Platythyrea parallela

Distribution
This taxon was described from Indonesia (Aru Island). It is also found in the Philippines, Mozambique, New Caledonia, New Guinea, India, Seychelles, Sri Lanka, Bangladesh, Viet Nam, Singapore, Malaysia, Fiji, Guam, Samoa, Micronesia, Northern Mariana Islands, and Australia.

Description
Brown (1975) provided the following:

Wilson (1958:151) began to assemble the synonymy of P. parallela during his study of the Melanesian ponerines, when he sank under it P. pusilla, P. coxalis, P. melancholica, and P. melancholica var. aruana, the last two names provisionally. P. melancholica originally was vaguely described by F. Smith in 1865 as a Pachycondyla, but Emery (1897: 556) surmised that the species was a Platythyrea, a placement in which he was followed by all later authors when the type of P. melancholica could not be located. However, the type is extant in the Oxford University Collection, where it was found by Barry Bolton, and I was able to examine it in London in 1973. It proves to be true Pachycondyla in the broad sense (Ectomomyrmex of recent authors), and so the species melancholica can be eliminated from consideration under Platythyrea. It will be dealt with in Part VI of this series.

Wilson was also able to compare digms of the parallela complex that I provided for him in 1955 with types of other related forms in European museums, and I have been able to study these and many others abroad and in MCZ in subsequent years, with the result that the synonymy of P. parallela is lengthened to include a surprising number of specific, subspecific, and varietal names.

As might be suspected from the length of the list of synonyms, P. parallela is a variable insect. The most conspicuous variation is seen in overall body size, as indicated by the length of the trunk (WL), which ranges from about 1.10 to about 2.50 mm in specimens representing more than 50 nest series assigned to this species that I have been able to examine carefully enough. In discussing individual synonyms below, I have given WL for at least one type specimen wherever I was able to make a suitable measurement. I am unable to split the size continuum into modal classes that seem to represent different species, and small, medium, and large forms are known to occur sympatrically at such relatively well sampled localities as the Cuernos Mts., near Dumaguete, Negros Oriental, Philippines, and on Ceylon. Within single nest series, however, size variation so far as known is relatively narrow in range.

Other variation, as noted by Wilson, occurs in the shape and size of the propodeal angles; in the length vs. width, dorsal convexity, and shape (of the posterodorsal margin) of the petiolar node; in the opacity of the background fine sculpture; in the number and distinctness of the superimposed coarser punctures; in the color of the legs, and other lesser traits. Some preliminary analysis of the length vs. width of the petiolar node by Wilson (1958:152) showed that the larger specimens tend to have disproportionately wider nodes. A rough check indicates that the additional data from the augmented series now available would probably widen Wilson's already fairly wide regression zone, but that the trend would remain the same. On the other hand, the sculptural variation does not seem to be closely correlated with size.

If more than one species is present in this complex, it will take much more refined analysis based on more and better material to reveal it. For the present, the burden of proof is shifted to the splitters, because I am considering the whole complex to represent a single species. Such a treatment only parallels the conclusions reached for other Platythyrea species in this study. My notes on the individual synonyms are repeated here, in the approximate chronological order of publication.

P. parallela type WL 2.15 mm.; tibiae dark. Sculpture weak, fairly smooth; larger punctures weak, obsolete on pronotum; tibiae dark. Type loc.: Aru Island.

P. inconspicua type: dealate queen, tibiae dark brown, WL 2.45 mm., is near the large end of the size range. On my request, Dr. Max Fischer of the Naturhistorisches Museum in Vienna kindly sent 2 specimens, each placed in the Mayr Collection as "inconspicua/G. Mayr Typus." Emery mentions these specimens as the two possessed by Mayr, who identified South American samples sent to him by Emery as P. inconspicua. One of Mayr's specimens now bears the tiny locality label, "Ceilan/Sichel," and on a larger label below this, apparently in Mayr's handwriting, "Cayenne/wolnicht/Ceilon." The writing is crude and was obviously done in haste; apparently Mayr decided, at some time after the original description, that the specimen was "wohl nicht" (probably not) from Ceylon, and its superficial similarity to the second "Typus," from Panama ("Pergand") further led him to believe that the locality of the first type should have been Cayenne. From this point, it was an easy step to identify Emery's specimens also as inconspicua.

