Kotatsu
Phthirapterist
No offense, but this question is basically as meaningless as "how do taxonomists recognize species?"; that is, it is impossible to give a general answer, as it will necessarily differ widely between groups of organisms. The reasons for this are manifold:
1. At the most basic level, two randomly chosen groups of organisms are likely to be so different in their biology, ecology, distribution, behaviour, and evolutionary history that they are essentially incomparable. This is true of all taxonomic levels, and perhaps more so in subspecies than in most other levels. This is because:
2. Subspecies exist less in the real world than do species. That is, while the concept of "species" has been discussed repeatedly from all kinds of angles over the last few hundred years, there has been less of a discussion on what is actually a subspecies. Although you may find taxonomists who claim that "species" do not exist, they are likely to mean that "species" do not exist as a natural entity, but are still meaningful as theoretical constructs that allows us to talk about biology. However, you may easily find taxonomists who claim that "subspecies" do not exist, and actually mean that the whole concept of "subspecies" is a meaningless null entity, that gives us no advantages in biology communication at all, and should be done away with. This leads to:
3. The fact that different groups of organisms have been worked on by different taxonomists (the days when someone like Linnaeus could hope to encompass all described species within a single book, let alone within a single mind, are sadly (??) long gone), and different taxonomists will have different outlooks on what approaches to low-level taxonomy is important or meaningful in "their" group. This is especially true in invertebrates, where you may find some groups having had only a handful of prominent experts working on them ever. Their viewpoints have thus had an inordinate influence on the present taxonomy of the group. Especially if:
4. At least one of these taxonomists have been a "rogue taxonomist", that is, a taxonomist who sets out with an agenda and reforms the taxonomy of the whole group to suit his or her preconceived ideas of what the taxonomy of the group out to be, regardless of actual data. Again, this may be more common in invertebrates, but Australian reptiles and amphibians have suffered greatly from this as well, due to three gentlemen (Wells, Wellington, Hoser) who are highly controversial within their research communities (or, rather, ostracized from them). Due to the taxonomic rules and the way taxonomy works, we can never ignore these rogue taxonomists, however, except when we manage to get a particular disruptive work suppressed by the Commission.
I could give you lots of examples of this. The Center for Entomological Studies in Anatolia (CESA) goes through synonymy lists systematically and renames all junior homonyms to names that include the phrase "CESA" (and are apparently in competition with other groups in Turkey who do the same) and publish this in their own, non-peer-reviewed journal. Colonel Richard Meinertzhagen, one of the greatest taxonomic frauds in the history of ornithology, habitually changed locality on labels of museum specimens to fit his ideas that the UK bird fauna was different from that of the continent, albeit only sub-specifically. Many Japanese bird field guides today recognize the Japanese populations of Holarctic or Palaeartic species as distinct subspecies (typically named japonica or japonensis or named after someone Japanese), even if these divisions are rarely recognized outside Japan. Some Chinese scholars, I have heard, do the same. In Australia -- where lots of these people seem to pop up -- there is a whole journal Calodema which is run by a crackpot taxonomist who feels he is prosecuted by the international scientific Gestapo, as his equally roguish friend Makhan calls it (see a post somewhere above).
I can take an example close at hand for my own research. Wolfdietrich Eichler in 1942 formulated the so-called Fahrenholz' Rule, which is often paraphrased as "parasite phylogeny mirrors host phylogeny". He then spent the rest of his professional career (the last paper he published was in 1982, I think) remodeling the taxonomy of chewing lice to fit this rule. Louse species that were found on multiple host species were split into different species on the flimsiest of evidence. Louse species that occur on different host families must belong to different genera. If the hosts are divided into subspecies, then so must the lice be. In the latest checklist of the chewing lice of the world, only 55% of the names he published are considered valid. Only three workers ever (two 19th century ones and one Russian) who published more than 50 names has a higher rejection rate.
Meanwhile, Zlotorzycka decided that while the lice of ducks are the same species, the fact that they occur on multiple species should be reflected in their taxonomy somehow, and she described numerous subspecies, all of which are now synonymized again. Timmermann proposed a system where "Kleinarten" (smallspecies) should be erected for the same louse subspecies occurring on multiple hosts. He proposed a quadranominal [or pentanominal (quintanominal?)] taxonomy, following this structure:
Genus -- denotes what "sort" of louse it is
(Subgenus -- denotes what supergroup (= often order) of hosts it lives on, but used sparingly)
Species -- denotes what group (= often family) of hosts it lives on
Subspecies -- denotes what subgroup (= often genus) of host it lives on
"Kleinart" -- denotes what species of host it lives on.
