Looks are deceiving: a cladistic analysis, three new species, and a new diagnosis of Paravima Caporiacco, 1951 (Opiliones: Agoristenidae)

A morphological cladistic analysis, consisting of 45 terminals and 58 characters was performed to evaluate the monophyly of the genus Paravima Caporiacco, 1951, and its relationship to other Leiosteninae. The analysis resulted in two most parsimonious trees, all recovering Paravima as monophyletic with the inclusion of two species formerly described in Avima Roewer, 1949, therefore, the following nomenclatural acts are herein proposed: Paravima plana (Goodnight & Goodnight, 1949) comb. nov. , and Paravima quirozi (González-Sponga, 1981) comb. nov. , Avima vigirima (Villarreal-M & Rodríguez-Manzanilla, 2003) is here considered as a junior subjective synonym of P. quirozi (González-Sponga, 1981) comb. nov. , Paravima acanthoconus Villarreal-Manzanilla & DoNascimiento, 2005 is here considered as a junior subjective synonym of Paravima propespelunca González-Sponga, 1987, and Paravima flumencaurimarensis González-Sponga, 1987 is here considered as a junior subjective synonym of Paravima goodnigh-torum Caporiacco, 1951. Additionally, three new species are described: P. lokura sp. nov. (Tamá National Natural Park, Norte de Santander department) and P. magistri sp. nov. (Los Tunos Natural Reserve, Cundinamarca department), both from Colombia; and P. totoro sp. nov. (Henri Pittier National Park, Aragua state) from Venezuela. Finally, an emended generic diagnosis, a key for all the species, and a distributional map are presented.


Introduction
The Neotropical family Agoristenidae Šilhavý, 1973, is an uncommon albeit diverse taxon (26 gen., 80 spp.), as is reflected for the small number of specimens deposited in natural collections, divided into three subfamilies, the most diverse being Leiosteninae Šilhavý, 1973, from res 2014; García & Villarreal 2020: fig. 9). Except for Avima Roewer, 1949, most of the genera of Agoristenidae are not very diverse, many of them being monotypic (14 gen.), and the phylogenetic relationships between them are poorly understood. Only three phylogenetic analyzes have been conducted with this objective (Kury 1997, Pinto-da-Rocha & Hara 2009, however, the monophyly of its genera has not been tested. In this work, a cladistic analysis of Paravima Caporiacco, 1951 is performed, to test its monophyly and evaluate its systematic position within the subfamily. Paravima, as currently diagnosed, is a small group of six species associated with forests in the central region of the Cordillera de la Costa in Venezuela. The genus was described by Caporiacco (1951) to accommodate the single species Paravima goodnightorum Caporiacco, 1951, from El Junquito, Venezuela. González-Sponga (1987 in his most comprehensive work on Leiosteninae, described the remaining species of the genus: Paravima flumencaurimarensis González-Sponga, 1987, P. locumida González-Sponga, 1987, P. morritomacairensis González-Sponga, 1987, and P. propespelunca González-Sponga, 1987, except P. acanthoconus Villarreal-Manzanilla & DoNascimiento, 2005, which was described 18 years later (Villarreal-M. and DoNascimiento 2005).
On the other hand, Avima is the most diverse genus of Leiosteninae, however, it is most likely an amalgam of unrelated lineages as partially suggested by , Garcia et al. (2022a, b), and the present analysis. The lack of ornamentation in area III has been referred to as the main generic diagnostic character of Avima (see García et al 2022a), ignoring a significant amount of hidden information in the morphological heterogeneity of the group. Some species currently located in Avima exhibit some aspects of external morphology (e.g. DS shape or eyes position) and genital morphology (e.g. stylus and lamina parva shape) that caught our attention, making us question its generic position. In addition, when compared to the generic nucleus recently defined by García et al (2022a) for Avima, it raises suspicions about a possible relationship between these species and those currently found in Paravima instead of being related to the Avima core. Avima plana (Goodnight & Goodnight, 1949) was described from the Cordillera de la Costa (Venezuela), and later erroneously recorded to Cueva de Benito (Mérida state), in the Venezuelan Andes, by Rambla (1978) (González-Sponga 1987Kury 2003). González-Sponga (1981) described Avima quirozi González-Sponga, 1981 from Yaracuy state (Venezuela), and Villarreal-M. and Rodríguez-Manzanilla (2003) described Avima vigirima Villarreal-M. & Rodríguez-Manzanilla, 2003 from Carabobo state (Venezuela), and commented about the possible relationship between these three species.
