Don't judge a plant by appearances ...

... writes Ken Thompson in the Telegraph. The article provides welcome support for those of us who believe evolutionary relationships should be of prime importance to classification.  Most of the time, related plants are also look-alikes, but not always.  Look-alikes aren't necessarily related.  Those who believe plants that look just alike should be classified together are at risk of grouping unrelated similar plants.  Thompson points out that DNA data have provided a powerful new tool in this research, and that DNA is the stuff that evolves, so it's fundamental to such studies.

Thompson finishes with a comment on the recent (Garnock-Jones et al., 2007) transfer of Hebe and related genera back to Veronica, where these plants were originally placed: "So, with apologies to the many dedicated members of the Hebe Society, hebes are gone, and they aren't coming back."

That change continues to be argued.  Most professional taxonomists, like David Mabberley and the experts at Landcare Research, have accepted it, but a few taxonomists and many field botanists and ecologists in New Zealand have not.

The change to Hebe, made in the 1920s by Oliver, Cockayne & Allan, and others, reflected the differences between the New Zealand hebes and the northern hemisphere Veronicas.  We now see these changes as reflecting adaptations of Veronica to life in New Zealand, especially the hebes' woodiness, more tubular flowers, and capsules flattened in a different plane.  These adaptations made many (but by no means all) of them look different from their northern relatives. Also (and secondarily, because this was discovered after the split) they have a different chromosome number (which is now interpreted as easily derived from common Veronica chromosome numbers, i.e., no mystery).  But although they look rather different, fundamentally they have a lot in common, as studies of morphology, chemistry, cytology, and genetics show.

Other distinctive groups within Veronica have also in the past been classified as separate genera, like the Eurasian Pseudolysimachion and North American Synthyris.  The problem is that when these and the hebes are removed from Veronica, the remaining Veronicas aren't a natural group.  Many of them are more closely related to one or other of these segregates than they are to the type species of Veronica.


Here's the problem illustrated.  These three species (let's forget about their names for now) are all related, but the middle one is more closely related to the right hand one (they share a more recent common ancestor) than it is to the left hand one.  That's a statement that's well supported by molecular systematics.

Three Veronica species (the left one courtesy Wikipedia commons).  The left and middle plants are herbaceous and have blue flowers; the right one is woody and has white flowers.

Now we can rewrite that statement, replacing "left", "middle" and "right" with their previous names:  "Veronica arvensis is more closely related to Hebe elliptica than it is to Veronica officinalis".  Does that statement make you feel uncomfortable?  Note also that the left and right species are the types of Veronica and Hebe respectively.  It would make more sense to transfer V. arvensis and a couple of hundred other Veronica species to Hebe (our statement would then read: "Hebe arvensis is more closely related to Hebe elliptica than it is to Veronica officinalis"), but that would really put the cat among the pigeons.

Although Hebe as a scientific name is no longer supported, we recommend keeping it as a common name (hebe), and gardeners can use it that way with cultivar names (e.g., hebe 'Blue Gem').  Very few have a different epithet in Veronica (e.g., Hebe elliptica reverts to its original name Veronica elliptica and H. societatis is renamed V. societatis).

At a more fundamental level this discussion is about where we should place the gaps in classifications. Some say they should be placed where lineages diverge. Others say they should be placed where apparently major bursts of evolutionary change have occurred, even if those fall within lineages and separate related species.  But so many biologists are already using phylogenetic thinking and phylogenetic classifications (in ecology, biogeography, evolution, physiology, drug discovery, plant breeding) that the question is already pretty much decided.

References

Garnock-Jones, PJ; Albach, D; Briggs, BG 2007: Botanical names in Southern Hemisphere Veronica (Plantaginaceae): sect. Detzneria, sect. Hebe, and sect. Labatioides.  Taxon 56: 571–582.