Biological taxonomy, the identification and classification of the various kinds of living organisms on Earth, is nowadays based principally on genetic analysis. The older basis of gross morphological characteristics is now seen as more of a supporting and confirming discipline than the defining one. In consequence, the evolutionary tree of life has become far clearer and more certain. But there is one embarrassing exception. Beyond the finer branches known as genus, the twigs and leaves of the tree are the various species, sub-species, breeds, varieties and strains we identify. Here, the term "species" has become cloudier, controversial even.
What may be described (somewhat incorrectly) as the lower forms of life are mostly single-celled and habitually clone themselves as the please. They do not reproduce sexually. For these, the genetic variations between individuals fall in number as you follow a twig down, eventually disappearing altogether. At that very tip of the tree, minor mutations are constantly occurring. Most soon die out or join forces through other, fragmentary gene transfer mechanisms, and are of only trivial significance to the biologist. The subdivision of such organisms into specific strains is therefore somewhat subjective, according to the accumulated genetic characteristics which become significant to us. The dividing lines between genus and species, and between species and strain, are useful only as rules for naming the individual strain. Characteristics of biochemistry, morphology and behaviour tend to dominate our decisions as much as genetics; what appears different in some significant way, is different.
For organisms which reproduce sexually, which is to say most plants and animals, the classification of species is fundamental. The parallels between genetic makeup and gross behaviour are exact and leave no room for subjective pragmatism. A few may be able to reproduce asexually when necessary, but that is beside the point; they can and do reproduce sexually as well. At least, that was the way it had been since Darwin, with the only advance being the discovery of the genetic mechanism.
Yet it is here that the worst controversies have arisen, with different camps vying for different definitions of a species to become the norm. Part of the problem lies in the fact that different biological disciplines focus on different aspects of the organism and like to tilt the definition their way. I suspect that part also lies in the nature of the scientific establishment. To discover a new species is a thing in your career, to discover a new sub-species or variety, rather less so. Even if cynical self-promotion does not carry the claimant along, excitement at the prospect will. So incumbents in each discipline, be it genetics or metabolism or morphology, push that much harder for their definition. The result is a plethora of claims to have discovered and named a new species, all under definitions running at cross-purposes with each other. The wider this muddle spreads, the easier it becomes for the more dubious definitions to remain in the mix. It is a downward spiral that the steadier voices now acknowledge is getting out of hand. The re-establishing of a sound, unarguable definition which can be consistently applied is long overdue. So this is the bit I want to focus on from here on.
I would argue that returning essentially to the old viewpoint is the only viable way out of the mess. A twig on the tree of life forms where a population is able to interbreed within itself and produce fertile offspring, but not with any other population. Barring future variation and selection, its gene pool is essentially stable. Note that it is not necessary for every mating of the two sexes to be fertile, only for fertile individuals to breed successfully with some others. Crucially also, the population cannot be subdivided into smaller populations which can interbreed. If we treat a population as a mathematical set of individual members or elements, partially connected by a breeding relation, we can make these conditions mathematically precise (specifically, the set is partially ordered into ranks by genetic sex, with the breeding relation applying to any given pair of members capable of breeding together. The breeding relation shall not apply to any external elements which might exist, nor shall the set be divisible into two such sets). This logical rigour is useful in evaluating other definitions to determine whether, and if so then how, they conflict.
A large number of "species", especially among those recently discovered, would on this basis be demoted. Besides the fuss this would create among their aficionados, how should they be reclassified? Genetic makeup can tell us little more than that the population is a subset of the main group, with a narrower genetic variation. The identification of this smaller subgroup is in fact made on two visible characteristics; geographical distribution and morphology. Firstly, the group must be geographically coherent, more or less isolated from the main population and with little chance to interbreed with it. This creates the conditions for the second characteristic to arise, which is the evolution of some distinguishing morphological characteristics. In the case of the Galapagos finches, it was famously the size and shape of the bird's beak. In other cases it may simply be skin colourings or markings, perhaps fur that turns white in winter or the number of flowers on a stalk. The population is not fully stable; if it were mixed back into the main group to continue breeding, its characteristics would rapidly dissipate. I will say that it is meta-stable. Traditionally, such populations have been classifies as sub-species.
We may be able to identify the genes which characterise the sub-species, but we should not make the mistake of regarding this genetic stamp as definitive of the given sub-species. The isolated geographical distribution and the evidence of some significant observable characteristic both contribute to the definition, while the genetic analysis is an afterthought, once the sub-species has been identified. We can say that a sub-species is a connected subset of members, sharing some other arbitrarily defined property which is not shared outside the subset.