In the present work, a phylogenetic analysis based on morphological characters was carried out. The monophyly of Paravima was tested and the relationship between the three previously mentioned Avima species and the genus Paravima was evaluated. So, after two new combinations, three synonyms, and the description of three new species, the genus Paravima passes from six to nine species, all of them included in an identification key and an updated geographical distribution map.

Material and Methods
Individuals of the species were imaged from varied sources. We mostly used a Nikon 5200, a Canon Pow-erShot S3IS, and a Sony Cybershot DSC-V1 camera attached to a stereomicroscope. Multiple images of each selected specimen at different focal planes were combined with Zerene Stacker of CombineZP (Hadley 2015). The holotype of Paravima magistri sp. nov. was photographed with a Leica M205C stereoscope attached to a Leica DFC450 digital camera, and combined with Leica Application Suite (LAS) software version 4.6.2. All resultant photographs were posteriorly edited in Photoshop CC 2014 software. All measurements are in mm unless otherwise noted.
The male genitalia illustrated with Scanning Electron Microscopy (SEM) was dehydrated with Critical Point Drying (CPD), sputter-coated with gold-palladium, and examined with a JEOL JSM-6390LV Scanning Electron Microscope at the Center for Scanning Electron Microscopy of Museu Nacional de Rio de Janeiro (MNRJ), with an accelerating voltage of 10 kV. Drawings were made using a stereomicroscope with camera lucida, and digitized with Inkscape 0.91 software. Some penis drawings have no scale. Color descriptions use the standard names of the 267 Color Centroids of the NBS/IBCC Color System as named in Centore (2016) and color description refers to specimens preserved in ethyl alcohol.
The distribution map was made with Quantum GIS 2.18.19 software (QGIS team 2020). The references to biogeographic provinces in the text and map follow the biogeographic regionalization proposed by Morrone et al. (2022). Geographic coordinates have been transcribed verbatim from the labels and are in different formats. When no original coordinates were available, those were estimated using Google Maps and placed between square brackets.
Patterns of description follow . Terminology for the integumentary ornamentation follows DaSilva and Gnaspini (2010). Chaetotaxy of penis lamina parva and truncus follows Kury and Villarreal (2015). Dorsal scutum outline types follow Kury and Medrano (2016), with the modifications explained in . The complementary descriptions of the species do not repeat the generic characteristics.

Remark:
The material marked with an asterisk (*) was destroyed in the fire at the National Museum / Federal University of Rio de Janeiro, Brazil (September 2018). There is no formal impediment per ICZN rules to describe a species whose holotype was lost before the publication of the description (Krell and Marshall 2017).

Character sampling
The phylogenetic analysis was performed using 39 somatic and 19 genital characters. All the logical descriptions of the characters were written according to the recommendations by Sereno (2007). Characters were mainly based on , with the addition of 11 characters and some adjustments according to the newly included taxa.

Phylogenetic analyses
Parsimony analyses were performed in TNT v.1.5 (Goloboff and Catalano 2016), using heuristic methods ("traditional search") under implied weights (Goloboff et al. 2008) and equal weights. Tree bisection-reconnection (TBR) and the following setting for starting trees (mult = tbr replic 1000 hold 1000) were used. Branches with no possible support were collapsed (collapsing "strict (= nelsen)") during and after the tree search. All characters were unordered and the multistate characters were treated as non-additive (Fitch 1971). The searches under implied weighting (IW) were run with the concavity constant (K) set to 1-500 (piwe = 1; mult = tbr replic 1000 hold 1000). Implied weighting in TNT weights the characters according to a concave function of homoplasy (K) that is set by the user and negatively correlates with how strongly homoplastic characters are down-weighted (Goloboff 1993). In this study, we used a TNT script (setk.run) written by Salvador Arias (Instituto Miguel Lillo, San Miguel de Tucumán, Argentina) to calculate the value for the concavity constant K, which best matches our data (for details see Goloboff et al. 2008), and additionally, four searches were conducted with TNT using IW with K concavity values of 1, 2, 4 and 8. In parallel, an analysis was conducted using equal weights (EW) and the results between both are compared. The characters were studied and optimized, and trees were edited in WinClada 10.00.08 (Nixon 2002). All discussion of the cladistic analyses was made under ACCTRAN optimization as a way to preserve the primary homology (de Pinna 1991). To establish group support, bremer support (Bremer 1994), relative bremer support (Goloboff & Farris 2001), and bootstrap values were calculated using TNT.

Character list
The data matrix (Table S1) consists of 45 taxa and 58 characters. The numbers in parentheses refer to the character numbering in . DS = Dorsal scutum; LP = Lamina parva; MS = Macrosetae.

Cladistic analysis
We obtained two parsimonious trees, using a value of K = 6.4844 whose strict consensus is presented in the figure  16 (L=189, CI=43, RI=73). Paravima was recovered as monophyletic, grouping all the previously described Venezuelan species, plus the species Avima plana and A. quirozi, together with the species described here, Paravima lokura sp. nov., P. magistri sp. nov., and P. totoro sp. nov. A sister group relationship between Avima venezuelica and Paravima was found under the following values of K: 2, 4, 6,4844 and 8 (Figs 16, 17A). When performing an analysis using equal weights, 138 equally parsimonious trees were obtained but, the hypothesis of internal relations within Paravima does not change in relation to another K values analyzed. However, the hypothesis of relationships with other Leiosteninae were modified, forming a large polytomy with several of the outgroup genera included (Fig. 17A). Some issues that could be influencing these results are discussed (see discussion for details).  Soares & Avram (1981).

Taxonomy
Diagnosis. The genus can be diagnosed within Leiosteninae by the placement of the ocularium, located close to the anterior half margin of the carapace and the absence of a longitudinal dorsal keel in the stylus of the penis, and additionally by the combination of characters presented in the description.
Description. DS outline Epsilon type 3, with posterior border distinctly wider than anterior border, giving an ovoid appearance. Ocularium low, very close to the anterior margin of DS, smooth (granulate in P. lokura sp. nov.), and with median concavity. Interocular distance half of carapace width (except in P. magistri sp. nov. that is one third). Mesotergum concolor with the rest of the DS (except in P. magistri sp. nov., P. plana comb. nov., and P. quirozi comb. nov., with a dark spot). Mesotergal area III with paramedian ornamentation (except in P. plana comb. nov. and P. quirozi comb. nov., smooth). Long legs (FeIV length at least 3.7 times DS length). Penis with tips of LP sharp, not differentiated and not curved; subdistal depression between MS-E2 and horned LP; malleus with three pairs of branched MS-A (except in P. magistri sp. nov. where it has two pairs); MS-C and MS-E1 absent; stylus sinuous (except in P. magistri sp. nov., P. plana comb. nov., P. quirozi comb. nov., and P. totoro sp. nov., where it is straight), and without dorsal keel or process.
Distribution. Venezuela: Cordillera de la Costa mountain range (Distrito Capital, Aragua, Carabobo, Miranda, Guárico and Yaracuy states); Colombia: P.N.N Tamá (Norte de Santander department) and San Antonio del Tequendama (Cundinamarca department). Paravima plana (Goodnight & Goodnight, 1949)   Diagnosis. Paired tubercles on the mesotergal area III mammilliform, with blunt tip shorter than the base (in P. locumida the tip is acute and slightly higher than the base; in the remaining species, the mesotergal area III is unornamented or exhibits conical or domed tubercles).
Description. See González-Sponga (1987). Here we offer a complementary description based on MNRJ 18301 Mesotergum ill delimited, divided into four areas: area I divided into two halves, areas II-IV undivided; areas I, II, and IV with a pair of granules, area III with a pair of paramedian mammilliform tubercles darker than the rest of the mesotergum and higher than the ocularium (Figs 1A-D, 2B, D, E, 3A-D). Posterior margin of DS substraight, smooth (sometimes with some granules). Free tergites I-III with granules (Fig. 2B). Anterior margin of coxa I with two rows of median irregular tubercles (Figs 1E, 2C). Cheliceral hand swollen. Legs increasing in thickness from leg I to leg IV; Fe IV three times DS length ( Fig. 2A); Pa IV with some granules (Fig. 3K); Ti IV thickened distally (Fig. 3K). Penis with small LP (height shorter than width), apex with anterolateral crescent-shaped corners (  Remarks. The female specimen MAGS 1115 (MIZA 0105921), identified by González-Sponga as P. flumencaurimarensis, seems to correspond to one of the female paratypes of P. flumencaurimarensis cited in type data above as MAGS, without number. The penis drawing (González-Sponga 1987: 473, fig. 604) shows one pair of MS-D and two pairs of MS-A, but after studying males from four localities (including paratypes of P. flumencaurimarensis and P. goodnightorum) (Fig. 3A-D, H-J), all exhibited two pairs of MS-D, and three pairs of MS-A. Likewise, we have found a subtle unnoticed variation in the color patterns of the dorsal shield ( Fig. 3A-D) and in the shape of the mammilliform tubercles of the mesotergal area III ( Fig. 3L-P). The right lateral SEM image (Fig. 3E) of the penis shows three MS-A, but the dorsal view (Fig. 3F) shows two pairs of MS-A on the left side. We consider the common in the genus is three pairs of MS-A, as seen in all other species of Paravima. The LP is slightly collapsed (Fig. 3E), this collapse could be an artifact of the preparation of the genitals, which can be reflected into slight variations in the drawings (Fig. 3H-J).
Description. See González-Sponga (1987). Here we offer a complementary description based on González-Sponga 1987: figs 616, 617: Anterior and lateral margins of DS smooth, posterior margin with some granules (Fig.  4A-D). Mesotergum ill delimited, divided into four areas: area I divided into two halves; areas II-IV undivided; area III with a pair of paramedian mammilliform spines (higher than ocularium) slightly darker than the rest of the mesotergum (Fig. 4A-C). Anterior margin of coxa I with a high proximal conical tubercle (Fig. 6E) Remarks. One male from MAGS 324 was photographed in 2017 (Fig. 4). However, in a recent revision in 2022, we found the material totally deteriorated and that all the specimens are useless for taxonomic purposes (a lot of fungus on it). We suspect that MAGS 324 corresponds to one of the male paratypes cited as MAGS in type data, without number. Of all the drawings of Paravima genitalia made by G-S, those of P. locumida mostly smooth. Area I divided into two halves, each one with one small tubercle; area II-IV undivided; II with a pair of medium-sized tubercles close to the medial axis of the body (Fig. 6A, B); III with a pair of conical wide tubercles close to each other (Figs 5D, E, 6B); IV with four small tubercles. Posterior border of scutum straight, with some tubercles. Free tergites I-II and anal operculum with some tubercles (Fig. 6A). -Venter: Coxae I-IV with some granules (Fig. 5C). Coxa I with one medial tubercle on the anterior margin, a longitudinal row of tubercles at the medial region, a group of three tubercles on the anterodistal margin, and two tubercles close to the posterodistal margin (Fig. 6C); coxa II longer than coxa I; coxa III longer than I and II; coxa IV backward projected (Fig. 5C). Sternites with a row of small tubercles each. Stigmatic area smooth. Stigmata large, oval and transverse (Fig. 5C). -Chelicerae: Segment I rectangular, with well-marked bulla (Figs 5B, 6A), one ectal subdistal tubercle, and two tubercles on the proximal border (Fig. 4B). Chelicera swollen (Figs 5B, E, 6B). Hand with some sparse setiferous tubercles of different sizes in the fronto mesal portion, and a frontal row of small setifer- ous tubercles extending from the mesal to the ectal face of the chelicera close to the base of the fingers (Fig. 6D). Fixed finger with the inner surface irregularly dentated. Movable finger with one trapezoid, sub-basal tooth and with the inner surface dentate (Fig. 6D). -Pedipalps: Longer than DS length, smooth. Tr ventrally with one subapical setiferous tubercle. Fe with a ventromesal row of five setiferous tubercles (the third shorter than the others), and one large ventroectal setiferous tubercle in the distal portion (Fig. 5C). Pa with one large mesal setif-erous tubercle. Ti ectal iII, mesal IIi. Ta ectal IIi, mesal IIi (Figs 5, 6B). -Legs: Leg I smooth, legs II-IV with minute granules. Leg I filiform, the rest, getting steadily thicker from leg II to IV (Fig. 5A). Fe IV three times DS length. Tarsal formula: 9(3)-9(3)/20(3)-20(3)/8-9/8-8.  (Fig. 6G). MS-D1-2 long, located in a vertical line on a keel between the LP and the base of the stylus (Fig. 6G). MS-E2 large and branched (Fig.  6F). Stylus sinuous, surpassing the lamina parva, narrower at distal region, tip irregular (Fig. 6G)

Derivatio nominis.
In the Tunebo language, spoken by the indigenous people that inhabit the region where the species was collected, lokura means spider. Noun in apposition.
Distribution. Known just from the type locality, PNN Tamá, Norte de Santander department, in the Páramo biogeographic province (Morrone et al. 2022) (Fig. 18). Diagnosis. Mesotergal area III with large paramedian spines (the remaining species of Paravima, except P. morritomacairensis, without ornamentation or with conical, mammilliform or domed tubercles). Can be distinguished from that species by the coloration of mesotergal areas I-IV darker than the rest of the DS (concolorous in P. morritomacairensis).  (Figs 7B, 8A). Mesotergum delimited, darker than the rest of the DS, divided into four areas, with some granules. Area I divided into two halves, each one with a pair of tubercles; area II-IV undivided; II with a pair of medium-sized tubercles close to the medial axis of the body (Figs 7B, C, 8A); III with a pair of high paramedian spines (Figs 7C, E, 8A, B); IV with four small tubercles. Posterior border of scutum substraight, with a row of tubercles. Free tergites I-II and anal operculum with a few tubercles (Fig. 8A). -Venter: Coxae I-IV with some granules (Fig. 7D). Coxa I with a longitudinal row of tubercles at the medial region, being the third larger than the others, three large tubercles reaching the anterodistal region, and one small tubercle close to the posterodistal margin (Fig. 8C); coxa II longer than coxa I; coxa III longer than I and II; coxa IV backward projected (Fig. 7D). Sternites and anal operculum with a few small tubercles. Stigmatic area smooth. Stigmata large, oval and transverse (Fig. 7D). -Chelicerae: Segment I rectangular, with well-marked bulla (Fig. 7B, C), two small ectal subdistal tubercles, and two separated tubercles on the proximal border (Fig. 7B). Chelicera not swollen, monomorphic (Figs 7C, E, 8A). Hand with some sparse setiferous tubercles of different sizes on the frontomesal region, a group of setae on the mesal face of the fixed finger, and some setiferous tubercles close to the base of the fixed finger (Fig. 8D). Fixed finger with the inner surface irregularly dentated. Movable finger with one trapezoid, sub-basal tooth and with the inner surface dentate (Fig. 8D). -Pedipalps: Longer than DS length, smooth. Tr ventrally with two subapical tubercles. Fe with a ventromesal row of six setiferous tubercles (the second and the third larger than the others), and one large ventroectal setiferous tubercle in the distal portion (Figs 7C, 8E). Pa with one large mesal setiferous tubercle. Ti ectal II, mesal iII. Ta ectal IIi, mesal IIii (Fig. 8E). -Legs: Legs I-IV smooth. Leg I filiform, the rest, getting steadily thicker from leg II to IV. Fe IV four times DS length (Fig. 7A). Tarsal (Fig. 8H). MS-D1-2 short, located in an oblique line on a keel between the LP and the base of the stylus (Fig. 8G, H). MS-E2 medium-sized and branched (Fig. 8F). Stylus straight, surpassing the lamina parva, tip truncate (Fig. 8H). -Color (in ethanol): Carapace reticulated dark brown (59) on light orange yellow (70). Mesotergum dark brown (59). Pedipalps and chelicerae deep yellowish brown (75). Posterior border and free tergites Dark Olive (108). Legs I-IV dark yellowish brown (78). (Fig. 19G) is quite different, with carapace, chelicerae, pedipalps, coxae and trochanters strong reddish brown (40); laterals of DS vivid orange yellow (66); mesotergum and medial region of free tergites brownish black (65); free tergites vivid yellow green (115); and the rest of the legs brownish black (65). -Female: Chelicerae of the same size as male (monomorphic). Femora slightly thinner than in male.

Derivatio nominis. The species name honors professor
Eduardo Flórez, curator of the ICN Arachnological collection. Professor, in Latin magister, (magistri in genitive case), in recognition of his excellent work in the teaching and development of arachnology in Colombia and his interest in the divulgation of scientific knowledge about the arthropods in the country.

Distribution. Known just from the type locality, RN Los
Tunos, Cundinamarca department, in the Magdalena biogeographic province (Fig. 18).

Figs 9, 18
Paravima morritomacairensis González-Sponga, 1987: 482, figs 618-623;Kury, 2003: 31. Diagnosis. Paired tubercles on the mesotergal area III spiniform, gradually sharpening (shared with P. magistri sp. nov., from which it can be distinguished by the absence of a dark spot in the mesotergal areas). The remaining Paravima species lack ornamentation or have conical, mammilliform or domed tubercles). González-Sponga (1987). Here we offer a complementary description (based on MIZA 0105923 and González-Sponga 1987: figs 622, 623): Margins of DS, chelicerae, and pedipalps variegated (Fig. 9A). Anterior and lateral margins of DS smooth, posterior margin with some granules (Fig. 9A, B, D). Area III with a pair of paramedian mammilliform tubercles (higher than ocularium) darker than the rest of the mesotergum (Fig.  9A-C). Anterior margin of coxa I with a high proximal conical tubercle (Fig. 9E)  Remarks. One of the males of MIZA 0105923 was dissected but the genital was not together with the male; another male was without genital operculum nor penis. MCNC material seems to be lost (not found in MAGS collection). We suspect that MAGS 504 corresponds to the material cited as MAGS, without number. The figures of the penis by González-Sponga (1987: figs 622, 623) does not show MS-D, but this is probably an omission (Paravima typically has two pairs of MS-D). Specimens of this species are extremely similar to P. locumida and a detailed review of the genitalia and variability of both taxa would be useful to determine their identity. (Goodnight & Goodnight, 1949) comb. nov.
Description. See González-Sponga (1987). Here we offer a complementary description based on (MNRJ 9328* and MIZA 0105868): Anterior and lateral margins of DS smooth, posterior margin with some granules. Margins of DS variegated; posterior region of carapace, areas I and IV, the laterals of areas II-III, and the anterior region of area V with a dark spot (Fig. 10A-D). Mesotergum delimited, divided into four areas: area I divided into two halves; areas II-IV undivided; areas I-IV with a pair of paramedian low tubercles, those of area III domed (Fig.  10B, D). Cheliceral hand swollen. Pedipalpal segments slender (Fig. 10A-D). Legs increasing in thickness from leg I to leg IV. Leg I filiform. Legs III and IV with granules (Fig. 10A, B). Coxa I with one proximal conical tubercle at the anterior margin and a longitudinal row of tubercles at the medial portion (Fig. 10E). Fe IV four times DS length (Fig. 10A).
