CYRTOSPERMA MERKUSII

Species

PROTOLOGUE:

Cyrtosperma merkusii (Hassk.) Schott, published in Oesterreichisches Botanisches Wochenblatt 7: 61 (1857).

Lasia merkusii Hassk., Catalogus plantarum in Horto botanico Bogoriensi cultarum alter: 59 (1844). Type: Hasskarl s.n., Java, cultivated in the Bogor botanic garden (holotype L).

Schott's 1857 note, under the heading "Aroideae, sketched by H. W. Schott", says that the plant Hasskarl first made known and placed in Lasia agrees entirely, in habit and in the structure of flower and fruit, with Griffith's later Cyrtosperma, and transfers it. He gives the genus in outline (spathe scarcely twisted, at length withering but persistent; spadix elongate; flowers hexamerous; berries smooth; seed reniform, margined and crested; albumen fleshy; leaves sagittate), contrasts Lasia (spathe tightly twisted, flowers tetramerous, berries muricate, seed almost without albumen, leaves at length pinnatipartite), and separates the two known species: Cyrtosperma lasioides Griff. with a long-cuspidate blade whose basal lobes are parted by a narrow sinus, a spathe 3–6 inches long and a spadix reaching the middle of the spathe; Cyrtosperma merkusii with a cuspidate-acuminate blade whose basal lobes are parted by a wide sinus, a spathe 10–12 inches long and a spadix two-thirds the length of the spathe.² Hay (1988) later showed the two to be one species.⁴

SYNONYMS:

HOMOTYPIC SYNONYMS: Lasia merkusii Hassk., Cat. Hort. Bot. Bogor. alter: 59 (1844)

HETEROTYPIC SYNONYMS: Cyrtosperma lasioides Griff., Not. Pl. Asiat. 3: 150 (1851), the type species of the genus (type Griffith 5950, Malacca, K); Arisacontis chamissonis Schott, Bonplandia 5: 129 (1857) and Cyrtosperma chamissonis (Schott) Merr., Philipp. J. Sci. 9C: 65 (1914), the name long used for the Pacific crop (type Chamisso 54, W); Cyrtosperma edule Schott, Bonplandia 9: 367 (1861) (type Seemann 653, Fiji, BM; K, GH); Cyrtosperma cuspidilobum Schott, C. dubium Schott and C. intermedium Schott, Ann. Mus. Bot. Lugd.-Bat. 1: 284 (1864) (types Korthals s.n., Borneo and Sumatra, L), the last also as C. merkusii var. intermedium (Schott) Engl. (1879); Cyrtosperma ferox L.Linden & N.E.Br., Ill. Hort. 39: 59, t. 153 (1892) (type a Kew plant from Borneo, K); Cyrtosperma merkusii var. giganteum Nadeaud, J. Bot. (Morot) 11: 116 (1897) (type Tahiti, Hitiaa, P); Cyrtosperma bantamense Koord., Bull. Jard. Bot. Buitenzorg sér. 3, 1: 160 (1919), nom. subnud. (lectotype Koorders 41645, Bantam, BO); Cyrtosperma nadeaudianum J.W.Moore, Bishop Mus. Bull. 102: 22 (1933) (type Nadeaud s.n., Raiatea, P)

ACCEPTED INFRASPECIFICS: N/A

DISTRIBUTION:

Malesia to the Pacific: Borneo, Caroline Is., Cook Is., Fiji, Gilbert Is., Java, Marshall Is., New Guinea, Peninsular Malaysia, Philippines, Samoa, Santa Cruz Is., Society Is., Solomon Is., Sumatra, Vanuatu. Introduced into the Line Is., Marquesas, Nauru, Tokelau-Manihiki, Tuamotu, Tuvalu, Vietnam and Wallis-Futuna Is.¹

DISTRIBUTION NOTE: Wild and little used in Peninsular Malaysia, Sumatra, Borneo and Java (very rare in Java);⁴ in 1925 Ridley called it, as Cyrtosperma lasioides, common in open swamps of the Malay Peninsula.¹⁶ The Philippine plants are the unarmed cultivated kind, perhaps escaped; both wild-type and cultivated giants grow in the Solomon Islands. Hay saw no wild material from mainland New Guinea, only cultivated plants on Kiriwina and Manus.⁴ Pollen of the species in early-Holocene, pre-human sediment on Pohnpei shows its natural range was wider than the wild collections suggest.⁶ The Sangihe, Talaud and Siau islands off North Sulawesi grow it as daluga.⁷

SPECIES DESCRIPTION:

Robust to gigantic herb to 4 m tall. Rhizome short and slender to very large, sometimes globose and weighing up to 70 kg, weakly to freely suckering. Leaves several, sagittate, rarely hastate; petioles c. 40 cm to c. 3 m long, heavily armed, often very heavily with stout conical spines toward the base, to unarmed in cultivated forms; blade c. 30 cm to 1.3 m long and to c. 80 cm wide, armed or not beneath, held more or less horizontally to vertically with the posterior lobes down; posterior lobes somewhat exceeding the anterior; primary venation of the anterior lobe curvinerved in small plants to pinnate in large ones; posterior midribs naked in the sinus for c. 5 cm.⁴

INFLORESCENCE:

Inflorescence solitary on a peduncle similar to but usually shorter than the petioles, like them with a geniculus at the top that turns blackish on drying. Spathe very variable in size, color and shape, 2.5 to c. 30 cm long, white, yellow, green or purple, the pale ones usually with purplish streaks outside, rolled back to deflexed to erect with increasing size, the erect ones somewhat constricted above the level of the spadix apex; spadix 2–24 cm long, equalling or exceeding half the length of the spathe, sessile to stipitate, the stipe free or adnate to the spathe. Flowers hexamerous; ovary 1–2-ovulate; stamens exserted from the tepals at male anthesis. Fruit reddish orange, sessile, 1(–2)-seeded. Seed campylotropous, almost circular in profile to shallowly kidney-shaped, 5–11 mm long, brown, with about three raised, warty longitudinal crests to sparsely and irregularly warty and faintly crested.⁴

ECOLOGY:

Lowland swamp forest, open swamps and cultivation, mostly near sea level; in cultivation dug pits and taro swamps on atolls and high islands.⁴ Herbarium records range from sea level to about 1,800 m, with a median near 10 m.⁵

DISTRIBUTION MAPS:

Hay 1988 Fig. 2 (Malesia; additional localities further east in Oceania).

ETYMOLOGY:

Hasskarl's epithet honors Pieter Merkus (1787–1844), Governor-General of the Dutch East Indies when the Bogor garden was catalogued; the 1844 entry itself carries no dedication. Griffith's lasioides means Lasia-like, after the genus in which Hasskarl had placed the plant. The generic name is from the Greek kyrtos, curved, and sperma, seed, for the strongly curved seed, as Schott spelt out in 1858.⁸

NOTES:

1. The giant swamp taro of the Pacific: babai in Kiribati, pulaka in Tuvalu, baba on Guam,⁹ ape veo on Raiatea,¹⁰ daluga in the Sangihe islands,⁷ sampi gede in Sundanese Java.¹¹ Grown for its enormous corm, in the past a staple of atolls where little else will grow; Hay judged it a crop becoming obsolete in favor of Colocasia and Xanthosoma, but one whose origin, variation and development deserve inquiry as an early Melanesian and Polynesian crop.⁴

2. The species as Hay circumscribed it covers a wide range of form. Spathe poise follows its size relative to the spadix: a small spathe is stretched during expansion and deflected, a large one stands erect. Wild plants are heavily armed throughout the range; the unarmed, massive forms are cultivated selections, whose seeds are larger, less ornamented and less curved and whose fruit set is lower. None of these characters correlates with another well enough to separate wild from cultivated plants formally.⁴

3. Where it was domesticated is unresolved. Either the species has a natural range split between West Malesia and Papuasia, like its closest relatives on Bougainville, the Louisiades and the Lorentz River, and was taken into cultivation in the Solomons or the west or both; or the seemingly wild plants of the Solomons are relics of the earliest, still spiny, introductions from the west, such as are grown in the Trobriands and on Manus.⁴

