Clustering Animals into Taxa
The non-professional often has only a hazy idea of what the different systematic units of zoology and botany mean – the genus, the species, and so on. In consequence, they are often confused.
The basic unit of the whole system is that of the species. It is the only one that has an objective definition. This definition is a biological and not a morphological one: in sexually reproducing animals, a species is a population whose members, in the wild, interbreed freely and produce viable offspring. What separates species and keeps them apart is a reproductive barrier.
In a way, all taxa are constructs, figments of the human mind. Delineating them always involves some degree of arbitrariness and hence is eternally open to discussion. But the biological concept makes the species less of a construct than all the other taxa. At least in theory, the biological definition provides an objective test that allows to tell conspecifics from members of other species: if two populations interbreed successfully, they are one species; if they don't, they are different species. The problem is that in many cases it is hard or downright impossible to apply the interbreeding test. Also, the results may be inconclusive even where it can be applied – that is, some interbreeding may be going on, yet it may not be the rule and its success may be doubtful. Though in practice the test may fail, the biological definition of the species as an interbreeding population has an enormous advantage: it explains why there should be any species in nature to begin with, rather than a continuum of forms.
The taxon just above the species is the genus. A genus unites a number of species that resemble one another due to their having a common ancestor. A number of related genera form a family. Related families are united into an order. Related orders form a class, related classes a phylum, related phyla a kingdom. Thus the perhaps commonest of all butterflies, the Small Cabbage White (napi), belongs to a genus of rather similar and related white butterflies called Pieris – which in its turn belongs to the family Pieridae (comprising a great number of morphologically similar butterflies, many of which are white or yellow), which belongs to the order of insects ('the segmented ones,' that is, those whose thorax is segmented into three parts), also called hexapods ('those with six legs,' namely one pair at each segment), which belongs to the phylum of arthropods ('articulate invertebrates with jointed limbs'), which together with some thirty or so other phyla makes up the kingdom of animals – all neatly and unambiguously arranged in one large hierarchy.
Down the ladder from the species, there is only the subspecies, often called 'race' in former times. A subspecies is an important variant form of a species that occurs locally, but consistently, due to isolation from other members of the species. A small local variation does not yet make a subspecies. Subspecies usually are thought of as new species in the making. The process will take a long time, many thousands of years. To start it, some population of an existing species has to become isolated geographically. Gradually this colony will develop characteristics of its own. Some of them will be due to adaptation to a different habitat, others to what is called genetic drift: chance mutations that are neither helpful nor deleterious and for that reason are not eliminated by natural selection. For a long while, this population could still interbreed with the species out of which it is evolving if brought into contact with it. But as time passes, all kinds of changes may accumulate, and one day the reproductive ties will snap and it will be a species in its own right. This point is not easy to determine. To prove that two different lepidopteral forms do not interbreed it must be shown that there are places where they fly simultaneously (the word is, where they are 'sympatric') and yet refrain from mating, or if they do mate that no viable offspring ensues. Or else breeding experiments would have to be carried out – but these are a practical option only for the few Lepidoptera whose biology and life histories are known and where fertilized eggs can be raised in the laboratory.
With the advent of molecular biology, a powerful tool has been provided that will eventually help to settle taxanomic problems like the distinction between subspecies and species. It consists in DNA mapping of the taxa involved and in looking for similarities and dissimilarities not in their morphology but in their genetic make-up. If this has not become general practice yet, it is for two reasons: the cost of sequencing and the necessity of analyzing DNA from living cells.
Given this state of affairs, it is obvious that the concept of the subspecies is a problematic one. In lepidoptery there always have been 'lumpers' and 'splitters'. Lumpers tend to place as many variant forms as possible into one taxon. Splitters tend to assign every variant form a name of its own. It often happened that even one slightly variant specimen with a different locality label was named as a 'race' or 'subspecies'.
Moreover, though the concept of the subspecies seems to imply that there is a model species on the theme of which different daughter subspecies are variations, there mostly is nothing of the sort. All there is are different peer subspecies intergrading into one another. In this case simply the first one discovered is arbitrarily promoted to the status of nominate species, specific and subspecific name repeating each other, as in Hyles lineata lineata, the American White-lined Sphinx. It is small wonder that there is a good deal of uncertainty whether some new insect is to be granted specific or subspecific rank. Would-be species have often been demoted to subspecies and subspecies promoted to species, as has happened with some of Nabokov's butterflies. The matter may never be settled definitely.
If subspecies are problematic, all infrasubspecific taxa are still more so. In the late nineteenth and early twentieth century, many lepidopterists have reveled in describing and naming all kinds of infrasubspecific would-be taxa: 'races,' 'forms,' 'varieties,' 'hybrids,' 'aberrations', 'seasonal forms'. Nabokov, though by temperament more of a splitter than a lumper himself, has often made fun of this craze which was fuelled by the commercial interests of insect dealers who liked to have as many different items for sale as possible. Once he ridiculed an amateur from Berlin-Pankow for naming eight 'races' of the Apollo Butterfly from his holiday hide-out in Tyrol. Most of the names thus introduced have silently disappeared. Races have been equaled with subspecies, and the Code for Zoological Nomenclature allows 'forms' and 'varieties' described before 1960 to be promoted to subspecific rank. In a number of cases, this has happened. Mostly it has not happened. All those former infrasubspecific names today have no standing whatsoever in taxonomy. If they are cited at all, it is without author and year; Nabokov also insisted on quotation marks. (As the purpose of this Guide is historical rather than taxonomic, quote-marks are omitted and authors and years are supplied wherever they could be dug up in the literature of former times.) Some contemporary handbooks don't mention them at all. They are not listed in the authoritative checklist of European Lepidoptera by Karsholt & Razowski.