Diagnosis. Paired tubercles on the mesotergal area III small and conical (in P. lokura sp. nov. large), the remaining Paravima lack ornamentation or have mammilliform, domed tubercles or spines).  (Fig. 11A, B, D). Mesotergum ill delimited, divided into four areas: area I divided into two halves; areas II-IV undivided; area II with a pair of low tubercles centrally located; area III with a pair of paramedian conical tubercles (higher than ocularium) darker than the rest of the mesotergum (Fig.  11A-C). Anterior margin of coxa I with a high proximal conical tubercle (Fig. 11E). Fe IV four times DS length. Malleus of penis with one pair of branched MS-B; three pairs of branched MS-A; two pairs of medium-sized MS-D; one pair of MS-E2 large and branched. Stylus elongated and sinuous, surpassing the lamina parva ( fig. 4).

Remarks.
We suspect that MAGS 103 corresponds to one of the male paratypes cited in type data as MAGS, without number. The synonym of P. acanthoconus with P. propespelunca is based in: (1) The drawing in the original description (González-Sponga, 1987: fig. 625) presents tubercles on area III with a slightly mammilliform shape which do not fully correspond to what is observed in the studied material (Fig. 11B-D). However, some unnoticed variations of the shape of these tubercles between different specimens were found (Fig. 12); (2) The schematic drawing of the genital morphology (figs 628, 629) shows a supposed reduction in the number of MS-A, absence of MS-D1-2, and a short and erected style, barely exceeding the height of the LP. Nevertheless, after examination of the male genitalia of P. propespelunca, we noted that this matches perfectly with that of P. acanthoconus; (3) Villarreal-Manzanilla & DoNascimiento (2005: fig. 3b) show that the pedipalpal femur of P. acanthoconus bears four spiniform tubercles (against five in P. propespelunca, very similar in its external morphology), but, after revision of new material, we noted that this is a variable character even in the same population or when comparing right and left pedipalp of the same individual, so, it is not a reliable character to differentiate species. Diagnosis. Mesotergal area III unarmed (shared with P. quirozi comb. nov.). The remining Paravima species have paired conical, mammilliform, spiniform or domed tubercles. Mesotergal areas with a semicircular or elliptical dark spot on the areas II and III.
Description. See González-Sponga (1981, 1987.  two-thirds of area III (Fig. 13A-E; figs 1a, b). Mesotergum delimited, divided into four areas: area I divided into two halves; areas II-IV undivided; areas I-IV with a pair of paramedian low tubercles, those of area III domed; areas II-IV with a wide hump (Fig. 13B-E; figs 1a, b). Pedipalpal segments slender ( Fig. 13C; fig. 3). Legs increasing in thickness from leg I to leg IV. Leg I filiform. Legs III and IV with granules ( Fig. 13C-E; fig. 4). Patella IV with some conspicuous distal tubercles ( fig. 4). Tibia III-IV distally thickened. Fe IV is four and a half times DS length. Penis with small LP (height shorter than width), apex with anterolateral crescent-shaped corners Remarks. The differences in shape and size of the dark spot in the mesotergal areas of P. quirozi seems to be populational: specimens from Maimire (Yaracuy) have more rounded spots, individuals from La Guáquira (Yaracuy) exhibit wider and shorter spots (Fig. 19A, B) and specimens from Carabobo have an intermediate state. On the other hand, after revision of the male genitalia of paratypes of P. quirozi, we observe the presence of two pairs of MS-D and three pairs of MS-A, contrary to the original drawings of G-S that lacks MS-D and show just two pairs of MS-A (González-Sponga, 1981: fig. 13). Besides that, one of the male paratypes of P. quirozi from Yaracuy (MAGS 294) exhibits a gap between MS-A1 (most dorsal MS-A) and MS-A2/MS-A3 (located more basally). Not finding genital differences between the males studied, we prefer to synonymize both species and understand spot shape differences as intraspecific