A REVIEW OF THE NEW WORLD ONYCHOPHORA WITH THE DESCRIPTION OF A NEW CAVERNICOLOUS GENUS AND SPECIES FROM JAMAICA*

Few animals are as avidly sought and as. rarely found by invertebrate zoologists in the American tropics as are onychophorans. Yet, attempting to determine any specimens which one is fortunate enough to find can be a frustrating experience because o.f the scat-tered and incomplete literature.


Psyche
September-D ecember (946) gave a key for species known from the Canal Zone and adjoining regions.
Distribution has been discussed by many, including Brinck (I956), Brues (923) and Clark (913a,93b,95a,b). The circumtropical distribution of the amily Peripatidae and the generally south temperate distribution of the family Peripatopsidae have led several authors to speculate that these distributions have been achieved through a Gondwana or continental drift mechanism (for instance, Cunot, 949; Hutchinson, 928). However, although continental drift may be a act, onychophoran distributions have not necessarily been controlled by it alone, and active, dynamic biological t:actors were probably more important in forming the distributions (Brinck,956 Darlingt.on, 965).
Since the summary papers by Bouvier and Clark, several species have been described and others have been placed in synonymy. The following key to the genera and list of species unites this dispersed literature and hopefully will provide a basis for encouraging further studies.
Key to the families and genera of New World onychophorans a. Sexual opening between last pair of legs ( indenting proximal margin f third creeping pad band on leg, usually dividing third pad into two segments (Fig. :3); four or five creeping pads; our oot papillae, 2 on each side of oot Oroperipatus 2b. Not the above combination; nephridial tubercles of ourth and fifth legs perhaps connected to third pad and indenting its proximal margin, but opening in space between third and fourth creeping pads, usually indenting fourth pad or rarely dividing fourth pad into two segments (Figs. 4, 7) usually four (rarely five or three) creeping pads (Fig. 4)  In the following list, references to Bouvier (19o5, I9o7) and Clark I913a, I913b, I9I 5a) are not individually given because they apply to almost all species. Also, dates and literature citations to descriptions o.f Bouvier (I9o5, I9o7) and earlier are not given, because they may be found in these works. The word "seen" indicates specimens from the Museum of Comparative Zoology and the Institute of Jamaica on which some data .are based. All new records reported here are based on specimens of the. Museum of Comparative Zoology. A history of synonymy of each species is not given because the intention of the list is only to. report presently accepted species and their documented localities.  Brues, 1925;Clark, 1913b;Clark and Zetek, 1946. M. guianensis (Evans). Guyana: east bank o Demerara River. 1975] Peck---New World OnychoIhora 347 M. insularis insularis 'Clark,x937. Haiti: between Jacmel and Tronin.
M. ohausi (Bouvier). Brazil: near Rio de' Janeiro; Petr6polis. M. kerrieri (Bouvier). Mexico (Bouvier). Ecuador Durn, on Guayas River. O. bimbergi (Fuhrmann), 9 3. Colombia: Amagatal, 9oo-8oo rn; Guaduas, 8oo m. Fuhrmann,914. O. bluntschli Fuhrmann,9'5 (Fig. 8). Legs and "antennae" not appreciably elongated. Ventral surface of left fourth leg of Oroleri#atus koe#ckei (after Zilch, 1954b) showing urinary papilla dividing the third band of the creeping pads, and four foot papillae (fp). 4. Ventral side of right fourth leg of Heteroleri/mtus engelhardi (after Zilch, 1954a) showing urinary papilla between creeping pads three and four, indenting pad four, and three anterior and one posterior foot papillae (fp). 5. Outer blade of right mandible of Sleleo#eri/mtus s#elaeus. 6. Inner blade of right mandible of . selaeus. long, with 23 legs on the left side and 22 on the right; a second 34 mm long with 22 pairs o.f legs, and the third 34 mm long with 23 pairs of legs. Widths 2 to 3 mm depending on preservation. All legs with four creeping pads; pad three broadly .connected to urinary papillae in legs four and five; pad four crescent-shaped; all eet with two anterior and one posterior papilla (Fig. 7)-Psyche [September-December Each body segment with 12 skin (olds; olds neither diverging nor i:using along lateral side above the legs. The skin papillae with round bases but o varying sizes and shapes; lateral papillae with more elongate and straight-sided tip, tips shorter and more rounded on dorsal papillae; one to. three but usually two. small papillae between major papillae. Mandible (Figs. 5,6)  Cave adaptation. The eyeless and depigmented conditions of the species are usually associated with troglobitic (cave-evolved) species, but are not limited to such species. These characters are also. encountered in forest litter-evolved species, and many litter inhabitants in the past were certainly ancestral to. presently troglobitic species (Barr, 1968). The legs and antennae o Speleoperipatus may be somewhat elongated (Fig. 8) as a specialization o.r cave lie, but measurements (in lie) o the appendages o this and epigean species are not available or comparison.
