First record of the mygalomorph spider family Paratropididae (Arachnida, Araneae) in North America with the description of a new species of Paratropis Simon from Mexico, and with new ultramorphological data for the family

Abstract A new species of the genus Paratropis is described from North America: Paratropis tuxtlensis sp. n., from a tropical rainforest in Veracruz, Mexico. This is the fifth Paratropis and the tenth paratropidid species described and the first North American record of this Neotropical family. The species is described based on adult males and females, and juveniles. The juveniles show ontogenetic variation in the number of cuspules on the labium and endites, and in the number and position of leg trichobothria. This is the second Paratropis species, and the third paratropidid known from both sexes. The scanning electron photographs (SEM) reveal new morphological data and contribute to the knowledge of the family.


Introduction
The spider family Paratropididae Simon, 1889 is currently composed of four genera and nine species, being one of the less known spider families in the infraorder Mygalomorphae (Platnick, 2014). The genus with the highest number of species is Paratropis Simon, 1889 with four species: P. papilligera F. O. P. -Cambridge, 1896, P. sanguinea Mello-Leitão, 1923P. scruposa Simon, 1889;andP. seminermis Caporiacco, 1955. The genus Melloina Brignoli, 1985 is composed of three species: M. gracilis (Schenkel, 1953); M. rickwesti Raven, 1999;and M. santuario Bertani, 2013. The genera Anisaspis Simon, 1891 and Anisaspoides F. O. P. -Cambridge, 1896 have one species each: Anisaspis tuberculata Simon, 1891 and Anisaspoides gigantea F. O. P. -Cambridge, 1896. Three of the nine previously described species are known only from males (M. gracilis, M. santuario, and P. papilligera); two are known from both male and female (M. santuario and P. papilligera); and the remaining four are known only from females. Although some species, such as the recently described M. santuario have recorded juveniles, juveniles paratropidid specimens have never been described.
The majority of the species have a natural distribution in South America, principally in Brazil, with other species distributed in Venezuela and Peru. The northernmost records of the family Paratropididae are in Central America (Panama) and the West Indies (St. Vincent) (Platnick 2014). Raven (1999) mentioned that spiders of the family Paratropididae are enigmatic because of our limited knowledge about their habits, natural history and biology. It is known that they can be found in tropical forests, under fallen logs and boulders in the ground. As Rick West said in a personal communication to Raven (1999), these spiders make no burrow and hide under objects in the top layers of the soil. One characteristic of this family is that usually the entire body (including legs) is encrusted with soil particles (Raven 1985), although in some species such as M. santuario the cuticle has soil only in very restricted areas on the carapace (Bertani 2013). The encrusted soil on the exoskeleton could provide protection from predators or serve as camouflage to deceive their prey; it is a fact that they are very cryptic, which coupled with lack of movement when exposed makes them quite difficult to find and collect.
In this work, we describe a new species of Paratropis from a tropical rainforest from Veracruz, Mexico, based on juveniles and adults. It is the second species for the genus, and the third species for the family Paratropididae for which adults of both sexes are known. Juveniles specimens for the family are described for the first time. New morphological characteristics, distribution records, and natural history observations are presented.

Materials and methods
The specimens were collected manually and deposited in ethanol (80%). Samples for future molecular studies are cold-stored in vials with ethanol (96%). The general description of the species and terminology of the chaetotaxy follows Raven (1999) and Bertani (2013) with some modifications. The description of the spinnerets follows Bertani (2013). All specimens are deposited in the Colección Nacional de Arácnidos (CNAN), Instituto de Biología, Universidad Nacional Autónoma de México (UNAM), Mexico City.
The specimens were examined, measured and photographed with a Nikon SMZ 645 stereoscope; measurements are in millimeters (mm). Female epigyna were dissected in ethanol (80%) and cleaned in KOH (10%) for 15 minutes. All structures photographed and drawn under the stereoscope were submerged first in gel alcohol (available commercially as a hand cleaner), the firm consistency of the gel allowing the immobilization and positioning of the structure. The structure suspended in the gel alcohol was then covered with liquid ethanol 80% to minimize diffraction during examination and photography. The morphological structures were dissected and cleaned (first with a needle and a fine paintbrush, then with an ultrasonic cleaner at 20-40 kHz to remove the soil particles encrusted on the exoskeleton); subsequently they were critical-point dried, and examined at low vacuum in a HITACHI S-2460N scanning electron microscope (SEM) to take the photomicrographs. All scale measurements on SEM photomicrographs are in microns. The map was done with ArcView GIS version 3.2 (Applegate 1999). The photographs and map were edited using Adobe Photoshop Version 7.0. Morphological abbreviations: ALE, anterior lateral eyes; AME, anterior median eyes; PLE, posterior lateral eyes; PME, posterior median eyes; PLS, posterior lateral spinnerets; PMS, posterior median spinnerets.

Family Paratropididae Simon, 1889
Genus Paratropis Simon, 1889 Type species. Paratropis scruposa Simon, 1889 Diagnosis. The genus can be diagnosed with the following combination of characters (after Raven 1985): 1) eye tubercle highly elevated, 2) transverse fovea, 3) narrow cheliceral furrow, with teeth on both margins in two juxtaposed rows, 4) endites with anterior conical projection, 5) Legs I of male without tibial spur, 6) paired claws of tarsus with one long tooth, 7) claw tufts absent, and 8) third claw absent on leg II (However, see Discussion concerning this character on the new species).  fig. 7); by the number of conical teeth in the cheliceral furrows, in P. tuxtlensis the promargin has 11 teeth and retromargin 9 (Figure 9), whereas in P. papilligera the promargin has 14 teeth and retromargin 10.
Coloration: The general coloration under alcohol is the same as soil particles encrusted on the body, which is pale brown (Figures 1-4, 6, 8). Chelicerae orange ventrally (Figures 2,7,9), becoming brown dorsally (Figures 1, 4), fangs of chelicerae dark reddish brown (Figure 2). The carapace has reddish coloration when the soil particles are cleaned. Sternum pale orange; endites and labium orange ( Figure 2). Legs olive color when soil particles are cleaned, becoming paler on tibia, metatarsi, and tarsi. The opisthosoma was difficult to clean, even with longer time in the ultrasonic cleaner, and the coloration could be similar to the carapace. Spinnerets pale yellow.
Chelicerae: Fangs wider than on male ( Figure 50). Chelicerae furrows with conical, wide, short teeth on promargin and retromargin. Right chelicera promargin with 13 teeth, retromargin with 11 teeth; left chelicera promargin with 12 teeth, retromargin with 11 teeth; teeth on retromargin wider than teeth on promargin; on both rows the proximal teeth are wider and longer than distal teeth.
Claws: Slightly longer than the male. Only the tarsi I and II with small, unpaired third claw (differing from the male, which lacks it in tarsus II, and from Raven's 1985 generic diagnosis). Palp tarsus with one single claw, without tooth.
Opisthosoma: Bigger than in male (Figure 51), genital operculum not visible due to encrusted soil particles.
Etymology. The specific name is an adjective and refers to the type locality: Estación de Biología Tropical "Los Tuxtlas", Municipio San Andrés Tuxtla, Veracruz, Mexico.
Distribution. The species is known only from the region around the type locality in the Volcan San Martin Biosphere Reserve (Figure 62).
Natural history. The specimens were collected in tropical rain-forest, under boulders on the ground (Figure 57). The holotype, two paratype males and one juvenile where collected near each other, within around 3 m 2 , in a zone with numerous small and big boulders on the ground. The specimens remained motionless when they were exposed by removing the rock that provided shelter, possibly as a defense mechanism because the soil particles encrusted on the body cuticle serves as camouflage with the moist ground (Figures 57-60). The type locality is at 1039 m elevation, and two adult females where collected nearby at 480 m.
Towards the end of spring (May 19, 2012), one paratypes female (CNAN-T0767) kept in captivity in the laboratory laid an egg sac ( Figure 61). The female kept her palps and legs in contact with the egg sac constantly. Twenty-three spiderlings emerged 38 days after oviposition (July 26, 2012

