On the status of Red-Seeded Dandelions (*Taraxacum erythrospermum*)

By Daniel Atha


Everyone knows that trust in science, scientists and scientific institutions is at a low point. It’s not surprising the public is skeptical. The masses have been excluded from the language and practice of science. The public has little say in the goals of science. And many of the products of science have made our world more dangerous and inhospitable.

What I love about natural history is that it is comprehensible to everyone with a desire to look. The natural world is there for everyone to enjoy and study equally. There are no barriers to access and the possibilities for learning are infinite. You don’t need a Hubble telescope or a particle accelerator to study natural history.

As naturalists we know that we would all be better off if more people had a greater appreciation for and understanding of Nature and the earth system processes that make life possible. iNaturalist is democratizing science by empowering everyone to participate in the practice and process of natural history science. Every observation records what we find interesting and meaningful. And the totality of our observations reflect our values individually and collectively.

The literature on species concepts in biology is vast. Generally, most people would agree that a plant species is a lineage that has diverged from a common ancestor and has some genetic and morphological cohesion. The offspring inherit genetic, anatomical and behavioral traits from the parents. Put more simply, a species is a reproductively independent lineage (Rieseberg et al., 2006) with corresponding morphological traits. Disputes arise over reproductive capacity and how well genetic markers (genotype) are correlated with morphological character states (phenotype). See Mayer and the biological species concept (Mayr, 1992).

We might successfully classify a species using one or the other or both genotypic and phenotypic characters without knowing everything about the basic life history and reproductive mechanisms of an organism. In fact, most species are named before a great deal is known about their basic biology. But when either data set or the data in combination are ambiguous, we must seek additional data sets, such as reproductive strategy, chemical productions, chromosome number, etc… See Stewart-Wade et al., 2002 for a broad overview of Taraxacum biology.

It’s also important to remember that science is driven by curiosity and scepticism. When it is based on observable facts and informed by mature and reasoned analysis, science can reveal processes and patterns that may be hidden to the casual observer. The basics of science are the formulation of a hypothesis, testing the hypothesis (including the null hypothesis) and repeatability. Much of what the public thinks of as science deals with abstract concepts and astronomical values. But we must remember that a flora of a given area is really a whole series of hypotheses. Our concept of a species is really just a hypothesis that an organism is definable by a set of traits and that we can tell it apart from others reliably by use of a key. And because we are dealing with plants in the landscape, our floristic hypotheses can be tested by virtually everyone, regardless of their training or prior knowledge. That’s the beauty of a flora and the fun of writing keys.

As scientists (and citizen scientists) we must be careful not to make uninformed judgements or appear overly certain about our interpretations of natural phenomena. In our haste, carelessness or ignorance, we may base a hypothesis on insufficient evidence or poor interpretation of the evidence that is available. In such cases, when the hypothesis is overturned by evidence that should have been considered or analyzed properly, the public’s trust is eroded.

The subject of this post, the common Dandelion, is an example where careful studies carried out by scientists over decades can and should help us interpret and appreciate what we can all see in our lawns and gardens every day-- the common Dandelion. My goal here is to determine (by empirical evidence) whether the Red-Seeded Dandelion is a species as commonly understood by botanists and the public or whether it is a color morph that arises spontaneously from a large pool of genetic variants-- similar to an albino Rat.

Like the common Brown Rat, the Common Dandelion (Taraxacum officinale in the broad sense) is nearly ubiquitous across North America in areas disturbed by humans. The Common Dandelion thrives with ample sun, moisture and nutrients, especially in lawns, garden beds and exposed urban soils.

In North America, nearly all Common Dandelion plants are triploid and apomictic (Lyman and Ellstrand, 1984), meaning that they have three sets of chromosomes and reproduce by parthenogenesis (a process called apomixis whereby seeds are formed without fertilization). In other words, they are clones. Common Dandelions with a “normal” complement of chromosomes (diploid, 2n=16) capable of sexual reproduction with other diploid Dandelions and with triploids (2n=24) are known from Europe, where the species is native. Diploids are extremely rare in North America (Verduijn et al., 2004; Lyman and Ellstrand, 1984) and most plants studied in North America are triploid clones, reproducing asexually by apomixis.

Red-Seeded Dandelions (given the name Taraxacum erythrospermum in 1822 by Antoni Andrzejowski and Joseph Besser.) is treated as a species in most of our floras, including the Flora of North America.

