Autumn flowering narcissus in the Northern Hemisphere 2018

[Maggi Young]

Comparison of seeds of  Nar. obsoletus and N. elegans from   Rafa

[Karaba]

 I am lost, what is clear is that although the sp # of Mallorca, Algeria ... is well known, as well as the N. obsoletus, both have been in the same bag or one as a subspecies of the other, when they are clearly two different species. I put a photo, well known for those who have not seen it, already with the names"

If even Rafa is lost, how should we be ?

I'm very surprised of the name retained for the multiflorus leafy autumn flowering species.

About Narcisuss miniatus/deficiens/obsoletus : it seems quite clear that the name deficiens Herbert, 1847 has the priority and that miniatus is a synonym. Here is a link to the description : https://archive.org/details/mobot31753002748462/page/n63 I don't now why Koopowitz et al., 2017 discussed about the miniatus/obsoletus controversy and totaly ignored the name deficiens that was bring back by F Casa (2008)

About elegans/obsoletus : both name should be forgotten ! and shouldn't be used anymore as so many confusion appeared in the various papers  . At the beginning, there's two plates, one from Desfontaines (https://gallica.bnf.fr/ark:/12148/bpt6k6514977k/f173 with description of the species here https://gallica.bnf.fr/ark:/12148/bpt6k6510373m/f309) and one from Parkinson (https://archive.org/details/paradisiinsolepa00parkrich/page/88) who was inspired by other plate (see Koopowitz et al. 2017). For both plates, we have very few information as Desfontaines called it Narcissus serotinus, believing that it was only variation of this species, and Parkinson wrote that it has only 2 flowers (but did he see it ?). From these plates, Haworth described Narcissus elegans (read https://www.biodiversitylibrary.org/item/15398#page/14/mode/1up and second edition http://bibdigital.rjb.csic.es/ing/Libro.php?Libro=7195 page 19) and Narcissus obsoletus (https://www.biodiversitylibrary.org/item/37886#page/163/mode/1up) without seeing the plants but only the plates !  The description he made of Narcissus elegans don't fit Narcissus elegans/obsoletus (as we know it) as he only described the plate (with lots of 'elegantissimus' !). For the plate made by Parkinson, lots of authors wonder if it's a Narcissus deficiens with 2 flowers and leaves or if it's a elegans/obsoletus with only 2 flowers. Because of the poor quality of Parkinson's plate, the name Narcissus elegans remains for the multiflorus autumn daffodils.

If there's now two species of multiflorus autumn daffodils, it seems difficult to me to assign names based on old plates drawn from plants that we don't now where they came from. Reading the description of the 2 species made by Rafa, it's a bit odd that Narcissus elegans (illustrated by Desfontaine as a big 7-flowers plants) is assigned to the less robust and Narcissus obsoletus (which have only 2 flowers on Parkinson's plate) is assigned to the more robust. As there is no type of both species, it should be better to abandon both names or at least obsoletus as Narcissus elegans is widely used as the multiflorus autumn daffodil with orange corona. obsoletus has been so many time badly used that keeping this name for renaming what we know as N. elegans seems to me a really bad idea.

(Gay, 1850 is also an old interesting paper : http://bibdigital.rjb.csic.es/spa/Libro.php?Libro=2279, page 80 to 91, in french)

[Rafa]

Dear friends,
Sorry for my long silence here.
Well, I think I am understanding this complex. This is my theory
I think we have 3 parents and 2 similar species

Description of the parents:

- 2 leaves when blooms, fluted, with strips, glaucous or green. Robust plant,  stem with strips glaucous. 1-21 (+?) flowers per stem, normally 8. Pure orange corona liek a cup, with the margin close, wide tepals -------> N. obsoletus
Distribution: at leastSpain, Morocco, Sicilly, Libia, Tunisia (?)

In my drawing you can see Mediterranean elements and Atlantic elements, this is because I consider the ecology is fundamental to understand the plants. So, for example you can see more localities of N. serotinus in Spain, Portugal and Morocco,

-  2 leaves when blooms, fluted, with strips, glaucous. Slender plant, stem with strips glaucous. 1-8 (+?) flowers per stem, normally 2. Yellow-orange corona, like a narrow ring, lobulate margin and open, narrow tepals---> N. elegans
Distribution: at least  Algeria, Balearic Island, Malta.

- Without leaf when blooms, glaucous stem without strips. Tiny and slender species, uniflower (with genetical exceptions, rarely 2), citron yellow corone with segments,  perianth tube like a bottle with segments Yellow/green.----> N. serotinus
Distribution: Portugal, Spain and Morocco. I am sure it is possible to find N. serotinus hidden in N. deficiens and N. miniatus localities, this is the case of Formentera.

