In Which I Reveal Just How Much of a Freak I Am

It's all there in the subtitle to this site. In the last few days I've decided to set myself a task that will probably be ridiculously time-confusing, gut-wrenchingly futile and will doubtless cause me to become even older before my time than I already am. But it's something that hasn't been done since 1923, and I think the time is ripe for it to be done again. I'm thinking of compiling an index for all described taxa of long-legged harvestmen. With a few thousand species involved, this is no small task.

But the thing is, and this is the freakish part, I actually really like nomenclature. Nomenclature is the specific part of the taxonomic process where the researcher sifts through the assortment of available names and works out which is the correct name to use for the organism sitting before them. It is important to distinguish the identification of the correct nomenclature from the identification of the organism itself - the nature of the specimen won't somehow magically change if the name attached to it does. Nomenclature is simply the system of labels that researchers have agreed to use in order to allow communication. As such, many people seem to regard the identification of the appropriate label as a somewhat arduous and uninspiring task, but personally I find it can be quite a lot of fun. As frustrating as past confusions can be, there is also something appealing in the challenge of sorting them out.

As a group, harvestmen have their share of nomenclatural challenges. I've just linked to my post on the mess that is Gagrella in which I just scratched the surface. There are no less than five taxa laying claim to the name Gagrella bispinosa as a result of its repeated use as a subspecific name. The oldest harvestman genus, Phalangium, was originally used by Linnaeus for pretty much any arachnid that wasn't a spider or a scorpion, leading to a fair number of homonyms spread between a number of orders. These are the sort of things I'd like to delve into for the next few years. Sure it's a big call, but if you can't be a little hubristic as a grad student, when can you be?

Gnah! Gagrella! Headdesk!

'Gagrella' splendens - is this the face of Evil? Photo by sswroom.

Please permit me to vent some frustration. I've written before about the ghastly legacy left to many areas of harvestman taxonomy by the work in the first half of the 1900s of Carl-Friedrich Roewer, henceforth referred to as the Antichrist of Arachnology, through his use of artificial classification systems and slipshod taxonomy. In the past, I personally have managed to remain relatively unscathed by the dark influence of Roewer, who did not deal much with the Australian opilionidan fauna. In the last few days, this has sadly changed dramatically. I have found myself wandering into the toxic wasteland that is Gagrella.

The Gagrellinae are a sizable subfamily found in tropical and subtropical Asia and the Americas. The centre of diversity for the subfamily is in Asia, whence about 700 species have been described. Many of these species are exceedingly colourful, strikingly ornamented and just downright remarkable. The problem is that most of the Gagrellinae, including Gagrella itself, have not really been revised since Roewer's massive investigation of the group, culminating in a monograph of the Asian Gagrellinae published in four parts over 1954 and 1955. As with other groups of harvestmen, Roewer used an artificial classification system based on characters such as the numbers of nodules in the legs or spines on the carapace - characters which have sometimes since been shown to not even be consistent between members of the same species, let alone the same genus (there are specimens of other taxa that, if one was to use Roewer's identification system, the left side of the animal would key out as one genus while the right side would key out as another). It is therefore quite likely that, were a full revision to be conducted, many species would have to be placed in different genera from their current position. With, as I said, about 700 species involved, this would be a mammoth task.

And yet there is a second layer of ghastliness, to make the problem even more difficult. Of the species assigned to Gagrellinae, a little less than 200 have been assigned to the largest genus, Gagrella. The type species of this genus is one Gagrella signata. However, when Roewer worked on the genus, he moved G. signata out of Gagrella and into another genus, Crassicippus, leaving the remaining species as Gagrella. Because the genus name is required to always stay with the type species, this actually meant that what Roewer called Crassicippus should have been called Gagrella, while what he called Gagrella should have been something else! Unfortunately, almost all authors following Roewer used his inaccurate sense of Gagrella, meaning that of those nearly 200 species, none of them actually belong to Gagrella unless they happen by coincidence to have been placed in the wrong genus and are actually closer to G. signata.


Unidentified South American Gagrellinae, congregating suspiciously. Photo by Bruno Buzatto.

Under strict application of the rules of nomenclature, what is now 'Gagrella' should actually be called Hexomma, that being the oldest genus name synonymised with Gagrella sensu Roewer in the past (Crawford, 1992). Unfortunately, the name Hexomma has been used as a valid genus in all of about three publications since it was first published back in 1876. Also, there are serious doubts about whether the type species, Hexomma vulcanicum, is actually identifiable - Roewer (1954), who may or may not have seen the type specimens*, claimed it to be based on unidentifiable juveniles, and the type specimen(s) seems to have since gone AWOL (Crawford, 1992).

