These two get mixed up a lot. Both are tall, spiny-trunked Pachypodiums from Madagascar with leaves only at the top, and young plants of both are often sold side by side. I’ve had both for a few years now: one big P. geayi, bought around 2022 when it was already eight or nine years old, a large old P. lamerei, two P. lamerei var. ramosum, two lamerei seedlings of my own and six geayi seedlings. Below is what I look at to tell them apart. It’s all from my own plants, so don’t read it as a botanical key.

Leaf width and colour

On grown-up plants, lamerei leaves are about twice as wide as geayi leaves. (On seedlings it’s a different story, which I come back to further down.) My geayi has long, narrow leaves that hang out from the top almost like a fountain, and they’re a darker, muted green. Lamerei leaves are a lighter green, and with light behind them they look a little see-through compared with the geayi.

My two var. ramosum have leaves a bit wider than plain lamerei, and a touch lighter in colour. The difference is small, so I wouldn’t use it on its own to say which one you have.

How the spines sit

Both species have spines in groups of three, so counting them doesn’t help. What helps is how they sit. On my old geayi, all three spines in a group are thick and about equally strong, and they point upward along the trunk. On lamerei, one spine in each group is shorter, and the two longer ones stick out flat to the left and right. So from a bit further away, the spines on a lamerei seem to lie sideways around the trunk, and on a geayi they mostly point up.

Trunk colour and texture

My old lamerei has a slightly lighter trunk than the geayi, and the ramosums are lighter again. The texture is harder to put into words. The geayi’s skin is smoother, but the trunk itself is lumpier, with more bumps and dips under the spines. The lamerei’s surface is flatter but looks more worn, almost weathered. On seedlings I can’t see any difference in the trunk at all, so I don’t use it until the plants are a few years old.

Var. ramosum and branching

Ramosum is supposed to be the branching form of lamerei, and mine started branching early. I bought my two at about four years old, and one of them already had a small new branch, so the branching probably began at around three. None of my other seedlings have branched yet. My old lamerei has two heads too, but only because I cut it back years ago, so a branched lamerei isn’t automatically ramosum.

Telling geayi and lamerei seedlings apart

Seedlings were the hardest for me, because most of the differences above don’t show yet.

Young lamerei leaves are much narrower than on an old lamerei. They’re about as wide as the leaves on my grown geayi, and geayi seedling leaves are the same width. The colour is the same too: lamerei seedlings start out dark green, very close to the geayi seedlings, and only get lighter as they grow.

The only difference I can see is length. Geayi seedling leaves are longer, lamerei ones shorter. So if you have a tray of young Pachypodiums of the same age and don’t know which is which, look at leaf length, not width or colour. On my seedlings, the length difference showed after about 18–24 months.

Growing geayi and lamerei

Both have been easy. I haven’t lost a single geayi or lamerei, old or young, including the winters on my unheated balcony, which sat mostly at 5–10°C (41–50°F) and close to 0°C (32°F) for a couple of weeks. Neither has flowered yet, and I don’t expect flowers for another ten years or so.

They grow in similar mixes: my geayi seedlings are in about 70% mineral and 30% organic, and the big geayi in 80–85% mineral. The full recipes are in my soil mix post. The one real difference in care is winter leaf drop: my old geayi loses about 80% of its leaves, the old lamerei 40–50%. I wrote about that in last week’s post on why my Pachypodiums lose their leaves in winter.

What’s next

In spring I want to take a photo of a geayi and a lamerei seedling of the same age next to each other. I didn’t have anything like that when I started. Next week is something completely different: my Welwitschia mirabilis seedlings, from the 2024 and 2025 sowings.

Subscribe now

I’m not a Pachypodium expert. I grow a few species, mostly lamerei, geayi and namaquanum, and I’ve had them through several unheated winters on my glazed balcony before moving everything into the greenhouse this year. This is what I’ve seen with them.

