This site is for people who like plants -- growers, enthusiasts, aesthetes, novices and professionals, those who appreciate wild things and those who appreciate the cultivated. I garden in Chelsea, and I've been visiting people's yards for 20+ years in the course of my work. My goal is to make this blog a community project, so if you share my interests, please consider becoming a participant and contributing content -- Guerin. Info:

Sunday, April 17, 2011

Why plants change their name

At the last meeting of the local chapter of the Rock Garden Society, Tony Reznicek addressed the issue of why the scientific names of plants continue to change. The talk was titled "Why plants change their names: Rock Garden Plants, Woodland Plants, and Modern Biology." I am posting the notes on his talk taken by Laura Serowicz. It's a fascinating story, IMHO worthy of publication in the New Yorker.


At the October 30, 2010 meeting Tony Reznicek (presented a) lecture discussing the elements of modern evolutionary biology generating the current change in some plant names. He joked that it is standard practice for Plant Systematists to change the names of plants as soon as they become familiar to people to keep us on our toes, and if they didn’t we’d decide that we didn’t need those experts.

To put modern systematic biology into perspective Tony gave us a little background as to how the science of plant systematics, the naming of plants and the understanding of their evolutionary relationships have developed through time. People have been naming plants since forever. Even as far back as the Neanderthals plants were used in burial ceremonies, and they must have had ways to identify them.

With just a few plants to name you didn’t have to be very scientific about it, but once you developed a large enough body of knowledge and were teaching, then you had to change. The monks in Renaissance Europe realized that they could teach students all year long if they pressed and dried plants while they were in flower and then taught students from those dried specimens in the winter. Thus dried herbarium specimens began in the 1400’s and some of those are still in the herbarium at Florence Italy, one of the centers of the Renaissance. At that point collections were simply used as a reference for identification and not to study the plants.          (continued, click link below!)

By the 1700’s things had developed to the point that there were extensive books about plants, almost all of which were written in Latin. At that time, still with a limited catalog of plants to name, the names evolved by laboriously describing the plants in Latin with a polynomial name encapsulating the name of the plant and the characteristics by which you told this plant apart from others. However, with the beginning of world explorations, samples of new plants were flooding in to Europe from all over the world and the more plants you have the more names and adjectives you need in your description to differentiate them. Soon the names became very long and the system became completely unwieldy. Linnaeus realized that to grasp the natural world complex polynomial descriptions didn’t work. He developed a hierarchical system for both animals and plants with large classes divided into smaller groups, using simple observable features such as with plants the number of stamens, stigmas and other simple things, creating the ultimate name as a binomial name with a genus and species. This was the beginning of the great era of describing and naming the flora of the world.

The binomial system is how we have done it since 1753, as it is easy and straight-forward. Linnaeus continued the practice of collecting herbarium specimens amassing a lot of specimens which he used as a reference collection and research tool. The vast quantities of new plants discovered throughout the world necessitated the organization of systematics as a discipline with scientific procedures and rules to follow to name newly discovered things becoming much more than just a convenient filing system. This period was the dawn of the English garden books. Curtis’s Botanical Magazine, still being published today, began in the late 1700’s. From the beginning it followed Linnaeus’ system and has maintained the quality of the illustrations, which were originally hand-colored copperplate engravings. It is one of the longest running serials and U of M library has a complete set. With this and other botanical works the Linnaean system of classification immediately had immense power over the naming of plants.

During the time of world exploration there were some remarkable finds. In eastern North America Franklinia alatamaha was found along the Altamaha River in Georgia in the late 1700’s. It was observed and seeds, seedlings, and presumably plants, were collected from it by the Bartrams of Philadelphia, with no understanding of the consequences, until the early 1800’s. They assumed that along the next river down there would be a lot more of it and “suddenly” that patch was no longer there. That location happened to be the only spot in the world for it. It has never been seen as a wild plant again and over-collection may have been cause for its demise. Fortunately they were able to get it into cultivation.

The Linnaean system, based on correlating and describing the patterns and discontinuities of physical variations observed in organisms, actually worked well but didn’t explain what accounted for that variation nor how that variation was transmitted. All of this was occurring before Darwin, Mendel, and before genetics was understood. Even Mendel’s elementary and primitive genetic experiments were ignored for half a century and in all the classification after Linnaeus there was no unifying theory that could tie them together. With Mendel’s concepts of inheriting natural variation and Darwin’s concept that natural variation is affected by selective pressures to change the forms of plants the science of naming plants evolved.

The Linnaean system assumed species were constant in space and time so that we only needed one or a few samples to represent that species, and herbaria only needed one specimen of everything (like a stamp collection) and did not need to represent the variations of a species. Descriptions were all written in Latin, so all educated people learned Latin. Even to this day, if you discover and name a new species of plant you have to provide a diagnosis or brief description of it in Latin in order to validate the name of the plant. This standardizes understanding the descriptions for all countries. You do not have to learn to translate all the world’s languages and scripts, only Latin. Just think of trying to read a description of a Chinese plant if it was only written in Chinese.

Today we have a lot of tools available to help with identification. We use many characteristics instead of focusing on a few like Linnaeus. There are a lot of statistical and quantitative analyses we can use and numbers of specimens are used to describe the species with large herbaria now having millions of specimens. What has really driven the changes in our understanding of plant relationships are three crucial things. 1) In the past 2 ½ decades we have been able to directly access the genome of an organism, and we can now look right at the DNA to see what makes it tick. The DNA molecule is a ladder-like structure with variations in the cross members of complementary base pairs that hold the double helix together. The language of the genetic code based on the four chemical bases [abbreviated as A, C, G and T, with A always pairing with T and C with G] and variations in how they are bonded is what generates individual organisms. 2) However, genomes are gigantic, with millions of paired bases in the DNA. Until the development of complex mathematical techniques we could not analyze them. 3) Now that we have more computing power with laptops and desktop computers than was previously only available to the wealthiest government research lab even in the 1980’s everyone can work with analyzing the millions of possible base pairs of all the species in the world that balloon to a virtually infinite size. The situation being compared to a game where every move you make opens paths for hundreds of additional possible moves and every one of those moves opens up possibilities for hundreds more and after many moves the possibilities could be in the gazillions. Even now computers cannot actually calculate to completeness all the astronomical numbers of possibilities. Mathematical techniques group the possibilities according to certain classes, look for parallel variation and try to summarize and contract the amount of variation that is explored. Even so, with all the species in the world you need powerful computers. Perhaps as computers evolve before this next decade is done we will be able to sequence the genomes of any organism we want on demand, and will be able to analyze hundreds of thousands of species of organisms and see how they all fit together into phylogenetic trees. Those three things have resulted in most of the changes we have in modern systematics.

In appearance, when you actually pull it out of the cell, extracted DNA is very uninspiring with (according to Tony) roughly the consistency of snot. There are various techniques for analyzing and comparing similarities in DNA. A common method, albeit with some faults, involves searching for short pieces of DNA that are unique to individual species, scoring them as 1’s and 0’s (presence/absence) and creating a 1-0 matrix. Postulating what may be considered primitive/original characteristics (sometimes using fossil records) and more advanced/later characteristics you can start linking organisms that share the advanced characters. Using a little bit of fancy math and complex mathematical algorithms you can generate a network diagram, a branching tree pattern (phylogenetic tree) that shows how organisms are most likely linked to each other, and forming the concept of a common ancestor and all of the descendants of that common ancestor giving a sense of evolutionary change. Using fossil records you can corroborate that these patterns are roughly correct [the phylogeny tree used in Tony’s talk is on the Angiosperm Phylogeny Website at and click on “Trees” at the top of the page.]

Botany teaches that flowering plants are divided into Monocots and Dicots. Cycads and Conifers are primitive. Ferns are even more primitive, and below that you get into the slime that covers your pots. Looking at the tree diagram you see the Ferns are primitive by following the fern branch all the way to the base, they are just after mosses and liverworts and so forth. Next are the few survivors from the great forests that were covering the world during the dinosaur era. Then, 60-80 million years ago, Cycads and lots of species of Ginkgo were all over the world as they were once one of the dominant trees, but now there are just a few survivors. A few scattered Cycads [Cycadales] are in the tropics, and Ginkgo [Ginkgoales] survives in eastern China on one mountain where there might still be a native stand of them, with all the rest only as cultivated stock. Pines, spruce and other conifers [Pinales] fared better, there are still large areas dominated by conifers. Primitive seed-bearing plants [Gnetales] that look a little more like flowering plants, have not fared very well, there are a few Ephedra out in the deserts, the famous Welwitschia in South Africa and a few Gnetum in the tropics. In reality almost everything that bears seeds is an Angiosperm (a flowering plant) and they have been so tremendously successful.