In floristics and systematic botany science, we use taxonomic keys (among other methods) to test our hypotheses. The key is a series of choices that are meant to be mutually exclusive. An organism (taxon) is supposed to fit the characteristics presented in one choice, but not the other. Sometimes key are long and have multiple series of choices that narrow down the options until we are finally presented with just two choices-- one leading to our organisms. Most keys are dichotomous, meaning that each set of questions consists of just two choices, as in the key below.

Every flora treats the "species" in nearly the same way. The detailed descriptions of the two entities (when provided) differ slightly, but the characters in the key here are the ones almost always used to distinguish the species.

A. Leaves usually deeply cut throughout their length, the lobes sharp and narrow, usually pointed back; flower heads smaller; floral bracts (phyllaries) hooded and/or with horns near their tips; seeds (cypselae) reddish or purplish at maturity..... Taraxacum erythrospermum.

A. Leaves less deeply cut, particularly toward the base, the lobes less sharp, not pointed back; flower heads larger; the floral bracts (phyllaries) not hooded and without horns near their tips; seeds (cypselae) brown, olive or tan at maturity..... Taraxacum officinale.

The material presented here demonstrates that these are continuous characters (subject to bias and observer interpretation) and are not correlated with each other or other traits. All research specifically designed to test the hypothesis that Red-Seeded Dandelions are a valid species have found that they are not.

Materials and Methods
Research on the leaf morphology of the Common Dandelions shows that leaf length to width ratio and degree of lobeing varies as the plant ages (Sanchez, 1971) and is influenced by light intensity (Wassink 1965; Sánchez 1967). Rounded leaf blades with less lobeing develop in low light and deeply lobed leaf blades develop in high light (Sánchez 1967; Slabnik 1981). Increasing light intensity increases the leaf lobeing and decreases the leaf length:width ratio (Slabnik 1981). These studies show that the character used by every flora writer to distinguish the two Dandelions (leaf lobeing) is not a trait inherited from the parents, but is rather is a response to environmental variables.

Another study found that achenes actually sort into seven color morphs, but these occur independent of other character states (citation needed).

A study of 20 individuals from several populations across Washington, Oregon, Idaho, Montana, Wyoming, Utah and Nevada representing plants identified as Taraxacum officinale and Taraxacum erythrospermum (Taraxacum laevigatum) found that there was no correlation among twenty-five character states of achenes, involucres, receptacles, leaves and phenolic (chemical) profiles. In fact, these characters showed more variation within populations than between them (Taylor, 1987, citation and abstract below). Achene color-- the single most consistent character state used to distinguish Taraxacum erythrospermum-- was found to be independent of any other character, including shape and degree of leaf lobing. As Taylor points out, it is improbable that the many previous studies have overlooked hitherto untried character combinations to distinguish these entities (e.g. micro-characters). Additionally, as an overwhelmingly asexual, apomictic triploid in North America, it is improbable that hybridization and introgression (back crossing) between putative Taraxacum officinale and Taraxacum erythrospermum has blurred species distinctions creating a continuum of character states bridging one species to the other. Finally, the author’s work and several works cited therein demonstrate a correlation between environmental stress and phenotypic (anatomical) expression including leaf lobing and achene color and that this variation is best explained by a single, weedy species exhibiting a range of phenotypic expression in response to environmental conditions.

Morphological and chemical (phenolic) variables were used in principal components and cluster analyses to determine patterns of variation among and within 22 wide-ranging populations of Dandelions. Intrapopulational morphological variation was as great as or greater than interpopulational variation. Morphological variables were poorly correlated, and plants failed to cluster into the two described species, Taraxacum officinale and Taraxacum laevigatum [Taraxacum erythrospermum]. Phenolic distinctions existed among populations but not between species-types, and chemical variables did not correlate with morphological variables. The data, therefore, suggest that morphological variation is largely due to phenotypic plasticity. This conclusion was supported by the observation of a strong relationship between microhabitat and morphological phenotype, with characteristics of T. laevigatum being expressed under conditions of environmental stress. The pattern of phenolic variability reflects the existence of chemical biotypes. -- Taylor, 1987

Another study of 518 individuals from 22 populations across North America found that there were up to 13 enzyme clonal profiles discernible within a single population of Taraxacum officinale, demonstrating considerable genetic variability within a single population (Lyman and Ellstrand, 1984). In fact, Taraxacum officinale was shown to have a total of 47 enzyme and morphological clonal types among all populations sampled and the highest number of clones per individuals sampled (0.091) of any plant known. By comparison, 108 clonal types have been identified by enzyme analysis for Oenothera laciniata, but only 0.051 clones per individuals sampled. Based on these data and the work of others cited by Lyman and Ellstrand, the authors suggest that multiple introductions of Taraxacum officinale from Europe and Asia have contributed to the high genetic diversity found in North American plants.

