Phyllotaenium lindenii André

Where it comes from

Kew gives the native range as Panama to Colombia, and describes it as a tuberous geophyte of the wet tropical biome.¹ That is the species. The plant on your shelf has a much smaller range than that.

There are two varieties. The pale-veined one — the one everybody grows — is var. lindenii, and Hay’s 2023 revision says it “is currently only known in nature from Antioquia east of Medellín in the western Middle Magdalena Valley”.² The plain green one, var. sylvestre, is the widespread plant, running through the Chocó and into Panama.³⁴ So the ornamental is the regional endemic and the one nobody sells is the common plant, which is the reverse of how this usually goes.

A tuberous geophyte is not a figure of speech here. It grows from a tuber, it dies back to that tuber every autumn, and it comes out of it again in spring.¹⁵ On a site otherwise full of evergreen aroids, that is the thing to know about it.

Naming and description

This plant has been formally placed in three different genera, and the nomenclature is worth setting out, because it is why it arrives from three nurseries under three names.

Kew accepts Phyllotaenium lindenii André, published in L’Illustration Horticole 19: t. 88 (1872).¹⁶ Phyllotaenium is a genus of exactly one species.⁷ It was reinstated by Alistair Hay in Aroideana in March 2023, on both morphological and molecular grounds — the paper argues the plant “is neither a Caladium nor a Xanthosoma” — and POWO and IPNI have adopted it.²⁸

The two names it is actually sold under are both synonyms, and both are formally illegitimate:

• Xanthosoma lindenii (André) Engl. (1878), the name it went by until 1981 — illegitimate, because an independent Xanthosoma lindenii T.Moore had already been published in January 1872.²⁹
• Caladium lindenii (André) Madison (1981), the name it is still most widely known by — illegitimate, because Caladium lindenii F.F.Stange got there first, and that one is probably just Caladium bicolor.²¹⁰¹¹

The wrinkle is that “Xanthosoma lindenii” is a legitimate name — under T.Moore, for the same plant.¹² It is simply not the accepted one. A label reading Xanthosoma lindenii is therefore not quite wrong and not quite right, which no label has ever managed to explain.

Horticulture has not followed the taxonomy yet, and that is normal rather than an error: the RHS still files the plant under Xanthosoma lindenii, lists Caladium lindenii as the synonym, and marks the name status “Correct”.⁵

The leaf is deep green, arrow-shaped rather than peltate, and cut through with broad ivory veins — the RHS describes “large, arrow-shaped emerald green leaves with prominent white veining”.⁵¹³ The whole varietal distinction turns on that veining.² The leaves rise separately on long stalks from the tuber; the RHS gives an ultimate height of 0.5-1 m and a spread of 0.1-0.5 m, reached in two to five years.⁵ For scale, its crop relative Xanthosoma sagittifolium reaches nearly 3 m with leaves up to 1.8 m long.¹⁴ This is a small plant from a genus of very large ones.

The epithet almost certainly honours Jean Jules Linden, the Belgian collector and nurseryman — L’Illustration Horticole, where André published the name, was Linden’s own journal.¹⁵ No source I could find states the dedication outright, so that is the evidence rather than a citation.

“Indian kale” is the RHS’s common name for it.⁵ It has nothing to do with India: the relative Xanthosoma atrovirens carries the fuller name “West Indian kale”.¹⁶ That the shorter name is a clipped form of the longer one is the obvious reading, but not one I could find written down anywhere.

Fun facts

• Its close relatives feed hundreds of millions of people, and this one feeds nobody. Xanthosoma is tannia, cocoyam, malanga, yautia and macabo; one review puts cocoyam as sustenance for over 400 million people, and West Africa alone at half of world production.¹⁷¹⁸¹⁹ Kew’s useful-plants field does flag a food use for P. lindenii itself,¹ but nothing else supports it — ColPlanta, Kew’s own Colombian resource, records no uses at all,²⁰ and the RHS rates the plant “harmful if eaten”.⁵ A food flag on a name that spent 150 years filed inside a genus of staple crops is the kind of thing that gets inherited rather than observed. Treat it as not edible. Nothing in this group is eaten raw in any case — the cormels are only food after thorough cooking, which cuts the oxalate by roughly half — boiling took 52 per cent out of the leaves in one trial, and drum drying about half out of the cormels in another.¹⁷²¹²²
• A divided tuber that does nothing for a month has not rotted. Cut pieces carry a dormancy of their own: one propagation trial found whole corms “the most effective as it will sprout early and gives the highest sprouting percentage”, and put the lag in cut material at around five weeks.²³ On top of that, cormels of cultivated Xanthosoma have a natural dormancy of one to three months.²⁴ Patience is the technique.
• Don’t give it Colocasia’s water. Most “elephant ear” advice online blends the two genera, and Colocasia is the paddy crop — it will sit in shallow standing water quite happily. This group is the upland one: tannia “requires well-drained soils and does not tolerate water logging”,²⁵¹⁸ and waterlogging is what Pythium myriotylum wants. A single isolate needs no help from anything else — one study produced typical symptoms within 48 hours on every inoculated plant.²⁶
• If you lift the tuber, warm and humid first, cool and dry after — which is the opposite order to most people’s instinct. Curing works at 86-95 °F (30-35 °C) and near-saturated humidity, and the cork layer that seals the wounds simply does not form at 77 °F (25 °C) or at 104 °F (40 °C). Only then does it go into store at 45-59 °F (7-15 °C); below about 45 °F (7 °C) it takes chilling injury — surface pitting, flesh discoloration and decay.²⁷ An unheated garage in a British winter is a real risk, which is a good argument for the RHS’s other option: leave it in its pot indoors.⁵
• The taxonomy is still moving. Hay suggests Xanthosoma alversonii, also from Antioquia, “may represent a third element in Phyllotaenium”, and flags the lack of accurate description of the floral structures as an open problem.² The same paper establishes a cultivar, P. lindenii ‘Cream Cheese’, an Australian chimaeric variegate — so anything sold as a variegated X. lindenii is probably that.²