Seedlings vs. older plants: why only some Pachypodiums drop leaves in winter

If your Pachypodium is young, it will probably keep most of its leaves through winter. My own Pachypodium lamerei seedlings did, every year so far. They lost a few leaves in spring, when the new ones started coming in, and that was it. The same went for the two Pachypodium lamerei var. ramosum I bought at around four years old. Here are my lamerei seedlings over three winters on the balcony:

Older plants are different. They lose leaves slowly over the winter, a few at a time, starting with the lowest ones in the rosette. My Pachypodium geayi, which was already eight or nine years old when I bought it around 2022, loses the most. By the end of winter it has dropped around 80% of its long silver leaves. My old P. lamerei loses less, maybe 40–50%, and keeps a fuller rosette at the top. In spring both grow new leaves from the top, and over the season the old leaves are gradually replaced by new ones.

This is what they looked like on October 17. On all the older plants it's the lowest leaves that turn yellow first, while the new leaves at the top stay green. The lamereis started earlier and are yellowing faster than the geayi, but they usually keep more of their upper leaves through winter. My old geayi starts later and ends up losing more. My young geayi seedlings don't lose many leaves at all, but in many species seedlings behave differently in their first years.

So an older plant losing leaves gradually in winter, with leaves yellowing from the bottom up, is normal. To check whether something is actually wrong, I look at the stem rather than the leaves.

Pachypodium namaquanum: leaf drop in summer instead of winter

Namaquanum works the other way around. Mine is about 11–12 years old, and I’ve had it for four or five years. It drops its crinkled grey leaves at the end of spring, stays bare through the hot months, and grows new leaves at the end of summer. So if your namaquanum loses its leaves in June or July, that’s its normal rest, not a problem. I’ll write a separate post about how I grow it next month.

My namaquanum seedling, now three years old, has kept its leaves all year so far.

How to tell dormancy from rot: squeeze the stem

When one of my Pachypodiums drops leaves, I gently squeeze the stem between the spines. If it’s firm all the way down to the soil, the plant is fine and just resting. If it feels soft anywhere, especially near the base, that’s rot.

This is how I lost one of my namaquanum seedlings. I had two of the same age, and both got through their first winter on the balcony without problems. In their second winter, one of them died. That winter wasn’t even colder; the balcony stayed at 5–6°C (41–43°F) or above. There were more humid days, though, and even with the seedlings kept dry, it wasn’t enough. I only found it in March. The base was soft and looked odd near the soil, but the top of the plant still looked almost normal, so I hadn’t noticed anything earlier.

My earlier Pachypodium losses happened faster. P. cactipes, P. horombense and one baronii (or windsorii, I’m still not sure which) all died during their first winter with me, before spring came. Looking back at all of them, I think the problem was humidity together with cold. Cold on its own, with dry air, my plants handled much better. I wrote more about this in my post on minimum winter temperatures for caudex plants and cacti.

Soil makes a big difference here too, because a mix that stays wet for days in cold weather is exactly what leads to rot. My big P. geayi is in a mix that’s about 80–85% mineral, my geayi seedlings are in about 70% mineral and 30% organic, and my P. namaquanum is in a fully mineral mix. I explained all my mixes in last week’s post on soil mixes for caudex plants.

Winter watering for Pachypodium

• P. lamerei and P. geayi: I water very little in winter. Just enough now and then that the roots don’t dry out completely.

• P. namaquanum: a bit more, around once every three weeks, and only on warmer days when the air is drier, so the pot dries out again quickly.

• Seedlings: the same as the adult plants of their species, but they get the warmest spot.

When to start watering again in spring

I start watering again depending on the weather. If early March is warm, I start then. If it’s still cool, I wait until we get days of 20°C (68°F) and above. New leaves appearing at the top of the plant are a good sign that it’s ready.

What’s next

This is the first winter my plants will spend in the greenhouse instead of on the balcony. I’m setting the heater thermostat to 8°C (46°F), and I’ll keep an eye on humidity as well as temperature.

Next week’s post stays with these plants: how to tell Pachypodium geayi and P. lamerei apart, since they’re easy to mix up when they’re young. Later I’ll also write about my P. namaquanum and its mineral mix, and in spring about growing Pachypodium from seed.

