MYRIAPODOLOGICA Virginia Museum of Natural History Vol. 5. No. 5 ISSN 0163-5395 1 March 1998 Studies on spiroboloid millipeds. XIX. Thyroproctus, an exceptional genus in the Rhinocricidae. By Richard L. Hoffman ABSTRACT A brief taxonomic review of Thyroprocius w'iih redescnptions and illustrations of the two included species, both endemic on Jamaica. Oespite the similarity of the genitalia to tliose of most other American species of Rlnnocncidae, the genus is provisionally kepi sepai'ate on the basis of the highlv' modified hypoproci to avoid e.xlension of the name Thyroproctus to include the numerous species cuiTently refeijed to Anadenobolus . PREFACE Species of the pantropical family Rhinocricidae have indulged m about as man>' modifications of basic peripheral structure as possible: great variation in the size and shape of antennae and the number of apical sensory cones; presence or absence of scobinae and, if present, their size, shape, and distribution; number of body rings; spiniform projection of the epiproct; presence of dorsal spines on the paraprocts; presence or absence of ventral tarsal pads m males; modification of anterior male coxae. By contrast, there is relatively little modification in basic gonopod pattern, and when it does occur, the greatest diversity' seems centered in the West Indies. What IS surely the most extreme body specialization so far known in the family is encapsulated in the second oldest generic name based on a rhinocricid; Thyroproctus Pocock, 1894. Derived from the Greek elements thyreos (large, oblong shield) and procios (anus), the name alludes to the shape of the hypoproci. precisely described by Pocock as “very large, thickened, and transversely elongate.’' The genus was based on a new species, townsendi, knowm from a sample taken at an unspecified locality on Jamaica. The original description was concise and very Scanned with permission by Virginia Tech Insect Systematics Group 2014 (www.jointedlegs.org) 56 Myriapodologica brief, with several drawings of the posterior end made with low magnification. Subsequently - two decades later - R. V. Chamberlin (1918) described a second species, ThyroproctiiS cinchoniamis, from a small series taken m the Blue Mountains by C. T. Brucs. The description of this species was fairly detailed, but without illustrations. Since more recent papers about Jamaican millipeds (e g.. Loomis, 1975) did not report any new material of either species, our knowledge of the genus has not materially advanced beyond what Pocock knew m 1894. The genus w'as not mentioned in Brolemann's essay on spiroboloid classification (1914). It was how'ever admitted to the Rliinocricidae in Attems' survey of Indoaustralian myriapods (1914) and his overall classification of 1926, in both of wiiich the modified hypoproct was used as the generic criterion. Ill 1952, I received an extensive collection of Jamaican millipeds from the Science Museum of the Institute of Jamaica, which included material of both T. townsendi and cmchonianus. Descriptions and drawings of both w'ere prepared for a planned regional s\iiopsis of the island's milliped fauna, an enterprise which eventually went aground on the reef of rhinocricid taxonomy and w'as ne\'er completed. Most of the identified material was returned to the Institute and my interest in Thyroproctus went into an eclipse. Incidentally, during a short visit to Jamaica in 1977, I picked up an adult male of w'hat appears to be townsendi and fortuitously had material of Thyroproctus available w'hen its generic status came into question in connection with another project. The identitV' of Rhmocricus itself, another enigma of long tenure, was settled fairly conclusively by me in I960, when the name w^as restricted to the type species pa reus Karsch of Porto Rico and several apparently congeneric species in Cuba, This exclusion had the effect of orphaning the great majority of species which had been described as Rhmocricus'" and never transferred to any of the scattering of allied genera (e.g., Eurhinocricus, Argentocricus) that