Everybody always knew that Magnolias and their allies were primitive -- they had fossils of Magnolia-like things going back into the dinosaur era, and at the end of the Cretaceous era the later dinosaurs actually ate Magnolias and were living in a flowering-plant flora. The traditional view was that all primitive plants were woody because woody plants fossilize more easily. Turns out there are also a lot of these primitive plants that were not woody so the new term for them is “Paleoherbs”.

So, in additional to the Magnoliids there were also these primitive herbaceous plants as well as woody ones, and all of these predate the split between monocots and dicots. This represents a huge change in our thinking about the ancestors of flowering plants. Up to this point, things corresponded with the traditional view taught in school, but looking at the phylogeny tree a little further some surprises show up. There are Monocots, but what we used to call Dicots are actually two different groups. One is a small group of plants, the “ANITA” group [Amborellaceae, Nymphaeaceae (Water lilies), Illiciaceae (Star Anise), Trimeniaceae (old-world tropicals), and Austrobaileyaceae (Australia/New Guinea tropicals), all descended from ancestors that predated the split between Monocots and Dicots. The most primitive branch of the “ANITA” group is Amborellaceae which has a single species, Amborella trichopoda, growing in the tropical forests of New Caledonia, the most primitive flowering plant known as yet, but some botanists think that possibly with a further looking at odd, unplaced, and poorly known plants from the tropics might turn up an even more primitive one. Amborella is a far cry from what people thought was primitive only 15-20 years ago, then thought to be small bushy woody plants with big floppy flowers and a dinosaurian look. Such plants were indeed there, but there was much more diversity and it is very possible that the most primitive things looked very different. With its small flowers, Amborella is certainly nothing like the Magnolias.

Nymphaeaceae (Water lilies) include Nymphaea. The odd thing about this group is that botanists always knew there was something strange about them, but did not understand the significance of things like petal number. For one example, everyone was taught that monocots have flower parts in threes and dicots generally have flower parts in fours and fives. The word “generally” was used because there were a few of these annoying dicots that had their flower parts predominantly in threes such as Schisandra, Nymphaea, Magnolia, Decaisnea (in the Lardizabalaceae family), and Asimina (Pawpaw). Illiciaceae include Schisandra chinensis that we can grow in Ann Arbor and Star Anise (Illicium verum; Zone 8). Trimeniaceae are old-world tropicals. The Austrobaileyaceae are mostly from the Southern hemisphere. Wild Ginger, Asarum, and its relative Saruma henryi are “Paleoherbs” with flower parts in threes like monocots but have other characteristics of dicots, such as two seed leaves, leaf venation, pollen structure, etc. Early botanists missed the clues that showed that these plants were different. Genetic evidence is very clear that these evolved before the split. Chloranthus henryi is also a “Paleoherb” in that it has one of the weirdest looking flowers (with parts arrayed in threes). It is part of the Chloranthaceae family which is tropical everywhere except in China and Japan where a few species range into the temperate zone. Tony has a couple different species of Chloranthus that set seed every year, which he sends into the NARGS Seed Exchange.

Even the monocots have a new twist. We thought that the most primitive monocots were things that looked a little like dicots with three petals like Sagittaria and Alisma [Alismataceae] a group which contains many aquatic plants. However, it turns out the most primitive monocot is Acorus the Sweet Flag [Acoraceae] used in medieval times in “rooms strewn with flags” (laid on the floor so they would release their fragrance as they were walked on). Acorus has a superficial resemblance to, but is not an Aroid (a member of the Jack-in-the-pulpit family) and is much more primitive. Botanists have long documented the fact that this was weird and did not fit into the Aroids, but they did not understand the significance of its differences.