*Roewer (1954) claims to have examined type specimens that he had borrowed from Vienna, but then eight lines later lists "Holotypus (Thorell) (Mus. Stockholm, Genua?)". As the Vienna specimens were correctly attributed to Doleschall rather than the later Thorell (albeit from the wrong publication), the Vienna record seems more likely to be accurate. Thorell later redescribed Hexomma vulcanicum as Gagrella vulcanica, and the Stockholm/Genoa specimen listing probably refers to the specimen used by Thorell for the redescription (and so not actually a holotype).

Why does all this matter to me? I am currently trying to write a description of a new species of Gagrellinae. Seeing as I am not in a position to conduct the full revision of the Asian Gagrellinae, my best option for now would be to follow the Roewer classification, despite its faults. Unfortunately, the species keys out to 'Gagrella'... What am I (or other workers on Asian Gagrellinae) to do about this huge ugly pile of taxonomic blancmange? Among the options:

1. Move everything currently in Gagrella to Hexomma: Probably not that good an idea, considering that Hexomma is not a well-defined genus and many (if not most) of the species will end up having to be moved out anyway.


2. Conserve the definition of Gagrella as used by Roewer: One option might be ask the ICZN to conserve Roewer's usage of Gagrella. Under normal conditions, this would be ideal, but in this case much the same issues apply as with the first option - what's the point with conserving Gagrella in its current condition if most of the species are probably going to end up having to move anyway?


3. Run screaming in horror from the entire concept of Asian Gagrellinae and end up crouched into foetal position and whimpering in the darkest recesses of the wet collection: At the moment, this option is looking increasingly attractive...

REFERENCES

Crawford, R. L. 1992. Catalogue of the genera and type species of the harvestman superfamily Phalangioidea (Arachnida). Burke Museum Contributions in Anthropology and Natural History 8: 1-60.

Roewer, C. F. 1954. Indoaustralische Gagrellinae (Opiliones, Arachnidae). (Weitere Weberknechte XVIII). 1. Teil. Senckenbergiana Biologica 35 (3-4): 181-236.

Taxon of the Week: Cynortula, Cynortula

Vonones ornatus, one of the few species of Cosmetidae found in the southern United States. Photo by Lynnette Schimming from Bug Guide.

The systematics of South American harvestmen have long been one of the taxonomic world's God-awful messes, with the painstaking work of Pinto-da-Rocha, Kury and associates only recently managing to go some way towards drawing it from the mire. The blame for this morass can be placed almost entirely with a single person - Carl-Friedrich Roewer, who described about a third of the world's total of harvestmen species, some 2,260 taxa. He was able to attain this prodigious output by employing a highly artificial mode of classification. Individual specimens were assigned to species and species assigned to higher taxa on the basis of quite superficial characters such as the number of sub-segments in the legs or the number of spines on the abdomen. Character systems such as genitalia that are now regarded as highly significant were not considered*. Many of the features used by Roewer have since turned out to be variable within individuals of a single species, and sometimes within a single individual - in the case of number of tarsal segments, more than one author (such as Hickman, 1939) has described specimens that have differing numbers of segments on the left side from the right, which would require that each side belong to a different genus, if not subfamily!

*To be fair, Roewer could probably be forgiven for his neglect of genitalic characters. While genitalia had been used by some authors in taxonomy by the early 1900s, the practice was not yet widespread and its importance not widely recognised.

It has to be said that the names he gave his excessive outpourings of taxa were not exactly inspired either. Many of them were derived by sticking some suffix or prefix onto a pre-existing name. For instance, from the original name Cynorta, he gave us Cynortula, Cynortoides, Eucynorta, Cynortella, Cynortellana, Cynortellina, Eucynortula, and I'll stop now before my head explodes. Trust me, there's a lot more. To quote Kury (2003), "The dreadful, uninspired and sometimes cumbersome names created by Roewer and Mello-Leitão and followers, and which are deformations of place names, people's names and (the worst!) pre-existing generic names, are best left alone."