N. obsoletus x N. serotinus---> N. miniatus
N. elegans x N. serotinus-----> N. deficiens

In my drawing you can see Mediterranean elements and Atlantic elements, this is because I consider the ecology is fundamental to understand the plants. So, for example you can see more and bigger localities of N. serotinus in Spain, Portugal and Morocco, and once it goes to the Mediterranean, seems to be marginal.

So finally I am thinking in support N. miniatus, but not because N. deficiens has short corona, one leaf when blooms etc... these are not valid characters because it is the normal variability, as they have an hybrid origin they have a range corone types, thay can have (or not) ONE, leaf when blooms... I am thinking in support N. miniatus, because there are N. elegans, N. serotinus in Balearic Islands so the hybrid (as Mediterranen enlement) must be the same like in Greece. Also in Malta there are the fertile hybrid with one parent. So, it must be another N. deficiens like, in countries that contain N. obsoletus and N. serotius or at least one of the parents together with the hybrid. This is the case of Morocco, where they are greowing all together, or in Sevilla that you can find N. serotinus together with N. miniatus.

Is necessary to say that in the speciation proccess of an hybrid in Narcissus genus, the hybrid normaly replace the ecological role of one or both parents. The most interesting case apart N. deficiens and N. miniatus, is N. piifontianus (fertile N. x perez-larae)

[Karaba]

Hi Rafa, thanks for this long explanation. At least, I can understand your point of view even if I'm not agree on several points.

First of all, I have also to admit that I don't have any field experience about these species. I can only speak through publication, picture on the web...

I take some time to read some different papers, old and recent ones. At the beginning, I was a bit upset by the choice of the name elegans and obsoletus for this new couple because it seems to me very ambiguous and not very relevant. I try to understad why it was chosen like this and why it cause some problems. I also try to understand your speciation theory and try to find proof (or not  )

About the names. The use of obsoletus is confusing because it has be used to name two different taxonomic entities. In last 10 years, the same Parkinson's plate is used as the type for Narcissus deficiens/miniatus (Diaz Lifante & Andes Camacho, 2007; Aedo, 2010) and for the commonly named N. elegans (Fernandez Casas & Pizarro Dominguez, 2007; Koopovitz et al, 2017). Now, you use again elegans but name an other taxonomic entity.

Back to the protologue of obsoletus and elegans, Haworth 1819 and 1831. Both have stalk with severals flower and leaves when flowering, but the holotype of obsoletus (Parkinson's illustration) is said to grow in Spain and elegans (Desfontaine's illustration) is said to grow in Algeria. So the choice : the one who grow in Spain should be named obsoletus and the one from eastern mediterranean, elegans.

The problem comes for me from the rest of the description from the protologue. Obsoletus is said to have only 2 flowers and a "exigua obsoleta lutescente" (narrow reduce and yellowish) corona which fit more with the description of "your" elegans. In the other hand, elegans is said to have 1-7 flowers by stalk, and a corona "minuta integra" (small and entire, not lobed), which fits, for me, more with "your" obsoletus (number of flower, shape of the corona). The description of obsoletus with only two flowers and a yellow corona (see Haworth 1831) even lead to the typification of Narcissus miniatus/deficiens with the Parkinson's illustration.

Indeed, it's quite difficult to know what Parkinson have illustrated copied because, with no doubt, its illustration is a copy of Rabel's illustration (https://gallica.bnf.fr/ark:/12148/bpt6k1511032x/f53.item), which is nearly a copy of Valet's illustration (https://gallica.bnf.fr/ark:/12148/btv1b8612033t/f73.item) as Koopovitz et al., 207 have already shown. From the description of Parkinson, it's possible that he has grown it as he has grown Narcissus viridiflorus but, he gave very little information. The description of the leaves is like the description of the illustration and don't fit with the thin leaves of any of the autumn daffodils. But he described the color of the corona as he has seen it : the cup is small and round, like unto the cup or crown  of the least Rufh Daffodil (Narcissus assoanus?) (which fit with "your" obsoletus), of yellow color at the bottom but toward the edge of a dune or sullen color (which is not a pure orange and fit with "your" elegans). He wrote that it comes from Spain but didn't say if it was rare or common or anything else. From the illustration and description of Parkinson, Haworth made an mix, describing the illustration (which, I repaet, is a copy of prvous illustrations) and mixing some of the element of the description.
At the end, I really have no certitude about which species have been described by Haworth and Parkinson

About Desfontaine's illustration, it's a bit easier. Desfontaine was thinking that the same species could have one or more flower by stalk and only one species was known at the moment. Its illustration is an autumn daffodil with 7 florets and he wrote that the plant was growing in Algeria. There's few useful information about the corona of the multiflower individual as he was thinking to describe only one species, Narcissus serotinus.