A close examination of the original (and only real) type of P. inconspicua shows that it is really an Old World tropical member of the genus, and in fact is a large specimen of P. parallela. In the first place, the masticatory borders of the mandibles are clearly denticulate or crenulate (in the American species, the border is toothless and cultrate). In the second place, the palpi are very short, and are segmented 3,2, as in other P. parallela examined, large and small. Mayr's Panama specimen, on the other hand, has much longer palpi (extending past the posterior margin of the buccal cavity), and the maxillary palpi are 6 merous, the labials 4 merous, as in other members of the punctata group. The Panama specimen is in fact a specimen of P. pilosula, and cannot be considered as in any way a type of P. inconspicua; it was never mentioned in the original description because it had not even been collected by then.

P. inconspicua is therefore removed from the synonymy of P. punctata, where Emery had assigned it, and is placed as a new synonym of parallela.

P. pusilla and P. coxalis were described in the same paper at a time when the correct ,generic placement of Ponera parallela F. Smith was still unknown, and by this time, Emery had already considered P. inconspicua to be a neotropical species (see above). Wilson (1968:151) examined the holotypes of Emery's two species in 1955 and left type compared digms in MCZ, one of which also compared with P. parallela, and the other with P. wroughtonii, its subspecies victoriae, and P. parva. His synonymy of pusilla and coxalis under parallela (1958) seems to me to be correct. The type locality of P. pusilla is Amboina, and for P. coxalis is Perak, Malaya.

The types of P. wroughtonii and P. wroughtonii victoriae were examined in the Forel Collection in Geneva. WL of a wroughtonii specimen was 1.36 mm, and victoriae is a little larger (MCZ syntype WL 1.40 mm). The differences cited by Forel to separate these forms are very vague, and I was unable to appreciate them in the type specimens; these both seem to me to be small sized parallela. The type localities for wroughtonii are Travancore and Mysore (S peninsular India); for victoriae, they are Bangalore (in Mysore), Barrackpore (N of Calcutta), and "Inde septentrionale." Both forms have the tibiae light yellow.

The types of P. subtilis and P. pusilla were compared directly in the Emery Collection (MCSN, Genoa), and I could find no important differences between them; the subtilis type is perhaps a little more heavily punctate on the sides of the propodeum and petiolar node. A syntype specimen of P. wroughtonii from "Bangalore" in the Emery Collection compares well with these types also. P. subtilis type loc.: Pulau Laut, SE Borneo (Kalimantan).

In the Kodeco Tract west of Batulitjin (SE Kalimantan), inland from Pulau Laut, in 1972, I collected several samples of the P. parallela complex, including workers of fairly large size (WL 1.81 1.90 mm) and one very small stray worker, the smallest I have seen (WL 1.11 mm), with a fairly short, heavily punctured node with distinct blunt teeth behind. While on the subject of size, I should also mention a single, very large, worker individual (WL 2.50 mm) taken by C. F. Clagg at 5000 6000 feet elevation on Mt. Apo, Mindanao. It is hard to believe that this relative giant is simply a large specimen of P. parallela, but it is almost exactly like that species; its eyes are a bit small (greatest diameter 0.13 mm) for such a large specimen, and its second funicular segment is about 1 1/2 times as long as broad (usually about as broad as long in parallela), but in other characters, including 3,2 palpal segmentation, it fits P. parallela.

P. coxalis var. tritschleri is a variety intermediate in size between the types of P. parallela and P. wroughtonii with the lobules of the posterior border of the petiolar node almost completely obsolete. Type locality Indrapura, Sumatra (MHN  Geneva).

P. coxalis var. javana is similar to the foregoing in size, but the tendency to form 3 low lobes on the posterior border of the petiole is better expressed. In javana, the larger punctures are weak and difficult to see. Type locality is Buitenzorg (now Bogor), Java. Both of the above varieties have reddish brown tibiae (MHNG).

P. coxalis var. annamita is a little larger than the foregoing (WL 2.12 mm), but seems to belong to the same species. Sculpture fairly weak; tibiae brown. Type loc.: Region of the Mois, central Cochin China. (MHNG).

P. wroughtonii subsp. sechellensis has WL 1.55 mm. It has weak, feebly shining sculpture, with small, widely spaced punctures; tibiae yellow. Type loc.: Praslin, Seychelles Is. (BMNH London.)

P. coxalis var. cylindrica is a medium sized queen (WL 1.67 mm) with larger punctures distinct, but reduced in size on pronotum; tibiae brown. Type loc.: Tandjong Slamat, E Sumatra. (MNHU Berlin.)