The above-listed genera belong to the family Cosmetidae, one of the largest harvestman families in the Neotropics. While still officially divided into two subfamilies, these are divided solely by whether the claws on legs III and IV are smooth or pectinate and this distinction is not expected to stand up to proper phylogenetic analysis when someone should one ever be conducted (Kury & Pinto-da-Rocha, 2007). While the work of Kury and associates has vastly improved matters with the Gonyleptidae, the other major Neotropical family of short-legged harvestmen*, the Cosmetidae remain almost untouched by modern researchers**.

*Harvestmen fall into three groups, the mite-like, long-legged and short-legged harvestmen. I covered mite-like harvestmen once before. Long-legged harvestmen are the daddy-long-legs type harvestmen. Short-legged harvestmen are generally more heavily armoured, and while they do tend to have shorter legs than long-legged harvestmen, they probably have what would be fairly long legs for any other group of animal. The first episode of Life in the Undergrowth included footage of egg-guarding behaviour in a short-legged harvestman.

**Fortunately, I have reasons, such as the publication of Kury et al. (2007), to hope this may change over the coming years.



The genus Cynortula Roewer, 1912, as it currently stands, contains 32 species from throughout tropical Central and South America, from Mexico and the Bahamas to Bolivia and Brazil (the illustration above, from Goodnight & Goodnight, 1947, shows Cynortula granulata from Trinidad). Lord only knows what will happen to this genus in the future, however. Roewer (1923) seems to have supplied the last description of the genus, and described it as "Schlanke Tiere mit langen, dünnen Beine. 1. und 3. Area mit je 1 mittlerer Tuberkel-Paar; 2., 4. und 5. Area und 1.-3. frei Tergit unbewehrt. 2. Chelicere-Glied auch beim ♂ klein und normal gebaut oder seltener beim ♂ viel dicker als beim ♀ unten oben das 1. Chelicere-Glied weit überragend. Beine: die basal Glied des 3. und 4. Bein auch beim ♂ von gleichem Habitus und gleicher Stärke wie die des 1. und. 2. Bein; Endabschnitt des 2. Tarsus 3-gliedrig; 1. Tarsus 6-gliedrig; 2.-4. Tarsus jeweils mehr als 6-gliedrig, variabel. Sekundäre Geschlechtsmerkmale des ♂ bisweilen am 4. Bein."* This roughly translates (if I translate it correctly through the gibberish of BabelFish) as "Slim animals with long, thin legs. 1st and 3rd areas always with 1 central pair of tubercles; 2nd, 4th and 5th areas and 1st-3rd free tergites unarmed. 2nd cheliceral segment of ♂ small and normally built or more rarely with ♂ much larger than ♀. Legs: basal segments of 3rd and 4th legs the same as 1st and 2nd legs; Final section of 2nd tarsus 3-segemented; 1st tarsus 6-segmented; 2nd-4th tarsus in each case more than 6-segmented, variable. Secondary sexual characteristics sometimes present in 4th leg of ♂." For those not familiar with variation in harvestmen, that's not a very impressive list of distinguishing features. In fact, in Roewer's key to the Cosmetidae, only one character is used to key Cynortula out from similar genera - whether the dorsal ornamentation is a tubercle (Cynortula) or a spine (other genera). Not convincing.

*If there are any German speakers reading this, I apologise profusely for the errors that I have no doubt are all through that. As a result of its publication not too long after the Great War, with materials in short supply in Germany, Roewer (1923) was condensed as much as possible for publication and hence is entirely composed in a series of arcane abbreviations. Any grammatical errors are therefore probably the result of my attempts to restore the description to a readable form.

REFERENCES

Goodnight, C. J., & M. L. Goodnight. 1947. Studies of the phalangid fauna of Trinidad. American Museum Novitates 1351: 1-13.

Hickman, V. V. 1939. Opiliones and Araneae. British, Australian and New Zealand Antarctic Research Expedition Reports Series B 4: 157-188.

Kury, A. B. 2003. Annotated catalogue of the Laniatores of the New World (Arachida, Opiliones). Revista Ibérica de Aracnología, special monographic volume 1: 1-337.

Kury, A. B., & R. Pinto-da-Rocha. 2007. Cosmetidae. In Harvestmen: The Biology of Opiliones (R. Pinto-da-Rocha, G. Machado & G. Giribet, eds.) pp. 182-185. Harvard University Press: Cambridge (Massachusetts).