4. Nicolson suggested the cultivars be surveyed for polyploidy; Hay found the chromosome and stem-morphology data too scant to settle it.⁴ Genetic work has since begun in North Sulawesi.⁷

5. Koorders could find no authentic material of Lasia merkusii at Bogor in 1919 and matched his Bantam plant to it on Hasskarl's description and the shared Sundanese name Sampi; his manuscript name bantamensis appears only in a footnote.¹¹

6. Engler's Pflanzenreich account of C. merkusii and C. edule mixed in New Guinea material that belongs to C. cuspidispathum and C. macrotum, and his fig. 6, captioned C. edule, is C. cuspidispathum.⁴

7. In cultivation the plant is the one Cyrtosperma most often seen, in botanic gardens and among growers, and the giant unarmed forms exceed 3 m.

STORY:

STORY TITLE: The crop that was planted in a hole

It is babai in Kiribati, pulaka in Tuvalu, baba on Guam, ape veo on Raiatea, daluga in the Sangihe and Talaud islands, sampi gede in Sundanese Java. A plant collects that many names only by being carried, deliberately, a very long way, by people who needed it.

What they needed was the corm. How they grew it, on atolls with no soil worth the name, was to dig: pits excavated down to the water, lined and fed and inherited. On islands that are rings of coral sand, this was not gardening but infrastructure.

Hay gave it two pages of economic botany in 1988 and closed on a note of regret, calling it a plant that deserved further inquiry as an early Melanesian and Polynesian crop, while recording that it was already giving way to Colocasia and Xanthosoma. Faster crops, smaller corms, less digging.

Then in 2011 Athens and Stevenson cored a swamp on Pohnpei and found Cyrtosperma pollen in sediment laid down before people arrived. Not carried there, then. Already present. Its natural range was wider than the wild collections had ever suggested, and the tidy west-to-east account stopped being the whole story.

Where it was domesticated is still open. Either the range was naturally split, West Malesian and Papuasian, and the plant was taken up more than once independently; or the spiny wild-type plants still growing in the Solomons are relics of very early introductions from the west. Both readings survive the evidence.

There is one more turn. This is the type species of the genus, though not under this name: Griffith founded Cyrtosperma in 1851 on a Malayan plant he called Cyrtosperma lasioides, which Hay later sank into merkusii. And it is the only one of the fifteen ranging west of Wallace's Line. The other fourteen are Papuasian. Which let Hay claim something almost no other aroid genus can, that this is the only oligotypic genus of Araceae more diverse east of the line than west.

The species that traveled furthest, fed the most people and carries the most names is the outlier in its own genus. Everything else stayed home.

REFERENCES:

1. Plants of the World Online, Royal Botanic Gardens, Kew — Cyrtosperma merkusii.

2. Schott, H.W. 1857. Aroideen-Skizzen. Oesterreichisches Botanisches Wochenblatt 7: 61–62 — protologue of the combination.

3. iNaturalist: Cyrtosperma merkusii

4. Hay, A. 1988. Cyrtosperma (Araceae) and its Old World allies. Blumea 33: 427–469

5. GBIF occurrence download for Cyrtosperma, 2026-09-05 (623 records)

6. Athens, J. Stephen & Janelle Stevenson. 2012. Pohnpei Coring Records: the Natural Distribution of Cyrtosperma chamissonis. Journal of Pacific Archaeology 3(1): 35–48.

7. Erlinawati, I., Abinawanto, Salamah, A. & Rugayah. 2018. Genetic diversity analysis of daluga (Cyrtosperma merkusii) using Sequence-Related Amplified Polymorphism in Siau, Sangihe and Talaud Islands, North Sulawesi, Indonesia. Biodiversitas 19: 2374–2380

8. Schott, H.W. 1858. Genera Aroidearum: 84, t. 84. Vienna

9. Merrill, E.D. 1914. An enumeration of the plants of Guam. Philippine Journal of Science 9C: 65

10. Moore, J.W. 1933. New and critical plants from Raiatea. Bernice P. Bishop Museum Bulletin 102: 22