Both genetic diversity, maintained by patterns of gene flows, and ecological diversity, maintained by patterns of matter and energy flows, are extremely dynamic in space and time. Any category at either ecological or evolutionary level, including the 'biological species' or the 'evolutionary species' concepts, are bound to be spatiotemporally fuzzy. Nonetheless, this fuzziness is useful, since it allows us to describe and quantify the heterogeneity and complexity of the measurable world. If we didn't analyse and break apart the continuum, we could not compare different levels of variation. Indeed, measuring implies defining units.

Intuitively, we grasp there is diversity and unity out of our mind, and inside our mind there is a mechanism called analysis, implemented by a serial processor, which allows us to describe both of them. This psychological mechanism may be peculiar, and can be biased (see the concept of "tyranny of the analytical mind", by Richard Dawkins), but we have tools (e.g., synthetic skills, deductive and probabilistic models) to cope with these problems.

This being said, there are measurable evolutionary consequences of a reduced gene flow, through pre- or postzygotic mechanisms. Driven by the evolutionary forces, divergence may lead lineages in different directions, and reproductive isolation mechanisms may lag behind other forms of differentiation. If environmental and ecological conditions change, reproductive isolates will likely be re-defined by introgression and hybridization, or rapid reductions of gene flow (e.g. acting on circles of species). I think we should still give names to evolutionary units that are ecologically, morphologically, and genetically distinct, especially when our priority is sustainably managing the rapidly declining ecological and evolutionary diversity.

Thanks Gianluca, that is useful. I agree that the "fuzziness" is very useful for understanding the evolutionary process - that's the fun part. I also agree that we still need to give names to evolutionary units that are distinct. So if the process if fuzzy by definition, but we need something distinct to describe it, doesn't that point to an objective species definition that is scientifically very clear about what we call a species or not. Some are concerned about the resulting species inflation, but tend to forget there has been major species deflation after the Biological Species Concept came onto the scene in the 1930s. To come back to my pig example, there were 48 SE Asian pig species in 1928, 3 in 1940, and now we are back to 11.

I observed the same trend in my main group of interest, the mudskippers. A drastic deflation in 1989 after a revision that also eliminated a lot of subspecies, then again an inflation, after several redescriptions and revisions. I think the lumping trend was excessive in vertebrates, as often revealed by molecular studies.

Hi Erik

Why do we think they are real? I guess because we haven't thought it through. Species are man-made concepts (abstractions).

I look at these topics from a plant sciences perspective, but the issues are shared. Taxonomic categories, species included, are human defined, not biological truths. They are a matter of practical convenience and, though they very often reflect genuine biological processes and patterns, they require drawing of sharp-lines across patterns of continuous change (how many generations back do we need to go to see where our ancestors were of a different species to us?).

There have been various efforts to develop taxonomy into a more formal science (i.e. with falsifiable hypotheses) which requires testable predictions, E.g., See Knapp S. 2008. Species concepts and floras: what are species for? Biological Journal of the Linnaean Society 95:17–25. But these relate more to phylogenies and their utility than to any reality for the labels.

There is a nice article by PF Stevens 2006. An end to all things? – plants and their names. Australian Systematic Botany 19:115–133. Who argues that species concepts and species names are "too valuable a communication device to be discarded" but we need to learn to recognize and accept the shortcomings.

So I guess the argument should be about which kind of definitions and criteria are most useful (and this may depend on purpose).

There are certainly plant scientists who appear to over emphasize the reality of plant species, for example in discussions about potentially invasive exotic plants.

See for example:

Article Of Cecropias, Snarks and Boojums

I think to a large degree we are learning to work around these things. For example, bracken fern (Pteridium aquilinum or Pteridium spp.) is near globally distributed but hugely variable. Is it one species or many? Whatever the answer we need to recognize and deal with the variation across sites. We cannot just assume that a study in Europe will apply in Africa. See, e.g.

Article A first look at the impediments to forest recovery in bracke...

Hi Doug. Good points and thanks for the input. If we can agree that species are primarily labels that help us categorize nature rather than distinct and fundamental evolutionary building blocks, the discussion can focus on how to best label them, i.e., formalizing the science, as you call it. If we are going to come up with falsifiable hypotheses regarding species identifies, we need to agree on a definition of a species in the first place. Whichever definition we pick, it needs to be characterized by variables that can be tested with data. And it should be possible for anyone to reproduce these tests given the same data. Variables such as natural reproductive isolation are not very useful then, because this can never be tested for all species, including for example asexually reproducing species, which probably makes up a large percentage of all species, or allopatric species that would never naturally reproduce.

Dear Erik,

I hope warmly that the querelle concerning the new Pongo species will help to make things more clear, even inside and around IUCN. Even recently I had an assistent editor of an important Conservation journal reducing our proposed ms on "taxonomy and genetic rescue" as a mere question of splitters vs lumpers. As you may well understood thinking to pigs and allies, we may easily assure the conservation of three or four 'species' around the globe; more these three or four species could be also on the rise being introduced to new regions and islands as it is the case with feral pigs or ancient introductions such as the Sardinian pig. We may even produce biodiversity while the true sad reality we are losing tens of ancient distinctive lineages around the planet. So I think that IUCN must decide if they need a scientific classification to seriously monitor biodiversity or a classification for the mass media.

Good point Gippoliti. I will take this up with the IUCN

All of this fuss is certainly due to the fact that the focus is on a large mammal and in particular, a primate. In ichthyology, it is everyday practice that new species are defined by taxonomists, based on essentially subjective morphological characters, and the history of the taxon. Of course ideas must be discussed and defended in peer-reviewed descriptions, but there are no objective criteria to name and define a species in the ICZN. And I do think things should stay this way.

As more information (molecules, ecology, physiology, anatomy, etc.) is gathered from different disciplines, descriptions may be updated, as we have a better understanding of what is the biological variability within the defined units vs. the variability between different units. Even when only considering phylogenetically-defined cladistic units, the distinction between a "genetic population" (e.g. using bayesian methods, or pairwise phist matrices among series of samples) and a biological species that more or less frequently hybridizes with other species, or an incipient biological species that is only partially isolated, is obviously fuzzy. And this is apparently not an exception, but rather the rule, especially in lineages with large and widespread populations.

As in any pedigree, it is not a question of distinction, but of proximity. This is also true for asexual evolutionary species.

After all, any pair of DNA molecules on the planet has an mrca.

I won't participate in further discussion, as I will travel soon and I am extremely busy. However, I really dislike when colleagues argue in favor of their preferred species concept using as a justification that such concept is better for conservation. What an indefensible argument! "taxonomy informs conservation, not the other way around". You cannot try to convince others about the presumed superiority of a species concept because you claim it to be more convenient for conservation efforts. The discussion of what is a species and how to define it is independent of whatever consequences such concept and methods might have.

Eliecer, I am unsure of where did you pick up the argument you mentioned. No one wrote that conservation informs taxonomy. On the other hand, essentially arbitrary taxonomic decisions can indisputably cause catastrophic conservation decisions.

I just wrote that we need to describe the ecological and biological diversity of the planet. The fact that we need to do this while we are losing it cannot be ignored. I did not write that we can "invent" this diversity, or pretend to see diversity where there is none, just because we want to conserve some biological matter, fidgeting with the available arbitrariness. Indeed, these forms of diversity can be objectively measured, and such measurements must be analysed, and peer-reviewed.

But the very borders of the categories remain fuzzy. This fuzziness appears to derive from the psychological bias of categorizing continuous physical phenomena (at least at the biological spatiotemporal scales). By definition, categories have exact, conceptual borders. Resultantly, the very definition of categories when describing taxa, ecotypes, ecosystems, populations, is somewhat arbitrary. It is this arbitrariness we are discussing about. We cannot hide in front of the responsibility of a choice, in the face of doubt. Do we want to risk and lose the perception of the diversity we've measured, or rather record and describe it as a sound taxonomic hypothesis?

I am not advocating the fashion of subspecies, or the resurgence of a new wave of taxonomic splitting. Descriptions can include varieties and be progressively enriched by contributions from many disciplines (physiology, anatomy, ecology, cytology, molecular phylogeography, etc.).

Further, one thing is the theoretical definition of the categories (e.g., the evolutionary species concept), one thing is its applicability. I think I should describe and name an organism I have in my hands that no one saw before, based on the information and data I have (e.g. morphology), rather than waiting to have all the molecular, phylogeographic, ecological, and life history information, in order to generate a better hypothesis matching this or that (more or less volatile) theoretical definition. And this is literally a question of life and death, if the organism in question is at risk of extinction.

I may add that this dilemma reminds me of a well-known philosophical principle: "pluralitas non est ponenda sine necessitate" (one should not propose more explanations without need), the Occam's razor, or principle of parsimony. Contrary to its popularised version, this does not mean that "the simplest hypothesis is more likely to be the correct one (i.e. not falsified)".

Rather, the Occam's razor is an economic principle. The key concept is the "necessitas", the need. We do not have infinite time and resources to test (falsify) all the hypotheses we can create. Therefore, we start with the simplest ones, that is the hypotheses requiring the least amount of resources. If we successfully falsify them, we can proceed with more expensive hypotheses (e.g. those in need of further ancillary hypotheses).

In other words, the scientific process is biased by our needs.

We are losing this diversity. We do not have the time to describe it, before it disappears. Therefore, we may strategically choose to start falsifying preliminary, simple taxonomic hypotheses (e.g. based on morphological data or DNA barcoding, that I also don't like very much), taking decisions based on the data we have. Further investigations (if they ever occur), will enable our successors to ameliorate the falsification process.

Because it is easier to study them that way. When Linnaeus first propose the Systema Naturae to standardize species names, I don't think he imagine there will be this kind of fuss. My zoology professor always say, "species is a hypothesis". It is not eternal. With the amount of information, some species are found as different, some are the same. He propose to do an integrative species concept; the culmination of all that we know.

But what are the things that define a species? They are indeed arbitrary, dependent on the expert currently active on the field. In terms of conservation, I think it always depends on what do you want to conserve. Evolutionary units? Ecological role? Genetic diversity? And the directive always come from they who has the most influence on opinion. In this context, IUCN. The kind of concept that is useful for conservation thus furthermore dependent on the aim of the specific conservation practice, I suppose. If you want to focus on ecosystem level, perhaps the ecological species concept. Evolutionary level, the genetic concept, etc.

I think the same struggle also happens with tigers, bears, and many charismatic mammals in the world. But that is how you do science, isn't it? We publish, we get comments, we discuss, we improve.

As long as there is no clear and objective way to define species we will keep on arguing. Why not use a criterion that is testable, so that we can debate the data rather than the resulting taxonomy. Why do some people protest against 4 species of giraffe or 13 species of klipspringer? What is the drama there? What rule apart from our own preferences and biases exist that says that a certain species group should only have a certain number of species? I find it strange that as scientists we cannot agree that that the best way to describe species is through testable hypotheses. Wouldn't that make taxonomy a proper science? And sometimes we have lots of data on lots of characteristics, from morphology, to genomes, to behaviour, while at other times we only have one distinct skull or tooth. The amount of information shouldn't matter, but only the statistical criteria through which we measure differences. And then we should accept that if one day we find a second tooth that dismisses our original species hypothesis, then we change the taxonomy again.

I will add, it is also hard when you see it with your specialist eye, with the organisms you know, understand how different the positions are of people working on different groups. That may be related to the state of knowledge of the group itself, so try to evaluate other groups with your may be little bit unfair to say the less.

I also have the problem to define a species, as I also found fuzziness in the borders to define what are species for many of the insects from mesoamerica. The demographic and evolutionary history of each group even when they are very close is different and therefore the form in which the evolutionary unit is formed is different as well. Of course, If we decide that is imposible to give a definition of species then we do not do anything. But we need to move foward and understand all the evolutive processes in our target group, therefore we need a "concept of species" that of course is human made. But we need to beggin from something. Therefore the experts in each group need to define they own species concept for the group that they work in with the knowledge of the history of this group. When i teach to students i love to give the example that the real world is much complex and we humans need to give "concepts" and I think Lewis Carroll do it very well in Alice Through the Looking-Glass.

“When I use a word,” Humpty Dumpty said in a rather scornful tone, “it means just what I choose it to mean – neither more nor less”

“The question is, “ said alice, “wheteher you can make words mean so many different things”

“the question is,” said Humpty Dumpty, “which is to be master, that’s all”

In conclussion. We still a lot for understanding the real biological and evolutionary world but we need some base to beggin and that is for example the species concept that each expert could give to the group that they work.

With all the interesting points in these replies, it seems (apologies if I misunderstood any answer) we have broad agreement that the species concept is useful for categorizing nature, and these categories tend to be applied with fuzzy boundaries. Maybe the categories will always be somewhat fuzzy. But as Eric suggested, it seems we lack a clear and objective way to define the species concept. Who has attempted that? Perhaps a first step could be to develop a clear and objective 'working definition', understanding that the definition will need to be further refined over time. Eric, can you propose a working definition to start with?

I definitely agree that the lack of a standardized idea of what a species really is severely hampers progress in biology. It makes especially arbitrary the studies that accumulate numbers of species from other studies of different taxa, e.g., the estimates of biodiversity and extinctions.

That said, I cannot agree with the view that species are inherently just an artificial human concept and nothing more. No matter what species concept we take, a population at some point in time can satisfy its requirements. From this perspective, species defined by different species concepts are by all means real entities as they are the populations that satisfy these concepts. Some concepts may be more broadly applicable to the populations at particular time moment, some are less, but that also doesn't make them just useful tools for cutting "nature continuum".

Species described by taxonomists are hypotheses, as already mentioned. Actually, each taxonomic species represents at least two hypotheses. First, that all specimens studied belong to a single metapopulation. While we can never be certain about the reality of this very metapopulation, we assume that populations and metapopulations in general do exist. Second, that this hypothetical metapopulation satisfies the requirements of one or more species concepts. Likewise, this can never be proved completely, but that doesn't mean that, in general, metapopulations that have properties of morphological, biological or phylogenetical species do not exist. These hypotheses are not usually articulated in papers, and that sometimes may cause misunderstanding.

Returning to the problem of standarts, I believe that we need a new code that will explicitly make taxonomy an evolutionary discipline and will give an obligatory definition to the term "species". While multiple classes of metapopulations must be preserved, only one should be called "species" to make comparison between different clades meaningful; otherwise we will always be comparing apples to oranges. Of all species concepts present today, I think the biological one is still the best from the point of understanding the evolutionary mechanisms and also from the point of decisive criteria which is much more difficult to imagine in other concepts. Since there aren't any reasons to doubt the reality of biological species in general, the difficulty of falsifying a hypothesis that some particular supposed metapopulation is a biological species should not be a reason to reject it.

For other types of species new terminology must be invented. Given that today structural units within metapopulations also have a big significance from the points of both conservation and understanding of evolutionary mechanisms, we also need a nomenclature for naming such units (note that "subspecies" also doesn't currently have a proper definition). As for asexual "populations", a genus is more suitable here that a species. Genera have also much more distinct boundaries than species. Therefore, a new code must also give a definition for genus.

Naming things is important so we can talk about a subject and have a hope of being understood. Also because in many cases the biological or other species concepts provide a relatively unambiguous insight into biological patterns that we care about.

Species are not "man-made concepts to help us categorize nature's diversity". Genera, families, orders, and classes are such man-made categories. But species are real. Of course, there are many species concepts out there. But the one that makes most sense, in my opinion, is to determine, if individuals and populations are able to reproduce and therefore share the same gene pool. Of course, this is difficult or sometimes impossible to determine. I study deep-sea polychaetes and breeding experiments are impossible in this field. That is why we use morphological traits or genetic markers as a proxy. But this does not mean that this species concept is not true. If the 18 species of hogs you mention share the same gene pool (determine with breeding experiments?), then they are not distinct species.

Ernst Mayr's concept of biological species is of a reproductive isolate "in natural conditions". For instance, if there are two species of flies, one nocturnal and the other diurnal, they may never reproduce "in nature" just because they never meet, but could readily interbreed and produce fertile offspring in the lab. That is, the concept does not necessarily imply the presence of post-zigotic barriers.

The biological species concept is fuzzy, since the concept of "nature" is fuzzy. In fact, the very concept of "nature" has no scientific meaning. Extracting man and everything man does, or the effects of man's action from the reality, and naming the rest "nature", is not only arbitrary, but practically impossible.

Further, the concept is fuzzy because "nature" (henceforth, reality) does constantly change in space and time. The two fly biological species may eventually meet, if not in a lab, making the two "biological species" instantly disappear into thin air: secondary contact, introgression and hybridization may reinforce or not the difference between genomes that have been separated for a certain evolutionary time. Another example I made, is the "circle of species". There is a continuous belt of populations of a species of gull (Laridae) in the northern hemisphere: there is gene flow between pairs of adjacent populations, but this flow decreases with the geographic distance between the pairs. Between the pair at the geographic extremes of the belt, gene flow is maintained only indirectly, through the other populations. Just cut the belt (e.g. extirpation of some of the populations in between) et voila': two new "biological species" are born, even if initially they likely have very similar genetic structures. Numerous examples are in the literature.

The borders are fuzzy. There is also plenty of incipient species out there, when ecological factors reduce the gene flow (e.g. the genetic and ecological polymorphs of the gastropod snail Littorina saxatilis). At the same time, there are examples of species that do not intebreed because they are separated by vast expanses of space (and sometimes, even of time), but are perfectly interfertile (I am thinking at two species of crickets). Another example of fuzzy borders is the discrimination of chronospecies in palaeontological series. Species are fuzzy entities in time and space.

This does not imply that the species concept has no scientific meaning, or that species aren't empirically real or measurable phenomena. Indeed, I think the evolutionary trademark of biological species is the coadaptation of the genetic variants (e.g. alleles in the nuclear genome), made possible by meiotic recombination. It is the possibility to continuously expose to selection different combinations of alleles in the diploid genome into different individuals, generation after generation, that makes possible for biological species to evolve as units, forming "gene pools" of genetic variants, that move though time in the adaptive landscape (fitness allelic contributions) in a coordinated way, like "gene rivers", as Dawkins depicted.

For asexual species, we do have a satisfactory definition, which is also valid for sexual species. It is the evolutionary species. In asexual organisms, evolutionary species are lineages of mothers and daughters, that do not form "gene pools and rivers", but evolve independently in the adaptive landscape. Also here, borders are fuzzy and inter-lineage gene flow can occur, through several different mechanisms (e.g. conjugation in bacteria).

Dear Doctors, OK for your discussions about orang-outan species. I am botanist and entomologist and to-day I have not access to modern genetic technics and computer gene-database, in the French laboratory where I was working during more than 50 years (because I am retired). I observe that species concept in the open field remains, although genetic help. I just finished to read the following paper: Sreedevi K., Meshram N. and Shashank P.R. 2015. In Chakravarthy ed. "New horizons in insect science. Towards sustainable pest management". Springer Editor, India. I conclude that for Vertebrate as for Invertebrate, species definition is the same. Don't remember Homo sapiens shares about 15% of its genes with Homo neanderthalensis! Without barcode, in the field I past year find one new species of Drosophilidae that I plane to describe based on genitalia and biological traits. Sincerely yours. André Panis [email protected]

I have read the discussion with interest and am happy to see that so many scientists have similar ideas. There are several species concepts that try to encompass what a species really is. There is no consensus. I can only see one logical conclusion: A species, just like every other systematic category, is a concept of our mind. We create it, because we cannot operate without names. In sexually reproducing species, populations have offspring with each other and to a lesser degree with other populations. Novel characters may arise and become fixed in the population; they may or may not move to the rest of the species. If separated from other populations (that may also occur sympatrically), hybridization may not be possibly any more, once the populations get together again. Or it may. There is no rule. If hybrids are disadvantaged, selection against hybrids may occur. Overall, after a few million years there is probably no more hybridization possible. None of that applies to bdelloid rotifers or any other parthenogenetic species. The argument that a population follows a particular species definition and therefore species are real makes no sense. The BSC can be applied to all sexually reproducing species. So can the phylogenetic species concept; or the evolutionary; or the morphological. It is logically not possible to make that a test of whether species actually exist. And yes, as it has been said, the "edges are fuzzy". We like to cheat, when it turns out that some of our preferred species do reproduce with each other. Again, to make it clear, the very fact that different species concepts exist and that there is no unified species definition (and really, cannot be) is clear evidence that species as such do not exist. There is no clear boundary - we are the ones who define the boundaries. We define boundaries, because we need boundaries, not because they are actually there. If we call both a bdelloid rotifer and a sexually reproducing organism a species, then those two cannot possibly be the same. The species definition in this case can only be a human construct and cannot be "real". To be fair, in many cases entities who look alike do not interbreed with other entities, and we call those species (I published several of those myself), but there is no selection to become a species. It is all time and chance. Take H. sapiens: We are the direct African descendants of H. erectus. When did our ancestors turn into another species? It is really irrelevant, if you simply consider certain genotypes spreading throughout Africa and replacing other genotypes, winding up with a much larger brain capacity and some other adaptations. Here quite clearly the species concept obfuscates what happened in evolutionary terms. But it gets better: H. erectus spread out of Africa several times. The newly emerged H. sapiens then interbred with the then-current Eurasian line of H. erectus, H. neanderthalensis. Further East the newly-melted hybrid did the same with the Denisovian line of H. erectus. What then is a species in modern anthropoids? It is a line in the sand that we draw, a necessary convention. Species never interbreed, that is always done by individuals. OK, enough written for now.