This long analysis lead to me to 2 possibilities :
-using obsoletus and elegans as Rafa but the "new" species don't fit the protologue of their names on several points and, even more, they fit to the protologue of the other species which is very confusing. In this case, maybe the definition of an epitype (see article 9.9 of the ICBN https://www.iapt-taxon.org/nomen/pages/main/art_9.html) would be useful
-using obsoletus for the eastern species because it fits the protologue of obsletus and it has the priority but I found no names for the spanish/moroccan species since, at least, elegans , oxypetalus and cupanianus are typified by algerian or sicilian individuals !

(about the speciation in a few days  )

[Rafa]

I think Desfontaines was mistaken and what he was actually discribing as N. serotinus was N. elegans. So if  the plant in this herabrium sheet with narrow tepals is N. elegans, the other species with whide tepals and  two leaves when blooms, is N. obsoletus. Number of flowers is not a valid character.
Diaz Lifante & Andes Camacho, 2007; Aedo, 2010 are completely wrong in my opinion.

There are two similar plants in my opinion

1. N. elegans
http://mediaphoto.mnhn.fr/media/1443467552047lamVn3XxNXUes2UF

[Rafa]

2. N. obsoletus
No mater how many times it was copied, the drawing in my opinion represents this plant

[Karaba]

Rafa, you have the chance to be able to choose the right pictures 

No problem about Desfontaines.
About Parkinson, even if the number of flower is not a good character, Parkinson said no more than 2 flowers and the corona is said to be yellowish, not orange. This description is also used by Haworth. We can argue a long time about what Parkinson has drawing, with the quality of the illustration, the character that don't fit only one taxon, the absence of details of the provenace (yes, he said it came from Spain) of the plants, it will difficult to be sure of what he draw (that's my opinion  )

But we will not give a definitive answer to this question here (or on Facebook), only a publication with peer review or a proposal analysed by the ICBN Board would be a reference that everybody can use.

I would also like to have a valid publication showing without doubt that there is more than 3 species. Pictures of extreme may be clear but I would like to know what goes in the middle. For the moment, genetically, there's no proof of any differenciation between eastern and western 'elegans' populations, nore between eastern and western miniatus/deficiens populations and the subspecies statut would maybe be more adapted.

[Rafa]

I agree It is difficoult place N. obsoletus, but for sure, Desfontaines it is clear to me, and Parkinson's plate really looks like the pure orange corone from Malaga, the only problem is that there is just one place knonw in Spain (peninsula) for this species, and there are few places in Spain (Balearic Islands) for the other species. If we have two different drawings and two diferent species, to me is more simple to associate Parkinson's plate to the orange coronaplant, as N. elegans is perfectly described by Desfontaines in his drawing, and there is not a third autumn species with two leaves and round corona. N. deficiens/miniatus, never have two leaves in bloom, if you find it, is because in the same tunic they are vegetative divisions too younger  to develop the stem, and they use to develop the leaf. Also it has always segmented corona in different shapes sizes and colorus.

Morphologically I think it's imposible to make differences between N. miniatus and N. deficiens, so I think it is more practical to consider N. deficiens as the closest plant to N. elegans, and N. miniatus the plant as closest N. oboletus. This is because In Mediterranean countries the plant that I call N. obosletus, is isolated without contact with N. serotinus and  the descendant. But in this Mediterranean countries you can find what I call N. elegans together with the descendant (N. deficiens, was described in Greece) and also there are hidden N. serotinus.
The only place that I know where N. obsoletus is in contact with N. serotinus and the descendant is Morocco.

For example If we found in Peninsula Iberica, N. obsoletus together with N. deficiens/miniatus, it would be logical to think it is N. miniatus, instead N. deficiens. And the same with N. elgans. And this is applicable to many cases in Narcissus.

I think it is neccesary to make DNA test and Alkaloid trace in:

N. elegans from Mallorca
N. obsoletus from Morocco
N. deficiens from Mallorca
N. miniatus from Morocco.

I think is not necessary to describe or invent any other name, the plants have been well known for centuries so concerning names N. elegans is clear to me and N. obsoletus, it is easier to use, to make the differences between both plants, it has been know for a very long time.

[Rafa]

N. viridiflorus

[Rafa]

N. viridiflorus

[Rafa]

N. serotinus
second picture N. x alentejanus

[Maggi Young]

Superb photos of these  gems in nature, Rafa!

[Gerdk]

Superb photos of these  gems in nature, Rafa!

Indeed - especially from the serotinus population with hybrids near Morón!

Gerd

[Karaba]

Thanks Rafa for this nice interlude in our boring discussion 

What are the yellow flowers with serotinus, Ranunculus sp. ?

