The Goblin Spider Genus Ischnothyreus (Araneae, Oonopidae) in the New World

ABSTRACT Although originally described from St. Vincent in the Lesser Antilles, the goblin spider genus Ischnothyreus Simon appears to be an Old World taxon that is represented in the New World only by two presumably introduced, pantropical, synanthropic species: I. peltifer (Simon) and I. velox Jackson. Two specific names based on New World specimens (I. barrowsi Chamberlin and Ivie from Florida, and I. indressus Chickering from the Lesser Antilles) are placed as junior synonyms of I. velox, which is newly recorded from Mexico, Panama, Jamaica, Hispaniola, Venezuela, Brazil, Madagascar, the Philippines, the Marshall Islands, Hawaii, the Marquesas Islands, and New Caledonia. A third species, I. browni Chickering, that is supposedly from Costa Rica was apparently based on mislabeled specimens that are actually from the Philippines. The type specimens of I. browni resemble those of the Seychelle species Ischnothyrella jivani (Benoit) in that the dorsal abdominal scutum of males is extremely weak and that of females is either greatly reduced or entirely lost. Both species nevertheless share the synapomorphies of Ischnothyreus, and the generic name Ischnothyrella Saaristo is therefore placed as a junior synonym of Ischnothyreus.


INTRODUCTION
, in the first paper dealing with the generic-level diversity of New World oonopids, described a total of eight new genera, each based on a species collected on the tiny island of St. Vincent in the Lesser Antilles. In three of those cases, the relevant type species have since been shown to be pantropical taxa that are probably of Old World origin, rather than native to St. Vincent: Pelicinus Simon (see Platnick et al., 2012b), Opopaea Simon (see Platnick and Duperre, 2009), and Triaeris Simon (see Platnick et al., 2012c). In the present paper we argue that the same is true for a fourth genus, described initially as Ischnaspis Simon, a preoccupied name that was quickly replaced by Ischnothyreus Simon (1893a). Simon (1891), when describing the type species, Ischnaspis peltifer Simon, on the basis of females from St. Vincent, already considered the species to be widespread, citing additional females from Sierra Leone and the Philippines. Simon (1893a: fig. 264) later added Sri Lanka to the list as well, supplying an illustration of a Sri Lankan male that he associated with I. peltifer. That male was misidentified, and belongs to a different genus, Camptoscaphiella Caporiacco (see Baehr and Ubick, 2010). The actual male of I. peltifer was first described, misplaced as a species of Dysderina Simon, by Bryant (1942), and it was first correctly associated with I. peltifer by .
Although Simons females from Sierra Leone and the Philippines were also misidentified, workers since Simons time have followed his lead, and have recorded I. peltifer from a wide variety of both Old and New World localities. However, Saaristo (2001) demonstrated that some of those far-flung records actually refer to a second species, Ischnothyreus velox Jackson (1908), which was originally described on the basis of specimens taken in British greenhouses; Saaristo (2001) showed clearly that I. velox occurs also at least in the Seychelles, and that both species have been found in British greenhouses.
Our studies suggest that both of the species treated by Saaristo are actually widely distrib¬ uted, pantropical, synanthropic taxa that have also been introduced into buildings in temperate areas of the northern hemisphere; I. peltifer has been taken even in Canada (in the African Pavilion of the Toronto Metropolitan Zoo)! The two species are sometimes even sympatric, and both have been collected in Florida, Mexico, Panama, Jamaica, St. Vincent, Venezuela, Brazil, Madagascar, Seychelles, and Hawaii. Interestingly, Simon (1896) had already concluded that there is more than one widespread species of Ischnothyreus, as he reported that specimens of Ischnothyreus lymphaseus Simon (1893b), originally described from Sri Lanka, had been taken in the greenhouses of the Museum National d'Histoire Naturelle in Paris. The basis for that conclusion is uncertain; the only vial currently in the Simon collection containing material said to be from those greenhouses includes three females only, and I. lymphaseus was described (and remains known only) from a single male. There is an additional vial in the Simon collection, labeled only with that species name, which includes one male and one female; whether that pair of specimens is from Sri Lanka, France, or somewhere else is unknown. Simon may have used that pair to match the sexes, but he seems subsequently to have realized that his identification of the Paris greenhouse specimens was incorrect; after studying the apparently conspecific specimens from British greenhouses sent to him by Jackson, Simon agreed that they represented a new species (which Jackson then described as I. velox), rather than I. lymphaseus.
New World Ischnothyreus have been studied seriously only by , who confined his attention to collections from Central America and the West Indies; he later (Chickering, 1969) examined a few specimens from Florida as well. In addition to noting many new records for I. peltifer,  described two new species. One, I. indressus from the Lesser Antilles, is placed below as a junior synonym of I. velox, but the second, I. browni, is more problematic. It was based on two males and one female that Chickering (1968: 83) indicated were "believed to have been collected by Dr. W.L. Brown, Cornell University, in Costa Rica, Rio Toro Amarillo, near Guapiles, Heredia, March 1966 also noted that "Because of some confusion in sorting there seems to be a slight uncertainty about the type locality...." Costa Rican oonopids have since been thoroughly collected by Carlos Viquez and his col¬ leagues, but we have found no additional specimens of I. browni from Costa Rica or anywhere else in the New World. We therefore suspected that the type specimens are actually from somewhere in the Old World instead. Our searches of available collections of Ischnothyreus from the Old World revealed apparently conspecific specimens from the Philippines.
It is conceivable that the types of I. browni were an accidental introduction from the Philip¬ pines, and were actually captured in Costa Rica. However, given the confusion mentioned by Chickering, it seems far more likely that his specimens are actually from the Philippines, and were simply mislabeled during the sorting process. Until and unless additional specimens of I. browni are found in the New World, we regard the Costa Rican record of the species as spurious.
The type specimens of I. browni are notable for their lightly sclerotized abdomens (figs. 152, 153,155). Chickering (1968: 83) described the males as having abdominal scuta "which are hardly discernible with borders very indefinite; dorsal scutum appears to reach only a little more than half way from base to posterior end"; he similarly (1968: 84) indicated that the female has the "dorsal scutum hardly discernible; ventral and epigastric scuta clearly visible." Chickerings speci¬ mens may be teneral (i.e., collected shortly after their final molt, before the dorsal scutum had time to become fully sclerotized). We have seen similarly teneral males of other Ischnothyreus species in which the dorsal scutum is scarcely detectable, even though the palps are well sclerotized; given that the palps are so heavily sclerotized, it isn't surprising that sclerotization of the dorsal scutum may lag behind that of the palps. In the case of older specimens in collections, one often cannot tell whether they are teneral or merely bleached from overexposure to light, and of course some specimens may be both teneral and bleached, making them exceedingly difficult to study.
Nevertheless, the appearance of Chickerings specimens of I. browni is strikingly similar to that of the Seychelle specimens originally described as Ischnothyreus jivani by Benoit (1979).
In his subsequent review of the Seychelle oonopids, Saaristo (2001) established a new genus, Ischnothyrella, to contain only I. jivani. This monotypic genus was based primarily on the sup¬ posed absence of visible abdominal scuta, although Saaristo also cited minor differences in leg spination, as well as male and female genitalic features that are unique to the type species and hence uninformative about its relationships. The types of I. jivani are badly bleached, and it is possible that they may be teneral as well, so Saaristo's claim that the abdominal scuta are entirely absent is suspect, and needs to be checked against freshly collected specimens.
In any case, though, as with numerous other monotypic genera erected by Saaristo, he seems here to have been so overimpressed by species-level autapomorphies (such as the puta-tive loss of the dorsal scutum) that he promoted a relatively autapomorphic species to an unreasonably high level, thereby rendering the group to which it actually belongs (Ischnothyreus, in this case) paraphyletic. Just as in the similar examples detailed in Platnick et al. (2011) for Brignolia Dumitresco and Georgesco and in Platnick et al. (2012a) for Orchestina Simon, Saaristo provided no putative synapomorphies uniting all the relevant species other than I. jivani and hence supporting the placement of that species as the sister group of all the others. We know of no such characters; Saaristo's artificial, monotypic genus is therefore posi¬ tively misleading phylogenetically, and is here placed in synonymy. Ischnothyreus species actu¬ ally vary widely in the extent of the dorsal abdominal scutum; although it is usually small, covering only about half of the abdominal length and width, it can be much larger (covering almost the entire dorsum; see Kranz-Baltensperger, 2011: figs. 8A, 8C, 30A, 30C) or much smaller (reduced to just a narrow strip over the cardiac area), and may possibly be lost entirely (at least in some females).
As thus relimited, Ischnothyreus is defined by obvious synapomorphies: the heavily sclerotized, "burnt" palps of males (figs. 77, 82), which are associated with an elaborate, internal skeletomuscular system (figs. 43-45; Dumitresco and Georgesco, 1983: pi. 17, figs. 4, 5)  palps occur only in the genus Brignolia, but those palps are differently constructed, with a distinct dorsal depression that does not occur in Ischnothyreus. Females of the genus Triaeris also resemble those of Ischnothyreus in having hypertrophied posterior genitalic elements that occupy most of the postepigastric scutum and involve external modifications of that scutum, but they do not have the highly "squiggled" ducts found in most species of Ischnothyreus. As argued elsewhere (Platnick et al., 2011(Platnick et al., , 2012c), Brignolia appears to be more closely related to Opopaea than to Ischnothyreus, and Triaeris appears to be more closely related to Zyngoonops Benoit than to Ischnothyreus.
We suspect that, as suggested by Ubick and Griswold (2011), Ischnothyreus is actually more closely related to the Asian genus Camptoscaphiella and the Malagasy genus Malagiella Ubick and Griswold (2011) than to either Brignolia or Triaeris. Serious consideration of that hypoth¬ esis must await study of the many undescribed Old World species of Ischnothyreus, but there do seem to be potential female genitalic synapomorphies uniting these genera. Baehr and Ubick (2010: 6) reported a slit-shaped external copulatory opening, situated near the anterior margin of the postepigastric scutum, in female Camptoscaphiella; a similar slit occurs also at least in I. peltifer (compare figs. 72, 73 with Baehr and Ubick, 2010: fig. 150). However, within Camptoscaphiella, the presence of that slit-shaped opening has been confirmed by scanning electron microscopy only in the females of Camptoscaphiella paquini Ubick, a Chinese species that differs significantly from its congeners in that the female genitalic ducts resemble those of Ischnothyreus in being highly "squiggled." It is possible, however, that both genera actually show a similar range of female genitalic structures, as we have seen undescribed Ischnothyreus species (from Africa and Sulawesi) in which the posterior genitalic ducts of females appear (at least superficially) to be nearly straight.
In at least two species of the similar genus Malagiella, the external copulatory opening is rounded rather than slit shaped (Ubick and Griswold, 2011: figs. 11, 177), and is seemingly not accompanied by the longitudinal row of pores along the midline that is found in at least some species of Ischnothyreus (figs. 73, 74) and Camptoscaphiella. Interestingly, though, Malagiella spe¬ cies show similar variation in female posterior genitalic duct arrangements, ranging from nearly straight to very sinuous (Ubick and Griswold, 2011: map 4), and in I. velox the external copulatory opening is rounded ( fig. 132) rather than slit shaped. Nevertheless, female genitalic structure (i.e., the longitudinal row of pores accompanying the copulatory opening) may actually support the monophyly of Ischnothyreus plus Camptoscaphiella, whereas the male palps instead clearly support the monophyly of Camptoscaphiella plus Malagiella, which are united by a greatly enlarged palpal patella (Baehr and Ubick, 2010: figs. 161, 164;Ubick and Griswold, 2011: figs. 7, 85- figs. 48, 49); in both genera, it seems uncertain whether this groove represents the epigastric furrow, a groove connecting the posterior spiracles, or perhaps even a fusion of both. Similarly, males of all Malagiella species, and all but two Camptoscaphiella species, have the dorsal scu¬ tum fused to the epigastric scutum, but this character varies widely within Ischnothyreus; only about half of the males described by Kranz-Baltensperger (2011, 2012 have those scuta fused. Thus, we can only concur with Ubick and Griswold (2011: 7) that these three genera con¬ stitute a monophyletic group, the "Ischnothyreus complex." A fourth genus, Aprusia Simon, from Sri Lanka and southern India, seems also to belong to this group, sharing a similar eye arrangement, leg spination pattern, and abdominal scutum configuration, but both the male and female genitalic conformation suggest that Aprusia is less closely related to Ischnothyreus than are Camptoscaphiella and Malagiella (see Grismado et al., 2011).
Our methods follow those of Platnick and Duperre (2009); all measurements are in mm.
A detailed description is provided for the type species, and only differences from that species are mentioned in the other descriptions. High-resolution versions of the images, many addi¬ tional images of the pantropical species, a sortable version of the geocoded locality data, and a distribution map for each species will be available on the goblin spider Planetary Biodiversity Inventory (PBI) project's website (http://research.amnh.org/oonopidae). and we have not yet found any characters that will reliably separate such females. Females of Ischnothyreus could also be confused with those of Triaeris, which also have the bulk of the genitalia occupying the postepigastric scutum and often including somewhat sinuous posterior ducts, but which differ from members of Ischnothyreus in having a greatly elongated, spinose patella on leg I (see Platnick et al., 2012c).
Description: See Saaristo (2001: 345); a full description will not be possible until the many undescribed Old World species are studied in detail.
Distribution: Ischnothyreus appears to be natively an Old World group. Numerous endemic species occur from West Africa to Yemen (Saaristo and van Harten, 2006), Nepal (Burger, 2010), China (Tong and Li, 2008), Malaysia (Kranz-Baltensperger, 2012), Japan, the Philippines, the Marshall Islands, and Borneo (Kranz-Baltensperger, 2011) and, to the south, from Angola, Comoros, Madagascar, and Seychelles east to Australia (Edward and Harvey, 2009), Fiji, and the Cook Islands. However, as detailed below, two of the Old World species have apparently attained pantropical distributions.