He quickly abandoned this when he realized that the head louse of the Whimbrel Numenius phaeopus would be Saemundssonia (Saemundssonia) scolopacis-phaeopodis scolopacis-phaeopodis scolopacis-phaeopodis, a name he called "a nomenclatorial monster".
We are fortunate that at the time, this Fahrenholz-based dogmatics was countered by Theresa Clay, Roger Price, Harry Hopkins, and others who synonymized most of these names, and brought louse taxonomy into the still problematic state it is in today, but which is at least dominated by morphological similarity rather than dogma. Many groups have had no such countering influence!
Another example of how personal bias, preconceived notions, and animosity between workers can be detrimental to taxonomy is in frog taxonomy, which -- as I understand it, but I don't work with it at all -- is being done largely by two international groups, who can't seem to stand each other. In a very petty article (I have forgotten the reference, but can look it up if asked), one worker sets out to question the very spelling of proposed genus names of North American Rana, because he feels they are inappropriate, or are based on the wrong Latin term. A review of the "Amphibian Tree of Life" ends with "The best solution would be for all serious workers on the Amphibia to simply pretend this book was never published" (my paraphrasing; can look it up if necessary); note that as several names that would fall under the Code are presented in the AToL, it is not permissible under the Code to simply ignore it.
Examples like these exist for virtually every group of organisms, so even if there had been a biologically meaningful definition of "subspecies" that could work for all groups of organisms, the very fact that different people with often contrasting conceptions of taxonomy have worked on different groups, together with the principle of priority that is so central to the Code, would mean that taxonomy would still be extremely hard to standardize.
Bearing all that in mind, I can give you an example of how *I* decided to keep three populations of wing lice of Curlews (Numenius spp.) as subspecies rather than species. This work is now in press, and I will denote the subspecies as simply A, B and C.
Taxon A lives on the Curlew Numenius arquata, taxon B on the Whimbrel Numenius phaeopus, and taxon C on the Far Eastern Curlew Numenius madagascariensis. I collected B from Sweden and Australia, and got some fresh material of A from a collection in Romania. I further collected microscopy slides of all three taxa from 7 museums in England, Germany, the USA and New Zealand, and got some additional material from a collection in Japan. I could sequence DNA from one B from Sweden, one B from Australia, and one A from Romania.
In several independent analyses of this DNA, the two B individuals, which were collected from different host subspecies, were identical in their mitochondrial DNA (nuclear DNA primers do not seem to work for this group, for unknown reasons). The A individual was almost identical, but differed by 0.5% (uncorrected p distance) from the other two.
I could also compare this with distances between other taxa in the same genus. Within-group distances varied between 0.0-0.6%, whereas between-group distances were between 11.1-20.6%. I did a literature survey of other published genetic distances between and within species, and found that within-species distances were 0.0-3.9%, whereas between-species distances were 3.1-29.8% (but generally in the range 9-20%), taken from seen other studies. Taxon A and B could therefore be considered the same species (provided we consider such comparisons meaningful, which I have argued previously in this very thread they are not!).
However, I found when studying the microscopy slides that there were consistent morphological differences between taxon A and B, and that taxon C was very similar to taxon A, but differed consistently in a few characters (mainly in male genitalia). I therefore judged that it was most convenient to call all three the same species, and divide them into three subspecies following host divisions.
That is one way to recognize subspecies, but in other groups, worked on by other people, this approach may be too "lumpy" ("These morphological differences are enough to separate them into different species!") or even too "splitty" ("Genetically they are virtually the same so they should be the same species and subspecies!").
There is no single answer to this, either. Apart from as outlined above, when it comes to more charismatic animals especially, there is also the matter of conservation politics. What is likely to get more conservation funding, a very rare species that is found only in X County, a semi-rare subspecies that can be found in X County, but whose other subspecies are distributed throughout the state, or a unnamed local morphological variant of an organism that occurs throughout the continent? Typically, the first.
Things like this has been controversial as well, especially when it comes to cases where there are very small genetic differences but large morphological ones. I recall a case (I can look up references in asked) of a group of field mice in the US which was first divided into a large amount of small species, then lumped together (whereupon it lost much of the conservation need and funding), then split again to establish a few very rare morphs as their own species to get funding for their conservation, and then lumped again by a rival author. These all used the same data set, I believe, but interpreted it differently, and there is a series of five or six papers on this issue somewhere in the piles on my desk. I also have a paper (I haven't read it yet) on orchid taxonomy, which basically says the same, and which calls for caution in describing new species.
Controversies like this will always exist, especially in organisms that are more plastic, such as plants. I can think of several cases where a single worker has a taxonomy that is at odds with everyone else, and which is based on the smallest of differences, often for political reasons. A researcher in Lund, Sweden, for instance recognizes about 5000 species of Hawkwort Hieracium in Sweden alone, most of which are genetically identical, but can -- so he claims -- at least very often be told apart (by him), at least if you know exactly where they were collected. They have ranges of a single meadow, often, and reproduce at least partially asexually. He is not taken seriously, but his names need to be taken seriously, unless they are suppressed.