October is when I look at every pot in the greenhouse. The Dioscoreas are pushing new shoots, the Pachypodiums will start dropping leaves in a few weeks, and this is the last moment to fix anything before winter. This year it matters more than usual, because it’s my first winter in the greenhouse instead of the balcony. Almost every winter loss I’ve had came down to soil staying wet for too long in cold, humid air. So before the thermostat goes on, I want to write down what’s in my pots and why.

I’ll link back to this post a lot. It has my own mixes plant by plant, the Japanese mixes I learned from growers’ blogs years ago, and a table comparing all of them.

How my mix changed: from 50/50 to 70/30

When I started, I used what most beginner guides suggest: about half organic, half mineral. My usual recipe was perlite 50%, peat 25% and ordinary potting soil 25%. In some pots I also put a layer of lava rock at the bottom for drainage. It looked fine for a few months.

Then late autumn came on the closed balcony. Humidity sat around 60%, temperatures were 5–10°C (41–50°F), and those pots stayed damp for days after watering. The surface went green with algae. Summer was no better, because the mix also dried too slowly when the plants actually wanted water and air at the roots. I lost plants in both seasons before I accepted that the mix was the problem.

Over the next couple of years I moved to 70% mineral and 30% organic as my starting point. From there I adjust per plant, and most of my pots now end up between 70% and 80% mineral. I also stopped putting lava rock in the bottom of pots. With a mix this open, I find it unnecessary. I mix by the scoop and by eye. The ratios below are what I aim for, but I’ll often add a handful more pumice to a batch if it feels too soft in the hand. If you want precision, measure by volume, not weight. Peat and perlite weigh almost nothing compared with pumice.

What goes into my caudex soil mix

• Perlite. I buy the coarse grade, 5 mm (about ¼ in) or bigger. Fine perlite packs down and floats to the top after a few waterings.

• Pumice. Plain local pumice in mixed sizes. Heavier than perlite, so it keeps tall plants steady in the pot.

• Peat and coco coir. These hold the water. I like having both. Coir rewets more easily after the pot dries out completely.

• Compost (humus). Only a small amount, for some life in the mix.

• Lava rock, zeolite and river sand. Only in my Pachypodium namaquanum mix. Sometimes a little vermiculite or charcoal in other batches, never more than a few percent.

• Top dressing. Most of my pots get a top layer of mixed mineral grit: pumice, lava rock and zeolite. That’s why the surface looks much the same in most of my photos, whatever mix is underneath.

• Slow-release fertilizer mixed in. Recently MagAmp K, before that a 10-11-10 cactus and succulent granule.

How to read the mixes below: all percentages are by volume. Mineral means perlite, pumice, lava rock, sand and the Japanese soils. Organic means peat, coir and compost. Slow-release fertilizer goes into every mix and isn’t counted.

Soil mix for Dioscorea elephantipes

My mature Dioscorea and most of my seedlings are in 80% mineral now, the same as most of my caudex plants. I have a lot of seedlings, though, and they aren’t all in one mix. The youngest get 70/30 with most of the organic part as peat, because their small tubers swell faster when the soil doesn’t dry out completely between waterings. Some older seedlings are still in the wetter mix from my Dioscorea post (perlite 40%, pumice 20%, peat 20%, coir 15%, compost 5%, so 60% mineral), and a few are in something more mineral than 80%. I’m moving them across as I repot, in July and August while they’re dormant. The full care routine is in my Dioscorea elephantipes post.

Soil mix for Pachypodium, Dorstenia gigas, Welwitschia and Fockea

Pachypodium namaquanum

This is the one plant with its own mineral-only recipe. My mature namaquanum spent two years in a Japanese akadama mix (more on that below) until it needed a bigger pot, and has been in this mix since. It’s the Pachypodium I’m most careful with, because I lost a seedling of it last winter to cold and humidity, even kept dry.

Pachypodium geayi

Seedlings go in my 70/30 base. My big geayi, bought when it was already about 8–9 years old, gets more mineral, around 80–85%. It also did well in the akadama–kanuma–perlite mix further down. It’s a forgiving plant compared with namaquanum.