In yet another study of 318 Dandelion individuals, Lynn Mertens King found 145 chloroplast (cp) and ribosomal (r) DNA profiles among them. In her work, King states….

"North American dandelions with red achenes do not form a natural group based on either rDNA or cpDNA, so lack of differentiation between North American aggregate species Taraxacum officinale and Taraxacum laevigatum [Taraxacum erythrospermum] in rDNA and cpDNA is consistent with Taylor’s (1987) observations based on morphology and phenolic compounds and suggests they are not separate evolutionary lineages." - King, 1993

Citizen science data from iNaturalist demonstrate that the characters most often used by flora writers to distinguish Taraxacum erythrospermum do not stand up to the test. The characters are: 1. leaf lobeing correlated with achene color; 2. floral bracts (phyllaries) with horn-like appendages correlated with achene color.

The following observations show with empirical evidence that seed (cypselae) color is not consistently correlated with either of these commonly used morphological characters. Rather it appears to occur randomly within populations of "regular" Dandelions. (Field studies here in New York City will be designed to test whether distinctly red seed color occurs randomly in populations and is correlated with any other objective character.)

The following observations were culled from a review of approximately 300 of at least 17,000 global Dandelion observations. https://www.inaturalist.org/observations/identify?page=102&verifiable=true&place_id=any&taxon_id=47602. Observations were reviewed quickly and some which are ambiguous may remain.

Seeds (achenes or cypselae) red but floral bracts (phyllaries) without horns

Seeds red but floral bracts without horns and variable leaves

Red and brown seeded plants growing in mixed populations

Seeds reddish brown

There are many observations with olive colored seeds and with highly divided leaves. Of the observations where you can see both the seeds and the leaves, there are just about as many with olive seeds and highly divided leaves. This is an excellent series of photos with olive achenes, highly divided leaves and phyllaries with small projections. https://www.inaturalist.org/observations/80414881


I have found no scientific studies that support the recognition of Taraxacum erythrospermum as a valid species-- only descriptive accounts using continuous, qualitative characters that are subject to bias. It appears that every study actually testing the validity of the hypothesis with objective and quantitative criteria found that there is no basis for recognition of Taraxacum erythrospermum as a valid species. Don't get me wrong. I love all plants, including the ones we call weeds and invasives. An artificial ranking as species does determine the inherent worth of Red-Seeded Dandelions as living beings worthy of our love and respect.

Based on the evidence presented here, it is clear that the common Dandelion, nearly ubiquitous across North America, is an apomictic, triploid species that has very high rates of morphological, chemical and genetic variability within and among populations, but especially within a population. The author has seen no evidence that the red achene character state is anything but a mutant color morph that may appear randomly wherever clonal Dandelions occur. Until there is convincing evidence for the validity of the species, the available evidence argues for treating the common, weedy Dandelion as a single species (Taraxacum officinale).

This is how science is supposed to work. Someone has a hypothesis. That gets tested independently and objectively and either the evidence supports the hypothesis or doesn't. To ignore the evidence and persist with the unsubstantiated hypothesis is not scientific at all.

Except where explicitly stated, the words and research in this article are entirely my own. The facts belong to everyone. Please acknowledge that you heard it here first.

Literature Cited
King, L.M. 1993. Origins of genotypic variation in North American dandelions inferred from ribosomal DNA and chloroplast DNA restriction enzyme analysis. Evolution 47: 36–151.

Lyman, J.C. and N.C. Ellstrand. 1984. Clonal diversity in Taraxacum officinale Compositae), an apomict. Heredity 53: 1–10.

Mayr, E. 1992. A local flora and the biological species concept. American Journal of botany 79: 1537–2197 https://bsapubs.onlinelibrary.wiley.com/doi/abs/10.1002j.1537-2197.1992.tb13641.x

Riesberg, L., T.E. Wood and E.J. Baack. 2006. The nature of plant species. Nature 440: 24–527. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2443815/

Solbrig, O.T. 1971. The population biology of dandelions. Am. Sci. 59: 686–694.