Great feedback Christoph. This is all fun. I guess, we would be a bit arrogant if we thought that we could capture the incredible complexity of evolutionary processes in simple and discrete boxes. But then again, taxonomy would become rather boring if all was clear and we had nothing left to debate :)

I think this is a good one:

Pante E., Puillandre N., Viricel A., Arnaud-Haond S., Aurelle D., Castelin M., Chenuil A., Destombe C., Forcioli D., Valero M., Viard F., Samadi S. (2015) Species are hypotheses: avoid connectivity assessments based on pillars of sand. Molecular Ecology, 24, 525-544

Seems like our MHC genes have a hierarchical taxonomic approach for dealing with pathogens :-)

https://phys.org/news/2017-11-fish-insight-evolution-immune.html

Dear Christof, let us disagree. The core of the problem is just the word "species" (without adjectives), which is now used simply as a token whenever we find any gaps between hypothesized populations. Since currently this word doesn't have a definition, it's not even a concept, just a token. On the contrary, species concepts denote very particular objects with defined properties. If we have every reason to believe that a particular population has the properties, defined by e.g. BSC, then it logically follows that biological species do exist, as exemplified by at least this very population. Of course, this remains a hypothesis, which is not problematic at all. BSC cannot be "applied" to all sexually reproducing species [sic]; only those populations that have the properties defined by it may be biological species. Otherwise, they are not biological species, but could be evolutionary or phylogenetic species or maybe they don't have the properties defined by any species concepts. What we need is to give the word "species" meaning by equating it with one of existing concepts. Many other current taxonomic "species" that do not satisfy the concept chosen for "species" definition won't be species anymore; we will have to invent other terms for such objects; this is not a problem, but a progress.

Of course, when you look at lineages along the timeline boundaries become especially fuzzy. This actually warrants not the abandonement of the concept of species as real objects, but a more accurate approach to words and names. For instance, from the point of BSC it just follows that none of the descendant populations of African H. erectus had actual reproductive isolation from each other, therefore we're all H. erectus and "H. sapiens" is simply a token without clear definition at this point. Of course, given the status of the word "species" today to think of us as H. erectus is "impractical" and will "never be accepted". But that is why we need granularity and formal names for intrapopulation units as well as acknowledgement that although they are not biological species (but may be other kinds of species), they are extremely important nonetheless. So, the core issue here is the usage of the word as a token and its status, not the reality of objects behind it.

Also a good point Evgeny, it is indeed part of accurate word use. If defined by the PSC, it becomes a phylogenetic species, if by the BSC then its a biological species. It does take us back to the initial question as to whether we can agree on what is a species and how do we define it? Taxonomy wouldn't be helped much if with each taxonomic decision we would have to clarify whether it was any of the following types of species (as recently listed by Groves et al.): Biological species concept (BSC), Cohesion species concept (CSC), Differential species concept (DFC), Evolutionary significant unit (ESU), Evolutionary species concept (ESC), Genetic species concept (GSC), Phylogenetic species concept diagnosable version (PSC1), Phylogenetic species concept monophyly version (PSC2), Phylogenetic species concept diagnosable/monophyly version (PSC3), or Recognition species concept (RSC). In fact, that is what we are doing now, but clearly that is just leading to more disagreement.

By the way, if H. erectus hybridized with sapiens, then we would all be H. sapiens not H. erectus as the former is the older name. More interestingly though, would H. floresiensis, which likely diverged from H. erectus and may well have hybridized, and H. habilis which was ancestral to H. erectus, and there may have been geneflow then also become H. sapiens? The point is that unless some sudden barrier came up between two populations of a species and split them into two for good, so that there was never and more gene flow, there has likely almost in all cases been some level of gene flow between two diverging taxa. When can we safely call them species then? Here are some wise words in this regard: Article Hybridization as an invasion of the genome

"Hybridization between species is commonplace in plants, but is often seen as unnatural and unusual in animals. Here, I survey studies of natural interspecific hybridization in plants and a variety of animals. At least 25% of plant species and 10% of animal species, mostly the youngest species, are involved in hybridization and potential introgression with other species. Species in nature are often incompletely isolated for millions of years after their formation. Therefore, much evolution of eventual reproductive isolation can occur while nascent species are in gene-flow contact, in sympatry or parapatry, long after divergence begins. Although the relative importance of geographic isolation and gene flow in the origin of species is still unknown, many key processes involved in speciation, such as ‘reinforcement’ of post-mating isolation by the evolution of assortative mating, will have ample opportunity to occur in the presence of continuing gene flow."

I particularly enjoyed the last two posts :-)

You are essentially debunking the BSC. I totally agree with you and with many posts: gene flow is not everything.

Reproductive isolation is only one of very many criteria which signal an independently evolving entity. Another would be the presence of a reliable autapomorphy (even if in the face of gene flow).

I like the proposal by Mayden (1997, already two decades ago!): The Evolutionary Species Concept (originally Simpson's, refined by Wiley) is the final philosophical solution to the "species problem". However, it is not at all operational, so one must apply any of the available operational concepts: if you demonstrate reproductive isolation, good. If you have other kind of evidence for separate evolution, also good.

The mere fact that a candidate species does not fit a given (operational) species concept does not refute its status, if this is is supported under other concepts.

As for the post questioning the scientific status of Taxonomy: please note that saying "this is a species" is always a hypothesis, which can be tested, as everything in science. I fail to understand why the taxonomical endeavour would lack "scientific status".

Hi Juan. Thanks for the input. I guess the problem with the scientific credibility or status of taxonomy is that good taxonomists should indeed know that all their taxonomic decisions are mere hypotheses assessed on the basis of currently available information. As such, good taxonomists should also be quite happy to see their earlier taxonomic hypotheses dismissed when new data become available. Many others however like to consider species as true and permanent, and taxonomy a fixed labelling exercise that shouldn't undergo constant changes. I guess it is up to taxonomists to explain their science better, for example, by being upfront in their publications and media work about the hypothetical status of their species. Once we reform ourselves and formally treat taxonomy as a science that works with falsifiable hypotheses and testable predictions, as Doug Sheil mentions above, we may find it easier to convey that hypothetical nature of our work.

Every scientific statement is ultimately "nothing but" a hypothesis. Even theories are nothing but hypotheses upgraded by the test of time, through numerous rounds of unsuccessful falsifications. But who can tell "how many experiments" are needed, in order to promote a hypothesis to a theory? It's mostly a question of social perception and politics. Yet the history of science teaches that eventually, many robust theories eventually were falsified. As Bachelard wrote: "the history of science is a precious graveyard of false theories". Science is the only method of inquiry to declare itself as intrinsically fallible, as Peirce made clear. This is why it can evolve. This is what we have learned through centuries, leaving behind the black holes of authority and the ipse dixit. This is its greatness.

I am not a taxonomist, but I sometimes find taxonomy useful. A little. More often, taxonomy appears to be an endless argument between lumpers and splitters. In past instances, this argument has been amusing, but other times dangerous. As Sheil and Padmanaba (2011) aptly illustrate, pragmatism is an essential component of invasive species management. Next to prevention, early detection and rapid response (EDRR) is the most cost-effective strategy for addressing species invasions. The "rapid" part of EDRR is a narrow window that seldom, if ever, allows for publication of extensive research, prior to a decision on controlling an invasive species (no scare quotes needed).

Likely the worst example of research leading to inaction is the case of Mikania micrantha in Florida. This species, one of the "world's worst weeds," was discovered in south Florida several years ago. For some reason (the details are murky), a decision was made to conduct research on the plant, rather than target it for immediate eradication--which at that moment was likely to have been successful. Ironically, I had just returned from Nepal, where I had witnessed the devastation caused by this species in Chitwan National Park and other conservation areas. Long story short: we are not going to get rid of this plant. The best we can do is contain it.

In the context of taxonomy versus management, it does not matter if M. micrantha is one or four species; if there are none or a dozen subspecies, varieties, or haplotypes; or, what is the specific genetic origin of the infestation(s). If any or all of that information had been available, it would not have affected the decision to (belatedly) control this species.

I find "species" to be a useful label for discussion, but my funding bias is toward operational control, not to perpetuate a ponderous, insoluble, academic exercise.

Species are units that actually exist in nature, so it is not, in any case, artificial subdivisions, invented by taxonomists. Indeed, the species is a very old notion that primitive peoples, hunters and fishermen having to exchange precise information on the prey to be captured or the plants to collect, already knew. What does the word species mean? Biology recognizes two main meanings, which can be called morphological (or phoenetic) and reproductive concepts of species. According to the morphological concept, species are defined by their form: individuals, in nature, are distribute into categories only divided by their phenotypes, the criterion of belonging to a species is then a sufficient morphological resemblance with other members of the species.

According to reproductive concept, species are communities of interbreeding organisms; an individual belongs to the species whose members can reproduce with him.

With triumph of evolution notion and genetic progresses, morphological criterion, which for a very long time was the most obvious, gave way to the criterion of reproduction . Indeed, morphologically identical animal populations are known which are intersterile. MAYR (1942) defined these populations as "twin species"

.

Currently, the most widely accepted biological definition of the species is that of ornithologist Ernst Mayr, who presents species as "groups of natural truly or potentially interbreeding populations, isolated from other comparable groups with which they do not reproduce ".

This statement takes into account structure in space (populations group) and natural delimitation criterion (intraspecific interfecundity and non interspecific interbreeding).

The way I see it is that the species concept is a useful framework for understanding biodiversity and evolution, just like the atomic models in physics. For me, whether species really exists or not (and if it exists, what are the boundaries - still no consensus) is essentially philosophical. So yes, I think taxonomic work is cool, but also I think the absence of any concrete set of boundaries for species delimitation is a also a problem.

Thanks Dr. Khang. I guess that at least those working with atoms have a clear shared understanding of what an atom is, based on the number of protons, and how one atom can be reliably differentiated from another, even though they may not be able to accurately predict the atom's behaviour. Such clear knowledge of identity is a luxury we don't have in taxonomy because too many are unwilling to define species based on criteria that can be tested, i.e., they are falsifiable in a Popperian sense. We can define species boundaries quite easily in a way that allows anyone to repeat and verify or dismiss our hypotheses, but there is significant resistance to the development of such objective species criteria, because of the associated fear for taxonomic inflation.

The species concept is an abstract idea that can be raised by a researcher through a work hypothesis that can be validated, corroborated or invalidated by another researcher based on morphological, genetic and biological evidence.

A recent publication in Systematic Biology may give a good summary of all perspectives considered so far about the species concept.

Article Biodiversity and the Species Concept - Lineages are not Enough

Defining species in a transparent way is the most important thing for scientific progress, I think, and this was certainly the case in the new species of Pongo. However, I would like to ask the authors of the new orangutan species (or any mathematician or statistician here) what are the implications of diagnosing a new species based on a N=1. What inference can we make about this new species' populations and their diagnostic characters?

Hi Guilherme!

I think historically many species were defined on the basis of N=1, since the taxonomists of the early days essentially collected huge numbers of samples in expeditions, and it would have been a luxury to collect multiple specimens of the same species. This seems to be the source of many "splittings" in the taxonomy literature, and hence subsequent efforts to correct such "mistakenly erected species".

Having said that, standards have of course changed now. I notice that experts in different taxonomic groups have varying standards on the degree of statistical evidence needed to justify the naming of a new species, and these standards appear to be inversely proportional to the size of the organism of interest (laughs - large size animals also have lower population size, and are often endangered)! You mentioned orang utan apes, which is a large-sized animal, so I would not be surprised if N=1 is accepted by experts in that field. Of course statistically one would be very dissatisfied with such a number, but I guess they weighted many other evidence to arrive at a new species declaration, not forgetting that the status of the species is open to challenge.

I have worked with some taxonomists studying very small fish parasites (in micrometers, and in abundance), and from my observation, taxonomists in that field would never accept N=1 (or any low N) if one is trying to describe a new species. Typically something like 20 or more is considered good because the experts would need to assess morphological variation captured in a principal component analysis plot, and small N gives problem to the ordination method due to requirement that N must be larger than the number of variables recorded (quite a few are used).

True, TF. And this consideration applies also to other biological sciences. Ecological works based on the behaviour of a single ant would be probably unacceptable, but the same may not be true for a dugong, or a blue whale. Scarcity and population numbers obviously play a role. Think about human paleontology.

I'd add a few impressions about the history of species. The taxonomic boom occurred in the colonialist era, when scientists got more regular access to tropical biodiversity, in a period when most had a rather deterministic (creationist) species concept. As a consequence, many described species based on personal observations of a very small number of specimens (not always providing objective measurements), and did not care very much about intraspecific variation. This apparently had two consequences: on one hand, the splitting tendency that has been described; on the other, rather accurate, though subjective, observations of people who tried to make sense of "the god's design". Then Darwin came. Rapidly, the distinction between variants and species became blurred. When the importance of polymorphism and variation was acknowledged, species became entities with fuzzy borders, and taxonomists started to sample more specimens per species to describe intraspecific variation, relying less on subjective differences based on observations. Sampling became a logistical and statistical problem, and the general trend switched to lumping, as testified by many taxonomic revisions of the 20th century. However, as more objective measurements and statistical techniques were added to descriptions, and even as genes and molecular techniques stepped in, many realised that several species based on the subjective observations of the 19th and early 20th centuries' scientists (early naturalists that normally had a lot of time to publish and had a lot of hands-on experience) actually corresponded to phylogenetic clades, and a new wave of splitting occurred. I think this latter wave is what taxonomists fear today, as Erik noted. In the past the splitting trend determined a state of flux caused by the ever-changing concept of species, and soon became clear that a "natural classification" might not be the one and only scientific priority, as Ruark reminded.

Hi Tsung Khanga and Gianluca Polgar!

Thanks for your input. I think it is pretty clear now in the 21st century that species need to be defined based on observations from populations. This said, we need to take into account variations and their range. Of course one may describe a new species based on a small N, science is free and should remain so. But the authors (in my opinion) should bear in mind the limitations of a small N. I would never diagnose a new taxon saying that its skull is 56.5 cm long (N=1), while its sister species has a skull of 61.2 cm, because in my mind I know that species are populations, and I think this approach is dangerously close to typological thinking. Remember that typlogy is what led to a 19th century splitting frenzy in medium/large mammal taxonomy.

Turning back to the original question [“why do many of us still think that species are real?”], the answer is very simple: because species are real! I am deeply astonished that in XXI century it might still be a question, although explained long ago by e.g. Simpson, Mayr, and innumerable others. The persistent doubts are evidently connected with another astonishingly inveterate misunderstanding: lack of clear distinction (and, consequently, application of misleading terminology) between general concept of species [what do biologists have in mind speaking of “species”] and practical (“operational”) criteria [allowing in a concrete situation to decide whether a particular unit is or is not a species]. This distinction (analogical to that – also frequently misunderstood – between the definition of a taxon and a key character used to identify the examined specimen), again, has long since been clarified by e.g. Simpson (1951, 1961), myself (Hołyński 1977 in Polish, 1992 in English), de Queiroz (1998), and again by me (Hołyński 2005, with extensive explanation of most frequent misinterpretations). “All modern species definitions either explicitly or implicitly equate species with segments of population level evolutionary lineages” (de Queiroz 1998), and so, as such lineages unquestionably do really exist, species are real.

Another question is, that in many situations it may not be easy to correctly recognize the “status” (species? subspecies? infrasubspecific variety?) of particular [group of] population[-s], and therefore in practice we must iteratively apply, as successive approximations, “three complementary definitions of species:

(1) evolutionary concept [a modification of Simpson’s (1951, 1961) evolutionary definition]: 'species are segments of phyletic lineages (ancestral-descendant sequences of populations), whose gene pools have differentiated beyond limits of reversibility'; this is the most general formulation, closest to the basic tenets and applicable in principle to all groups of organisms, but transgression of the limits of reversibility is very seldom directly demonstrable in actual cases, so usually we must try the second-choice

(2) reproductive criterion [an adaptation of Mayr’s (1940) 'biological' definition]: 'species are groups of interbreeding populations, reproductively isolated by intrinsic mechanisms from other such groups'; among allopatric, allochronic, parthenogenetic, &c. populations interbreeding obviously does not occur, and the proof or disproof of intrinsic reproductive isolation is either [almost] never possible or irrelevant, so this definition can be actually applied to co-existing bisexual forms only; however, even in overwhelming majority of such instances we have no data on the reproductive isolation as such – this can only be inferred from the observed phenotypical (usually morphological) hiatus, what in fact means recurrence to the only universally serviceable

(3) morphoevolutionary diagnosis [a re-formulation of Hołyński’s (1977, 1992) 'operational' definition]: 'species are discrete groups of populations, showing – at least in one class (sex, caste, developmental stage) of individuals – consistent unique combinations of morphological characters'.

For less differentiated populations Amadon’s (1949) rule should be applied: if more than 75% of specimens are determinable, we have to do with a subspecies; if less, the form at issue does not warrant taxonomical recognition at all” (Hołyński 2005); the currently most popular 'PSC' (Phylogenetic Species Concept – Cracraft 1983) is essentially identical to the (3) above.

Two minor – but also frequently misapprehended – confusing factors should be mentioned: the “fuzziness” of taxon borders, and the role of molecular criteria. The former has been discussed in my polemics with Podani (2009; 2010a, b) – Hołyński (2011) [by the very fact of evolution “the boundaries are ‘fuzzy’ because the true relationships between natural taxa are fuzzy” “nothing would have pleased creationists more than the discovery that ‘crisp boundaries’ separate all, or the majority of, taxa: this would be the best imaginable proof that evolution does not occur!”, “the boundaries between colours, languages, oceans, biogeographical regions, between fluids and gases, planets and stars, &c., &c., &c., are all ‘fuzzy’ (even if artificially represented as sharp) because such is the ‘nature of the Nature’ that natural boundaries are rarely ‘crisp’!”]. In the case of species, “fuzziness” of course occurs at the origin of new species (when it is just being separated from its “sister” or “mother”), so the boundary between them is then naturally not sharp, but when the process of speciation has been completed, the continuum is broken (in principle, forever). As to the DNA sequences, they can sometimes supplement the phenotypic evidence (or, when the latter is not available, serve as preliminary “proxies”), but themselves cannot be decisive in resolving taxonomical problems (see detailed discussion in Hołyński 2010).

All my publications quoted above:

1977. Taksonomia form allopatrycznych [Taxonomy of allopatric forms]. Biul. Inf. PTE 20: 64-67

1992. Allopatry and the definition of species. Proc. IV Eur. Cgr. Ent. [1991] 1: 399-403

2005. Philosophy of science from a taxonomist’s perspective. Genus 16, 4: 469-502

2010. Taxonomy and the mediocrity of DNA barcoding: some remarks on Packer et al. 2009: DNA barcoding and the mediocrity of morphology. Arthr. Syst. Phyl. 68, 1: 143-150

2011. Philosophy, evolution, and taxonomy, or: what biological classification is for? (practicizing biologist’s comments on some recent papers by Podani). Munis Ent. Zool. 6, 2: 525-534

can be found and downloaded from RG, and in the references you may find the details of others.

Thank you a lot Pr. Roman Bohdan Hołyński for your relevant and interesting answer to the question asked.

I totally agree with everything you have written.

Best regards

Excellent input Dr. Hołyński ! Thank you. And indeed, creationists would have loved sharp boundaries.

I also agree with Roman Hołyński. Species are real.

Fuzziness is not a criterion to distinguish real things or processes from bona fide artificial entities. The original question falls into the so-called sorites paradox, which is a fallacious argument (also known as continuum fallacy).

The proposed "solution" to define species thanks to an a priori criterion doesn't make the concept "testable". This way of thinking leads to circumscribe artificial entities that make no sense. The realist approach to the species phenomenon is to use pluralistic and flexible criteria because they are emergent systems that don't fit into a monothetic category (the species category is polythetic).

Trying to simplify the species concept into a monothetic category is unscientific because it is like trying to fit the square peg of an ideal into the round hole of a complex reality. For example, the so-called phylospecies concept do so by assuming axiomatically that after a cladogenesis neither of the two daughter lineages can be recognized as the same species as the mother lineage, which is blatantly wrong. That's why I hold that nominalism (sometimes called "scientific" nominalism) is a dangerous fashionable philosophy because it tendentially leads to idealism, and thus to non-science.

I don't understand why scientific realism is so poorly known by scientists.

As I said, the fact the entities are fuzzy does not imply they are not real. The same is true for all measurable phenomena, from atoms to coastlines. And I strongly agree with Damien's statements about nominalism and realism. I am not a professional philosopher of science, but this distinction sounds like one of the pillars of scientific reasoning.

Once Darwin realized the fluidity of species he also became a nominalistic and left it to the specialist to decide whether to call his/her taxon a species or another unit. This is the way we are all are going except those who wish to stick rigidly to the biological species concept which, as many readers have already indicated, is impractical for palaeontological samples, most marine invertebrates, museum specimens and asexually reproducing species. While a scientific methodology to determine whether a taxon represents a species or not would be useful, it seems unlikely that one will emerge anytime soon. One will have to marry all the different species concepts to arrive at a scientific and practical definition and whether we can achieve this looks beyond the scope of current day taxonomists or systematists. While gene sequencing and phylogenetic analysis is useful it has its limitations and cannot stand alone. I work on sea cucumbers where I look mostly at museum specimens for which my descriptions of new species, while based mostly on the morphological species concept, I also look at biogeography. In the latter case slightly differing allopatric polulations I refer to as subspecies rather than a variety but in sympatric populations I regard the differences as indicating a forma/variety, thus attempting to make my determinations and/or descriptions as objective as possible. I have described several n=1 new species because of gross morphological differences from congenerics. I rather do so not to own the species but to indicate that such variants are distinct enough to be termed species and that more specimens will definitely justify my action. The problem here is to define slight or gross differences which, in the absence of a scientific method, are only subjectively arrived at. These do require scientific definitions and this is what the arguments are all about.

I am in league with Prof Guanluca in thought.

I was offline a few weeks and now see the richness of views in the diverging opinions and arguments. I feel provoked to offer a footnote to my previous comments above to restate a few points. These seem central to the disagreement and have not been answered by those asserting that people "think that species are real because species are real" (e.g., best presented by Roman Bohdan Hołyński above).

1) What we are debating:

No one disputes that the patterns of variation that are addressed by taxonomy are real. That is not the dispute. The question is whether the man-made designations that we call "species" are "real". So quoting de Queiroz (1998) to highlight that most modern species concepts are attempts to match categories to lineages proves little about the reality of these entities (and much about intention). We can accept the lineages and still reject the reality of the hard edged categories and the labels.