Dorstenia gigas

With this one, more mineral wasn’t better. I had it in a 100% mineral mix for a long time. It grew, but very slowly, and only kept going if I fed it all the time. Dorstenia gigas likes more water than most caudex plants, so I’ve recently moved it to 20–25% organic. It’s too early to show results, but the soil now stays usable between waterings instead of drying to bare grit in a day.

Welwitschia mirabilis

My Welwitschia seedlings germinated in this mix and are still growing in it. It’s simple and they’re doing well. I’ll probably repot the four of them next spring.

Fockea edulis and Euphorbia obesa

I’ve changed my mind on these two. I had my Fockea edulis in a mix with 30% organic, and I now think that was too much. One of them was in the akadama–kanuma–perlite mix. That plant died in a humid balcony winter, so I honestly can’t say whether the mix helped. Euphorbia obesa is the other one. My losses were in summer, in a 70% mineral mix. My feeling is it wanted something grittier, but the losses could have had another cause. Both now go into 80% mineral.

Cactus soil mix: Ariocarpus and Astrophytum

Ariocarpus

Ariocarpus has a thick taproot that rots easily, so nothing in this pot holds water for long. Mine was sold to me as A. retusus about five years ago. Now that I know the species better, it looks like a hybrid to me.

Astrophytum

I usually split the mineral part roughly between pumice and perlite, depending on what I have.

Seed sowing mix for caudex plants and cacti

For seed I use shallow rectangular bonsai-style pots. I used to sterilize the mix before sowing. After a few rounds I couldn’t see any difference in germination or losses, only more work, so for the last couple of years I’ve stopped. What I do instead is treat the trays with Physan 20, because constantly wet seed soil grows algae fast. Back when I was propagating a lot, I also rooted cuttings in 100% perlite several times with good results, but I don’t do that any more.

Japanese mixes with akadama and kanuma

Before I settled on my own mixes, I learned a lot from Japanese growers’ blogs. Their recipes are built on two soils that are standard in Japan and rare elsewhere. Akadama is a granular volcanic clay. Each grain soaks up water, while the gaps between grains stay open for air. Kanuma is lighter, softer and acidic, also volcanic. The recipes are written in parts in Japanese; I’ve converted them to rounded percentages.

Akadama–kanuma mix for caudex plants and cacti

All the cacti I tried it on liked it, mostly Astrophytums, and my Pachypodium namaquanum spent two years in it.

Akadama–kanuma–perlite mix for caudex plants

It worked well on the large Pachypodium geayi.

Akadama–kanuma–pumice mix for difficult cacti

This is the first mix without the compost, with pumice added. It suited my Ariocarpus.

The percentages surprised me. These recipes sound rich because compost is in most of them, but the compost is a small share. All three are 90% mineral or more. Japanese growers keep plants this mineral and rely on akadama to hold water inside the grain. I got to 70–80% mineral by a different route, through losses on the balcony.

I stopped using them because I can’t get kanuma any more, and all the Japanese soils are expensive when you’re filling dozens of pots. I only ever had a few litres of each, enough for a handful of plants. For most growers outside Japan, perlite and pumice with a bit of peat and coir do the same job at a fraction of the cost.

All my soil mixes compared

What’s next

The greenhouse heater will hold the minimum at 8°C (46°F) this winter. That’s close to where the balcony sat most winters, minus the couple of weeks near 0°C (32°F), but humidity in a closed greenhouse can still be high. I’ll be watching which pots stay dark and damp longest after watering, and I expect the Dorstenia and the young Dioscorea seedlings to be the ones to keep an eye on. In spring I’ll repot the four Welwitschias and report back on the Dorstenia’s new mix. Minimum winter temperatures for each of these plants are in my winter temperatures post.

If you grow any of these in a mix that works for you, especially Euphorbia obesa or Fockea, I’d really like to hear what’s in it.

Subscribe now

Most of my other caudex plants are slowing down now, and the Pachypodiums are getting ready to drop leaves, while the Dioscoreas are just starting to grow.