Kury, A. B., O. Villarreal-Manzanilla & C. Sampaio. 2007. Redescription of the type species of Cynorta (Arachnida, Opiliones, Cosmetidae). Journal of Arachnology 35 (2): 325-333.

Roewer, C. F. 1923. Die Weberknechte der Erde: Systematisches Bearbeitung der bisher bekannten Opiliones. Gustav Fischer: Jena.

Cyphophthalmids Wait for the Mountain to Come to Them

Carl Zimmer beat me to it. I was planning to announce the recent pettalid work after the paper arrived in the mail last week, but it seems I've been scooped. But because good work always deserves a second look, I'll write on it anyway. Besides, I was at least able to pinch the photo from Carl's site.

Pettalidae are a family of Cyphophthalmi, what are called the mite-like harvestmen. Cyphophthalmids are a fairly small group as far as is known, with probably less than fifty described species, but the number of species has been rapidly increasing in recent years. Though they are divided into about five families, cyphophthalmids are a fairly conservative bunch in appearance - the photo above is of Pettalus cf. cimiciformis*, but it is fairly typical of the group as a whole. They are quite distinct from other harvestmen (in fact, it is generally agreed that they are the sister-group to all others), and rather than having the spindly build of more familiar members of the order, cyphophthalmids are minute, stocky armoured tanks. If you look closely at the picture above, you may see a light spot on either side of the body that looks a bit like an eye, but it is in fact an ozophore - a raised mound bearing the opening of a stink gland. Except for members of the family Stylocellidae, cyphophthalmids have been described in the past as eyeless, but SEM studies of Pettalidae have revealed minute (often lens-less) eyes hidden on the side of the ozophore (Boyer & Giribet, 2007).

*For those who aren't already in the know, the 'cf.' in the name stands for the Latin confer (compare). In this case, it indicates that the animal in question is very similar to Pettalus cimiciformis, but is not definitely a member of that species.

The really interesting thing about cyphophthalmids (beyond their own intrinsic charm, of course) is their distribution patterns. Each of the various families has a definite, disjunct distribution (Boyer et al., 2007). The family Stylocellidae are restricted to south-east Asia. The Sironidae are found in what once was Laurasia - Eurasia and North America. The Neogoveidae are found in Florida, tropical South America and tropical West Africa - the tropical parts of what once was Gondwana. Two genera placed in their own families, Ogovea and Troglosiro, are found in West Africa and New Caledonia, respectively. And Pettalidae has a classic Gondwanan distribution, found in southern South America, southern Africa (including Madagascar), Sri Lanka, Australia and New Zealand (see Carl Zimmer's post for a map).

I think I should say something here about "Gondwanan" distributions. Science has a tendency to go through fads like any other aspect of human culture. For many years, most organisms showing what we would now call a "Gondwanan" distribution were interpreted as relicts of a former world-wide distribution. As acceptance of "continental drift" and recognition of the previous existence of Gondwana increased, more and more researchers considered its potential significance for modern biogeography. Needless to say, the significance was especially apparent to workers in the southern continents, doubtless not without some aspect of asserting the importance of the all-too-often neglected Southern Hemisphere biota relative to the Northern Hemisphere. Gondwanan origins became the next big thing for everything from birds (Cracraft, 2001) to beeches (Linder & Crisp, 1995) to butterflies (de Jong, 2003). In the last few years, the pendulum has begun to sway the other way, probably towards a more reasonable median.

The idea of a Gondwanan distribution for a given group of harvestmen particularly merits a critical look. The fossil record of harvestmen is pretty abysmal relative to the age of the group, but what record there is speaks of a remarkable degree of morphological conservatism. The Carboniferous long-legged harvestman Brigantibunum is almost indistinguishable from modern taxa (Dunlop & Anderson, 2005). The cyphophthalmid Siro platypedibus Dunlop & Giribet, 2003, from Bitterfeld amber (probably Oligocene or Miocene in age) is so similar to modern species that it is included in a modern genus.

In order to test whether the distribution of Pettalidae is an actual Gondwanan distribution as opposed to a relictual one, Boyer et al. (2007) tested the phylogeny of the family with just about every morphological and molecular method imaginable. They demonstrated that most of the cyphophthalmid families were monophyletic, with distribution matching phylogeny (the exception was the Laurasian Sironidae, which came out paraphyletic to the northern Gondwanan Neogoveidae and Stylocellidae).