Seagulls represent a classic example, where a great many populations are currently classified as species, yet interbreed wherever they encounter an adjacent population. There is at least one chain of populations around the globe, each of which is clustered around some limited geographical feature, and which can interbreed with its neighbours, except in one instance where the two neighbours can not. In the picture just described, these populations are definitively sub-species; even though the two end-of-line populations cannot interbreed, they form subsets of the greater population which can.
Perhaps the most contentious example of all is that of early Man. Traditionally the Neanderthals are classified as a separate species, Homo neanderthalis, from us, Homo sapiens. Yet we now know that the two populations interbred extensively and some of our most useful traits were acquired in this way. I have argued already for the Humanity of the Neanderthals. And other populations, notably the Denisovans and a still-unravelling chaos of semi-isolated African populations have recently been added to the mix. Following the present argument, all are at most sub-species of Homo sapiens; H. sapiens sapiens, H. sapiens neanderthalis, H. sapiens denisova and so on. In fact one might even play Devil's advocate and suggest that, given the significant range of skeletal morphologies seen between modern humans, or even more noticeably between our pet dogs, the real challenge to this classification is to show that these various groups of early humans were not merely an assortment of races, at the same level with say "black" Africans, "white" Europeans, "aboriginal" Australians and "yellow" Orientals.
Unlike living species, the Neanderthals and Denisovans can be identified only by unique characteristics in their fossil remains. Their geographical distribution and genetic makeup are then inferred from the fossil data. However variations in bones between individuals can be misleading. There are discernible differences between the various living races, although any identification of race from the skeleton is far from infallible. At what point do we compare the bones from two individual subgroups and declare, "these are not just widely differing subgroups from the same genetic group, but they are distinct genetic groups unable to breed with each other"? H. floresiensis, the "Hobbit" of Indonesia, is debatable on this point, but those who left their genetic imprint in our own makeup are definitively sub-species of H. sapiens. Yet their bones have, traditionally, been interpreted differently.
That tradition was created in Western society, during a time when racism was the norm, black Africans had barely been liberated from slavery, and prejudices against them and other ethnic groups was rife. Not only were anatomical features such as skull size and shape, nose profile and so on, regarded as significant but so too were cultural patterns such as musical styles or customs of worship. The poor "ape man" did not stand a chance. Today we know that if a Neanderthal dressed in jeans and a T-shirt and walked down the street, they would be on the short and stocy side and with an ugly face, but not enough for more than the most inquisitive passer-by to cast them a second glance. Yet the heritage of racism and prejudice lingers on today in a diehard refusal to accept them as human, and it is high time we did the right thing and eradicated it from our taxonomies. We must abandon the finer details of skeletal physiognomy as a guide to genetic status, just as we once abandoned phrenology as a guide to the intellect, and stick with the genes themselves.
Speciation occurs only when the ability to interbreed is finally broken, no intermediate individual is left who can bridge the genetic gap, and the species set is divided into two discrete parts.
On the other hand, one might wish to argue that the idea of all the interbreeding populations is so well established as individual species that it is too late to row back; after all, it is only a matter of what we choose to call things. This raises two problems. The first is, in that case what would we call the "supergroup" which does form a breeding set as defined? We have no term between genus and species. Would we weaken the definition of genus to embrace them, thus creating new genera in their tens of thousands? Would we define a species to embrace both the supergroup and the subgroup, and resign ourselves to the naming chaos we have already fallen into? Or would we invent a new term for the intermediate level, a super-species or sub-genus perhaps? But we already have "sub-species" well established, why not just roll it out a bit more consistently? The second problem is that many breeds and variants of some species would no longer stand as such. When you consider the range of morphologies of domestic dogs, both fossilised and alive today, and compare that range with the similar range of morphologies in early Man, there is no justification for labelling the one set as distinct species and the other as mere breeds. It is clearly unworkable to redefine every breed of dog as a distinct species and, once one has ejected the shadows of racism from the debate, there can be no justification for treating early Man differently.
There may be less glamour and self-satisfaction in discovering or classifying a sub-species than a fully-fledged species, but there can be a great deal more rational rigour. If biology has to rewrite all its textbooks and one's own name slips down the ranking in the list of the great and good, it will not be for the first time. Failing to deal with it is what has got us into the present mess. It is high time we did so.
Updated 1 May 2022