2) Reality and science:

Ultimately nothing in science is certain outside of mathematical theorems. Science is an approach--a quest to investigate and best understand reality but never to have certainty what is real (all answers are provisional and all can be updated as information improves). In some areas the weight of evidence is sufficient that we might concede the point and accept something as "reality", e.g., that gravity is an attractive force, or that a tiger is phylogenetically closer to a leopard than to a mosquito--but formally such arguments remain as hypotheses and calling them "reality" is just a convenience for things that appear sufficiently probable given our knowledge (not a separate category of scientific knowledge).

3) Conflicting views:

Most taxonomists have the humility to acknowledge that their species designations are hypotheses (and thus may be false). I would argue that something that "may be false" cannot also be "real" in the way that most of us use that term. (Even if we suspect that some or many populations appear accurately represented by the hypotheses -- this view may need to be updated based on new information). I do see that not everyone accepts this logic (see, e.g., Hołyński 2005) but I for one am unconvinced that taxonomy is a special case with a deeper insight into reality than other sciences (indeed the opposite seems to be the case).

4) Fuzzy thinking:

Accepting that species are "fuzzy" does not address the problem of how such categories are still "real". The idea of the categories is a human made abstraction--if it doesn’t fit what nature presents us with then the approach is flawed. While this is no surprise if we already accept they are merely conceptual constructions that we are trying to impose on real patterns of variation it does pose a challenge to those who want to claim that "species are real". The question that several of us posed--"how many generations back do we need to go to see where our ancestors were of a different species to us?"--is a genuine question that deserves a meaningful answer.

Happy New Year!

Excellent response Douglas, which sums up pretty much the reasons I posed the question in the first place. Accepting that species aren't real in a biological sense but merely our hypothetical approximations of current reality should force us to be explicit about this hypothetical nature in our taxonomic descriptions. We then also need to accept that a species today may not be a species tomorrow, either because new data or species definitions provide different insights, or because the conditions that drive genetic variation have changed and a different biological entity arises.

This realization is important. Calling something "real" indicates, at least to the outsider, an impossibility of change. But of course change does occur and we need to be alert to how that change affects our taxonomic hypotheses. All this gets a bit philosophical. Many things we consider real are in fact hypothetical and often time-bound categories. A van Gogh painting is one because we assume he painted it, not because we know for sure -- and so on.

What does all this mean to taxonomists? For a start, we should be upfront about the hypothetical nature of our taxonomic conclusions, the species definitions used, and the caveats in our data (i.e., just good science). We should also remind ourselves of the fluidity of evolutionary processes. Life did probably not originate only once and then kept splitting off into species in a neat bifurcation process. But instead there was a much more intricate process involving bifurcation as well as reticulation resulting in new biological entities. Taxonomy is one of the building blocks of understanding this complexity, and being very clear about the limits of this science can only help in increasing our understanding of evolution.

Happy New Year too.

Wouldn't a test for reproductive isolation still result in a hypothetical outcome? Such a test of pre or postzygotic reproductive barriers would still be based on a sample, and based on the outcome of the test we would hypothesize that the observed extent of reproductive isolation in the test population was a good representation of the total wild population of the species. To be certain that full reproductive isolation had occurred, and that the species was thus real according to the Biological Species Concept, we would have to mate every individual of the two species to check for viable offspring. This is was certainty requires after all, and of course it is not realistic. So there will be no certainty.

Also, as the orangutan species are allopatric such hybridization experiments would have to be conducted in a non-natural experimental setting, for which we would not know how well it represented natural reproductive isolation. The BSC revolves around reproductive isolation between species under natural conditions. As recently stated by Groves (DOI 10.1007/s10592-017-0976-07), "interbreeding under human control does not give any indication of species status even under the BSC, especially in allopatric populations."

As to Pongo tapanuliensis, the data indicate that gene flow between species completely stopped 10-20 kya. It remains unclear whether that was because of natural reproductive isolation, or whether it was induced by the arrival on Sumatra of modern humans and their hunting, isolating the two Sumatran species in their respective allopatric ranges.

Finally, I don't agree that "reproductive isolation does not change."What evidence do we have for that? The Eurasian Wild Boar, Sus scrofa, and the endemic Javan Warty Pig, Sus verrucosus, co-existed on the island of Java for at least 70 ka, apparently without any detectable gene flow between the two, and the two retaining their very distinct morphologies. It seems that only in recent decades the two species have started to hybridize when Sus verrucosus was significantly reduced through hunting and habitat loss. If this can be confirmed through further genetic research, it would indicate that reproductive isolation once existed but that because of changed ecological conditions and altered population dynamics this isolation was broken. Reproductive isolation under natural conditions could depend on those conditions.

«Darwin declared vain the researches to which gives rise to the term species. Indeed, this notion escapes any attempt of framing : resemblance, progeny and interfertility are insufficient to describe what is a species, a concept certainly useful, but blurred.

Advances in chemistry and computer science have led the analysis of a "specimen" up to its ultimate degree, now DNA sequences are the true barcodes of species identification. DNA sequences of such specific nucleotides are available that it becomes conceivable to attach to each taxon, whatever its rank, from the phylum to the individual, its own label».

The species: between stability and evolution

Philippe Lherminier

Pour la Science > N°350 - décembre 2006 >

Dear Robert,

I simply reported, for discussion, ideas developed by Philippe Lherminier in an article published in "Pour la Science> N ° 350". And since they are not my own ideas, I putted them, according to my french culture, in bold and in quotation marks.

I hope that now it is clear for you and all those who follow the answers to this question.

Best regards

I share Erik Meijaard's skepticism about the "real" nature of species. Species evolve, so species delimitation is always bound to be a decision that we humans need to agree on.

However, I do not share his assertion that the recent paper by Nater et al. 2017, in which a new species of orang-utan is described, has employed "a scientifically, objective, robust and replicable way to make taxonomy scientific," as he puts it. The genetic analysis performed was well-executed and very useful, but it provides better evidence for population fragmentation and bottlenecks than for biological species.

As Darwin himself argued, "reproductive isolation" (or, as he called it "hybridism") has always been a somewhat poor correlate of what we mean by species. By biological species, we generally mean biologically relevant taxa that can contribute to local biodiversity in sympatric overlap, as well as to spatial genetic diversity. Sympatry is a more objective and conservative criterion than mere absence of current gene flow, or divergence in different geographic regions. The sympatry criterion was also that intended by the promoters of the mid-20th Century "biological species concept". For more discussion see: A species definition for the modern synthesis http://www.ucl.ac.uk/taxome/jim/pap/mallet95tree.pdf

In the case of the new orang-utan the very high estimates given by Nater et al. of "cross-coalescence rates" (i.e. gene flow) of 0.294 and 0.860 across Lake Toba in Sumatra, until a few tens of thousands of years ago, suggest a highly continuous population. The current divergence found in these populations is largely due to recent bottlenecks and a lack of recent movement, as well as a paucity of specimens (only 2!) in the new taxon. There is no indication whatsoever of any genetic incompatibilities that might separate biological entities that might be able to overlap in sympatry. In addition, strong links between Sumatra and Borneo suggest that the orang-utan as a whole might even be considered a single species across both islands.

Unless the goal of conservation is to preserve small, inbred and probably genetically doomed populations, I suggest that conservation needs would be far better met by considering the new so-called taxon to be part of a much larger orang-utan population on Sumatra, albeit one that happens to have had the misfortune to have become very fragmented.

Some twitter skepticism of the new species: https://twitter.com/WTF_R_species/status/926249485000376321

Dear Oum, thank you for clarifying this, I have changed my post.

Best regards, Robert

As I understand it, *all* known orang-utan populations should be protected. Designating this population as a new species has no biological justification, and might distract conservationists into withdrawing funds from another orang-utan population.

The question about the “reality” of species reminds somewhat on the definition of light in physics. Some characteristics of lights are best described as particles whereas others are best described as waves. The advantages to consider species as discrete units, e.g. particles, are employed in taxonomy and many other biological disciplines including ecology and conservation genetics. Unfortunately, the concept of species as real discrete units is incomplete because it does not integrate the continuum issue because species are not Linnean units that are constant. Instead, species are part of a evolutionary process and thus they are not discrete units despite some characteristics are best described as particles constant for some time. Some long living species may display characteristics such as conservation of genotypic and phenotypic features as well as reproductive isolation for extend periods of time but we can not make the assumption that all “species” observed at the current state of evolution display this probably rare conditions. This fact is obvious if we study species in the process of establishing species boundaries. But we may also need to consider situation in which evolutionary units, “species s.l.”, may never have a complete reproductive isolation but still maintain some uniqueness in genotypic and phenotypic features. A concept considering species in an evolutionary continuum similar to the definition of light as wave instead of particles may capture some of these characteristics better but theoretical biologist may still need to improve the concepts required to establish a definition that can be than transformed into application to recognise evolutionary units.

These theoretical issues concerning the definition of species has unfortunately some rather sever consequences for taxonomy and especially conservation biology. I am a bit afraid that these consequences have not been sufficiently explored yet especially in the context of prioritising species in conservation programs. Programs aiming to conserve rare species are arguably frequently confronted with the problem of species definitions because many rare species may be be not clearly distinct evolutionary units. Much more research is arguably needed to explore the consequences of conservation biology.

Without believing that species are real, how can we claim to discover new species, claim that a species is endangered and apply legislation?

Two concepts that try to resolve the “real species”-problem, including any "continuum issues":

• “A species is “entity composed of organisms which maintains its identity from other such entities through time and over space, and which has its own independent evolutionary fate and historical tendencies.” (Evolutionary Species Concept; Richard Mayden)

• “Species are segments of [meta] population-level lineages.” (General Lineage Concept, Kevin de Queiroz)

Operational criteria for delimiting “real species” are “contingent biological properties of species (e.g. internal reproductive isolation, diagnosability, monophyly)” (Naomi 2011).

Thanks James for your input. Based on current information, I am comfortable with Pongo tapanuliensis as a distinct species, and not just because it helps focus conservation efforts on a highly endangered taxon (this is already happening through a new action plan supported at high government level for the specific management of this species). But, as many have said, the goal of taxonomy is not to support conservation but to objectively categorize nature's diversity.

I think that the small Pongo tapanuliensis is a remnant of what was once a much more widely distributed species, possibly including the now extinct orangutans from the southern half of Sumatra (P. palaeosumatrensis) and also Java (P. duboisi). Together with P. pygmaeus from Borneo these orangutans might have been the original Sundaland orangutan clade. I have hypothesized that P. abelii, the second orangutan in north Sumatra is a new-comer to Sundaland, possibly more closely related to the now extinct orangutans from the Asian mainland (P. weidenreichi etc). The current taxonomy captures this considerable evolutionary diversity, which dates back 3-4 million years, quite well. Isn't that one of the objectives of taxonomy; to ensure that we don't overlook unique evolutionary lineages and let them go extinct before we even recognize them?