We have learned some really interesting things, developed some new reasons to grow odd-ball plants, and that has made quite a difference in our understanding of the relationships of plants. There is still a lot of work to do as can be seen by the number of “Unplaced” listings on the phylogenetic tree [see tree on APG website above] before everything is laid out firm and in full. Even from when Tony did the slide for this talk a couple years ago to the current tree on the APG website there have been several changes made. So that is the main large-scale backbone of the changes happening in plant systematics, but what really concerns us as gardeners more is when plants change their names and/or families. It happens mainly because we can now understand with much more skill the actual relationships of plants.

Tony gave us a small classification lesson using a simple tree with two main branches originating from a common ancestor. Off of one branch are A and D, off of the other branch are C, E and B. Let’s say we are grouping them by habitat and that A and D are woodland plants, E and B grow in the open, but C diverged from E and B and also went back into the woods. If you look just looked at superficial features such as the habitat, what would happen is that you might think that C is closely related to A and D because they all occur in the woods, therefore they probably all have things woodland plants have to have such as broad leaves for catching more light, etc. By understanding where the branching link up (in this case A and B split before C branched off from B) we can avoid being fooled by superficial resemblance. That is what is generating some of the bumpy parts that result in name changes now that we have a better understanding of their evolution. If you were to actually lump C, which is descended from a different lineage, in with A and D, the group you would form botanists call Polyphyletic (meaning “from multiple ancestors”) and that is a no-no in modern systematics.

As another example, using the same diagram, if you split E into a group, like a genus (with a unique feature), and grouped C and B into another genus that is also bad, because E is nested inside the other two, it is derived from the same ancestor as C and B. If you start splitting those out then the group becomes arbitrary, meaning that you don’t have any more uniform rules for forming groups and the groups cannot be predicted from the diagram with consequences. If, for example, you are going to use these groups to do something like predict biochemical relationships, or whether they might have the same compounds for medicinal use or other kinds of economical use, or whether they might show similar characteristics of resistance to insects or so forth. You can see that your predictions would be thrown off if your groups didn’t include everything, because those characteristics would also show up in another group. We would call that group Paraphyletic (because it contains some but not all members descended from a common ancestor). These terms, as well as Monophyletic (meaning “from a common ancestor” and including all descendants from that ancestor), are often seen in botanical texts and will probably soon creep into gardening books. The bottom line is that if a traditionally recognized group (like Aster or Anemone) does not all come out together in one spot when the phylogeny is reconstructed then there is a problem. Most botanists feel that what they need to do is fix that and rework the relationships until they do work back to a common ancestor.

The phylogenic tree for Rhododendron shows that what we used to call Ledum is right in the midst of all the Rhododendrons. In fact, people who worked with rhododendrons knew this because Ledum would cross with some species of Rhododendron, and now that the genetic evidence is so solid the names need to be changed on all Ledum. So Ledum groenlandicum is now Rhododendron groenlandicum. Menziesia is also found within the Rhododendron phylogeny and also has been crossed with Rhododendron. Our crowberry, Empetrum, now sits right in the middle of the Ericaceae phylogeny rather than in its own family of Empetraceae. It is just a wind-pollinated heath, not insect-pollinated, so it lost its petals since it didn’t need them to attract insects. It looks just like a heath and its fruit tastes like a heath so everyone is happy with that change. The bottom line is that the genetic makeup [genotype] trumps the external appearance of the plant [phenotype], which makes sense as it is your genes that determine who you are related to, not your appearance. Before we had access to the genetic code a lot of taxonomic decisions basically had to be made on appearance because that is all we had, and that is the reason why so many names have changed.

Not only are we now able to access the genes but we have the capability of translating that access into information about relationships. That has caused some major revisions to a few groups such as with the Asters. If you do detailed molecular phylogenies of the group of plants surrounding Asters you will see that the traditional Eastern North American Asters come out in various places along the tree which is a problem. You have two choices here in order to make a monophyletic group (a group that is all descended from one common ancestor and include all the descendants of that ancestor), which is what we would like. What you would have to do is lump everything into one big group or else split up Aster into smaller common groups . It was split, and action which many botanists refer to the “Aster Disaster”). The next version of the Michigan Flora will include nothing under the genus Aster. The former members of the genus are now Symphyotrichum, Eurybia, Oclemena, Canadanthus, Doellingeria, and Sericocarpus [Flora of North America also follows this splitting up]. True Asters are actually Old-World/Eurasian species with one Arctic-alpine species native to North America.

