Rhopalostylis: The scientist and the species

Keith Boyer’s book from 1992, Palms and Cycads Beyond the Tropics, did more than any other publication to draw the attention of palm enthusiasts to the variability amongst a number of populations of Nikau Palms. He distinguished between mainland forms and the more robust palms on a number of islands off the east coast of New Zealand. Plant scientists in the past had noted some of these differences also.

If we turn to the botanists, then the Auckland MA thesis by Fritha Stalker was the first “thorough taxonomic revision” of R. sapida.

This is surprising, she wrote, in 1998, as a number of botanists had commented on the variability in the species.

 

[Since our magazine is not an academic publication, I’ve eschewed the notion of giving lots of page references to the works I’ve used, even to individual quotes.]

 

Rhopalostylis had been placed in a subtribe, the Archontophoenicinae, along with six other genera. These are, Archontophoenix, Hedyscepe, Chambeyronia, Kentiopsis, Mackeea, and Actinokentia from Australia, Lord Howe Island, and New Caledonia. Rhopalostylis is considered to be closest to the first three of these genera, and closest of all to Hedyscepe.

 

There is a native Nikau Moth (Doxophrytis hydrocosma) which can destroy up to nearly a quarter of the seeds that set. That, by itself, is not enough to seriously affect reproduction of the species. However, it is known that introduced rats and possums, plus feral pigs, in conjunction can have a real impact on reproduction rates. The palms can live for perhaps as much as 250 years, so the impact of these introduced predators is not fully calculable yet: “... mature nikau are most common in areas also occupied by mature kauri 400–600 years of age”. The major dispersal agent was the NZ pigeon. Subspecies of Hemiphaga on the Chathams, and extinct ones on Norfolk, and probably on Raoul, presumably performed the same function there. These subspecies showed adaptation through beak sizes to the local seed sizes.

 

Fruit size and shape have often been cited as crucial in making distinctions in Rhopalostylis. Some botanists, such as Cockayne in 1902, had seen the Chatham Islands palm as having more in common with the Norfolk Island one than with the mainland Nikau. Stalker’s discussion of the literature also indicated that there was disagreement about height variations and other aspects of the palms in the genus. Palm material is often of poor quality when observed in most herbarium collections, the size of the fronds in particular being difficult to accommodate.

Stalker worked with specimens of these palms at Auckland University, and she also used seedlings from some of the different forms. She thought that the Chatham Island palms had more in common with the Kermadec ones than with mainland or inshore island populations, conceding, however, that larger numbers of specimens would result in better analysis.

 

Analysis of fruit sizes does tend to separate out the Norfolk, Kermadec and Chatham populations. They are larger than the mainland ones. Analysis of seed shapes was not really helpful, Stalker felt, because clear distinctions did not emerge. She did warn, though, that fruit collection (and later seedling analysis) is often based on one, or very few, fruiting trees in one locality because of travel and collection difficulties. The results of seed analysis by others “have not been substantiated in this study and could be difficult to establish and to use in the field.” She was more inclined to give weight to accounts that differentiate by analysis of seed colours, although that needed more work.

 

Stalker also did chromosome analysis of root tip material from seedlings, and she seems to have been the first to do this with Rhopalostylis and Hedyscepe. All populations had the same chromosome number (2n=32), as did Hedyscepe. The latter did differ in the number of nucleolar organizer regions and in the size and length of chromosomes. Another part of her analysis suggested that the Chatham Islands palm ought to be given distinctive standing because of characteristics such as “highly elaborated scale margins [which] give a silvery appearance to the mature rhachis.”

 

Her conclusions were that she thought that Rhopalostylis had one species in the genus (ie, R. baueri, a name which occurs before R. sapida in the botanical literature, and would thus take precedence). She would divide it into four distinct taxons based on Norfolk, Chatham, Kermadec and mainland forms. These would be regarded, she thought, as subspecies. This work, it is important to note, has not been published in a scientific journal, and has no definitive standing when the question of speciation in Rhopalostylis comes up.

 

The suggestion that petiolar scales could distinguish between the forms was taken up by Joshua Salter and Catia Delmiglio in an article published in 2005. One of their concerns (or their sponsor’s) was that the wrong forms (or even Hedyscepe) might end up being used if juvenile plants with indistinguishable characteristics were chosen for “restoration plantings”. “A further complication is the presence in garden centres of palms labelled as R. baueri (P. de Lange pers. comm.), and the possibility that specimens of the close relative Hedyscepe will also become available.” The thought did cross my mind that some people ought to ask around a bit more widely than their nearest garden centre.

 

“In juvenile Rhopalostylis, the petiole and rachis are ‘clad with peltate scales’ ... which form a pale grayish to brownish indumentum with darker brown speckles. The indumentum of mature leaves lacks the speckled pattern, appearing from a distance like a pale brownish bloom. This dense covering tends to disappear on older leaves, so that, on a fully mature palm, the rachis and petiole of the outermost leaves (including the crown shaft formed by the expanded base of the petiole) are green. Unlike its mainland relatives ... the indumentum on the mature leaves of the Chatham Islands variant ... is silvery grey to white, leading to the suggestion that petiolar scale morphology may have taxonomic value (Stalker 1998) and, thus, may prove helpful in identifying juveniles.”

So they studied a small group of juveniles in pots and mature plants at Auckland University, along with the same lone, but easily available, Hedyscepe that seems to be standing in for the whole species. Their results hardly improved the situation where clear distinctions are being sought between the different forms. Hedyscepe was easily distinguishable in some of their analyses, but “no particular features of scale morphology were found that set apart the Chatham Island juveniles from those of mainland New Zealand.” In fact, there were “great differences ... even in the scales of different leaves of the same juvenile.” There was also great variability in indumentum thicknesses across their complete range of Rhopalostylis samples, but one population could not be reliably separated from the others. Scale density (ie, not scale morphology) “may be a useful character” to separate the Chatham Islands form from the others, and also to differentiate juvenile mainland plants from the “island” ones. However, I didn’t think that the latter suggestion was all that much of an advance if ‘eco-sourcing’ is your problem, given the big visual differences between Nikau from the inshore islands, from the robust East Cape forms, from the more southerly mainland locations, and from the more shaving-brush like forms of the northern mainland. As seems to be the way of all academics, they argue in a mode I recognise all too well from my own social science background, that more research is needed (and doubtless somebody to fund it!)

 

Also appearing in mid 2005 was a multi-authored article on selected vascular flora of Norfolk Island. The section on Rhopalostylis was written by Peter de Lange. The Norfolk Island’s R. baueri and Raoul Island’s R. cheesemanii (ie, from the Kermadecs) were distinguished by Beccari in 1917 by means of fruit shapes (broad ovoid versus globose). However, it is clear that trees can be found on each island with fruit that have the shape supposedly diagnostic of the species on the other island. Raoul’s fruit may be a bit bigger, “but there is considerable overlap.” Beccari also stressed the shape of the seed hilum “but all specimens examined from both islands had narrow to broad linear hilums”, with “no clear-cut difference between seeds from either island.”

 

Sykes in 1977 had not been as impressed as Beccari was with the differences between the two populations. He turned them into separate varieties within R. baueri, rather than keeping them as two separate species. Now de Lange had looked again at the “minor differences” which Sykes had relied on and thought them relatively insignificant. “Accordingly we reduce R. baueri var. cheesemanii to synonymy within R. baueri.”

 

Another researcher in this area of Nikau speciation is Peter Lockhart of Massey University. He has not published the results of work he did in 1999 using techniques of molecular biology. When I first contacted him, he was a bit hazy on the details, but subsequently he reviewed the results with the person who did most of the lab work, and he did get back to me. He wrote: “With DNA fingerprinting (the AFLP {Amplified Fragment Length Polymorphism} method ) we found very little genetic variation among Nikau accessions sampled across New Zealand, including the Chatham Islands. This is in contrast, for example, with the amount of variation we have seen for other tree species such as silver beech.The distributions of silver beech in New Zealand are not as ancient as often thought. Our molecular clock studies suggest that the extant distribution is only a few million years old. Thus the distribution of Nikau must be even more recent, [and] it could have established very quickly following the last glacial period (less than 15,000 years old). Our genetic studies on the species on Raoul Island and Norfolk Island show that these are similar, and different from Nikau in New Zealand. However, we don’t have dates for their divergence times.” A little later he added: “p.s. Just checking through some files here — there may be some suggestion from our studies that accessions from Banks Peninsula and the Chathams are closely related — but this would need following up.”

 

There will undoubtedly be some sort of a follow up. Peter de Lange is finishing his PhD thesis soon. He tells me that after that has been done he intends to move on to studying, in particular, the Chatham Islands palm and its place in relation to the other forms. He also wants to consider whether R. sapida “is better treated as a subspecies of R. baueri or a full species.” He is sure that there is “a lot more work to be done on Rhopalostylis — it’s not as straightforward as people think”. In the meantime, “I can say that the genus is very good and distinct. Its nearest relative is indeed Hedyscepe, and then probably Howea.”

 

So there you have it. As of February 2007 there is no separate Kermadec species or subspecies. Since de Lange’s work was published in 2005 there has only been Rhopalostylis baueri. If you wish, then designate your plants by the seed source, if you are sure of it — viz, R. baueri “Norfolk Island” and R. baueri “Raoul Island”. Raoul actually is the island in the Kermadec group which has the palms, but R. baueri “Kermadec Islands” will probably be chosen by some people also. At least for now, R. sapida is a separate species. We ought to welcome future developments in nomenclature within Nikau Palms, perhaps with Peter de Lange galloping to the rescue, even if it results in short term nuisance value for those who resist learning new versions of old names. We can hope that next time, if there is to further change, it will be based on modern science and much more informed judgments than Beccari and the other classic botanists dealing with these palms could come up with by relying on their limited samples.

 

References

 

Fritha Stalker: “A Revision of the Genus Rhopalostylis H. Wendl. and Drude (Arecaceae)”, M.Sc. thesis, Auckland University, 1998.

Joshua Salter and Catia Delmiglio: “Exploratory study of petiolar scales in Rhopalostylis (Arecaceae)”, NZ Journal of Botany, 2005, v.43, pp.631-646.

  1. J. de Lange et al: “Vascular Flora of Norfolk Island: some additions and taxonomic notes”, NZ Journal of Botany, 2005, v.43, pp.563-596.

Peter Lockhart: personal communications, 6 February 2007.

By John Prince
From Issue 103 • Autumn 2007

 

This article differs slightly from that published in NZ Palm and Cycad magazine.