There are higher taxa, that is, taxa above the generic level, and as they have changed greatly during the last decades, they are bound to confuse the non-scientist. A family may be a very large taxon, one that counts many genera of varying similarity. It is tempting to fill the gap between the family and the genus with additional groupings, taking into account the fact that some of the genera within a particular family deserve to be bundled together because they have an especially close relationship which they don't share with the rest. This may be done by grouping them into a subfamily (with an ending in -inae), a supertribe (ending in -idi), a tribe (ending in -ini) or a supergenus. All these groupings remain constructs and open to discussion.
Nabokov was sceptical about them. "A family falls into groups and it is a purely subjective matter whether we call each such group a 'subfamily,' a 'tribe' or a 'genus' … The purpose of subgeneric and subfamilial classification is not to make things easier or harder for the collector; its sole purpose is to bring out affinities and distinctions by forming and separating groups of species and groups of genera." In his drafts for The Butterflies of Europe, Nabokov meant to avoid subfamilies and tribes altogether, replacing the names by numbers where need arose. However, he intended to systematically apply supergenera and subgenera throughout. The supergenus is special in that it imposes a new (super-) generic name, followed by the old and conventional generic name in parentheses which now becomes that of a subgenus. So Plebejus (Lysandra) coridon reads as 'the species coridon of a supergenus Plebejus which before the introduction of the supergeneric name was thought to belong to the genus Lysandra which now is only one of its subgenera'. For instance, on account of their structural relatedness he planned to unite the blues genera Cupido and Everes into one genus, Cupido (the older name that had priority), and to split this into the subgenera of (Cupido) and (Everes). In this notation, the Little Blue Cupido minimus would have become Cupido (Cupido) minimus and the Short-tailed Blue Everes argiades would have come out as Cupido (Everes) argiades. In general, lepidopterists have not been inclined to alter Linné's binominal scheme, rather designating supergeneric similarities and distinctions by forming subfamilies, tribes and subtribes which do not entail any change of name.
The higher classification of butterflies, skippers and moths had always been handled in a more or less intuitive or even haphazard way. To Nabokov, it was of little concern. "It matters little whether we place Danaus or Charaxes (with Apatura) or Satyrus in separate families or arrange them as subfamilies within Nymphalidae. The same in regard to 'Riodinidae' or 'Riodininae in Lycaenidae.'"
Things changed when entomologists began to ask whether the higher classification of insects that had been employed traditionally was really reflecting the order of nature – that is, if the higher taxa were monophyletic, descended from a common ancestor and excluding those that were less related. The seminal study was conducted in 1958 by Paul R. Ehrlich (who was an entomologist before he became a human ecologist). Ehrlich recognized only five families of butterflies: Papilionidae, Pieridae, Nymphalidae, Libytheidae and Lycaenidae. In his scheme, the Brassolidae, Danaidae, Ithomiidae, Morphidae and Satyridae were included in Nymphalidae. Subsequently, the Libytheidae were also incorporated in Nymphalidae and the number of "natural" butterfly families thus reduced to four. Though Ehrlich's classification was widely accepted, lepidoptery moves slowly, and many authors continued to treat the ex-families as valid families, especially Satyridae.
There have been further proposals, notably by Ackery, de Jong and Vane-Wright, based on recent morphological and molecular data. If implemented, a rearrangement and a further graduation of several nymphalid and lycaenid subfamilies would ensue, pushing some subfamilies and everything below down one level. Even if recent proposals such as these are ignored, the families and subfamilies are not quite any more what they were in Nabokov's times, and there will be more change. The search for a few morphological or molecular characters that are so typical of a higher taxon that a lineage could be based on them involves huge amounts of data that have to be analyzed statistically, and the results are not always conclusive. The last word has not been spoken, and further rearrangements are to be expected.
 Nabokov's Butterflies, p. 583
 Ole Karsholt & Józef Razowski, The Lepidoptera of Europe: A Distributional Checklist, Stenstrup, Denmark (Apollo) 21996
 Nabokov's Butterflies, p. 575, 576 (draft notes for "The Butterflies of Europe")
 Nabokov's Butterflies, p. 575 (draft notes for "The Butterflies of Europe")
 "The Comparative Morphology, Phylogeny and Higher Classification of the Butterflies (Lepidoptera: Papilionoidae)", The University of Kansas Science Bulletin (Lawrence, Kansas), 39 (8), 1958, p. 305–70
 Riek de Jong, Richard I. Vane-Wright & Phillip R. Ackery, "The higher classification of butterflies (Lepidoptera): problems and prospects", Entomologica scandinavica (Copenhagen), 27 (1), 1996, p. 65–101.– Phillip R. Ackery, Rienk de Jong & Richard I. Vane-Wright, "The Butterflies: Hedyloidea, Hesperioidea and Papilionoidea", in Niels P. Kristensen (ed.), Lepidoptera, Moths and Butterflies, vol. 1: Evolution, Systematics, and Biogeography, Handbuch der Zoologie / Handbuch der Zoologie, vol. IV, part 35, Berlin (de Gruyter) 1999