late spring to summer | Above 20 °C | Full ration, staple, once to three times daily |
| Heatwave | Unusually hot | Feed lightly, and not late in the day |
| Early autumn | 15–20 °C | Normal food while the water is warm, easing as it falls; this is the conditioning window |
| Late autumn | 10–15 °C | Taper; wheatgerm if you use it; last full maintenance |
| Winter | Below 8–10 °C | Most keepers feed nothing. Keep a hole in the ice, leave the deep water alone |
| Early spring | 10–15 °C | Restart small, every 2–3 days |
| Late spring | 15–20 °C | Build back to normal over 2–3 weeks; watch closely |
Frequently asked questions
At what temperature should I stop feeding goldfish?
Around 8 to 10 °C is the line many keepers use for pond fish, and below it most feed nothing at all. Digestion slows steadily as the water cools rather than stopping at a line, and goldfish eat less of their own accord as it falls, so by that point very little is going in; stopping is a precaution rather than a measured threshold. Between roughly 10 and 15 °C, feed sparingly, and a wheatgerm-based food is the traditional choice; above 15 °C return gradually to normal.
Isn't it cruel not to feed them all winter?
It feels that way, but a goldfish in cold water has little appetite of its own. Goldfish are temperate fish whose wild ancestors ate little or nothing in winter and lived partly on reserves built in autumn, and an adult in good condition generally comes through the winter on those reserves. Feeding a fish whose appetite and digestion have both dropped gains little, and uneaten food fouls the water. The fish may still beg; begging is foraging behaviour, not a reliable hunger report.
What is wheatgerm food for?
It is the traditional cool-water food, and makers market it as easier to digest in cool water. That is a manufacturer claim: there is no controlled evidence for it, and one major brand's wheatgerm food has the same minimum protein as its staple, so the idea that it is lower in protein does not always hold. Keepers use it as a shoulder-season food for the cool weeks at either end, when feeding is light anyway, and not as a winter food.
We had a warm week in January. Should I feed them?
No. Air temperature and water temperature are different things, and a large body of water lags the air by days — a mild week may raise a pond by a degree or two while the fish remain well inside the range where they eat little and digest slowly. Feeding on the strength of a warm spell also means the food is still being worked through when the cold returns. Measure the water, not the forecast, and wait for a sustained rise rather than a pleasant afternoon.
Does any of this apply to an indoor tank?
It applies wherever the water actually goes cold, which indoors is a question about the room rather than the season. A heated living room never enters these bands and the feeding routine does not change all year. A conservatory, a porch, an unheated hallway or a house whose heating goes off for a fortnight can drop a tank into the low teens, and the same logic then applies on a smaller scale — feed less, and judge it with a thermometer rather than a calendar.
When do I start feeding again in spring?
When the water has been reliably above about 10 °C for several days, not on the first warm afternoon. Start small — a fraction of a summer ration, once every few days, on a wheatgerm food if you use one — and build over two or three weeks as the water climbs through the teens. Spring is a high-risk period: disease in carp-family fish often appears as the water warms, and fish that came out of winter in poor condition are more vulnerable, so a heavy early feed is best avoided.
Should I feed more in autumn?
Feed well in autumn, which is not quite the same as feeding more. The fish is building reserves it will live partly on for several months, and while the water is still warm enough for it to eat and digest comfortably it should be having a full, good-quality ration. As the water drops through the teens, taper rather than stopping abruptly, and many keepers switch to a wheatgerm food for the last few weeks. Autumn matters for how a fish comes through winter, and many keepers overlook it.
Where should the thermometer go in a pond?
At the depth the fish are using, which in winter is the bottom. A floating thermometer reads the surface film, and in a deep pond in January the surface can sit near freezing while the bottom holds around 4 °C — a difference big enough to change what you do. Use a probe on a lead, or a thermometer you can lower and leave for several minutes, and take the reading at roughly the same time of day each time so you are comparing like with like.