Yes. People have been identifying species without DNA since Carl Linneaus (1735). Vertebrate paleontologists like myself favor measurements and configurations of teeth and skulls (as they have typically evolved many derived characters), but as with all such work it depends entirely on the organisms being studied. There are many such papers coming to my ResearchGate feed every day. Although I have yet to formally define a new species myself (but I am forever hopeful), I've used morphological characters to confirm other workers' diagnoses: articular facets of Miocene camel metapodials to distinguish between the genera Alforjas and Procamelus (the closest I've come so far), the number dental cusp lophs within a 10 cm span in Mammuthus, skulls of the lungfish Gnathorhiza, teeth of Permian labyrinthodont amphibians, and many, many others. All of this work is posted here on ResearchGate. Keep in mind that statistics are helpful to sort through "noise" in large samples, but they are not required by any means. All you need is an exhaustive review of already known but similar species, a thorough systematic description of the characters that identify it as a unique genus and/or species, and to follow the naming and publishing conventions laid out in the International Code of Zoological Nomenclature. Good luck!

Maybe, this work will help: Keat-Chuan Ng C, Aun-Chuan Ooi P, Wong W, Khoo G. A Review of Fish Taxonomy Conventions and Species Identification Techniques. Survey in Fisheries Sciences. 2017; 4 (1) :54-93 DOI: 10.18331/SFS2017.4.1.6 URL: http://sifisheriessciences.com/article-1-103-en.html

As stated by Hannan and Alex, you can do it without a barcode, but your case will be stronger if you can also place your new species in a phylogenetic context. Check GenBank which molecular markers have been used for the genus in question, so that there is data available for comparison. You can always use CO1, which is the classic barcode gene, but many studies are done with other genes that are often more informative, like cytochrome b or ND2.

Best

Urban

Remember a species is a population of interbreeding individuals isolated from other populations, which therefore evolved its own way. What will show up in morphology and behaviour. Such statements are enough to define a species.

DNA is a nice add-on, if you can prove the stated divergence exceeds intra-specific variability. DNA in fact is much more useful to set up phylogenetic trees showing relationships between different groups. Results obtained this way are objective scientific facts everybody can check.

Conversely, introducing barriers in a phylogenic tree between such monophyletic groups to define taxa is a purely subjective approach. Nature is a continuous process, ignoring barriers human mind needs to introduce for its better comfort, you may call understanding.

That's why modern workers prefer to refer to "clades" of undetermined status, rather than to usual taxonomic categories. Going back to species concept, yes it can be defined the usual way. Which is the only one available to paleontologists. Remember that introducing fossil species in a phylogenetic tree is essential to understand evolution.

All these answers are pretty good. Of course you can describe a species without using any particular character system, as long as you provide a diagnosis that separates your new taxon clearly from others, meet the criteria of availability and other rules in the ICZN Code (for a fish).

DNA is a good source of data, but many species are described based on morphology only, as paleontologists and classical taxonomists regularly do.

And of course, it is possible to place taxa in a phylogenetic context without DNA sequences - that's how we know birds are dinosaurs, for example, and that is what Willi Hennig did, after all.

Andrew: "Birds are dinosaurs" (because they are considered descendants of dinosaurs)? If so, birds - like dinosaurs and all vertebrates (including man...) are invertebrates; but, like invertebrates and all metazoans, they are protozoans; but, like protozoans and all eucariotans they are procaryotans; - and so down to coacervates! I cannot imagine what could be the sense of such "classification" (unless we believe that dinosaurs - and vertebrates, metazoans, eucariotans &c., &c., &c. - have been separately created and have no ancestors)...

Birds are dinosaurs in the same way that you and I are primates. Birds are the sister taxon of oviraptosaurs, and both of those are more closely related to a Tyrannosaurus than a Tyrannosaurus is to a Triceratops. The two groups (birds, dinosaurs) represent different levels in the classificatory hierarchy, or a small twig attached to a bigger branch on the tree of life.

Paraphyletic groups are not evolutionary units.

Ancestors are reified theoretical constructs inferred from empirical evidence. Taxonomically speaking, there is no fossil that can be demonstrated with any degree of empirical certainty to be the ancestor of anything.

Andrew: "Paraphyletic groups are not evolutionary units" - a cladistic dogma without relation to reality: paraphyletic groups are exactly as evolutionary units (being results of evolutionary development) as holophyletic ones!

"there is no fossil that can be demonstrated with any degree of empirical certainty to be the ancestor of anything" - whether or not we know which particular species (or individual fossil) was the ancestor, we nevertheless know that some ancestor had existed, and in overwhelming majority of situations we know to which higher group it belonged (otherwise what would be the meaning of the definition of monophyly: having common last ancestor?).

I fully agree with the various comments about describing species without DNA 'barcoding' as I do it all the time. In morphology (which is really morphogenetics) I look for consistent patterns of differences and uniquely shared similarities to propose species that may be subject to any kind of future 'testing' (in the open sense of evaluation and corroboration or rejection rather than falsification). It is quite possible to designate different species categories even where breeding is not isolated (i.e. hybridization occurs). I recall a classic case with a frog species where adjacent populations can all 'interbreed' but the most distant populations cannot. And I agree with the view that a phylogeny need not only come from DNA, or that DNA trees are necessarily correct when in-congruent with morphology (as in human and great ape relationships). I would add the view that species need more than a diagnosis of difference, but also at least one uniquely derived characteristic to support a phylogenetic separation.

The phenetic methodology, along with other data such as ethology, proteomics, geographical distribution is sufficient to differentiate species. Only in case the previous methods are not conclusive, a comparative study of DNA sequencing can be used. On the other hand, the study of genetic material can show us species of extraordinary genetic resemblance, but which are actually cryptic species. In summary, all the methods are compatible, but I doubt that only one genetic barcode is sufficient since the same codes could occur in two populations apparently of the same species but that separates them from a physiological or ethological barrier that de facto prevents reproduction between these two populations.

Only in the case of the organisms of the Monera and Protista kingdoms does it seem that genetic barcodes could be more useful and determining, but the problem of the correct definition of the species concept will always be present, but this is another matter.

Regards

I am in general agreement with the points made by José although I would distinguish between phenetics as a measure of overall similarity (whether morphology or molecular) and phenotypic in reference to morphology (although DNA sequence studies are also molecular morphology and therefore phenotypic at the molecular level.

This might help: Article Should we describe genera without molecular phylogenies?

Yes, of course.

As asked ("can we..."), my answer is - YES, we can.

yes, we can

Yes, of course, but it must be well justified. Integrative analysis combining morphological, anatomical and molecular analysis is desirable, but sometimes it is not possible. Best wishes, Andrzej

Morphological study for species identification is the mirror image of species identification by DNA barcoding. Yes, we can describe a new species without DNA barcoding, but if DNA barcoding facilities is available, then we can use this molecular tool to confirm the identification of new species done by conventional Taxonomy.

M. A. Azadi Rather than barcoding confirming the identification of new species done by 'conventional' Taxonomy, I would suggest that it is just a potentially corroborating source of evidence. Otherwise the implication is that barcoding (a misleading term in of itself) is the arbiter of what is a species or not. DNA differences may or may not correspond to a morphogenetic study without automatically prescribing a species status - in my opinion.

there are two ways :

the first defines new species based on minor differences between organisms.

the second tends to ignore minor differences and to emphasize major similarities.

Surely, you may describe new species on the basis of their morphological characters if that organism has well supportive morphometric and anatomical characters which may clearly differentiate from previous ones. But DNA barcoding is wide acceptable tool to supports morphological taxonomists in various angles: First, based on keys made by morphological characteristics all stages of life of organisms cannot be identified so keys have limited role for species identification. Second, morphological identification tools often become unable to differentiate cryptic species. Third, phenotypic plasticity often leads to incorrect identification. Fourth, misdiagnosis is common during identification of species on the basis of taxonomic keys because it needs high level of expertise that cannot be common.

DNA patterns can certainly be correlated with a morphological entity as an aid to identification - where that resource is available. Cryptic species are just species not previously recognized. Such 'cryptic' species are often discovered through morphological means. DNA patterns can also point to unrecognized species but levels of similarity or difference still require evaluation. There is no species 'essence'. If the necessary expertise for using keys is not common, neither is the expertise and resources for using DNA criteria all that common. At least a key does not require additional money to use.

DNA bar coding is a recent development in genetics in which a short DNA sequence is read from any genetic sample. It is used to obtain taxonomic information about unidentified organisms. Comparing the genetic code of organisms by DNA bar coding has made species identification easier. DNA bar coding can provide a more specific analysis to identify a new species. So, we can not describe a new species without DNA bar coding.

Munira Nasiruddin DNA sequencing does make ID easier where the sequence is correlated with a known taxon and the morphology is not readily distinct or requires some other special treatment (such as dissection). But it is my impression that DNA sequencing does not provide a 'more specific analysis' to identify new species, at least not necessarily more specific than morphology, and the identification of new species requires a prior context by which to judge the significance of sequence similarity and difference, particularly as there is no single sequence that defines any species. At least that is my current understanding (and happy to consider correction on that).

We exactly CANNOT identity new species based on barcoding! Barcoding data could have a signal for existing of a new species, it's true. Barcodind identifies some BINs, nobody known what is it, and they change all the times after adding new sequences. Regading these strange products as something species-like is clearly contradicting to the main approach declared in the Preambule to the International Code of Zoological Nomenclature created for STABILITY.

As per ICZN, till nor there is not provision to exclusively describe any species new to science based on DNA bar coding. I fully agree with Dr. Kotov on this issue. DNA bar coding or molecular systematic provides supplementary material to support morpho-taxonomy. A balance is desired to be maintained between 'reverse taxonomy' vs molecular taxonomy'. If not done so, it hay seriously affect nomenclature stability. I strongly believe that even various cryptic species have certain micro or macro-morphological differences. One certainly has to figure out these important differences and not to rely exclusively on bar coding or other modern techniques. In this context, I strongly opine on approach of integration of 'eco-systematics'.

Since it is usually difficult to obtain direct evidence of the presence or absence of reproductive isolation, it is common to determine the conspecificity or heterospecificity based on the consistency among multiple indirect evidences from an integrated approach.

Phylogenetic analyses and species discrimination analyses using sequence dataset of multiple genetic markers and or genome-wide sequence dataset usually play a central role in such integrated approach. On the other hand, the reliability of conventional COI-based DNA barcoding (DNA barcode gap) is limited. COI-based DNA barcoding is convenient and useful in the practical field (for example when identifying unknown biological contaminants in an imported product), but should not be overly emphasized in species discrimination. This is my feeling based on my recent experience when dealing with ants (Formicidae).

yes you can do it without DNA barcoding if there are important and new morphological differentiation between these species

Yes it's possible to identify new species from morphological characters especially when there are visible differences with related species.