are currently considered valid. How to accomodate these homeless species is a difficult problem, already mentioned in my ‘"Classification” (1980: 76). As an interim solution, Mauries (1980: 1088) has adopted the third oldest rhinocricid generic name, Anadenobolus Silvestri, 1897, to encompass at least the Antillean species that were displaced by the 1960 fixation oi Rhinocricus . The most vexing problem has not been settled, how-ever. In virtually all other heiniinthomorphous millipeds, genera have been defined almost exclusively on characters of the male genitalia. Since these appendages tend to observe a uniform basic configuration throughout the Rhmocricidae, various authors have resorted to every imaginable combination of the structural variables mentioned m the first sentence (as w'ell as some rather negligible gonopodal differences) to achieve something like generic groupings. The phylogenetic artificiality' of these menages IS not difficult to contemplate. But not even gonopodally-based groups are quite satisfactory'. Are the “simple” attenuated posterior gonopods that appear in the West Indies {Rhmocricus), South America {Andocricus), and Papuasia {Acladocricus) the descendants of a common ancestor, or homoplasic reduction products? Other characters do not offer much of an insight into this problem. Hoffman : 1 hyroproctus 51 The gonopods in Thyroproctus (Figs. 1. 2) are notable for their unn^ersality: the\' could without any reservation be regarded as strictly congeneric with the species from Guadeloupe that Maurics considered to be Porat's politus, the type species of Afiadenobolus.U rhinocricid genera are to be gonopod^based, Thyro- proctus IS tlie correct genenc name for the majority' of New World rhinocricids. Looking at the posterior end of a specimen of townsendi, however, may engender some hesitancy. Here is a structural departure of considerable magnitude, not approached anywhere else in the family or order (Figs. 3, 4). Perhaps it has greater systematic importance than mere nmnbers of antemial sensoria, or dorsalK acuminate paraprocts Perhaps an e.\ception, a pureh’ subjective dispensation, can be justified to maintain this small Jamaican endemic group as an entity unto itself Strict observation of “gonopod genera"" is an option with its own built-in set of problems. In the belief that only a consensus of concerned investigators will solve the dilemma, I propose here to take the stand that Thyroprocius merits special consideration as a valid local genus, the recognition of which will not pre-empt the validity of other Antillean genera with similar gonopods. For the benefit of present and future colleagues I provide here a se\ erc]y condensed version of nw' original accounts of the genus and its two species, based on Jamaican material. Such information is in any e^'enl necessary for e\ aluating the zoological and nomenclatorial status of these animals. Thyroproctus. Thyroproctus Pocock. 1894, Joum. Linnean Soc. London, 24: 506. Type species: T. townsendi Pocock. b>' monotypy, Diagnosis:A rhinocricid genus characterized by modification of the temunal segment: Iwpoproct enlarged, incrassate. transverscl)' elongated, preceded by a prominent deep lrans\erse groove: paraprocts flat, without niargination, ly ing at nearly a right angle to median body axis: epiproct not distomcdially produced, its edge nearly transverse (Figs. 1, 2). Antennae with four apical sensory cones. Segments with two transverse sulci, the posterior indistinct, anterior prominent and deeph’ impressed entirely around the segment, curv'ed fonvard in front of ozopores. Gonopods similar to those of most other rhinocricid genera (Figs. 3, 4), the telopodite of the posterior gonopods distally laminate and broadened beyond midlcngth: solenomere nearly as long as acropodite. Distribution: The genus is endemic to the eastern third of Jamaica, Remarks: Thyroproctus is one of the most disjunct of all spiroboloid genera in the modification of the posteriormost segments. The two species are very similar in terms of