Yes our sample size is as small as it can get (n=1 for the morphology), and with a population of only 800 animals, we won't be able to quickly add more, but statistically it is distinct to animals of similar age and sex in the other species, so for now it is perfectly fine species according to the Phylogenetic Species Concept. If we find more specimens and it turns out that our species hypothesis was wrong, then we need to revise the taxonomy and I need to rethink my evolutionary hypotheses.

As to reproductive isolation, maybe Pongo isn't in a particular hurry to develop it and evolutionary change happens in other directions. Maybe retaining gene flow helped the different species select for genes that benefited their individual survival (as has been demonstrated in pig species - Article Genome sequencing reveals fine scale diversification and ret...

I must said I'm really skeptical of overlumped classification well below the species level, as in the last felid taxonomy proposed by IUCN. According it the extinction of Javan and Bali tigers is simply a non-extinction. This is a wonderful way to solve conservation problems in the Anthropocene!

This is a long thread and I'm sure I will be replicating other's responses. However, species are indeed real population level phenomena. They are in fact the only "real" taxonomic rank. Species concepts and species delimitation has enjoyed a bit of a renaissance in the past decade or so with the emergence of unified species concepts (de Queiroz 2007 Systematic Biology 56(6):879-886) and coalescent approaches to species delimitation (Jackson et al. 2017 Systematic Biology 66(5):799-812; Rannala 2015 Current Zoology 61(5):846-853; Carstens et al. 2013 Molecular Ecology 22(17):4369-4383; Carmago et al. 2012 Evolution 66(9): 2834-2849). There is real progress being made in this field such that we no longer have to throw up our hands and consider species as some arbitrary or otherwise intractable problem. I think people have been too quick to characterize species problems as such.

I think part of the issue perhaps is the lack of communication between systematic biologists and conservation biologists. Species boundaries is really a problem that population geneticists and systematic biologists are equipped to tackle but there is maybe a lot that gets lost in translation with other biologists.

I would also add that maybe in conservation we get too hung up on species. I think everyone will agree there is value in conserving diversity within species and as such we should focus attention on distinct, ecologically important populations regardless as to whether they might represent a species lineage or not.

An interesting new paper on how commonly introgression (i.e., hybridization with backcrossing) occurs in birds: Article Avian introgression in the genomic era

One key line for the current debate: " the genetic basis of speciation is highly species-specific and context-dependent." and also "... avian speciation, [which] turns out to be more complex than the classical Mayrian triumvirate of allopatric, sympatric and parapatric speciation."

Douglas Shell (10 days ago):

“Ultimately nothing in science is certain outside of mathematical theorems” – sure, I am emphasising this fact at every possible occasion, it was also discussed in the paper mentioned by Douglas in the following quotation: “I do see that not everyone accepts this logic (see, e.g., Hołyński 2005) but I for one am unconvinced that taxonomy is a special case with a deeper insight into reality than other sciences”! In fact, also mathematical theorems are not certain because they have been conceived by humans, and errare humanum est… But if nothing is certain, and – as argued by Douglas – “something that ‘may be false’ cannot also be ‘real’ in the way that most of us use that term”, then nothing in this world is real and any discussion on reality of anything (including, of course, species) is pointless. But fortunately “most of us use that term” not in the abstract philosophical “absolute” sense (such use in itself would also be pointless…), but as meaning something like “according to our present knowledge this-or-that is real”. Moreover, as regards species, the discussion is not about whether this or that particular form is or is not a separate species, but if species category is something real, i.e. if some groups of existing organisms do represent “segments of phyletic lineages (ancestral-descendant sequences of populations), whose gene pools have differentiated beyond limits of reversibility”, or such lineages are only created by taxonomists' imagination – if at least some such groups do exist (even if we would be unable to prove this in any concrete case!) the species are real (in the same sense as the mathematical theorems mentioned by Douglas…)!

And one more important remark (already not addressed to Douglas): [most of] biological (and scientific in general) concepts, theories &c. do not primarily address the – in fact “unnatural” – conditions created by man: the world, including living organisms, and the natural phenomena and their interrelationships, exist and “work” already for billions of years, and human influence is but a trifling moment of confusing factor activity. We (taxonomists and most other biologists) are not primarily interested in how this or that population behaves under the circumstances created by man (e.g. in destroyed habitat or in the presence of alien – introduced or genetically modified – organisms), but in the general structure of the living world and general mechanisms of its evolution. So, even if two particular populations that hitherto behaved as separate species of pigs or birds now started to intermix, this fact has no relevance not only to the question of the general reality of the category of species, but even to that of these concrete populations having been “valid” species or only mistakenly considered as such: to quote Erik’ quotation from Colin Groves “interbreeding under human control [and generally under unnatural, man-made conditions – RBH] does not give any indication of species status even under the BSC”!

I appreciate Roman Bohdan Hołyński's (RBH) response above (many thanks!). We agree on most of the fundamentals but not (and on this we also agree) how the word "real" is used. Let me dig a little deeper into our disagreement.

RBH quotes me when he says: "“something that ‘may be false’ cannot also be ‘real’ in the way that most of us use that term”, and responds with his own interpretation "then nothing in this world is real and any discussion on reality of anything (including, of course, species) is pointless." (my bold)

I think this conclusion is profoundly wrong. It is useful to postulate what is real even while accepting we may be wrong. This is in fact how science works. We can accept that some reality exists even when we acknowledge that science only provides our "best current interpretation" of it. Note that we are not arguing if the species concept is useful or not. We can agree it is useful even while accepting that any such categories are abstract and may (at least in principle) be found to be wrong. There are many good reasons to improve our interpretation of reality, and that is what science does. Science has provided, and continues to provide, many valuable and useful insights. We benefit from the best picture of reality we can derive. So - no not "pointless".

RBH also talks about lineages and says that "if at least some such groups do exist (even if we would be unable to prove this in any concrete case!) the species are real (in the same sense as the mathematical theorems mentioned by Douglas…)!"

No, hypotheses and theorems are different. Please note the "if at least some such groups do exist " at the start of RBH's statement. It is a conjecture based on a hypothesis. Theorems are different. They are neither hypotheses nor conjectures and by definition they will not be falsified. We know theorems exist and are true in a fundamental sense that we cannot achieve for species (no need to discuss human errors ... that just muddies the water and is a different class of problem, e.g., Pythagoras' Theorem is true for a geometric space that accords with Euclid's axioms ... end of story). The belief in species, according to any given species definition, remains a hypothesis (as most taxonomists acknowledge).

We could spiral off here into the philosophy of " universals" and the existence of "properties" (here the property would be the species designations shared among individuals). I will keep it brief (others may be able to add): RBH appears to argue from the realism perspective (presumably favouring Aristotle and his "natural kinds"); I can accept some form of abstract realism for logic and mathematics, but find it inappropriate for science addressing the nature of the real world where the nominalism of William of Ockham (and many others) is better suited (i.e. all general properties are abstract and exist in the mind). The point is that our knowledge of external reality (including their properties) can always be revised as information and understanding improve (such knowledge exists only in our minds).

My conclusion: species are hypotheses. They are useful, but they exist in our heads. We can be fairly confident (given sufficient evidence) that these hypotheses represent something useful in the real world but they are not inherently real.

Yes, fully agree with Doug's conclusion!

To Eliécer E. Gutiérrez (2 months earlier; sorry, I'm a bit late here in this forum!) ... who wrote:

"However, I really dislike when colleagues argue in favor of their preferred species concept using as a justification that such concept is better for conservation. What an indefensible argument! "taxonomy informs conservation, not the other way around". You cannot try to convince others about the presumed superiority of a species concept because you claim it to be more convenient for conservation efforts." ... this received a lot of applause and recommendations ...

I would say: Whatever motivation you might have to use whatever species concept, if you have good reasons and evidence to call something a species, then just go ahead and do so! If it is eventually accepted by those people working on those organisms, then you probably did the "right" thing, right!? Of course, you shouldn't walk around lowdly proclaiming that conservation was your motivation and only a fool would do so. But if this actually was your motivation and you succeed in establishing an endangered organism as a species on scientific grounds, then: why not!? You may have saved a life! ... as a sincere scientist!

I noticed in this long conversation that there was no reference to the species issue concerning individuals and classes that was at the center of species discussion back in the 1980's. It seemed to me that most, if not all, the discussion was about species definitions - treating species as a class concept and all the tangle it produces (as evidenced in the nature of the discussion). If species are individuals (this is a species concept) then species are spatio-temporally bound entities - they are not defined, but diagnosed. Diagnosis is something quite different from defining and it is used quite adequately for recognizing entities as fluid as diseases. When it comes to the third orangutan, this is a particular diagnosis of reality with which anyone may agree or disagree according to whatever criteria one may wish to apply (and to confuse things even more, I would argue that there is a fourth orangutan in the room - us :) but not really joking.

Hi, John. How is it down under? You dismiss the previous arguments as "treating species as a class concept". So species as "spatio-temporally bound entities:" this solves the problem, eh? But if they are "individuals" then we should all be able to recognize them. The problem is that different people recognize different species, from the same data, and philosophizing about their nature hasn't yet helped anyone to come up with a workable criterion.

Hi Jim,

May be well enough, but I got exiled to the N. Hemisphere a few decades back :)

I guess my point is that it doesn't really matter with species. And it's not a problem over whether people recognize the same thing or not. Just as medical people may disagree over their diagnosis of a medical condition, the disagreements have nothing to do with the reality of the condition itself and medical science has not fallen down over the fact that a real, existing condition, can be the subject of disagreement. Of course sorting out the 'best' answer (diagnosis) may or may not be easy. Churchill (1949) might have had it right to say "There are always several points of view that may be right, and many which are plausible". Personally I think the designation of a third orang species is pushing things a bit, but it really does not bother me either way.

In reply to: 9 days ago

Gippoliti Spartaco

Added an answer

I must said I'm really skeptical of overlumped classification well

below the species level, as in the last felid taxonomy proposed

by IUCN. According it the extinction of Javan and Bali tigers is

simply a non-extinction. This is a wonderful way to solve

conservation problems in the Anthropocene!

This is most unfair on the IUCN. No-one at IUCN would argue that the extinction of Java and Bali tigers is just fine. To lose these populations is as much a loss of biological diversity whether we call them species or not. The loss of any population of an endangered species is a major loss in the IUCN's view and should be avoided at all costs.

With regard to comment by Josef Niederle,

Paraphyly is where a group is defined by leaving out taxa that also share the same unique common ancestor (e.g. making Primates to include all primates except humans).

Hybridization occurs between entities that are considered different. If different then hybridization, if same then no hybridization. We are all hybrids of our parents.

Whether or not inclusive monophyly is regarded as absurd, most evolutionists seem to favor that approach than deliberately leaving out some members of the phylogenetic sequence.

Regarding the assertion that no universal species concept can exist, I would beg to differ. The concept of species as phylogenetic individuals is a universal.