Solbrig. 0.T. and B.B. Simpson. 1974. Components of regulation of a population of dandelions in Michigan. J. Ecol 62: 473–486.

Solbrig. 0.T. and B.B. Simpson. 1977. A garden experiment on competition between biotypes of the common dandelion (Taraxacum officinale). J. Ecol. 65: 427–430.

Stewart-Wade, S., S. Neumann, L. Collins and G. Boland. The biology of Canadian weeds. 117. Taraxacum officinale G.H. Weber ex Wiggers. Canadian Journal of Plant Science 82: 825–853.

Taylor, R.J. 1987. Populational variation and biosystematic interpretations in weedy dandelions. Bulletin of the Torrey Botanical Club 114: 109–120.

Verduijn, M., P. Van Dijk & J. Van Damme. 2004. The role of tetraploids in the sexual–sexual cycle in dandelions (Taraxacum). Heredity 93: 390–398. https://doi.org/10.1038/sj.hdy.6800515

Wassink, E.C. 1965. Some Introductory Notes on Taraxacum officinale L. as an experimental plant for morphogenetic and production research. Mededelingen Van De Landbouwhogeschool, 65–16. Wageningen: Veenman.

Posted on Απρίλιος 05, 2021 1130 ΜΜ by danielatha danielatha


Oh good! I'm sold. I always found the supposed distinction pretty confusing, and my very limited experience with them did seem to show that the correlation between seed color and other traits just wasn't there. Anyway, I just have a couple of comments.
1) You may want to separate the links and the semicolons with a space. Right now a bunch of the links go to a "404 page not found" page. It's easy enough to delete the ; at the end of the url, but some people may be confused by that.
2) some of the observations you linked to are from Europe or other places with other native Taraxacum species. POWO lists about 30 trillion accepted Taraxacum species (well, ok. "just" 2364. See http://www.plantsoftheworldonline.org/taxon/urn:lsid:ipni.org:names:328762-2), and I'm nowhere near familiar enough with the genus to know how reliable the IDs are. How sure are we that some of those observations aren't of other species? I saw at least one observation that only had one ID, by the original uploader. I'm not terribly comfortable with using those unvetted observations to draw any solid conclusions.
3) In real life, I have strong lumper tendencies. For the purposes of iNat though, I think it's almost always best to retain as much information as possible, and to have a strong splitter bias. If we want to change the taxonomy that iNat uses, I suggest we treat the red-seeded variety as a form, and not just lump everything into T. officinale. That preserves the distinction between the two forms, and facilitates future research into this issue.

Αναρτήθηκε από davidenrique πάνω από 2 χρόνια πριν

Oh, one last comment/suggestion: it might be a good idea (and fun/engaging) to make a project for this. We could let people know exactly what to take photos of, and maybe upload the observations with specific tags/observation fields. Given how ubiquitous dandelions are, as well as the fact that they're currently blooming and setting seed, I think we could get a ton of really high quality data that you could use to get a paper published.

Αναρτήθηκε από davidenrique πάνω από 2 χρόνια πριν

I suggest we treat the red-seeded variety as a form, and not just lump everything into T. officinale. That preserves the distinction between the two forms, and facilitates future research into this issue.

@davidenrique the potential problems for this are that it's deviating so far from POWO and anything in the literature; it would basically be creating an entirely new taxonomy for the genus for iNaturalist. We would have to decide for the foms which name to use for each section/macrospecies, potentially create varieties for all the microspecies, etc....

Oh, one last comment/suggestion: it might be a good idea (and fun/engaging) to make a project for this.

What were you thinking would be the qualifications for observations in the project? I've created a couple collection projects to try to sort out observations that have been identified to microspecies: https://www.inaturalist.org/projects/dandelions-of-north-america, https://www.inaturalist.org/projects/dandelions-of-europe
However, to focus on just high quality observations of the genus in general, I'm thinking we would need either a collection project that only collects observations from specific users, or a traditional project.

Also note there have been additional responses to the original post on the forum: https://forum.inaturalist.org/t/overlooked-dandelion-diversity-in-bc-and-everywhere-in-north-america/3808/85

Αναρτήθηκε από upupa-epops πάνω από 2 χρόνια πριν

Hi David @davidenrique

Thank you very much for sharing your thoughts. I genuinely hope to engage people in a discussion of the scientific method, what is a species and how everyone can participate in scientific discovery and natural history research. I will use your very helpful comments (which touch on all these) to clean up and strengthen this post to present a more concise and accurate analysis. As you suggest, broken links and ambiguous or inappropriate data inclusion do not help. I forgot about the forum topic. Maybe we should move this conversation there.