P. pusilla var. australis is a small form (not measured) similar to wroughtonii types; tibiae with yellowish extensor surfaces. Type loc.: Blackall Range and Tambourine Mt., S Queensland. (MHN Geneva.)

Platythyrea parva is like the wroughtonii types, but head a bit broader; tibiae with extensor surfaces lightened and yellowish; WL 1.34 and 1.40 on holotype and paratype measured. Type loc.: Darwin, N Australia. (BMNH London, MHNG and MCZC.)

P. pusilla var. egena: WL 1.51 mm; tibiae more or less pale. Type loc.: Singapore. (MNHU Berlin.)

P. coxalis var. philippinensis: WL 1.75, 1.80 mm; tibiae brown. Type loc.: Los Bafios, Laguna Prov., S Luzon (MNHU Berlin, MCZ). Dr. J. W. Chapman assembled large series of P. parallela in the Philippines, although many of the specimens were molded or otherwise damaged during World War II. Most of his samples come from the Cuernos Mts. near Dumaguete, Negros Oriental, 1500 to 4000 feet elevation. These vary somewhat in body size and have the legs brown, reddish brown, or with tibiae somewhat lightened in color. Other Philippine samples are from Luzon: Baguio (Chapman); Los Bafios (F. X. Williams); Mt. Makiling (A. L. Cambre); Mindanao: Davao City (A. Reyes), Momungan (no collector), Miatan, Katipunan (D. Empeso); Romblon, Cabrador I. (L. Morato). Borneo: Kuching, Sarawak, (Hewitt); Sandakan (C. F. Baker); Mt. Dulit, N. Borneo, 3000 ft (E. Mjoberg). Gilolo I.: Kalam (C. S. Banks). S. Celebes: Balampesoang Forest, near Tanete (W. L. Brown).

P. cephalotes: WL 2.00 mm; tibiae pale. Weakly shining in most lights, but some small punctures extend onto pronotum. Petiolar node L 0.63, W 0.60 mm, posterodorsal border of node sinuate as seen from above. Eyes a little smaller than in New Guinea parallela. Type loc.: "Walde be.i Trial Bay," New South Wales, Australia. (MNHU, Berlin.) Additional material from N.S.W.: Jenolan Caves (J. C. Wiburd), Bulga (W. W. Froggatt), Dorrigo (W. Heron), Royal National Park, 25 m elevation (E. S. Ross and D. Q. Cavagnaro), as well as Brisbane, Queensland (H. Hacker) is much like the cephalotes type. Smaller specimens come from the McIlwraith Range, Cape York Peninsula, Queensland (P. J. Darlington).

P. melancholica var. aruana was synonymized with parallela by Wilson provisionally because the type was unavailable. The type locality is Wammar in the Aru Archipelago, and Aru is the type locality of ''parallela'. Upon reading Karavaiev's description again, I can only conclude that Wilson's synonymy is reasonable even in the absence of the aruana type.

P. pulchella: the type in NM Basel is an alate queen from Medan, Sumatra (Corporaal). Node rather short, only slightly longer than broad, but with a sinuate posterior margin as seen from above. Scapes and tibiae whitish, prominent against black of remainder of body.

P. pusilla var. pacifica was put into the synonymy of P. parallela by Wilson and Taylor (1967:20), with good reason in my opinion. Type loc.: Apia, Upolu, Samoa.

P. ceylonensis: WL 1.86 mm; tibiae brown. Type loc.: Ceylon. (BMNH London.) This last synonym in the long list is just a fairly large parallela queen and workers. Other specimens from Ceylon are from Yakkala (KLA Perera), and Soans and I took a series of slightly larger body size (WL 1.70 1.74 mm) in Kerala State, S India: Peria Reserve, Cannanore Dist., 4 April 1969, tropical evergreen forest, ca. 900 m. Other samples from the SE Asian mainland are: Thailand: E slope, Doi Sutep, 260 m, 15 July 1962 (WL 1.47 mm) and Khao Yai National Park, 750 m, 26 July 1962 (WL 2.06 mm), E. S. Ross and D. Q Cavagnaro. Malaysia: Pahang, below The Gap, about 850 m (R. H. Crozier), in hill forest. The MCZ also has a sample from Wai Lima, Lampong, S. Sumatra (Karny).