Discussion
The spider family Paratropididae Simon, 1889 was considered as being monophyletic and as the sister group of the family Theraphosidae Thorell, 1869 (Superfamily Theraphosoidea) based on morphological evidence by Raven (1985) and Goloboff (1993). However, molecular data are in conflict with morphological evidence (Hedin and Bond 2006;Bond et al. 2012). This enigmatic family appears to be difficult to place phylogenetically, due to the available material for molecular phylogenetic analyses being inadequate to evaluate the monophyly of the family. The last phylogenetic analysis made by Bond et al. (2012) using total evidence, found to Paratropididae as the sister group of the clade Bipectina (Goloboff 1993) as was found by Hedin and Bond (2006).
The monophyly of the four genera that compose the family has never been tested. Even the monophyly of the subfamilies proposed by Simon (1889) (Paratropidinae), and Raven (1985) (Glabropelmatinae) have never been tested; although characters as the single tooth on the claws, the steeply elevated eye tubercle, the absence of a tibial spur on leg I, and the lack of claw tufts have been useful to diagnose Paratropidinae; whereas in the case of Glabropelmatinae the presence of claw tufts, a normally elevated eye tubercle, and shorter anterior maxillary lobes are diagnostic.
We report the ontogenetic variation for Paratropididae juveniles for the first time. The number of the cuspules on labium and sternum is considerably lower in younger instars than on adults. We found two juveniles that we hypothesize belong to two different, not necessarily successive instars; the smallest had fewer cuspules on labium and endites than the larger juvenile and adults. It seems that the number of cuspules in each instar is incremented with growth, and the adult instars have the highest number. In addition, there is ontogenetic variation in the number of leg trichobothria in each instar, with fewest trichobothria on the younger instars and the highest number on the adults. Although we hypothesize that the juveniles belong to two different, not successive instars, based on size differences, both specimens had the same number of trichobothria on tibiae and metatarsi. However, there is allometric growth reflected in the trichobothria position on metatarsi: in the smallest juvenile one trichobothrium is located on the median third and the other one in the apical third; whereas in the larger juvenile both trichobothria are located in the apical third. Although on both juvenile specimens the trichobothria number on tibiae and metatarsi was the same, the number of trichobothria on the tarsi is different in the larger juvenile, having one more trichobothrium.
The genus Paratropis Simon, 1889 was diagnosed by Simon (1889) based only on the combination of somatic characters, some of them shared with other genera in the family. The diagnosis was based on the shape of the ocular tubercle, size of eyes, shape of the endites and labium, presence of scopules on labium, leg formula, and the presence of clubbed setae along the opisthosoma. However, Simon did not describe any sexual characters; although Paratropis scruposa (type species) was described based on one female, he did not describe the spermathecae, which have been traditionally useful not only at level species but also at generic level in the taxonomy of spiders.
Although Raven (1985) did not diagnose the genus, he mentioned the combination of some somatic characters, as Simon had done, that could be used to characterize it. In the key to genera he mentioned that Paratropis can be identified by the third, unpaired claw absent on leg II (Raven 1985;page 122). However, the female of the new species described here has tarsi I and II with a small, unpaired claw, which could call into question the validity of that diagnostic character for the genus, or could lead to the recognition of a new genus in the family Paratropididae. Because to the discrepancies mentioned above, robust cladistic analyses based on morphological and molecular data are necessary to test the validity of Paratropis and the other genera in the family. It is urgent to collect the males of most of the species, because Paratropis tuxtlensis is only the second species of the genus, and the fourth species of the family, where the male is known. The structures of male palps might provide additional information not only related with the diagnoses at species level, but also potential phylogenetic information within the genus and the relationships of the genera of Paratropididae.