(If you want to keep up with the taxonomy in real time, check out the U-M Herbarium's new on-line resource at

Another nasty change is to our shooting stars, Dodecatheon, now seen as highly-derived members of the genus Primula that have evolved buzz-pollinated flowers for bees. They evidently were derived from American primulas relatively recently. [A few of the species received new species names when changed to Primula since the former species names were already used for other Primula species. New names are listed as synonyms in the Flora of North America.]

Actually many of the name changes can be quite useful. For example, we can look at the genetics of some of the classic disjuncts between Eastern North America and Asia and start to understand when it happened, how long they have been separated, and so forth. For example, the American Liriodendron tulipifera and Asian L. chinense are so closely related that they will hybridize with each other, and we can see that they are closely related because nothing else looks like a tulip tree. With big-leaf Magnolias, one of the most interesting things is that even though they all look alike when you look at the genetics M. macrophylla has nothing to do with all the other big-leaf Magnolias. It is related to Mexican species. But our other native big-leaf Magnolias, M. tripetala, M. fraseri, etc. are very close relatives of the big-leaf species in Asia. It is remarkable and seems almost unbelievable but long before the article that came out with genetic evidence for this, Tony visited Phil Savage, the Magnolia breeder in Bloomfield Hills, to look at and talk about the Magnolias. Phil told him that M. macrophylla was weird because it would not cross with any other Magnolias except with extreme difficulty, and it was obviously very isolated genetically. So in many cases we knew about these puzzles except that we couldn’t really understand the significance of it. On the other hand, if you grow M. tripetala from the Appalachian and M. officinalis from sub-tropical China together you often can’t even get pure seed because they hybridize so readily with each other. People have been hybridizing disjunct species for a long time. One of Phil Savage’s big exploits was the breeding of yellow Magnolias, with most yellow Magnolias being derived from hybrids of American and Asian Magnolias that are closely related.

With Shortias if you look at the phylogeny of the Diapensiaceae family you can see there are some distinctive primitive North American lines and a bunch of Asian ones, but the two closest relatives are the Japanese Shortia uniflora and the eastern North American S. galacifolia. One of the easiest to grow of all the Shortias is their hybrid, S. × intertexta ‘Leona.’ From the phylogeny you can see that those two species are closely related despite that they were from opposite sides of the world.  

Calycanthus floridus and C. sinensis (or Sinocalycanthus sinensis) are another example of this. There were no Calycanthus known to occur outside of North America until the 1960’s when C. sinensis was found in the mountains of eastern China as a very rare isolated plant. It turned out to be a great garden plant and the hybrids (such as C. × raulstonii ‘Hartledge Wine’) are pretty spectacular too and grow like weeds.

With all this molecular technology we can now predict close relatives with more efficiency, and it will be a big help to breeding programs. In particular we can probably make some good guesses on plants that are on opposite sides of the globe but might produce some interesting results. Ancient floras that are now separated because of climate change and other changes in the planet probably 20-25 million years ago were once joined in a band of similar vegetation across the northern part of the globe. In the end, once we get over the bump of all those name changes, not only will there be a lot more plant genera, but we’ll actually be able to turn some of this information into some very practical uses, hybridizing being just one of them. Other things we can use that information for include studying the ecology and evolution of plants. With genetic knowledge you can look at dating, to figure out how long ago flowering plants evolved. Dating some of these past events with some precision is going to be another thing that we will be doing a lot more of in the future, and this is really important for plants because many plants just don’t fossilize (animal bones take a long time to rot and thus fossilize, but plant fossils are rare). The name changes are just one minor bit of discomfort. Tony thinks once we get through a fairly thorough survey of the plant kingdom genetically we can actually look forward (after we get over the name changes) to much more stability in both names and relationships/family compositions.

No comments:

Post a Comment