The general region o Jamaica rom which 8peleoperipatus co.rues is known to have several species o seemingly troglobitic invertebrates (personal observations), all o which have been derived rom litterinhabiting ancestors. But collecting in orest litter in Jamaica is inadequate so it is not known with certainty that the seemingly troglobitic species live .only in caves. In my field work on Jamaican cave faunas, I have made some 7o visits to. some 5o caves in search of their invertebrates, and Russell M. Norton has made some 2o visits to some 2o caves (many of the same caves in which I have collected). Since many of the other troglobitic species have been encountered in many of these caves over much .of Jamaica and Speleoperipatus has not, it is possible that this genus may now be limited to a single cave system. And since, it may exist only in small numbers, future collecting of it should be restrained, in order to preserve this unique genus from extinction.
Discussion. The genus and species is close to both Epiperilatus and Perilatus (s.str.) in its skin papillae; and to these and Macroperipatus in the number of feet papillae (2 anterior and posterior in all these genera), and four foot creeping pads (as in all these genera). However, the new genus differs from these genera in having some four fewer (at least) pairs of legs, and in the mandibles.
In some respects, the new genus and species resembles Typhloperipatus williamsoni Kemp (I914) from the. Himalaya foothills of eastern India (near Rotung, on the Dihang River, Assam, near Tibet). Both are eyeless, and the latter has only I9 to. :o pairs of Psyche September-Deeember legs. The latter species was taken abundantly in leaf litter in scrubjungle at from I2OO to 2000 eet elevation. However, eyelessness and loss of legs are characters that can be convergently acquired, and the two genera are presently placed in different subfamilies, based on the number of i:oot papillae (three and more in Peripatinae of the New World and equatorial Africa, and only two in the Eoperipatinae (containing Eoperipatus and Typhloperipatus in the Indo-Malaysian area). Kemp (I94: 49I) lists other features uniting Typhloperipatus with New World genera more closely than with Malaysian Eoperipatus. but this relationship should be reviewed.
A low number o'f legs may be a primitive i:eature in the Peripatldae. In comparison to the 22 or 23 pairs in S. spelaeus, 23 (to 27) pairs occur in Mesoperipatus tholloni of equatorial Africa, and 23 (to 29) pairs occur in some species groups of Oroperipatus. However, in all other New World genera, 27 pairs seems to be the lowes number.
In conclusion, the genus Spele,operipatus, because o.f its intermediate characteristics involving Asian and _African as well as New World genera, seems to have originated from an ancient lineage of onychophorans in which the characters were not .as well segregated as they are at present. In this case, Speleoperipatus. Mesoperipatus and Typhloperipatus of the Peripatidae can be regarded as relicts of earlier more widespread groups. The eyelessness of Typhloperipatus shows that this feature can be associated with species that do not live in caves. Speleoperipatus may thus have become eyeless before it entered a cave habitat, but until it is found in a non-cave habitat (such as forest litter or talus slopes), it should be considered as a troglobitic cave-limited species. It is worth noting, in closing, that the two families oi: the Phylum have once each, and independently, entered and permanently occupied a cave habitat and may have become restricted to, and specialized for, this cave habitat.