According to the ICZN, species can be named using whichever character evidence you want, as long as they are diagnosable and published in a correct manner. Species that I have seen described based only on molecular evidence usually accomplish the "diagnosable" part by having a short diagnosis section listing the actual bases which are present at the species-diagnostic sites in one or more gene sequences.

According to the ICZN, species can be named using whichever character evidence you want, as long as they are diagnosable and published in a correct manner. Species that I have seen described based only on molecular evidence usually accomplish the "diagnosable" part by having a short diagnosis section listing the actual bases which are present at the species-diagnostic sites in one or more gene sequences.

Hi

For a new species, you need to write a key to separate it from the close species. Therefore, you some morphological characters to use them in this key. Otherwise, how people can separate this species.

regards

As Gunnar said it is possible,these papers  are on gastropods but still the medhod should be usefull:

Article How to describe a cryptic species? Practical challenges of m...

Article How to use CAOS software for taxonomy? A quick guide to extr...

What Gunnar says is true the Code allows that method, but, as taxonomist I have to point out the problems that it will mean for taxonomists in the future. In an ideal world a species may always be identified by easy-to-observe morphological characters. In this case, I recomend you to make an effort to give at least a morphological diagnosis to distinguish each species. Molecular biology is a fantastic tool, but it is not available for everyone.

As Gunnar and Manuel pointed out, there is no specification in the Code which type of characters you need to use for species delineation and description. If your species are cryptic and morphology does not provide evidence for them being seperately evolutionary lineages, but other lines of evidence do - then it is perfectly justified to rely on these (i.e. in your case molecular characters and geographic ranges?). As in all species descriptions you should however make sure that your species delineation is well justified and that ideally it should not rely on only a single line of evidence (i.e. check whether results are supported by different molecular markers). And make sure that molecular-based species descriptions follow the same rules as morphological ones (deposition of type material etc.). I agree with Manuel, that we should avoid to create a world of parallel taxonomies (one molecular-funded, one morphologically). So even if the morphological data of your new species is not diagnostic, still provide the morphological characters traditionally used for your group. Then your species can be at east identified morphologically to a certain level/ cluster of species.

Here is also an interview of Alessandro Minelli on this subject: http://www.biomedcentral.com/biome/molecular-vs-morphological-alessandro-minelli-on-identifying-cryptic-species/

If you are searching for a program, which helps you to extract diagnostic characters: SPIDER in R might be helpful as well but it has less options when compared to CAOS.

https://www.researchgate.net/publication/266973173_Formal_Revision_of_the_Alexandrium_tamarense_Species_Complex_%28Dinophyceae%29_Taxonomy_The_Introduction_of_Five_Species_with_Emphasis_on_Molecular-based_%28rDNA%29_Classification

Article Formal Revision of the Alexandrium tamarense Species Complex...

In principle, defining a species on molecular characters alone  is not very different from defining them on internal genitalia alone, which is very common in entomology.  Both methods lead to great inconvenience for later taxonomists who may not have the equipment or skills to access the essential characters - but I dont see any solution to these problems. I just hope it doesnt reach the point where every taxonomist has to have his own sequencer, because that requirement would eliminate a large proportion of the currently active workers. 

A paper you might be interested in is "Need morphology always be required for new species descriptions?" by Cook, Edwards, Crisp and Hardy (2010; Invertebrate Systematics).

What about the type material in molecular based descriptions? If it is an extraction, would I be allowed to use it? and what would happen when the extraction runs out? There are still many practical problems to solve, I think.

Dear Ainara,

There is a  paragraph on the ICZN webpage entitled "Can DNA be a type specimen?" that addresses some of your questions.

Ideally, you would  deposit your extraction in a museum that is set up to deal with and store DNA of type specimens. 

Also, sequences from the type would be deposited (i.e. submitted to Genbank) and linked back to the type (be it a voucher of the specimen or the DNA extraction or both).

http://iczn.org/content/can-dna-be-type-specimen

It is important to make a distinction between (i) a molecular diagnosis, based on DNA data obtained from an archived type specimen, and (ii) a 'type specimen' consisting solely of a DNA extract, or a DNA sequence.

The former is acceptable by ICZN. The latter should definitely not be considered acceptable. A bunch of DNA fragments in a tube is not a type specimen (in fact, the solution most likely also includes human and bacterial DNA). A DNA sequence is just information, a string of letters, i.e. an interpretation of a tiny bit of information contained within the genome of a type specimen.

In other words, a diagnosis is based on information (evidence of whatever kind for the existence of a new taxon), whereas type material should be a physical body (a specimen to which the name is formally attached).

Joshua´s question is not essentially different from, for example, differentiating cryptic species using sonograms of songs alone, a task which was quite common on frogs or birds. As extensively explained in this discussion, good practice tells you to deposit the "voucher" for the DNA sequences (i.e., the type specimen). You might have sequences of several specimens indeed, but you should identify one of them as type (cannot imagine how to exactly proceed with dinoflagellates...). This is the very sense of using name-bearing types: when you describe a species you state in words what you see on it as "diagnostic features" (morphological or other) -- but nobody knows what kind of characters will be discovered to use in the future, so the only objective way to know (and test) what a given species is, is to make available to future researchers the actual specimen on which you based the name, so that they can check it by themselves (not only believing in the diagnosis you provide).

Dear Joshua,

maybe this article can be of great help.

Best regards

Ferruccio

Obviously you may use different data, e.g. sound or molecular, etc, as diagnostic characters, but for the type specimen, necessarily you must choose an actual specimen and use  the molecular evidances preserved with it.

To support Penelope's answer (and please read the link to the ICZN webpage http://iczn.org/content/can-dna-be-type-specimen, which she provided!

It is of course always better to have a specimen as type than only a DNA extract. But, there are cases (e.g., of minute cryptic and co-occuring species) - in which the species diagnose is based on molecular characters and you need to do the sequencing of all your collected material to ensure conspecifity and because of their small size this might again require to sacrifice the entire specimen . Of course it should be documented prior to DNA extraction and the taxonomist should always attempt to combine different character sets, but it does not violate - and it might be appropriate - to use extracted DNA as type material. PCR products and sequences, however, present copies/ diagnostic characters and can be used for describing species but may not serve as type material.

The problem is when the specimen becomes consumed by the procedure, and only an DNA extraxt is left. As Katharina says "it should be documented" before. Since I am not familiar with microscopic species, I would like to know: is the extraction then pooled, so you are no longer able to identify the exact individual for reference? Can in this case be applied the concept of Hapantotype – A special kind of holotype in the case of extant protistans, which can consist of more than one individual (Article 73.3)? If you previously document (i.e., microphotograph) the specimens, you might select as "hapantopypes" those in the picture, although then consumed and with only a little DNA left. This way you might be able to comply with the Code, I think. How about cloning them, can this be another way to proceed?

In my experience, there is no need to pool specimens. I am not very familiar with the Hapantotype-concept for protists, but pooling specimens for DNA extraction seem problematic to me in case later on, there arises doubt in conspecifity- how to seperate them again? With modern methods, DNA extraction of one specimen should be suffient. One can then use a little bit of the original DNA and amplifiy it to have sufficient DNA for e.g. targeted PCR amplification and use the remaining (non-amplified!) original DNA as type.

Thanks Katharina. Well, if a single specimen can be isolated, I wonder if it could be documented (photgraphed) before it is "destroyed" to obtain DNA. This would mean, the set of photos belong of the (no longer existing) type -exept for the remaining DNA, I guess. This seems useful to ensure a documented link between the DNA info and morphology of that single specimen denoted as type. Not existing types is not precluded by the code (remember there was a case of a endangered monkey whose "holotype" was left alive in nature, while a couple of photographs was the only document used to describe it as new). However, since typification is strongly recommended, I imagine that first documenting the individual and then using it for DNA extraction might be a valid strategy.

I'm maybe replying a bit late, but we had the discussion in my lab few years ago following this paper:

Leaché, A. D., & Fujita, M. K. (2010). Bayesian species delimitation in West African forest geckos (Hemidactylus fasciatus). Proceedings of the Royal Society B: Biological Sciences, rspb20100662.

Their approach can probably be used in your case.