On the Origin of Flowers

by Jennifer Frazer on December 15, 2009

A water lily. There is an evolutionary secret staring right at you in this picture. Can't see it? Read on to find out what it is, then come back here and click on the photo as many times as you can to see it in extreme-close-up-o-vision.

A few posts ago, I told you that star anise is interesting because it belongs to a group of plants that split from the rest of the other flowering plants early on, and that for a long time, scientists felt plants like these retained a lot of the features of the first flowers.

Scientists think that long ago, petals, carpels (the girl bits), and stamens (the boy bits) evolved from leaves. In other words, the leaves were the raw material upon which evolution acted to create a specialized cluster of sexual structures we now call flowers. By this reasoning, a flower is essentially a bunch of colored, modified leaves that the plant has packed together with its gonads at the end of a stem (a pedicel (ped’-i-sell)) for the purposes of luring insects to move pollen from flower to flower, i.e., reproduction. So yes, when you stick your nose in to inhale the perfume of a delicate flower, you are essentially shoving your sniffing nose – doggy-greeting style — into the flowers’ unmentionables.

The 3% of plants that are basal angiosperms (the group in which star anise fits) seem to have a lot of the characteristics we’d expect in first flowers. Their flower parts are usually physically separate and little changed from each other,  like magnolias or star anise or the water lily above – as if each one evolved from an individual leaf. This is not true of highly derived (changed a lot) flowers like orchids,


wiki_Nep_northiana_pitcheror this crazy pitcher plant, Nepenthes northiana, with the digital camo in the back of the pitcher and a convoluted and unified structure. Although not a flower, the pitcher is, believe it or not, a modified leaf.

Thus petals are essentially colored leaves, and the stamens of many of these groups, particularly water lilies, are essentially petals/leaves with pollen-making chambers on the ends. Some seem to defy classification as either petal or stamen. Click on the picture at the top of the page and keep clicking on it to magnify it. Look carefully at the stamens at the very edge of the bunch — they are more like purple petals with pollen making slits half way down them than true stamens.

In this photo, notice the transitional stamens toward the outside of the center – they also look like a cross between a petal and a stamen, although in this case, the pollen-making chambers (anthers) are at the tip of the petal rather than halfway down.


For a while, some scientists felt the first flowers looked very much like magnolias, hence an early term for the basal angiosperms, magnoliids. But that view is changing. Showy magnolias were probably a very early specialization.

Wiki_Amborella_budsScientists have compared plant DNA the same way they compare flower and plant body parts and have concluded the living flower lineage that split the earliest from the rest of the living flowering plants is the diminutive (and threatened) Amborella, pictured here.

Very closely related to Amborella at the base of the tree are the water lily family, Nymphaceae (nymf-a’-see-ay), and the Austrobaileyales, the group that includes star anise (see the tree here). Plants in these groups often possess characteristics of both the two major flowering plant groups that comprise the other 97% of plants, the monocots (grasses, grains, lillies, orchids, etc.) and the eudicots (roses, apples, poppies, maples, etc.), giving scientists more confidence that they inhabit a special place at the base of the flowering plant tree.

Now scientists believe the first flower was likely to have been small, green, simple, and inconspicuous – not unlike Amborella. Nova did a special on one Chinese aquatic fossil candidate for “the world’s oldest flower” in 2007 (see a great short slide show on it here). Though it’s interesting that water lilies, also at the base of the tree (look for Nymphaeales), are also aquatic, the status of the fossil — Archaefructus, or “ancient fruit” — as “oldest flower” has been challenged. There have been other “most ancient flowers” (Bevhalstia pebja) in the past and will surely be new contenders in the future.

So, four weeks and three posts later, it may finally be becoming clear why star anise is such a surprising plant! In addition to the gorgeous and pungent whorled fruits you can use to flavor stir fry pork or decoupage your 70’s hippie bus, it is a member of one of the most ancient flowering plant lineages on Earth and could save you from one of the planet’s newest viruses.

{ 6 comments… read them below or add one }

Psi Wavefunction December 15, 2009 at 7:40 pm

Interestingly, by overexpressing a small handful of genes you can induce floral-like organs to pop up out of leaves (Honma & Goto 2001 Nature); these genes include a family of transcription factors that’s actually found all the way over in opisthokonts as well, and has been exapted for reproductive use even in the gymnosperms. These genes duplicated and diversified, allowing more ‘room’ for screwing around and ending up with the vast diversity of flowers we enjoy today. Actually, the genetics of floral development is kind of cool; there’s a nice open access paper on the evolution of these genes (and, of course, flowers and pre-flowers) here: Chanderbali et al. 2009 PNAS

(haven’t had the chance to read it in detail, but has a really nice diagram, which I used in a class presentation. I insist on always putting stuff in an evolutionary perspective, especially with developmental/molecular geneticists watching =P )

Psi Wavefunction December 15, 2009 at 7:43 pm

PS: The more complex flowers are actually in the boring lowly grasses et al. (monocots). In fact, these lowly boring grasses seem to be perhaps among the more complicated and convoluted of plants, both genetically and structurally! (even their stomata are complicated — they have special cells around them that open and close the stoma, as opposed to the guard cells doing this all on their own. Oh, and the only reason I know this is because I happen to work on stomata at the moment, so stop looking at me funny! =P )

For a self-proclaimed NON-plant biologist, I think I’m responding too much to plant topics… >_>

Jennifer Frazer December 15, 2009 at 9:38 pm

You are totally right about the grasses. But grass flowers are boring to look at, IMO. One day I’ll do a grass flower post, though. We’ll get all freaky wit’ tha glumes and lemmas. [makes grass gang signs] [Awkward silence] Yeah. Anyway, I think I could do a post that simply said “Grasses Have Flowers” and I’d surprise 99% of the population. Grasses have flowers? Where are they? : )

George Shepherd January 4, 2010 at 10:40 am

Jennifer, while I have to agree that 99% of the population would be surprised to learn that grasses have flowers, I’d like to register a protest about your comment that “grass flowers are boring to look at”! Have a look at:
or even
and see if you still maintain this! They’re really quite beautiful. I think it’s all a question of how closely you’re prepared to look….

Barbara Collura December 22, 2010 at 8:43 pm

Dear Jennifer,

Have you heard or seen a show/video about the origin of flowers. I saw a how on TV but didn’t write the name down. It shows a place in China where all flowers were to have started. Could you help me find this video?

Thank you,

Barbara Collura

Jennifer Frazer December 22, 2010 at 10:14 pm

I think this might be what you are looking for, Barbara: http://www.pbs.org/wgbh/nova/flower/
It was a NOVA special — you should be able to watch it on the PBS web site. Hope that helps!

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Previous post:

Next post: