Koji, tempeh and natto

Three organisms that are not woodland mushrooms: a mould on rice, a mould on soy, a bacterium that pulls threads.

They are not woodland mushrooms

Koji, tempeh and natto share a page only because all three transform a cooked food in a few days, and because none of the three makes a cap. They are different organisms, with different temperatures, and mixing them up is how you get a tray you can no longer read.

White koji is Aspergillus oryzae, a mould. Its work on rice is to make enzymes, above all amylase, which break starch into sugars. Tempeh is Rhizopus oligosporus, another mould, and its work is to bind soybeans into a white cake. Natto is Bacillus subtilis, a bacterium, and its work is to make the beans pull threads. Three names, three signs of ready, three ways to fail. None of them is grown in a chip bed and none of them is fruited by cutting a slit in a bag of straw.

This page stops at the first transformation: rice that smells of sweet chestnut, a tempeh cake that is firm, beans that pull threads. Miso, soy sauce, sake, garum are later fermentations, with salt, sometimes with yeast, and with times of months. They do not come out finished from the forty-eight hours. And nothing you read here is a talk about the body. They are foods, when they succeed, and they are discarded, without tasting, when the signs are the wrong ones.

Koji rice, still in grains

White koji wants steamed rice, the grains whole and separate. You rinse, you soak, you steam until the grain is cooked at the heart and is not glue. Boiling it in too much water makes a porridge. In a porridge there is no air between one grain and the next, and the mould needs surfaces and oxygen. A raw grain in the centre, hard and white, is the opposite mistake: the enzyme does not enter where the water never arrived. Spread the rice. Let the free water leave. The grains should be moist and matte, not shining with broth.

Then cool below 35 °C before the mould touches the rice. Above that threshold you kill the inoculum at the moment you believe you have seeded. The pile lies: the surface is lukewarm and the heart is still too hot. Spread it, measure, do not trust one finger. Salt does not go in during these hours. Salt comes later, if one day you make miso. Salting now slows the mould you are trying to grow.

Hold the rice near 30 °C for 36–48 hours. In that window the mould runs, it releases amylase, starch becomes sugar, the smell moves from boiled rice to sweet chestnut, and the grains bind in a white felt. Growth heats. A thick pile, even in a cool room, passes 38 °C at the heart. At that point the enzymes you wanted are damaged, the run slows, and the smell goes from sweet to flat or sharp. Mix when it passes 38 °C. Mixing dumps heat and brings inside the grains that were left on the outside. A small pile in a perforated tray needs you less. A kilo closed in a deep bowl needs you without question.

It is ready when the felt is white, the grains are still visible, and the smell is sweet chestnut. Not when it is grey, not when it stinks. If you leave it longer, the mould sporulates and may haze with a pale dust. That dust, if the smell stays chestnut, is the same Aspergillus you inoculated, arrived at the end. For a first batch you do not need to get there. Stop at the white felt. A saturated green, a black, a pink, a smell of ammonia, of rot or of solvent are not “sporulation a little further on”. They are another organism, or a mould that has lost. You do not taste to decide. You close it and you discard it.

The white koji slant is the mother for the fridge. It is not rice already inoculated and it is not a meal. It lasts for months in the cold, between 2 and 8 °C, never in the freezer. When you need it, in still air, you take a trace and mix it into rice that is already below 35 °C. You do not need to empty the tube. Recap it and put it back. A tube opened over a steaming pot is a tube you have given to the kitchen.

Koji Bianco Aspergillus oryzae Slant Agar | Coltura Premium€17,90

The white koji plate is the same mould, already grown, so you can see a clean white colony before you touch the rice. A green edge or a wet film is not used. From a clean plate you take a fragment and mix it into the grains. Left warm, the plate sporulates and dries. Its place, afterwards, is the fridge.

The character you want in this first rice is the saccharifying one: more amylase, more sugar. That is the work called amilo, the work of amazake and of the first step of sake. Amazake, in short, is finished koji mixed with more cooked rice, held warm enough that the enzymes work and the mould does not keep growing. It is not a second koji. Sake is further on again: after the sugars a yeast comes in, and it does not finish over a weekend. Learn the felt and the chestnut smell on rice. The rest makes sense only once that sign is clear to you.

Proteo is not the first batch

A proteolytic strain is selected to break proteins, not to turn rice into a sweet. It is for projects such as garum and meat koji, where the substrate is not a grain and the reference smell is not chestnut. The proteo slant is that tool, later. The word “green” in the name is the name of the line, not the colour of a batch that worked. A rice, or any food, that turns a saturated green is discarded.

Do not start there. On meat or fish you do not learn the sign of ready, and a protein ferment done badly is not corrected by optimism. Salt, long times, and a method that does not fit on this page: the first batch is rice and white koji. When you can recognise the felt, the sweet smell and the moment to mix for heat, you will have a criterion. Without that criterion proteo is only a harder tube.

