Journal Number 88
September 2003

AUSTRALIAN NOTES

Robert Brown's Caladenia and Pterostylis Revisited
By Dr Stephen Hopper
Excerpted from a paper that first appeared in the June issue of the ANOS Victorian Group Bulletin.
Dr Hopper is Chief Executive Officer, Botanic Gardens and Parks Authority, Perth.

In the bicentennial year of Robert Brown's historic Australian landfall with the Flinders expeditions three major accounts of his Caladenia and allied genera were published independently by several workers (Szlachetko 2001, Hopper & Brown 2001; Jones et al 2001).

While there is some commonality among these treatments, especially the latter two that draw upon recent DNA sequence studies, there are nevertheless significant differences in generic and some species concepts as well as errors in the interpretation of the International Code of Botanical Nomenclature, mainly to do with typification.

Moreover, the situation was exacerbated when attempts by Szlachetko (2001) and Jones et al (2002; Jones & Clements 2002; Clements & Jones 2002) to rectify mistakes made in their respective earlier papers introduced even more errors or perpetuated others. Consequently, major nomenclature confusion has been generated, and is in need of resolution.

As in previous works (Hopper & Brown 2000, 2001), in a recently prepared paper submitted to the scientific journal Australian Systematic Botany, myself and Andrew Brown argue for retaining Caladenia in the broad sense, largely reflecting Brown's (1810) original concept save the removal of Leptoceras Hopper & AP Br. Thus Caladenia remains a large Australasian genus of terrestrial orchids, with 243 species and six subgenera.

We see little merit and much unwarranted nomenclature upheaval in further splitting Caladenia as advocated by Szlachetko (2001) and Jones et al (2002; 2002; Jones & Clements 2002). There are plainly no compelling phylogenetic reasons for such a split once Leptoceras, Praecoxanthus, Pheladenia, GIycorchis and Cyanicula have been removed from Caladenia.

The genus Caladenia as we circumscribe it has been affirmed in several DNA sequence studies as a monophyletic clade (all taxa included share a single common ancestor). Monophyly is regarded by most taxonomists today as an essential first criterion in the formal recognition of taxa if a predictive evolutionary classification is to be achieved.

Unfortunately, while rigorous science enables the question of monophyly to be resolved in a given
group, how to name groups of taxa within a monophyletic clade remains more art than science.
For example, in the case of our concept of Caladenia and Pterostylis, the key question for the taxonomic community is whether there is merit in retaining a broad concept of each genus with subgenera and sections within, or in elevating these subgenera and/or sections to the rank of genus.

Although favouring retention of a broad concept of Caladenia with six subgenera we accept that history will be the final arbiter on such a vexing question of rank. Hopefully our forthcoming paper, by clarifying and correcting key points of typification and nomenclature, will be helpful for either choice, broad or narrow, relating to circumscription of the genus.

For example, the valid type for Caladenia is C. carnea R.Br., not C. flava, as argued by Jones et al (2000), while that for Caladenia sect. Calonema is C. longicauda Lindl., not C.,filifera Lindl. as proposed by Jones et at (2001). The genus Jonesiopsis Szlach. and generic combinations Phlebochilus (Benth.) Szlach. were validly published. These conclusions, at variance to those of Jones et al (2001), render many of their taxa and combinations superfluous....

How should circumstances pertaining to the proposed splitting of a monophyletic genus such as our concept of Caladenia or Pterostylis into several genera best be handled? We consider nomenclatural stability to be of fundamental importance to avoid great discredit on the discipline of plant systematics for what is arguably perceived as needless change.

For guidance, we turn to the Preamble of the International Code of Botanical Nomenclature
(Greuter et al 2000). This is the international rule book for all botanical taxonomy, revised every
six years at the International Botanical Congress. We also note that most botanists follow a
conservative path of minimal taxonomic change consistent with the principle of monophyly.
The same should apply with Australian orchids.

As stated in its preamble. the ICBN "aims at the provision of a stable method of naming taxonomic groups, avoiding and rejecting the use of names which may cause error or ambiguity or throw science into confusion". It also, significantly, argues that "next in importance is the avoidance of the useless creation of names" and "The only proper reasons for changing a name are either more profound knowledge of the facts resulting from adequate taxonomic study or the necessity of giving up a nomenclature that is contrary to the rules.

Thus all taxonomic works should aim for stability, using valid published names consistent with scientific understanding (monophyly) wherever possible. To do otherwise is to indulge in the "useless creation of names" which the ICBN specifically seeks to avoid and reject.

Clearly the contribution of science in delivering "a more profound knowledge of the facts" is central in considerations about nomenclature change. The combined emergence of cladistic methodology and DNA sequence analysis have recently introduced unprecedented rigour and repeatability into the science of systematics, removing it from "a system that depends upon a whim (masquerading as authority) and accidents of history (Chase 1999).

For example, the difficulty of character choice and definition, prevalent in all studies reliant on morphological, anatomical or ultrastructural characters, and undoubtedly the cause of much futile argument regarding systematic relationships, rarely applies in DNA sequence studies, except for the choice of genes to be sequenced. The presence or absence of pairs along aligned DNA molecules can be rigorously and independently tested, and has been in many studies.

For example, in the case of Caladenia and allied genera, similar patterns of relationships have emerged in a number of independent molecular phylogenetic studies that have investigated the same or different gene sequences, both chloroplast and nuclear (e.g. Kores et al 1997, 2000, 2001, in prep.; Cameron et al 1999; Jones et al 2001, 2002).

Thus for Caladenia we agree with the removal of species in the segregate genera Cyanicula, Glycorchis, Pheladenia, Elythranthera, Glossodia, Praecoxanthus, Leptoceras and Adenochilus, all of which are sister to the monophyletic major radiation of Caladenia as largely encompassed in Robert Brown's original concept of the genus.

Having removed the above genera, further splitting of Caladenia is not needed to satisfy the criterion of hypothesised monophyly, and depends therefore on a judgement of appropriate rank. Nothing new is gained in terms of scientific understanding of phylogenetic relationships by elevating the six major clades of Caladenia to generic rank. Indeed, arguably the plethora of new generic names would obscure relationships, leading to a less predictive classification in the hands of most nonspecialists.

In such circumstances, nomenclatural stability emerges as a most important consideration in our view, to minimise inconvenience to and confusion of the users of taxonomic names and to maximise information retrieval and understanding from the literature.

Freudenstein and Remus (1999) aptly surmised: "As there are no rules for assigning rank to taxa we can only follow guidelines of striving for an internally consistent system, hopefully one that will disturb the stability of past nomenclature as little as necessary".

Similar arguments relate to the recent proposals to split Pterostylis into up to 16 segregate genera (Szlachetko 2001; Jones & Clements 2002). Molecular data unequivocally demonstrate that Pterostylis sens.lat. is monophyletic. A classification based on the principle of nomenclatural stability would therefore argue for retention of Pterostylis in the broad sense, with recently identified subclades within the genus classified as subgenera/sections. The latter approach is again consistent with the Preamble of the International Code of Botanical Nomenclature, and is the one Andrew Brown and I favour.