I'll address your suggestions now.

semicolons replaced by spaces and em-dash. All links are working.

I agree that it is probably a good idea to remove the European (and probably northern Canadian) observations as there are at least a few other generally recognized species that could be involved. On the subject of using only observations with "verified" or Research Grade identifications. I think that's circular reasoning. Either way, if we chose observations by the identification, we are biasing the data. The point should be that these are essentially random samples examined without regard to identification and analyzed for the character states under review. Do you agree?

Thanks for sharing your inclination on the lumper-splitter spectrum. I sometimes think my intuition on the taxonomy of a species or genus is the "right" one, but I try to keep an open mind. I also know that there really is no "right" answer. Are there 2,364 "species" of Taraxacum? I don't think so, but it really comes down to how we define a "species". If the methodology and criteria used to define 2,000 species of Taraxacum were applied objectively and honestly, we would probably have as you say a trillion "species". Good science defines a method and sticks to it. And I agree with you that the Red-Seeded Dandelions are interesting and our classification should not obscure their uniqueness and prevent or hinder further study. So I agree that recognizing them as a form of Taraxacum officinale is a very good solution. That's what the data from high-quality research suggests. Problem is I don't find a validly published name at that rank. https://www.ipni.org/?q=genus%3ATaraxacum%2Cspecies%3Aofficinale. Also, iNaturalist often follows POWO and there is not a forma name in use. One consolation is that the original identification as Taraxacaum erythrospermum in iNaturalist is preserved and findable by a variety of search methods. One can always apply that id and then apply another and both will be preserved. One could add the name into the comments section or annotation section as well.

Yes!! I agree that this research is perfect for a community science project. Would you like to help design, manage and promote it? We are approaching the end of Dandelion season here, but there is still some time. And there is always next year...... Yes, a publication is a worthy goal. The peer-review process will be very interesting. Some that are smarter and more erudite than I am will find major weaknesses with everything I have said so far and that will be a challenge. That might result in removing the stuff about what is a species and some of the editorializing about public perceptions of science and just sticking to the core questions. Let's see.

Thanks again for sharing your thoughts and I Iook forward to continue the discussion.

Best wishes,


Αναρτήθηκε από danielatha πάνω από 2 χρόνια πριν


the potential problems for this are that it's deviating so far from POWO and anything in the literature; it would basically be creating an entirely new taxonomy for the genus for iNaturalist.

Oh yeah, I agree... I wasn't suggesting that we DO change iNat's taxonomy, I just wanted to say that we should keep the distinction in some way if/when that happens. I'm totally agnostic about what name should be used or exactly what criteria we should use to justify a change in iNat's taxonomy.

I'm thinking we would need either a collection project that only collects observations from specific users, or a traditional project.

Yeah, that's along the lines of what I was thinking. Only observations from members would be allowed, and the criteria for inclusion would be that certain traits would be photographed (seeds, leaves, phyllaries) and maybe a couple of other things would be noted (whether it was growing fully exposed in full sun or in a shaded area, in rocky or loamy soil, in a mowed yard or unmowed area, anything that might be of interest). The problem is that almost nobody systematically takes photos of all the relevant traits, and people's often blurry and haphazard photos aren't very good for answering these kinds of more technical and nuanced questions.


The point should be that these are essentially random samples examined without regard to identification and analyzed for the character states under review. Do you agree?

That's a great point. I completely agree- if we restrict the sample to areas where there's only these two dandelions. My concern was just that there are so many species of dandelion that some may be easily confused with T. officinale/erythrospermum in other parts of the world.

Would you like to help design, manage and promote it?

I'm not sure how useful I'd be in designing or managing the project, since I have very little knowledge of Taraxacum taxonomy/ecology and haven't really managed projects before, but I'd be happy to help out in some capacity. I can certainly make some observations- I already went out earlier and checked out the dandelions around my place.

Αναρτήθηκε από davidenrique πάνω από 2 χρόνια πριν

Προσθήκη σχόλιου

Συνδεθείτε ή Εγγραφή για να προσθέσετε σχόλια