In English you’ll see it called elephant’s foot or Hottentot bread. Older books list it as Testudinaria elephantipes. In Japan it’s 亀甲竜 (kikkōryū), “tortoise-shell dragon.” It’s worth searching the Japanese name. Japanese hobbyists have written month-by-month growing diaries for this plant in far more detail than anything I found in English, and much of what I do started with them.

My plants, and why it grows in winter

I bought my mature D. elephantipes years ago as a 5-year-old plant from a grower in France. It’s now about 8–9 years old, with a caudex 13 cm (5 in) across.

I also have one D. sylvatica and a lot of seedlings. There are 10+ at about two and a half years old, 25+ at a year and a half, and a younger batch sown around a year ago. About 50 more seeds went in a month ago and are coming up well.

The oldest photo I have of my plants is from October 2023 on the glazed balcony: two young tubers, each with a thin vine twisting around a wire hoop. That’s where I learned what this plant wants. Everything is in the greenhouse now.

The main thing to know is that D. elephantipes is a winter grower. It comes from a winter-rainfall region, so it grows in the cool months and sleeps through the heat. Treat it like a Pachypodium and you’ll be watering it at exactly the wrong time of year.

My year with it looks like this:

• May · I stop active watering. Leaves yellow, then the vine browns and dries.

• Summer · It sleeps, and it’s just a caudex on gravel. I keep giving light surface watering, weekly or every two weeks. Seedlings get a bit more.

• Late August to September · New shoots. This year the seedling tubers pushed shoots first, in early September. The mature plant followed a little later.

• October to April · Growth, except for a pause in December and January. In the coldest weeks the vine just holds still. Through winter I water lightly every 2–3 weeks, always on a dry, sunny day.

A tip for new owners: you can count the age of an older plant. It sends up one new vine per year from the top of the caudex. If you buy a plant with old cut stubs on the crown, count them. One Japanese grower did this and worked out her plant was starting its fifth growing season.

Watering D. elephantipes through dormancy

Most care sheets say to stop watering completely in summer. I don’t quite do that. Some Japanese growers do exactly that, with no water from June to early September. They only cut off the dead vine in August, once the leaves come away with a gentle shake. The caution is understandable. One grower’s first plant was watered right after she bought it, went soft within two months, and when she pressed it the inside was mush.

Others are more nuanced. Older plants get no water, or a light misting once a month, in summer. But seedlings under two years old get water about every two weeks, ideally on a cool evening. Plants three or more years from sowing can handle a fully dry summer.

My greenhouse summers are very hot: around 35°C (95°F) during the day, big fan on most of the day. In that heat, a completely dry pot worried me more than rot did. So I do something in between and give just a little water on the surface. That keeps the fine roots alive, but the caudex never sits in wet mix.

A few small things helped me. I water around the caudex, near the pot rim, not over it. If you wet the tuber every time, a thin green film of algae can start growing on it. Once the vine is up in autumn the pot dries fast, so expect to water to drainage every 7–10 days or so.

In winter I only water in the morning, on a warm day, never in the evening. A pot that stays cold and wet overnight is how you lose plants. My balcony taught me that the hard way with Adeniums and Pachypodiums, not with Dioscorea.

And if you’re not sure whether a dormant plant is still alive, press the caudex gently. It should feel hard, like wood. Soft or spongy means rot, and more water won’t fix it.

Soil mix for Dioscorea elephantipes

Dioscorea wants a mix that drains well but isn’t bone dry. One point from the Japanese growers I hadn’t seen anywhere else: if drainage is too fast, the tuber swells more slowly. Young plants especially need some water and nutrient retention. This isn’t a cactus mix.

Mix 1 · Japanese style

• Akadama 1 : pumice 1 : peat moss 1 (perlite can stand in for pumice)

• At least three components. A single-ingredient mix ends up too extreme one way or the other.

• Fine-to-small grains for seedlings in 6–9 cm (2½–3½ in) pots, small grains from 12 cm (5 in), medium grains above 15 cm (6 in).

• From 12 cm (5 in) up, a layer of large pumice in the bottom.

