Updated: Dec 30, 2018
Some people reckon that taxonomists change the names of things too often. Others reckon they're too conservative.
In the case of this story, it's taken a long time for Australian taxonomists to reach a final conclusion, but the time has come to change the name of a common and often spectacular desert-dwelling wildflower.
This is a complex story, with some twists and turns along the way, and one that nicely illustrates some of the complexities that arise around the apparently simple question "what's it's name?"
The genus Calandrinia has always been something of a biogeographical oddity. The first-named species (the type species of the genus) is widespread in western North and South America. Around 14 species occur in the New World, mostly along the cordillera of the Rocky Mountains and the Andes.
When similar-looking plants were collected in Australia in the 19th Century, European botanists concluded that they belonged in the genus Calandrinia too. Over the last 200 years, more than 75 species of Calandrinia have been named in Australia. Interestingly, given the alpine habit of most New World species, most species of Australian Calandrinia are desert annuals. Many flower spectacularly after desert rains, painting sandy desert areas pink with their blooms.
This odd distribution - occurring in Australia and the New World but nowhere in between - has made taxonomists suspicious for some time that Calandrinia may not be a natural genus.
More than 30 years ago, in 1987, Australian taxonomist Roger Carolin first proposed that the Australian and New World species of Calandrinia should be recognised as separate genera. Under the rules of nomenclature, the name normally stays with the type species, meaning that Calandrinia would be the correct name for the New World species, and another name would be needed for the Australians.
However, despite making this suggestion, Carolin didn't follow through by changing the Australian species names. That was left to an American expert in the genus, Mark Hershkovitz, who coined the name Parakeelya and renamed all Australian species known at that time under his new genus.
But there was a catch.
A very odd, very rare plant called Rumicastrum chamaecladum occurs in the Stirling Ranges and around Esperance in Western Australia. Only a few specimens of Rumicastrum have ever been collected. It only flowers if the spring following a summer bushfire is unusually wet, so it's only been seen by a few botanists who happen to have been at the right place at the right time.
Almost since it was first discovered, Rumicastrum has been something of a riddle. It was first described in the saltbush genus Atriplex, but was later recognised as very different from Atriplex, so was transferred into its own genus in the saltbush family Chenopodiaceae.
This changed in the early 1980s when an expert in the family Chenopodiaceae at the Western Australian Herbarium, Paul Wilson, looked closely at specimens of Rumicastrum and quickly determined that it wasn't a saltbush at all, and in fact clearly belonged in the same family as Calandrinia. Roger Carolin, the Calandrinia expert, agreed, and went so far as to suggest that it could be a very, very odd Calandrinia. Again, however, he left this as a suggestion only.
This uncertainty around the relationships of Rumicastrum, unfortunately, meant that Australian taxonomists were in a quandary. There were now three possible names for the Australian species of Calandrinia: Calandrinia (the traditional name, and the one that would apply if the Australian and New World species were considered a single genus), Parakeelya (the name given by Hershkovitz), and Rumicastrum (the correct name under the rules of nomenclature if Rumicastrum was indeed an odd-looking Calandrinia; this is because Rumicastrum was named long before Hershkovitz's Parakeelya, and the rules state that the earliest name is the one that should be used).
It's very likely that taxonomists would have remained uncertain, but for the invention in the late 1970's of a cheap and effective method for sequencing DNA. DNA sequence data is almost perfect for helping resolve tricky situations like this.
But again, there was a catch: the only specimens of Rumicastrum were very old and were unsuitable for DNA extraction. For this reason, early DNA analyses of the Calandrinia problem could not answer the question as to whether Rumicastrum or Parakeelya should be used as the correct name for the Australian Calandrinia. Because of this uncertainty, Australian taxonomists continued using Calandrinia, naming more new species (there are now nearly 50).
But now, finally, we are confident of the relationships of all the species in this group. A PhD student in the US, Lillian Hancock, has published a complete DNA phylogeny of all the species, including Rumicastrum (luckily, this flowered shortly before Lillian commenced her PhD, and good material for DNA extraction was collected). Lillian's work has demonstrated, finally, that the New World and Australian species of Calandrinia are best regarded as separate genera, and that Rumicastrum chamaecladum is indeed an odd member of the Australian group.
So - should we start calling the Australian species Rumicastrum? Again, there's a catch.
Strictly speaking, as mentioned above, Rumicastrum, as the first-named genus, should be the correct name, and all the Australian species should be transferred into Rumicastrum. But the rules also provide an exception. Given that many of the Australian species have already be transferred to Parakeelya, it would in many ways be better to use that name. If we could use Parakeelya, only the relatively few species named in Calandrinia since Parakeelya was erected, would need to be transferred (plus of course Rumicastrum chamaecladum). This would be simplest, and would result in the fewest synonymous names.
Many taxonomists also feel that Parakeelya is a particularly apt name. It's used as a common name for the whole genus, and derives from South Australian and Central Desert Indigenous names for some species (parkilja is the name in the Simpson Desert Wangkangurru language, and baragilya in the Guyani language from the Flinders Ranges).