If you go back for collecting seeds of Narcissus viridiflorus, please, take also seeds of Acis autumnalis 

[Karaba]

Back to the boring taxonomic discussion

I agree It is difficoult place N. obsoletus, but for sure, Desfontaines it is clear to me, and Parkinson's plate really looks like the pure orange corone from Malaga, the only problem is that there is just one place knonw in Spain (peninsula) for this species, and there are few places in Spain (Balearic Islands) for the other species.

And the pure orange corone population from Malaga is only nown since 2007. Fernandez-casas and Pizarro-Dminguez (2007) didn't mention any historical herbarium sheet of this species in Spain (peninsula) before 2007. From Balearic Island, the only sample is from 1935.

The nomenclatural dilemna is also part of the taxonomic problem. The use of the names elegans and obsoletus results from the supposedly fourth white autumn flowering daffodil. I say "supposedly" because if I agree that there are some morphological differences between eastern and western populations, I need more proof of the species statu.
At least 2 DNA and karyological analysis have already been managed.

Karyology :
western and eastern populations have the same number of chromosomes : 20 (Diaz Lifante et al. 2007). For the sicilian population, they infer the species from the karyotype but the species and the karyotype has been confirmed by Kamari et al. (2013).

DNA :
Even if that was not the aim of the publication Marques et al. have done a phylogenetical study of these species. As it was not the aim of the publication, the sample protocol from the eastern populations is not optimized but it gives some useful information.

First paper and most useful is Marques et al. 2010 with a organella and nuclear DNA study of Narcissus cavalinesii, elegans, miniatus/deficiens and serotinus. The sample include 2 eastern populations of N. elegans (from Sicily and Sardegna), 3 of miniatus/deficiens (one from Greece, one from Sicily, one from Sardegna), and 3 of serotinus ! (from Malta, Crete and Ionian Islands !), and several western populations from all species. (Marques et al, 2017 use the same samples than in 2010 so I won't talk about this paper)

I feel a bit embarrassed that there's no more information about some of the sampled population and/or there is some mistake :
- it was the first time that a Sardinian population of N. elegans is formely sampled (even if the presence of N. elegans appear on several maps, it is not reported on this Island by Kamari et al. 2013) but this sample is located in Sicily in Marques et al. 2017 !
- the sicilian population from Narcissus elegans comes from the botanical garden of Lisboa and it has some strange behaviour (I suspect that it doesn't come from Sicily)
- this is, as far as I know, the first published coherent localisation of Narcissus serotinus out of southern Spain and Morocco (I say coherent as the karytype is reported to be 10 and at least the maltese population has a serotinus haplotype).

Haplotypes (DNA from organella) are separated in 3 main groups : Cavanilesii, Miniatus-Elegans and Serotinus. This supposed that the miniatus species came mainly from crosses with elegans as the mother. There are some oddities as 4 populations of elegans are in the cavanillesii groups  : Sicilian (in fact from BG of Lisboa...), 2 moroccan and the one from Malaga. Elegans populations from Majorca have some pecularities but inside the miniatus-elegans group. The sardinian population of elegans is with the other elegans-miniatus populations. The eastern populations of miniatus are with the other populations of miniatus (n east/mest distinction). The 2 greek 'serotinus' populations are also in the miniatus-elegans group haplotypes but have some quit distincts ones.

Ribotypes (DNA from nucleus) : the tree published is even much more difficult to describe and analyse. Its quite funny as lots of sympatric populations groups together even if they are from different species. Even elegans and cavallisnesii seems to exchange some genes. Sardinian population of elegans is closes to the other populations of this species, the sicilian one is quite isolated (but see previous remarks about this population). Several elegans populations are indide the cavalinesii main branch but these are only moroccan populations. miniatus ribotypes are difficult to analysed since this species is quite dispersed in the tree (multiple hybridisation events ? retrocrossing ?...). Part of the sicilian population is isolated in the main cavalinesii branch, the other part of the sicilian population and the sardinian one is withe the maajorcan and spanish populations inside the serotinus branch. So, still no east/west distinction. (note that the 3 eastern serotinus populations are grouped together inside the serotinus branch...)

From these studies, it comes that eastern and western elegans population have the same karyotype, so they should hybridise nearly freely when they meet (between Morocco and Algeria). DNA studies show no differenciation of the studied eastern elegans and miniatus/deficiens populations from the western populations. That why I think that it's better to use a subspecies level inside the elegans species and to gather the eastern and western populations of minitaus/deficiens under the same species.

I would keep elegans because there's no confusion about what Desfontaine has drawn. And I would drop obsoletus because it's too confusing, as it's difficult to know what Parkinson has reproduced  and the name has been typified and used in several Flora (at least Flora iberica and Flora gallica) for Narcissus miniatus/deficiens, writing the synonymy would be terrible.

Sorry for all the people that don't like long writing and don't care about the names..