Mix 2 · What I use

Perlite 40% Pumice 20% Peat 20% Coco coir 15% Humus 5%

Akadama is great, but for most people outside Japan it’s hard to find and expensive. Perlite, pumice and peat with a few additions does the same job well, and every component is available almost anywhere.

Pot size and planting depth:

• A tuber of 1–2 cm (½–¾ in) goes in a 7.5 cm (3 in) pot. A tuber of 3–4 cm (1¼–1½ in) goes in a 9 cm (3½ in) pot. After that, leave about a finger’s width between caudex and pot wall, and go up roughly one size a year. Too big a pot holds water and rots roots.

• Naturally about half the tuber sits below the soil. In one Japanese grower’s trial of 12 seedlings, shallow-planted ones developed deeper tortoise-shell cracks, while buried ones stayed smooth and pale. If you want the classic plates, raise the tuber a little each repot.

Feeding and pest care

Dioscorea puts on a lot of vine in a short season, and the tuber only grows if the vine is well fed. What I use:

• Slow-release fertilizer in the mix: a cactus and succulent slow-release fertilizer (10-11-10), and more recently Magamp K, a slow-release granule popular with Japanese growers.

• Liquid feed once a month during growth: a cactus and succulent liquid fertilizer (2-7-7). I stop for the two cold months, December and January, when the plants stall anyway. Seedlings benefit most. Mature plants need less.

• Avoid organic fertilizers in a humid greenhouse. They encourage mold.

On pests and disease:

• Seedling trays get Physan 20. Their soil stays constantly wet, which invites algae, and Physan keeps it in check.

• Pesticide twice a year on the whole collection, in March and October. On Dioscorea, watch for aphids on new shoots, spider mites on the leaves, and scale on the caudex.

Repotting in dormancy (July–August)

I repot while the plant is dormant, in July or August, before the new shoot comes. A vine broken during repotting may not regrow that season. Japanese growers who tried spring repotting moved to late summer for this reason. By spring their roots had grown round and round the pots and were pushing out of the slits.

Seedlings grown one per pot make noticeably bigger tubers than ones sharing a pot. I didn’t do this well. I left mine sharing trays and round pots much longer than I should have, and you can see in the photo below how the vines end up tangled together. Next summer the 1.5-year-olds finally get their own pots.

Light, and looking after the vine

It doesn’t want blazing sun. Japanese growers use about 50% shade from April to September and full sun from October. Under strong spring sun, seedlings’ leaves can blacken, with dark veins showing. The caudex itself burns more easily than the leaves. A healthy vine covers it with leaves, so that’s another reason to let it grow.

I don’t use shade cloth. My greenhouse roof already cuts about 30% of the light, and where the plants sit there’s usually at least another 30% shading for much of the day. That has been enough. Under grow lights in winter, around 3,000 lux works. Leaves grey if the light is too close.

The tuber only grows if the vine does, so don’t damage it. One Japanese grower snapped a seedling’s stem moving a pot, and that plant fell clearly behind its siblings for the rest of the year. Another measured her vine at about 3.5 m (11½ ft), then snapped the main stem while re-winding it. So give it a trellis early, let it find its own way, and don’t untangle it later. That’s what the red trellis in my photos is for.

A few things worth buying if you’re starting out

• Shallow rectangular bonsai-style pots for sowing. That’s what you see in most of my seedling photos. They give the tubers room and dry evenly. Ordinary 3-inch (7.5 cm) pots work fine too; I’ve sown in those as well. Once seedlings are big enough, move them on to one plant per pot.

• A slow-release cactus fertilizer plus a low-nitrogen liquid feed. See the feeding section above.

• Physan 20, or a similar algaecide, for seedling trays.

• Fan. Still, cold, damp air is the enemy in winter, and moving air keeps a hot greenhouse bearable in summer.

Can D. elephantipes take cold? What my balcony showed

For several winters these plants lived on my closed, unheated, north-east-facing balcony. It sat mostly at 5–10°C (41–50°F), with a couple of weeks near 0°C (32°F) and about 60% humidity. That’s right in the middle of their growing season. Adeniums and several Pachypodiums didn’t survive those winters. I didn’t lose a single Dioscorea. They simply stalled in December and January, then picked up again as the days warmed.

That fits what Japanese growers report. A 5°C (41°F) minimum is recommended, and a few hours near 2°C (36°F) is survivable. Best growth comes between 10 and 20°C (50–68°F). Below about 10°C they hold still rather than die.

What’s next

This is the first winter these plants spend in the greenhouse, and the first time they’ll get heat during their growing season. A heater with a thermostat set to 8°C (46°F) goes in within the next 4–6 weeks, along with window insulation. I think the December–January stall might be shorter this year. I’ll photograph the mature plant and the 2.5-year-olds every month to see.

The new seedlings get their own post: Growing Dioscorea elephantipes from Seed. For why my thermostat is set to 8°C, see my post on minimum winter temperatures for cacti and caudex plants.

Further reading (Japanese): Dioscorea care guide · seed-sowing diary, month by month · one grower’s year with a mature plant

Subscribe now

What my balcony winters were like

The balcony is closed, glazed and faces north-east, so it never froze outright, but it never got much sun either. A few shelves ran under grow lights to make up for it. Most of the winter it sat at 5–10°C (41–50°F), with a couple of weeks close to 0°C (32°F). Humidity stayed around 60% the whole time, because a closed glass box in winter doesn’t dry out.

I kept almost everything bone dry, as most guides advise. That was enough for the cacti, but not for the tropical caudex plants.

What died, and why

Adenium. Every single one, over several winters: Adenium obesum subsp. arabicum seedlings and mature A. obesum plants. With my Adeniums the caudex went soft from the base up, while the top stayed firm for a while and still looked fine. By the time the top gave way, it was too late. I don’t have a single photo of them left.

Pachypodium. I lost several Pachypodium rosulatum var. cactipes, a P. horombense, and one that was either baronii or windsorii (I was never fully sure which). These two are among the least cold-tolerant Pachypodiums, and in my experience anything kept below about 8°C (46°F) in humid air was in trouble.

Fockea. I had three mature plants and now have one. I also lost both of my Fockea seedlings.

Cold and damp together

Over several winters and several species, I saw the same thing. Adeniums and Pachypodiums died when humidity and low temperature came together. Keeping the soil dry wasn’t enough, because the air wasn’t dry. At 60% humidity and near-freezing temperatures, the roots weren’t active and the tissue stayed cold and wet for weeks. A cold night in dry air seemed fine for them, but weeks of cold, damp air were not.

The cacti were very different. Almost all of mine took 0°C (32°F), kept dry, with no problem. The only one I remember losing to cold was Ortegocactus macdougallii, one or two plants.

Minimum winter temperatures by plant

I checked my experience against PukuBook, a Japanese succulent database that lists a cold-hardiness figure for each species. Those figures are survival limits for a dry plant, not comfortable temperatures.

• Adenium obesum: listed 10°C (50°F). The highest on this list, and I lost every one I had.

• Pachypodium baronii and P. windsorii: listed 5°C (41°F). Mine died in damp air somewhere below 8°C (46°F). I trust my losses over the listing.

• Pachypodium rosulatum, including var. cactipes: listed 5°C (41°F). Stop watering once nights drop below 10°C (50°F).

• Euphorbia obesa: listed 5°C (41°F). Only a light surface watering every few weeks in the cold.

• Fockea edulis: listed 0°C (32°F). About where my balcony got to in the coldest weeks, and I lost two of my three there.

• Dioscorea elephantipes: listed 5°C (41°F). It grows in winter, so it needs light and water while everything else rests.

• Most cacti, kept dry: 0°C (32°F) in my experience. Fine for nearly all of mine. Ortegocactus macdougallii was the exception.

My plan for the first greenhouse winter

In the greenhouse I can control the temperature, which I never could on the balcony. Nothing is installed yet, so this is just the plan.

1. Insulate the windows. I haven’t settled on the material yet and am still asking around. If you’ve insulated a small greenhouse, I’d like to hear what worked.