One infamous case is the scholar who split the Birches Betula into hundreds of different species based on leaf shape. Another researcher, who didn't think this was a very good character as the leaves are very variable, took leaves from different heights of the same single tree and sent them to the birch scholar for identification. He sent them back, all properly identified. The second scholar noted this down, mixed the leaves, and sent them back again, and got the collection back with different names attached to all the leaves. This work was never published, as it would seem too much like a personal attack, but of course the taxonomy was ludicrous. Still, unless it is suppressed, it cannot be ignored.
The same goes for European orchids, and there was a worker in Canada who believed he could identify about 65 different species of Canada Goose Branta canadensis (he could also identify the geese in his home area to an individual level by plumage alone!). Naturally, different conservation plans would have to be drawn up for a world in which the Canada Goose in the Great Slave Lake is its own species than in one where all Canada Geese are the same.
In a more controversial example, both England and the Netherlands recognize three species of Redpoll: Common Redpoll Carduelis flammea, Hoary Redpoll Carduelis hornemanni, and Least Redpoll Carduelis cabaret. As many as seven different studies have shown that these three are genetically identical both in mitochondrial and nuclear DNA, that morphometrics overlap, that not all individuals can be placed in the "correct" species on plumage characters, and that there is interbreeding in southern Norway. There is no talk of synonymizing them, however, and in my country, some people are even suggesting we should accept the Dutch/English approach and split them. Twitchers are hopeless.
In other cases, these changes -- whether splitting into new species or lumping into the same, but in either case removing the subspecific level -- are indeed due to sloppy taxonomy in older works. If you look at pre-Soviet Earthworm taxonomy (and why shouldn't you?), a pattern will emerge where it looks like the few workers in that field simply walked from St. Petersburg to Vladivostok and sampled every river they came across. Every sample was then given its own name -- specific or subspecific -- based, so it seems, on the principle that it couldn't possibly be the same species in two different water catchment areas.
My old supervisor was once sent a large sample of aquatic worms from Lake Baikal -- together with Lake Biwa and Lake Ochrid the Mecca of the worm world -- which he was asked to identify. As it is very hard to get permission to collect in Russia, he was delighted, but his joy quickly died when he realized that the worms were mainly immature or poorly mounted on microscopy slides. In short, as the genital elements are usually vital to determine a worm to species level, he had to send the whole collection back, having identified only a very small part. The Russian then sent the whole collection to another Russian (1), who after some time had identified them all, and described a large number of new species, most of whom could in no way be identified from the illustrations and the text.
However, large parts of this change is of course because the ideas of taxonomy, our understanding of evolutionary processes, our technology, and the formulation of the Code changes. We have better microscopes now, and DNA technology, and SEM microscopy, and proper sound recording equipment, and so on. We can base our judgements on where to draw lines on more lines of evidence today than before, and we simply have more data than we did 100 years ago. We no longer believe that two taxa can be "bridged" by a third taxon, and thus all three should be the same taxon, as was common well into the 20th century. International cooperation and increased flow of information has caused the nationalistic taxonomy -- where Meinertzhagen could claim that the British birds were subspecifically different from the French -- to become almost extinct.
I don't know if I answered your question -- or even addressed it as you intended -- but at least I escaped having to read this stupid book on didactics for some time, so thank you!
---
(1) I hate to generalize, but just as Australia seems to have a high concentration of taxonomy rogues, and Japan and China have high concentrations of nationalist taxonomist, Russia seems to have the highest concentration in the world of crap taxonomists. Much of this is perhaps because the Soviets had a different approach to whether or not you should publish quality or quantity than the West did. Eichler, for instance, worked in East Berlin, and in his obituary over Eichler, Emerson suggests that the sloppiness of Eichler's work was mainly due to a political pressure to publish more than his colleagues in the West did. Since he also had an agenda which the West didn't recognize (Fahrenholz' rule), it is easy to suggest that he also wanted to emphasize the supremacy of the Eastern, Fahrenholz-based, taxonomy over the Western. We will never know if this hypothesis is at all true, but as a rule, Soviet Bloc taxonomy is generally crap, despite them supposedly having better microscopes and stuff than the West did. While I'm denigrating whole nationalities, let me just add that Italians are also crap at taxonomy and science in general. I firmly believe that there is no manuscript that is so idiotic, self-contradictory, or flat out wrong that there isn't an Italian journal that will publish it. John Davison's God-based front-loading hypothesis was published in an Italian journal, for instance. I could tell so many stories about how Italian taxonomists work, but this post is already getting very long...
