I was motivated to ask based on the answer to another SE Biology Question that uses the following argument to conclude that two varients of snake are indeed a different species:

To test whether [Dendrelaphis ashoki and D. girii were a separate species of snake] they (Vogel and van Rooijen, 2011) compared 9 individuals of D. ashoki with 34 individuals of D. pictus, and found statistically significant (p<0.05) differences in five morphological characters....


The authors didn't rely just on morphological evidence: they note that D. ashoki, which is found in the Western Ghats, is also geographically isolated from other populations of D. pictus, which are found in Northeast India, Myanmar and South East Asia.

Minor differences in morphology of a snake: Why designate it a different species?

My question is, why can't we on the bases of an analogous argument conclude that (say) Eskimos & Kalahari Bushmen are, forget the race debate, but two entirely distinct species?

I mean if "number of infralabials touched by the first sublabial" can count as a legitimate species-distinguishing morpohological factor, then so can heights, eye colors, hair texture, skin color etc. right? And we won't even have to fish too hard to gain statistical significance given how different Eskimos are from Kalahari Bushmen morphologically.

Further, the geographical isolation must have been pretty strong too.

  • 1
    $\begingroup$ The quick answer is "because not all characters are equally informative". A great example of this is classifying dogs on the basis of their coat colours -- I don't think you'd think of a fawn-coloured chihuahua as similar to a fawn-coloured Great Dane. However, this is a great question for delving into derived characteristics, and I'm hoping someone comes up with a really great, detailed answer! $\endgroup$
    – Gaurav
    Commented Mar 4, 2017 at 5:52
  • $\begingroup$ @Gaurav Thanks for humoring all my questions. To pick your brain a bit more, when you say "not all characters are equally informative" that begs the question (to a layman like me): Informative of what? What's the basic attribute that we are getting down to here? In other words, the "morphology" is means to what end? $\endgroup$ Commented Mar 4, 2017 at 5:57
  • $\begingroup$ @DavidBlomstrom Thanks! Is it though? Then it would be more straightforward but then why do we get so much elaboration & statistical significance testing of morphological differences & geographic isolation? $\endgroup$ Commented Mar 5, 2017 at 2:40
  • $\begingroup$ @curious_cat I may be wrong, but I think the general idea is to look at overall similarity, i.e. over all possible criteria. Of course that's impossible, but I think the traits that are chosen as relevant are those that have been observed to correlate with differences in other traits. So when mammals are defined based on milk or a specific bone shape instead of, say, color, it's not because traits were picked arbitrarily, it's because it was observed that plenty of very different animals share a color and nothing else, but all animals that give milk are also similar in many, many other ways. $\endgroup$
    – Oosaka
    Commented Mar 6, 2017 at 14:09
  • $\begingroup$ @RozennKeribin Thanks. When you talk about features like "animals that give milk" sure that's a substantive criterion. It's at the level of individual species-defining criteria that I'm skeptical. e.g. When two snake species are differentiated on the basis of the feature "number of infralabials touched by the first sublabial" can we be sure this wasn't something picked arbitrarily? Is that a relevant trait for snakes? $\endgroup$ Commented Mar 6, 2017 at 14:18

2 Answers 2


I think that species delimitation is not a science, or at least not a coherent one. Different fields use different criteria, for practical or historical reasons. For instance, in plants, loads of fertile hybrids exist between species. It seems that botanists do not split based on the traditional biological definition of what a species is: they apparently tend to split more (see the splitters and lumpers notion).

My intuition is that humans are generally interfertile. There even seems to have been interfertility between H. sapiens and H. neandertalensis which have been (or maybe still are?) considered as distinct species. But how are the actual genetic exchanges across the human populations?

In my opinion, if it were proven that genetic exchanges between human populations living very far apart and having evolved very different ecological adaptations (like populations from the arctic and from the Kalahari) were almost reduced to zero, then it would not be absurd to call them different species. Humans would form a continuum that can be considered a unique species (if interfertility is to be taken as a criterion), but with isolated extremes being in the process of forming separate species, somewhat like the extremes of a ring species. Maybe there even is some degree of reproductive incompatibility between very divergent human populations. The thing is that it is not ethically acceptable to experiment with this. We need to rely on "natural" experiments happening. Globalization could potentially tell us in the future, by facilitating encounters between individuals from very distant populations, whether such reproductive incompatibilities exist or not. That said, I doubt that the origins of parents are recorded in a scientifically usable manner when a couple has problems having children.