2. Run an electric heater on a thermostat set to 8°C (46°F). Every mature plant I still have came through balcony winters near 0°C, so 8°C in drier air should be a comfortable margin. The thermostat gets bought before November.

3. Keep the air moving with a small fan during the day. The big summer fan on a timer is too much now. The goal is just to stop cold, damp air settling around the pots.

4. Give the seedlings the warmest spot. The mature plants are proven at these temperatures; this year’s seedlings aren’t. If the warmest corner isn’t enough, they come indoors under grow lights, the same kind of setup I ran on the balcony. Adeniums or baronii would need that treatment too, since 8°C is below what they want.

What’s next

I’ll log temperatures through the winter and report back in spring: what the heating actually held, what it cost to run, which plants came through, and what I’d change. If you’re also facing your first heated winter, tell me about your setup in the comments.

Next post: Dioscorea elephantipes, which grows through the winter.

Subscribe now

In autumn most of my greenhouse slows down. The cacti and most succulents grow less every week, and by November most of them have stopped. Dioscorea elephantipes does the opposite. This is the winter-growing caudex from South Africa whose tuber is covered in corky, cracked plates like a tortoise shell. Just as everything else winds down, a thin vine pushes out of the top of each tuber, and they grow straight through the cold months.

My oldest seed-grown Dioscorea is almost three years old. I watched it go from a flat, papery seed to a tuber you can hold in your palm. Growing plants like that from seed is mostly what this newsletter will be about, and they’re the plants I’m proudest of.

How it started

I didn’t start this hobby. In early 2020, during the pandemic, my spouse began bringing home cacti, succulents and all sorts of other plants. The desert plants pulled me in, and before long I was the one looking after them.

By 2022 there were 60 or 70 plants on the balcony. Most were common ones, and I’ve since given most of them away. One that stayed was an Ariocarpus retusus, and it flowered that October.

The buying phase

For the next couple of years I bought plants, lots of them. Boxes of cacti arrived, and nursery trips came home in shopping bags.

A mature caudiciform looks impressive, but you don’t know much about how it grew. The Dioscorea I bought from a grower in France was already about five years old when it reached me. All that growing happened somewhere else. Almost anything is for sale now if you can afford it, and over time buying just got less interesting for me.

Why I switched to seed

I started sowing as an experiment, and over several years the results have been mixed. Some trays failed outright. Others worked better than I expected: Dioscorea elephantipes, Dorstenia gigas, Welwitschia mirabilis and several aloes have all come through for me. I’ve enjoyed watching them grow much more than unpacking a finished plant.

The first seed trays

I sowed my first serious batch on January 2, 2024, in a covered propagation box with every pot labeled. That batch included Welwitschia mirabilis, Dioscorea elephantipes, Aloe dichotoma (now Aloidendron dichotomum), Agave horrida and Moringa drouhardii. Germination took anywhere from 3 to 14 days depending on the species, and some took longer. The Operculicarya never came up. Many of the cacti I sowed were lost too, because seedlings need daily attention that work trips didn’t allow.

Most of what I knew back then came from Japanese growers’ blogs. They had far more species-specific detail than anything I found in English. They gave germination temperatures for individual species, and the soil mixes people actually used, which I replicated with ingredients I could buy locally. For hard seed, they showed exactly how growers handle scarification and cold stratification. That kind of detail is still hard to find in English, and I’ll be passing a lot of it on here.

What I get from growing from seed

For every plant in my seed trays I know the sowing date, where the seed came from and every repot. I also get to see the caudex slowly build up over the seasons, which you never see when you buy a plant.

Today: a small heated greenhouse

The collection outgrew the balcony, and everything now lives in a small greenhouse that I heat through the winter. The mature plants are there too, but I usually check the seed trays first.

What’s coming

In the coming months I’ll write about how I heat the greenhouse and what it costs, Dioscorea elephantipes care through its winter growing season, the soil mix I use for caudex plants and seedlings, and growing Pachypodium from seed. Many of these plants have been photographed since 2022, so you’ll often see the same plant across several years.

If you grow caudiciforms, or want to start them from seed, subscribe and follow along.

Subscribe now