gonopod structure, but differ substantially in a number of peripheral characters to be assurred of full specific status. Although considerable work remains to be done in defining their ranges, it appears at present that cinchonianus IS more partial to higher elevations in the Blue Mountains north of Kingston, while townsendi occurs at lower elevations further to the east. 58 Myriapodologica Two interesting cases of mimicr>' may be noted; Anadenobolus newtomanns and Thyroproctus cinchomanus have similar striped color patterns and appear to be sympatnc: A. excisiis and T. townsendi are both unifonn blackish and occur together at the easternmost part of the island. It IS also noteworthy that relativch extensive collecting throughout Jamaica in recent years has not disclosed additional species of this genus (e g.. Loomis, 1975). KEY ^rO THE SPECIES OF THYROPROC TUS Uniform black; metazona of each body segment elevated and imparting a distinctly monilifomi or annular appearance; scobinae present; adults longer thn 35 mm townsendi Body dark laterally, dorsum lighter witli black middorsal stripe: metazona not elevated, the body therefore parallel-sided; scobinae absent; adults less than 35 mm long cinchonianus Thyroproctus townsendi Pocock Figures 1-4 Thyroproctus townsendi Pocock, 1894, Joum, Linnean Soc. London, 24: 506, pi, 37, figs. 6, a-c. HoloUpe o' (Brit. Mus. Nat. Hist.) labeled only ‘Uamaica^^ (T. Towmsend). Diagnosis; The unifonn black color, annulate appearance of the segments, and presence of scobinae readily distinguish this species. Description (specimens from Com Puss Gap, St. Thomas): Color entirely dark brownish-black to piceous; legs and antennae tan; front of head light brown with median frontal black mark. Body stout, W/L ratio about 1 1%; head and collum much wider than succeeding segments, body otherwise nearly equidiametric back to last several segments. Front of head generally rounded and smootli, slightly flattened between antennae and minutely papillose in males, finely striate transverse in both sexes. Clypeal groove distinct. Labral setae varying from 6-6 (male) to 10-10 (female). Head strongly margined below antennae; latter small and stout, not extending back beyond middle of collum, 6th article largest, 4th and 5th somewiiat smaller than proximal three; antennomercs almost glabrous, only a few' scattered setae present except on distal half of 6th and over entire 7th. Four apical sensory cones. Front surface of antennal socket transverse striate. Ocellaria normal in size and shape, with about 25 ocelli m six row^s. Collum smooth, broadly convex, appearing slightly inflated, lateral ends rounded, with w'dl-defined marginal sulci extending up to level of middle of ocellaria. Second segment prolonged cephaloventrad well beyond middle of ends of collum, but not deeply excavated; surface of lateral ends with a few sharply Hoffman: Thyroproctus 59 P’IGS. 1-4. Thyroproctus townsendi Focock. 1, posterior end of body, lateral aspect. 2, postenormost segments, ventral aspect. 3, anterior gonopods, antenor aspect. 4, left posterior gonopod. mesal aspect. Specimen from Bath Spa, St. Thomas Pansh, Jamaica 60 Myriapodologi ca defined longitudinal striae. Succeeding segments finely shagreened, othenvise smooth: primary^ sulcus indistinct. Metazona strongly elevated, imparting an evident annular appearance. Secondaiy sulcus deeply impressed entirely around segment, cure ed forward m front of ozopores. Lower sides of segments with a few indistinct oblique striae. Epiproct veiy short, thickened, broadly truncated, equalling but not exceeding paraprocts, latter small, flat, almost transverse. H>poproct greatly enlarged, Lransversly elongated, pi eceeded by a very deep groove. Sternal aiea of penultimate segment wnth large conspicuous median projection. Legs short, stout, podomeres in order of decreasing length, 2-3-6-4-1-5, tarsal claw long and slender; ventral tarsal pads present in males. First tw^o pairs of legs of males reduced m size, somewhat incrassate. Coxae of anterior legs without lobes or other modification. Gonopods of the form shown m figures 4 and 5 . Scobinae present from 8th to about 2 1 st segment, small, transversely elongated, about five or six times as wide as long, separated by about tliree times a scobinal width. Scobinellae rounded, finely striate, about 1.5-2 diameters apart, present back to 28th segment. Variation: In the series from Com Puss Gap, three females range from 30.3 to 37.8 mm in length, with 40, 40, and 42 segments: the single male is 37.0 mm long, with 40 segments. In specimens from Priestman's Ri\ er, six males have 39-41 segments (avg. 40.3) and four females 38-41 (avg. 39.7). Variation in segment number appears to be minimal. Distribution; Eastern part of the Blue Mountains, probably also the John Crow- Mountains. The following 16 specimens w^ere examined: Sr. Thomas Parish: Corn Puss Gap, north of Bath, \ &2 ? ,25 August 1952, R. P, Bengiy. Portland Parish: Priestman’s River, 7cf, 49, 27 November 1951, G. R. Proctor & R.L. Hoffman. 5 miles w^est of Priestman’s River, 19, 5-7 Februaiy 1952, C. B. Lewis. Bath Spa, lo". 5 Febmarv' 1977, R, L. Hoffman. Thyroproctus cinchonianus Chamberlin Thyroproctus cinchonianus Chamberlin, 1918, Bull. Mus. Comp. Zoof, 62: 205, Male holotype (MCZ) from Cinchona (Plantation), St. Andrew Parish, Jamaica (C. T. Bmes). Diagnosis: Readily separated from townsendi by the attractive lineate color pattern, smooth-sided body, absence of scobinae, and much smaller size. Description (specimen from Clydesdale, St. Andrew Parish): A generally pale species with dark sides and a black middorsal stripe bordered each side by a light band. Under magnification, each segment shows the metazona to be mostly yellow'-, the mesozona dark up to level of ozopores, dorsad to which is it yellowish white except for a broad cmciform black mark with its base pointed caudad and apex extended anteriad onto prozona. Front of head pale brown with a black epicranial Hoffman; Thyroprociiis 61 mark, diverging ventrally around median clypeal suture; a black margin above each antennal socket; eyes, ffont edge of collum, middorsal stripe, and caudal half of epiproct black. Legs and antennae yellow. Structurally very similar to umnsendi except that meta/:ona are not elevated above level of mesozona, and no scobinae can be detected. Small rounded scobincllae are present back to the 29th segment. V.AJOATiON: Eight specimens (itemized below) appear to show no geographic variation. Four males vary’ from 25.8 to 32. 1 mm m length, with 37-39 segments. Four females range from 24.3 to 28.6 mm in length, with 35-4 1 segments. DisTRiBi TioN; Western and middle parts of the Blue Mountains. The following nine specimens have been examined: St. Andrew Parish: Clydesdale, 19,19 August 1949, R. P Bengry: Shooters Hill. 29 9 {date lost). G. R. Thompson, Newcastle Road, 8 December 1946, Thompson. St. Thomas: ravine near Abbey Green, 1 o". 19. 16 December 1951, G. R, Proctor; W'hitfield Hall, lo", 13 Mav 1950, Bengiy Portland Parish: Haycock Mountain, above Balcarres P. O., 19,11 June 1953, Proctor. RliFERENCES Chamberlin. R. V. 1918. The Chilopoda and Diplopoda of the West Indies, Bull, Mus. Comp. Zool.. 52; 151-262. Hoffman. R L, 1960 'Studies on spiroboloid millipeds. V.The correct identity of the genus Rhinocncus, based upon a study of its type species. Proc. Biol. Soc. Washington, 73; 5-14, Hoffman, R, L. 1980. Classification of the Diplopoda. Geneve, Museum d' Histoirc Naturelle, 236 pp. Loomis, H. F. 1975. New millipeds in a noteworthy collection from Jamaica. Florida Entom., 58; 167-185. Mauries, J.-P. 1980. Diplopodes Chilognathes de la Guadeloupe et ses dependances. Bull. Mus. Natn. Hist. Nat. (Pans), (4) 2 (A4): 1059-1121. Pocock, R. 1. 1894 Contributions to our knowledge of the artiiropod fauna of the West Indies. Part III. Diplopoda and Malacopoda, wJth a supplement on the Arachnida of the class Pedipalpi. Joum. Linnean Soc. London. 24; 473-544 Address of the author; Dr. Richard L, Hoffman Virginia Museum of Natural History Martinsville, Virginia, 24112, LISA