To add another level of interest to the whole deal, most of the genera within Pettalidae each have separate geographic distributions (Boyer & Giribet, 2007). Chileogovea in South America, Purcellia and Parapurcellia in southern Africa, Pettalus in Sri Lanka, Karripurcellia in Western Australia, Austropurcellia in eastern Australia. The exception is New Zealand. New Zealand has a remarkable diversity of Pettalidae, with more described species than everywhere else combined, in three genera. But let's look a little closer. In the South Island of New Zealand, the genus Rakaia is concentrated in the east, while the genus Aoraki is found in the west from Mount Cook* north. New Zealand actually lies on the boundary between the Indo-Australian and Pacific plates, and if you were to map the distributions of the genera, you would see that Rakaia is mostly found on the Pacific plate, while Aoraki dominates on the Indo-Australian!**

*The Maori name for which just happens to be Aoraki. Not a coincidence - the genus was named after the mountain.

**I know, I said three genera. The third genus is a single species, Neopurcellia salmoni, in the southwest of the South Island.

REFERENCES

Boyer, S. L., R. M. Clouse, L. R. Benavides, P. Sharma, P. J. Schwendinger, I. Karunarathna & G. Giribet. 2007. Biogeography of the world: a case study from cyphophthalmid Opiliones, a globally distributed group of arachnids. Journal of Biogeography, in press.

Boyer, S. L., & G. Giribet. 2007. A new model Gondwanan taxon: systematics and biogeography of the harvestman family Pettalidae (Arachnida, Opiliones, Cyphophthalmi), with a taxonomic revision of genera from Australia and New Zealand. Cladistics 23: 337-361.

Cracraft, J. 2001. Avian evolution, Gondwana biogeography and the Cretaceous-Tertiary mass extinction event. Proceedings of the Royal Society of London Series B – Biological Sciences 268: 459-469.

Dunlop, J. A., & L. I. Anderson. 2005. A fossil harvestman (Arachnida, Opiliones) from the Mississippian of East Kirkton, Scotland. Journal of Arachnology 33: 482-489.

Dunlop, J. A., & G. Giribet. 2003. The first fossil cyphophthalmid (Arachnida, Opiliones) from Bitterfeld amber, Germany. Journal of Arachnology 31: 371-378.

Jong, R. de. 2003. Are there butterflies with Gondwanan ancestry in the Australian region? Invertebrate Systematics 17: 143-156.

Linder, H. P., & M. D. Crisp. 1995. Nothofagus and Pacific biogeography. Cladistics 11: 5-32.

What is a Daddy-Longlegs?

"Daddy-longlegs" is one of the worst animal names there is. The name is widely used and generally recognised, but causes endless confusion because it is applied to no less than three very different animals. In a brief attempt at disambiguation, these are the animals involved:

First are harvestmen of the order Opiliones (picture from UMMZ). Harvestmen are often confused with spiders, but the body is not divided into a cephalothorax and abdomen, the opisthosoma (the posterior part of the body corresponding to the abdomen) is externally segmented, the chelicerae (mouthparts) are pincers rather than fangs, and harvestmen do not produce silk. The name "daddy-longlegs" as applied to harvestmen usually refers to the group known as "long-legged harvestmen" (Palpatores). There is some uncertainty about whether Palpatores are a monophyletic group, but that's a subject for another time.

Second are spiders of the family Pholcidae (picture is from Iziko Museums of Cape Town - the object the spider is holding is the egg-sac, which is carried by the female until the eggs hatch). Pholcids are true spiders, and so have a divided body, an unsegmented abdomen, fangs, and produce silk. Here in Australia and New Zealand, the 'daddy-longlegs' that are almost ubiquitously found in houses (particularly bathrooms) are pholcids, most often the introduced Pholcus phalangioides. Offhand, there is a common belief that daddy-longlegs (either pholcids or Opiliones) are "the most poisonous spiders in existence, but their fangs are too small to pierce human skin". I have come across this story many times, and have even been assured of it by people who really should know better. This story is absolute bunkum. The University of California, Riverside site has more info.

Finally, the third group accused of being 'daddy-longlegs' are crane flies of the family Tipulidae (picture from Wikipedia). Crane flies look a bit like giant mosquitoes, but they are not blood-suckers. They are a large family - the adults are nectarivores or do not feed, while the larvae, commonly called leatherjackets, feed on vegetation. Crane flies are easily distinguished from the other 'daddy-longlegs' - the wings are a bit of a give-away.