  • 1
    $\begingroup$ Thanks. If species definition is not a science is it even relevant to modern biology? e.g. When someone conducts a statistical analysis of a sample of snakes and publishes an article that purports to establish a different species of snake based on some set of distinguishing features like "number of infralabials touched by the first sublabial" is that useful? Or does it just get the guy a publication? Do we make progress by defining Dendrelaphis ashoki seperate from D. girii? $\endgroup$ Commented Mar 6, 2017 at 14:22
  • $\begingroup$ @curious_cat Giving distinct names to distinguishable subsets of a species is definitively useful. Giving these names species status may not necessarily be a scientific progress, but it can have practical implications. I remember having heard at a conference a case where the issue was whether some particular edible bivalve on the coast of Brazil and Africa was or not a single species. I don't remember the details, but this had implications on whether or not the species could be used in aquaculture or something like that, due to Brazilian environmental laws. $\endgroup$
    – bli
    Commented Mar 6, 2017 at 14:45
  • $\begingroup$ Though globaluzation has/will also increased gene flow :p. $\endgroup$ Commented Mar 9, 2017 at 5:44

Welcome to the difficult concept called "Species".

Let me ask a simple question are tigers and lions different species?

They look different (Lion is a solid light brown. Tiger is a darker brown with stripes. Male lions have a mane. Male tigers do not)

They are socially different. (Lions live in prides. Tigers are solitary)

They are behaviorally different. (Tigers climb, Lions do not. Tigers love water and swim. Lions do not. Tigers always attack from behind. Lioness have elaborate ambushes as they hunt in groups.)

They live in different environments (lions in Savannah, Tigers in forest)

They rarely interbreed.

Yet when they do...tiger-lion offspring are viable. The males are sterile. But the females are fertile. So gene flow between the two population is possible although limited.

So are lions and tigers different species?

This is the problem with the concept of species.

A lion and tigers are clearly different species of animal.... but why? Is it because they look different? Because of historical reasons? Perhaps because they cannot freely interbreed. Or is because they do so rarely or with difficulty?

If yes, to rarely... how difficult/rare does it need to be for it to be a different species not not simply different populations?

Biology as a field has not come up with a comprehensive answer. The botanist have their own criterion (Because plants can tolerate polyploidy relatively and thus make hyrbid easily). The microbiologist have their own (because DNA can move from wildly divergent bacteria very easily). And so forth.

As to your question (paraphrased) "why scale structure and not hair colour?" Not all biological traits are equal in the ability to change. Some traits are virtually impossible to change (number of limbs) while others are relatively easy (fur colour).

The reason boils down to the genetic programming of those traits. How interdependent one trait is with other traits. So an extra set of limbs... would also require changes in the brain to control said limbs, it would also change the walk cycles of body for a 4 now 6 legged animal. etc.

Hence traits like fur color is not useful as it changes easily with the time.

Thankfully, we now have genetic testing... and we can determine the degree of genetic flow between two populations to help make the final call if a population is one species or possible two or more.

But genetic testing takes time and a lab. So if there is some visual aid that would be helpful in the field. And these is where counting scales comes in.

  • $\begingroup$ Thank! Compelling explanation. What is your opinion on the alternative: We keep speciating on minor differences because it gets people publications and taxonomists a way to justify their existence? $\endgroup$ Commented Mar 6, 2017 at 16:38
  • $\begingroup$ The best I alternative I like, is that two species is defined as two population having biological or behavioral barriers that limit gene flow between the two to a level below Y% per generation (And there is big debate of how much Y should be). However... this definition would make every cell of some bacteria (which rarely exchange DNA or mate) different species. So it does not quite work either. $\endgroup$
    – JayCkat
    Commented Mar 6, 2017 at 22:12
  • $\begingroup$ I like this quantitative approach you mention but it will be a cataclysmic change in the current taxonomy I am sure. Taxonomists won't like a DNA-based definition, right? It would put them out of business. The scale-counting on snakes would become redundant. $\endgroup$ Commented Mar 7, 2017 at 2:34
  • 1
    $\begingroup$ Actually they are slowly going out of business as it were. DNA based definition is moving in. This is why we now consider the orangutan to be 2 species even when morphological they look nearly the same. The problem with a DNA based definition is that it was very expensive to do until very recently for a single genome. Far outside the budget of most labs. Probably still is as you have to genome sequence not one animal but a population. Funding to study a specific plant or animal species is also very limited. So counting scales if backed by DNA sequencing (even if limited) would have its place. $\endgroup$
    – JayCkat
    Commented Mar 7, 2017 at 15:53

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .