Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 82, 1954-55
This text is also available in PDF
(181 KB) Opens in new window
– 613 –

Further Notes on Gastrodia. R. Brown and Spiranthes L. C. Richard.

[Read before the Auckland Institute. October 21, 1953. received by the Editor, November 16, 1953.]


This paper discusses the pollination of the New Zealand species of Gastrodia, and proposes following Van Steenis in using the combination Spiranthes lancea (Thunb.) B.B.S. in place of Spranthes sinensis (Pers.) Ames.


The majority of Gastrodia species have a normal, more or less elongated column. Only 3 species—cunninghamii Hook. f. and minor Petr. from New Zealand, and orobanchoides F. Muell. from the New Hebrides, have the column conpressed I consider this reduction of the column to be complementary to the requirements of self-fertilisation In cunninghmii the column is foreshortened in front and compressed, concertina-fashion, into a series of rolls at the back. In the bud the column is more or less erect and the posterior rolls tightly compressed, but after anther-dehiscence (which generally coincides with the opening of the flower) these rolls expand, provting the anther on the fixed front of the column, and forcing it forward and down until the protruding pollinia are pressed against the stigma. In minor the process is similar though not so forceful. The posterior rolls are hardly developed, and while the anther is pushed downwards, it does not (or did not in those specimens I examined) descend actually on to the stigma. The pollinia, loosely suspended about 2 mm. above the stigmatic plate, probablv fall the remaining distance. In either case the result (self-fertilisation) is equally assured. Gastrodia minor appears to be derived from cunninghamii by an allround process of degeneration. The habit is much less robust, the flowers fewer and smaller, and the pollinating mechanism less virile. In spite of its origin it is a good species from a taxonomic point of view, occurs with but little variation from National Park to Otago and breeds true to type. Although the two species are sometimes to be found side by side, the restrictions of self-fertilisation effectively prevent hybridisation. Gastrodia cunninghamii in its turn is related to the New Hebrides orobanchoides F. Muell.—indeed Hooker once considered them identical—but they appear to be sufficiently distinct to warrant specific rank and probably represent divergent lines from a common Papuan ancestry, which also produced (or rather retained in the original column-form) Gastrodia papuana Schlti. The third New Zealand species Gastrodia sesamoules R. Br, is obviously windborne from its Australian home, and is related to and probably had a common origin with the Javanese Gastrodia crispa J. J. Smith.


In the Dutch journal Blumea 6: p. 358, 1950, Backer, Van den Brink and Van Steenis record the results of their study of type material, lodged in the herbanum of the University at Uppsala, to support the new species described by Thunberg in the Florula Jaucanica in 1825.

They say of the descriptions— “they are too short and vague to allow a definite opinion, but the type speeLmens are well preserved and complete, so that nearly all the species can be very easily identified”.

One of these plants, Ophrys lancea, had already been described by Swartz in 1800. The type material— “clearly represents Spiranthes australis Ldl.

Thumberg's name, being validly published and supported by a genunine lectotype has priority under the Code and the species now becomes—

– 614 –
  • Spiranthes lancea (Thunb.) B.B.S. Blumea 6: 361, 1950.

  • Sp. sinensis (Pers.) Ames Orch. 2: p. 53, 1908.

  • Sp. australis (R.Br.) Ldl. Bot. Reg. t. 823, 1824.

  • Neottia australis R.Br. Prodr, p. 319, 1810.

  • N. sinensis Pers. Syn. 2: p. 511, 1807.

  • Ophrys lancea Thumb. ex Sw. Vet. Akad. Handl. Stockh. 21: p. 223, 1800.

  • In Trans. Roy. Soc. N.Z. 79: p. 398, 1952, I wrote—

  • “the genus Spiranthes is at present being revised by the Canadian botanist L. A. Garay, who advises the writer that the species sinensis is highly compound and that the New Zealand form will probably be reduced to varietal rank.”

Garay finally came to the conclusion that while Sp. lancea is indeed highly compound, its variations are slight and seldom stable, and the there is nothing to be gained by splitting the species.

In this I agree with him. Sp. lancea in the broad sense occurs in Korea, China, Tibet, India, Siam, Japan, Formosa, Ceylon, Borneo, Buru, Sumatra, Java, New Guinea, Australia Tasmania New Zealand and New Calednia. Although the nomenclatural type was taken from Java, the species appears to have originated in Asia as an offshoot of the Eurasian

Picture icon

A.—Column and labellum of Gastrodia crispa (after Garay).
B.—Column of Gastrodia cunninghamii before dehiscence (after Irwin).
C.—Column and labellum of Gastrodia cunninghamii, showing column after dehiscence (after Irwin).
D.—Column and labellum of Gastrodia sesamoides.
E.—Flower of Spiranthes lancea (cf. Trans. Roy. Soc. N.Z. 79: t. 73, fl-a. 1952).

– 615 –

Sp. aestivalis (Lam.) Rich, to which it is closely related. The ecological differences between bogs in, say, Tibet and Java and Tasmania are quite pronounced, and it would be reasonable to expect plants growing in them to show some epharmonic variation. A comprehensive range of specimens from over the whole area does in fact show up apparently distinct local forms. But experiments with plants in cultivation over four seasons proves that these changes (largely vegetative) are related to the degree of heat and humidity prevailing in the late summer and autumn, when the new plants are developing, and can be intensified or retarded by regulating the environment. In the tropics the flowers open fully and are crosspollinated by insects. In south-eastern Australasia the flowers tend to become progressively self-fertile (eventually cleisgamic) and open only at the tip. The resulting narrow-tubular flower is superficially distinct, but by flowering the plants under glass it is possible to make the flowers open fully. The differences between the New Zealand and the tropical forms becomes immediately less obvious. (Incidentally the illustration in Trans. Roy. Soc. N.Z. ibid. t. 73. fig. 1-a, b, c represents a fully opened flower—artificially induced—such as will not be found in the field in New Zealand). The reduction of the rostellum and labellar calli, used by Schlechter (Bot. Zentrbl. Beih. 37: p. 350, 1920) to separate the Australasian subforms, is due in my opinion to degeneration following self-fertilisation as the species extends southwards into cooler regions. Thus we find the degenrate form in Tasmania, and in New Zealand to the south of the Egmont-Ruapehu line, while the normal form occurs in the North Auckland peninsula, the Australian mainland and New Caledonia.

Oakes Ames, discussing Sp. sinensis in 1908, lists no less than 17 primary synonyms and a further 17 secondary synonyms (recombinations of the originals). Since then another 6 synonyms have been added to the list, making 40 in all. This multiplicity of names and the diversity of opinion they represent is surely additional evidence of the extreme plasticity of the species.