I'm afraid, in this discussion we - to use Johnson's (1968) formulation - “have perhaps already ventured too far into the philosophical chamber of horrors”; I am biologist, not philosopher, so am neither competent nor particularly interested in, and therefore will not comment or use in my argumentation, questions like e.g. whether in philosophical sense species are or are not "individuals" (in biological sense individuals are my neighbour's dog, a flea in its fur, or myself, but not the taxa Canis familiaris, Pulex irritans or Homo sapiens...). The reality of species is important to me (as biologist) only as far as it influences the interpretation of observed biological phenomena and facts (of course, keeping in mind that "facts are no more - but no less! - than highly corroborated hypotheses): as regards this discussion, the structure of biodiversity. Well, theoretically the "structure of biodiversity" is itself "but a hypothesis": maybe biodiversity is an unstructured aggregation of individuals (in biological, not philosophical sense)? But my (and of course not only mine) taxonomic experience does not allow me to accept this "null hypothesis": I see that biodiversity is indeed highly structured, among others being composed predominantly of separate, hierarchically arranged, definable supraindividual units termed "taxa", those at one particular hierarchical level being called "species" - in the "biological commonsense" this means that species is (as a category) and are (as particular units) real!

Once more to Douglas: I'm sorry, I had apparently misunderstood your formulation that "something that ‘may be false’ cannot also be ‘real’": you (as I see now) wished to say that nothing can be at the same time false and real, but I interpreted this as if you say that it is impossible to be simultaneously falsifiable and real, i.e. that "may be false" is synonymous with "assuredly [by definition] not real"...

Johnson L.A.S. 1968. Rainbow’s end: the quest for optimal taxonomy. Proc. Linn. Soc. NSW 93, 1: 8-45

I can sympathize with Roman’s desire not to comment on philosophical questions about individual and species, but philosophy does lie behind all perspectives on species, especially about their being real for if real, what kind of reality? The point about the species as individuals concept is that species cannot be defined by an spatially unrestricted quality. If they were they could not evolve (a defined essence cannot, but its very nature, evolve and why a chair is a chair whether it is found on earth or Mars). But they species can be diagnosed for particular places and times. So if we take the new orangutan species (population, taxon or whatever) its meaning lies in its spatiotemporal boundaries by which its physical (genetic and epigenetic) characteristics have their existence. As these parameters may change the orangutan species will evolve.

John: "if real, what kind of reality?" - the biological commonsense "kind of reality" as outlined in my last comment: when we see that something does exist, it is real! As a simple biologist, not philosopher, I feel not quite clear what do you mean by "spatially unrestricted quality", but whatever it is, coaltits, tigers, Colorado beetles, Scotch pines do exist as distinct units of biodiversity (taxa), = are real!

No argument about existence being real from a practical viewpoint. But what kind of reality is the question behind all science. For example, the real nature of lightning was not self evident. It took the development of certain concepts along with investigation to get to our present understanding. The same goes for species. "Common sense" may not always be the best guide since it has a rather fluid nature. Common sense tells me that the sun moves and the earth does not, for example.

Spatially unrestricted qualities are definitions that are not constrained by a place and time. For example, a chair can be defined and recognized wherever it is. One may specify one or more features for a species (or any other entity) but the 'definition' is meaningless without reference to a place (large or small) and a time. At least that is how I look at this. If the authors had described the third orangutan species without specifying its location and distribution much of the significance of the diagnosis of the taxon would have been non-existent. It makes a big difference that it is located where it is as opposed to somewhere else.

The analogy between a species and a chair is good. We generally know a chair when we see one. And we generally know a species when we see one. But boundaries are fuzzy. Is a chair with a leg missing still a chair? How many legs does it need to miss before it stops being a chair? I looked at Wikipedia and found the following definition: A chair is a piece of furniture with a raised surface supported by legs, commonly used to seat a single person. Chairs are supported most often by four legs and have a back. however, a chair can have three legs or can have a different shape. Pretty fuzzy too - see the almost-not-a-chair here: https://www.yankodesign.com/2017/11/16/the-almost-not-a-chair-chair/. Formally speaking, anyone designating an object to be a chair is posing a hypothesis based on a definition and matching observed characteristics.

Of course, chairs do not generally reproduce (at least not mine at home, as far as I have so far observed), but they do "evolve" through a selection mechanism that favours certain forms of chair over others. Those changes and our increased understanding of the many variety of "chair" push us to rethink our chair definitions. A Neanderthaler might have called his favourite contemplation rock a "chair", but for us that same rock is now just a rock, as it doesn't have legs and therefore cannot be chair.

Chairs and species are real because we generally recognize them when we see them. But they are also hypothetical in terms of their definition. Defining them with 100% accuracy and in all their many forms is unlikely to be possible.

My question was originally about the definition part of the species issue. How can a new orangutan species be dismissed by some if it is clearly hypothesized as such based on a clearly defined species concept? Just because some do not like our species concept does not mean they can dismiss the species definition based on that concept. It would like arguing with the Neanderthaler that his rock was not a chair. A wack over the head with his or her stone hammer would have been the likely response.

That makes species and chairs human-made constructs, based on our observations and definitions, even though for all means and purposes those species and chairs have biological and physical reality.

Not sure these Saturday morning ramblings are useful, but it was fun trying to work what a chair was (whilst sitting on one) :)

One may specify one or more features for a species (or any other entity) but the 'definition' is meaningless without reference to a place (large or small) and a tim""""

"One may specify one or more features for a species (or any other entity) but the 'definition' is meaningless without reference to a place (large or small) and a time" - in my opinion, rather the opposite is true: if I say that orangutan in Sumatra is an orangutan, but if found in Amazonia (or even on Mars or Proxima Centauri...) it is not orangutan, then my definition of orangutan is really meaningless! If I say generally that species in, say, Europe is a section of phylogenetic lineage but species in Australia is something else, then my concept of species makes also no (at least biological) sense. Place (type-locality) and time (usually recent) are indeed included in the original description, but they are but "supplementary" data (as for species - the situation with subspecies is somewhat different, but this is another, here irrelevant question): if indeed I find an orangutan in Amazonia, it is still an orangutan, I will only have more trouble in explaining how it got there... Of course I should verify whether it is indeed an orangutao or only something similar, but that verification must be done according to the features of the animal itself, not just on grounds of its having been found outside the Malay Archipelago!

By the way:

1). Do your commonsense really (based on all the knowledge you in XXI century have) tells you "that the sun moves and the earth does not" [i.e. that this is the simplest interpretation congruent with observations concerning other planets, stars &c.]? You (like, unfortunately, many others...) seem to mistake commonsense with superficial appearances evaluated with brain "switched off".

2). If you are not looking for simplest interpretation, "Uncle Albert" (Einstein) would assure you that both interpretations are true: Earth moves in relation to Sun, Sun moves in relation to Earth!

I'm curious if anyone has a view regarding my conclusion explained above :

Species are hypotheses. They are useful, but they exist in our heads. We can be fairly confident (given sufficient evidence) that these hypotheses represent something useful in the real world but they are not inherently real.

It seems to me that this simple summary escapes some of the major disagreements (due to lack of clarity regarding what we mean by "real"). Please let me know!

Another point that has not been sufficiently stressed in this forum is that, regarldess of the many philosophical/practical definitions of species, there is an inevitable subjective component in taxonomy. To verbalize what I mean, Simpson is certainly more eloquent than me:

"even if interrelationships in a group of animals were completely known and even if there were complete agreement about the scientific principles to be applied, innumerable different classifications could be made consistent with those interrelationships and valid under those principles. Selection among those alternatives is decidedly an art."

Douglas - I probably agree with you, but it depends upon what do you mean by inherently real?

Guilherme - again, generally I agree, but mainly just as long as we have not such absolute knowledge - if we have, "if interrelationships in a group of animals were completely known and if there were complete agreement about the scientific principles to be applied", then the information content (predictive value) of each classification would also be known, and selection should be done on that (already not "art" but objective calculation!) basis.

Robert:

1). "When considering the two morphologically and genetically differentiated populations that are still able to reproduce (“hybridize”) as (two) “species”, how do we acknowledge the importance of the reproductive isolation of the third population terminologically?" I do not really understand the problem? If the two populations can (in natural circumstances) reproduce, they certainly do, and so they are not really "morphologically and genetically differentiated", and are not separate species (if they are partly - at the level above 75 % - differentiated thay are subspecies).

2). "An orangutan in Amazonia is not an ordinary orangutan" - so, what it is??? A separate species to be described as Pongo amazonicus??? Had Neil Armstrong "speciated" into a non-human taxon while on the Moon?

Doug, I agree with that statement: " Species are hypotheses. They are useful, but they exist in our heads. We can be fairly confident (given sufficient evidence) that these hypotheses represent something useful in the real world but they are not inherently real. "

Species are hypotheses, and they are real - there is no contradiction between these statements! Every statement, every concept, every term is, in fact, a hypothesis ["facts … are nothing more than highly corroborated hypotheses” – Eldredge & Cracraft (1980)], so either the "state of being a hypohesis" contradicts real existence (and so nothing in this world is real), or there is no contradiction and no grounds to question the reality of (admittedly hypothetical, but well corroborated) species!

Additional term for a population that cannot reproduce with others is not needed: this is just the species; additional term for a population that can hybridize but is partly (at above 75%) differentiated is also not needed: it is a subspecies.

Real or not, our ideas of what constitutes a species and what it means can have lethally real consequences. Take the case of 'Tobin' who had the misfortune to have parents of different species. Tobin was considered such a blasphemy for conservation theology that he was given a personal name (Tobin), duly photographed and then executed. All in the name of species 'purity'. If species do not behave themselves watch out! If that reminds anyone of actions carried out by a certain political philosophy I would not be surprised. Tobin is probably still preserved somewhere in a collection.

Sorry, I do not know: who (or what?) is (was) 'Tobin", and what is his (her? its?) relevance to the question of reality of species?

Tobin was a hybrid of two different species of robin in New Zealand. The relevance to the reality of species is that real or not, the consequences of one's view on the reality of a species can have lethal consequences for animals if they find themselves on the wrong side of that species definition, concept or whatever.

Thanks for the information! However, I still do not understand the case. Occasional interspecific hybrids (e.g. among ducks) are rather frequent in nature and nobody, to my knowledge, considers them "blasphemy for conservation theology" (or for anything else) - why just this one became so disturbing??? Perhaps it was a hybrid between a rare native bird and an introduced alien, and conservationists (if so, justly!) wished to prevent the "genetic contamination" of the already otherwise also endangered population? But this does not seem to have much to do with reality of species: alien, artificilally introduced species are everywhere considered one of the main cause of the deterioration of natural communities.

Roman,

Thank you for the thoughtful questions.

Ah yes - 'genetic contamination' is indeed the philosophy of species as pure breeds that was behind it all. In the name of species purity the hybrid was executed. The hybrid was so disturbing because it confounded purity. Just like ethnic intermarriages often arouse so much bigotry from both sides. Hybrids are always a threat to the theology of purity, whether in science or in society.

In the tobin case it involved two endemic species where eggs of one rare species were transferred to the nests of another more common species to rear so the rare species females would lay more eggs and further increase the population. One of the transferred eggs that hatched later mated with the foster species and a hybrid resulted. Tut, tut. Can't allow that. So are species real. yes, real enough to execute transgressing hybrids.

Too much is made of reproductive isolation as the key to evolutionary divergence, and too little of hybrization as a speciation mechanism. I quote the first lines of a 2005 paper by Bell and Travis that reflect this (Trends in Ecology and Evolution Vol.20 No.7 July 2005: 358-361): "Hybridization between members of conspecific populations and closely related species is common and has important evolutionary consequences [1]. One possible consequence is transgressive segregation (Box 1), which

produces hybrid phenotypes that exceed the phenotypic range of the parental species and which might be a common phenomenon [2]. Transgressive hybrids can be sufficiently divergent from both parental species to establish a new hybrid species that can coexist with both

parental forms." What does this mean for taxonomy? According to the BSC, the parent taxa cannot be "species" as they hybridized under natural conditions and produced fertile offspring. I still think such a taxonomy does not accurately reflect the evolutionary processes that took place. Don't we now have 3 taxa, all on their own evolutionary paths?

Maybe Tobin should just have been left alone.

because it real .

Sorry again, but I think here two serious mistakes are being made: confusion of natural and artificial (induced or at least provoked by man) hybridizatiion, and unrealistic anthropomorphization. As seen from John's explanation, "Tobin" was a result of human intervention: unintended creation by man of unnatural situation, with the result opposite to the desired. On the other hand, comparison of killing one - in fact, somewhat "monstrous" [how would we evaluate a hybrid between man and, say, a gorilla (some Godzilla)?] animal with racist (intraspecific) segregation between humans is rather unjustifiable. It is very nice, and I am happy to see, that people start to notice that animals are not "things" that can be killed, tortured, destroyed at will, but we should not fall into the opposite nonsense: otherwise we should not only prohibit hunting or angling (what, indeed, should be strongly restricted to real needs, killing just "for pleasure" should not be allowed), but also fishery, breeding animals for meat, killing mosquitoes, fleas or other parasites, production of silk, &c., &c., &c. - in fact follow the life of digambara ("clothed with air") jainists (but even they eat something, and why, for that matter, "Tobin" should have the "right to live" but rice or potato should not? - only because plants represent another evolutionary branch than animals do?). It is strange that extermination of entire species does not raise opposition from many people (questions like "why to protect any insect" are not rare even in the "columns" of RG), but the death of a single abnormal individual does and induces so drastic pseudo-analogies!...

Like much of what was said about the inconsistencies of attitude about species and individual organisms. I too have seen the outrage over treatment of individual animals (no more killing lobster in boiling water for example) but indifference to the species survival and even comfort.

Interesting to refer to the Tobin as an "unnatural situation". To me this reflects a moral judgement that human action is unnatural whereas anything of non-human action is natural. If this were to be the case then, for example, all of the Amazon rainforest was unnatural from the time of first human colonization many thousands of years ago.

I would add that in no way was I implying that Tobin had a 'right' to live, was just disconcerted about the rationale for the decision by conservation scientists.

What would happen if one or more captured northern Sumatran orangutans (I understand it is illegal but still happens) was brought south and then one or more escaped into the wild and interbreed with the new species and had viable offspring. What would conservationists do?

That would depend on which conservationist is involved. When peregrine falcons disappeared from the eastern U.S., conservationists had no problem bringing peregrines over from Europe to release. When bald eagles declined from eastern U.S., conservationists went to Alaska to secure birds for the reintroduction. We now have reproducing bald eagle and peregrine falcon populations throughout eastern U.S. but their genetics is not quite the same as was found in the native birds.

Irrespective of the arguments being presented here being correct or not, it should be noted that debates such as this one are very often being distorted and used by people who understand very little about taxonomy to derogate this science, and justify the deprivation of resources and positions to taxonomists. Taxonomy is the science that underpins all of biology -- for correct names are necessary to communicate correct science -- and yet it is being less and less valued based on arguments more political than scientific. I would like to see 'anti-species' people conduct biodiversity surveys in species-rich areas, when the survival of several narrowly endemic species could depend on good quality taxonomy being available and adequately applied.

In my opinion, the so-called "active nature protection" (as in the case of Tobin, Peregrine Falcons or Bald Eagles) is generally (except very rare specific situations) a misconception (in fact, creation of "unnaural nature") - but this is a different question.

As to "that human action is unnatural whereas anything of non-human action is natural. If this were to be the case then, for example, all of the Amazon rainforest was unnatural from the time of first human colonization many thousands of years ago", this is almost so but not quite so. Until humans lived like other species, hunting and collecting what they needed to survive, they were practically as natural element of biocoenose as foxes, eagles or alligators; only since the beginning of agriculture and industry, with their destruction of environment on the scale much greater than that caused by all the other living beings together, their activities evidently cannot be considered natural (if that word has to mean anything in similar contexts). Of course the transition was gradual and not coordinated, so it is not possible to say the date of "unnaturalness" (as almost always in this world, the borders are fuzzy...).

The way I see it, there are two problems here.

1. The codes of nomenclature assume that every organism can be referred to a species (or, for botanists at least, be given a hybrid formula). The boundaries of species may be fuzzy or not in reality – the codes don't care, they assume the boundaries are not fuzzy at all.

Ernst Mayr once wrote (or so I read in secondary literature) that asexual organisms "do not form species". That's obviously true under his preferred species concept. And yet, the Codes force us to pretend otherwise. We should simply be naming smaller and smaller clades, and no species at all, in such cases; but the Codes won't let us. (Not as long as the International Code of Phylogenetic Nomenclature hasn't been implemented; and that one won't deal with species names at all.)

2. Of the many, many different species concepts, many describe kinds of entities that really exist in nature. But they all describe different ones. They are not different definitions of the same kind of thing; they have nothing in common except the word "species" in their names. Ideally, we'd have separate nomenclatures for all of them (or at least several of them) so we could talk about them all; after all, as someone has mentioned above, different ones are useful in different contexts. It would be great if we could name the reproductively isolated "species" and the morpho"species" and the phylogenetic "species" that the Third Orangutan belongs to. And yet, the Codes force us to pretend otherwise. If we name one such entity, we can't name the ones that overlap with it, because the Codes don't allow for a single organism to belong to more than one "species". That makes talking about them difficult.

BTW, many species concepts describe entities that are neither individuals nor classes in the philosophical sense, but "homeostatic property clusters". But I don't think this matters much.

David: The Code of nomenclature is, of course, a convention to allow us to know what we are speaking about, the first condition sine qua non for this is that each taxon (species in particular) must have one (and only one!) official name and each officially accepted name must be applied to one (and only one) taxon.

"Of the many, many different species concepts, many describe kinds of entities that really exist in nature. But they all describe different ones. They are not different definitions of the same kind of thing; they have nothing in common except the word "species" in their names" - no, there is only one generally accepted concept of species (in short: a section of a separate evolutionary lineage), there are only a few (by far not "many, many"!) commonly applied criteria to assess whether or not a given population (or, usually, group of populations) matches this concept (i.e. is a species) or not; in principle all these criteria should give the same (except for somewhat fuzzy border situations) result: if the population under study is morphologically distinctive ("morphoevolutionary" = "phylogenetic" criterion) then it is reproductively isolated ("biological" criterion) [otherwise intermediate or mixed characters would have occurred]; if, in turn, it is reproductively isolated, it represents a separate evolutionary lineage! Of course there are some confusing situations (e.g. single gene or "supergene" inheritance like eye colour in humans, or purely phenotypic differences like degree of sunburn) which could be mistaken for taxonomic characters, and anyway sometimes we erroneously evaluate the interpopulational relationships, or misinterpret the criteria as independent "concepts" (errare humanum est - and of course taxonomists are not exempted from this "original sin"), but in principle all should lead to the same conclusion (and have been designed just for that purpose)!

So Mayr is, strictly speaking, right: asexual species do not form species (or, in other words, every individual, being by definition reproductively - and, thus, evolutionarily) isolated, is a different species. But it would be practically impossible (and would make no sense...) to name each individual as separate species, so a "subsidiary" conventions have been accepted to divide them into "manageable" clusters and conventioally consider them species (not being a specialist of any asexual group, I do not know concretely what are the accepted criteria, but this does not matter in our context).

David: The Code of nomenclature is, of course, a convention to allow us to know what we are speaking about, the first condition sine qua non for this is that each taxon (species in particular) must have one (and only one!) official name and each officially accepted name must be applied to one (and only one) taxon.

“Of the many, many different species concepts, many describe kinds of entities that really exist in nature. But they all describe different ones. They are not different definitions of the same kind of thing; they have nothing in common except the word ‘species’ in their names” – no, there is (at least in zoology) only one generally accepted concept of species (in short: a section of a separate evolutionary lineage), there are only a few (by far not “many, many”"!) commonly applied criteria to assess whether or not a given population (or, usually, group of populations) matches this concept (i.e. is a species) or not; in principle all these criteria should give the same (except for somewhat fuzzy border situations at the “roots”) result: if the population under study is morphologically distinctive (“morphoevolutionary” = “phylogenetic” criterion) then it is reproductively isolated (“biological” criterion) [otherwise intermediate or mixed characters would have occurred]; if, in turn, it is reproductively isolated, then it represents a separate evolutionary lineage (i.e. fits into the species concept = is a species)! Of course there are some confusing situations (e.g. single gene or “supergene” inheritance like eye colour in humans, or purely phenotypic differences like degree of sunburn) which could be mistaken for taxonomic characters, and anyway sometimes we erroneously evaluate the interpopulational relationships, or misinterpret the criteria as independent “concepts” (errare humanum est – and of course taxonomists are not exempted from this “original sin”), but in principle all should lead to the same conclusion (and have been designed just for that purpose)!

So Mayr is, strictly speaking, right: asexual species do not form species (or, in other words, every individual, being by definition reproductively – and, thus, evolutionarily – isolated, is a different species). But it would be practically impossible (and would make no sense...) to name each individual as separate species, so “subsidiary” conventions have been accepted to divide them into “manageable” clusters and conventionally consider them species (not being a specialist of any asexual group, I do not know concretely which are the accepted criteria, but this does not matter in our context).

“It would be great if we could name the reproductively isolated "species" and the morpho"species" and the phylogenetic "species" that the Third Orangutan belongs to…. If we name one such entity, we can't name the ones that overlap with it, because the Codes don't allow for a single organism to belong to more than one "species". That makes talking about them difficult” – no, that makes meaningful talking about species possible: otherwise we would never know whether difference in opinion reflects different interpretation of facts, or only different terminology used! As said above, the reproductively isolated "species" and the morpho"species" and the phylogenetic "species"”, if they are real species (a sample of evolutionary lineage), mean effectively the same, so the “third orangutan” either is a separate species (and so fits all these criteria) or is not a species at all (i.e. does not “belong” to any)! If one entity overlaps the other, they are simply not separate species, so there is no need (or sense…) to name them separately (as species – of course they can, if appropriate, be named as subspecies, varieties, morphs or anything else if they match the respective criteria).

“We should simply be naming smaller and smaller clades, and no species at all, in such cases; but the Codes won't let us” – the Code does not prohibit to name varieties, subvarieties, infrasubvarieties, cultivars or anything we wish, only such names do not denote species (or other phylogenetically meaningful units) therefore the Code simply does not officially recognize them: they remain “out of the scope” of the Code.