Tempeh: dehulled soy, surface-dry

Rhizopus oligosporus is not Aspergillus. It does not want koji’s 30 °C for two days on rice. It wants soybeans that are dehulled, cooked, dried on the surface, in a thin perforated layer, around 30–32 °C, for about a day.

The hull comes off because it stops the mycelium welding one bean to the next. The bean is cooked through, then dried on the outside: a cloth, a fan, until the surface is matte and does not drip. The inside stays cooked and moist. If you inoculate wet beans, bacteria make a jelly before Rhizopus makes a cake. If you dry them until they wrinkle, the mould has no water. Matte, not dry.

The Rhizopus oligosporus liquid culture is the seed. Do not pour the whole syringe over a handful of soy. A few drops, or a dilution in a little boiled and cooled water, is enough to wet the beans without undoing the drying you just did. Ten millilitres is for more than one batch. The rest goes back in the fridge. Spread the beans in a layer a few centimetres thick, in a container or a bag with holes. Rhizopus wants oxygen. A sealed jar heads toward ammonia. The holes are the method, not an optional. The thin layer is how heat gets out: this mould too, if the mass is too deep, cooks itself in the middle.

Rhizopus oligosporus Slant – Starter per Tempeh€17,90

At about a day, look. A ready cake is white, firm, it holds when you lift it, and it smells of fresh mushroom, not of waste. If at twenty-four hours it is still a pile of beans, give it a few more hours, not another forgotten day. Stop at white. Black, on this mould, arrives when it sporulates because you left it too long, and by then it is no longer the batch you wanted. You do not learn by tasting how to separate that black from a black that is another organism. If it is green, black, pink, if it smells of ammonia, of rot or of solvent, it is not tasted. It is discarded. A soft cake is not tempeh. It is soy that did not bind.

It is eaten cooked. Raw tempeh is not the destination. And do not incubate it in the same crate where you fruit mushrooms: spores travel, and a Pleurotus does not improve inside a bag of soy, nor does Rhizopus improve an oyster block.

Natto: about 40 °C, and the threads

Bacillus subtilis does not make felt. It makes a mucus that pulls. Cooked soybeans are inoculated and held around 40 °C for 18–24 hours. That is warmer than koji and warmer than tempeh. At 30 °C it may not thread in a day. At 50 °C you can kill the bacterium. A shallow container, covered but not sealed: a little air, no draught that dries the beans. Perforated film or a cloth is better than a lid screwed down hard.

The natto slant is the mother, a tube, not a fruiting syringe. You take a trace into a spoon of boiled and cooled water, mix it into the beans, and the tube goes back in the fridge. Do not salt. Salt, now, stops the bacterium.

Spore Natto Bacillus subtilis – Slant Agar Fermentazione€17,90

It is ready when the beans are intact and, stirred, they pull threads. The smell of proper natto is strong. That is the point that confuses. A successful batch is pungent with fermented bean, with a clean note that resembles ammonia and is not rot. If you have never smelled natto, make a little the first time and compare it with a shop natto before you scale up. Failure is a different smell: putrid, solvent, or a dirty ammonia together with collapsed beans, a pink, a black, a mush. You do not taste that to “see if it is only the natto smell”. You discard it. Threads, intact beans, and a smell you recognise: that is the sign. Without threads, after twenty-four hours at the right temperature, it is not a shy natto. It is a batch that did not start, and stretching it forever in the warm will not correct it. You start again, with beans cooked through and with a real trace of culture.

Natto too is cooked or used the way natto is used, in small amounts if you do not know it. It is not a preparation for an effect. It is a thread-pulling bean. If the tray seems wrong to you, the decision is already made: out, without a test spoon.

Keep the three apart

The temperatures tell the separation better than the names. Koji near 30 °C, and you mix if it climbs past 38. Tempeh at 30–32 °C, a thin layer, one day, very white. Natto around 40 °C, less than a day, threads. A single badly set appliance fails all of them. A thermometer in the food, not in the room, is the tool. A room at 30 °C can mean a heart of rice at 40.

Spores and bacteria stay in the kitchen after you finish. Wash what touched the batch, and do not use the same cloth to dry the tempeh beans and to cover a mushroom block. Keep the mothers capped, labelled, in the fridge. A tube with no name, in a year, is not a reserve. It is a risk.

The first useful success is small. A handful of rice with the right felt teaches more than a kilo that overheated. A thin tempeh cake teaches more than a dense brick that went black in the centre. A cup of natto that pulls threads teaches more than a pot that spent the night at 50 °C. Scale later, when you recognise the sign of ready without negotiating with a smell that does not convince you.

What this guide is not

It is not a key to the woods, it is not a garum manual, and it is not health advice. It is the short cycle of three organisms: a mould on rice, below 35 °C at the start and near 30 °C for two days; a mould on soy in a perforated layer for about a day; a bacterium on the beans at about 40 °C until they pull threads. Green, black, pink, ammonia, rot, solvent: discard without tasting. Sweet chestnut, the white cake, the thread. The rest can wait.