variations. Villarreal-M & Rodriguez-Manzanilla (2003: 181) suggested a morphological similarity between Avima vigirima syn. nov. Paravima plana comb. nov. and Paravima quirozi comb. nov., which appears reflected in our analyzes (Figs 16,17). Posterior border of scutum straight, with a row of small tubercles. Free tergites and anal operculum with some tubercles (Figs 14D, 15A). -Venter: Coxae I-IV with some granules (Fig. 14D). Coxa I with one medial tubercle on the anterior margin, a group of five tubercles on the proximal region, three large tubercles reaching the anterodistal margin, and two tubercles close to the posterodistal region (Fig. 15C); coxa II longer than coxa I; coxa III longer than I and II; coxa IV backward projected (Fig. 14D). Sternites and anal operculum with a few small tubercles. Stigmatic area smooth. Stigmata large, oval and transverse (Fig. 14D). -Chelicerae: Segment I rectangular, with well-marked bulla (Fig. 15A, B), one small ectal subdistal tubercle, and two tubercles on the proximal border (Fig. 14B). Chelicera swollen (Figs 14C, E, 15B). Hand with one tubercle at the mesal subapical region, a group of nine transversal tubercles going from ectal to mesal region, and a group of setiferous tubercles close to the base of the fingers (Fig. 15D). Fixed finger with some teeth on the inner surface. Movable finger with one trapezoid, sub-basal tooth and with the inner surface dentate (Fig. 16D). -Pedipalps: Longer than DS length, smooth. Tr ventrally with one subapical setiferous tubercle. Fe with a ventromesal row of five setiferous tubercles (the basalmost larger than the distalmost), and one large ventroectal setiferous tubercle in the distal portion (Fig.  14C, D). Pa with one large mesal setiferous tubercle. Ti ectal iII, mesal IIi. Ta ectal IIi, mesal IIi (Figs 14B, D). -Legs: Legs I-IV smooth. Leg I filiform, the rest, getting steadily thicker from leg II to IV (Fig. 14A). Fe IV four times DS length. Tarsal counts: 6(3)-6(3)/?-14(3)/7-7/7-7. -Penis: LP short and depressed, half-moon shaped, with anterolateral corners dorsally pointed (Fig. 15E-G). Malleus carrying three pairs of branched MS-A and one pair of branched MS-B (Fig. 15G). MS-D1-2 long, located in a vertical line on a keel between the LP and the base of the stylus (Fig. 15E, G). MS-E2 large and branched (Fig. 15F, G). Stylus sinuous, surpassing the lamina par-  Derivatio nominis. Totoro is a character in the Japanese animated fantasy film My Neighbor Totoro (directed by Hayao Miyazaki and animated by Studio Ghibli), being a friendly wood spirit in post-war rural Japan. For us, the paramedian armature of the new species resembles the ears of the charismatic Totoro. We take advantage of exalting the excellent work of Studio Ghibli with this tribute. Noun in apposition.
However, the internal relationships within Paravima have not been fully resolved (Figs 16, 17), mainly due to the limitation of not being able to study the genitalia of some species and the low degree of detail of previous descriptions. This led us to focus our analysis on the ge-neric relationships of Paravima with other Leiosteninae and the generic rediagnosis.
Paravima magistri appears as sister group to the rest of the species, supported by (a) the proportion of interocular distance vs width of carapace; (b) the unmodified shape of the spines of area III and (c) MS-A quantity (only two pairs vs three pairs). The remaining taxa are clustered in a polytomy formed by two groups of species and two isolated species. The more specious group gathers most of the species with area III armed, from sectors further east of Cordillera de la Costa in Venezuela (except P. totoro). P. plana and P. quirozi, the unarmed species, are clustered for sharing, in addition to the absence of ornamentation on mesotergal area III [17(0)], the elongation of the base of the LP [47(0)] (Fig. 16).
Due to the unarmed mesotergal area III, these species were historically related to Avima. However, our study confirms the degree of homoplasy of this character. Other morphological similarities, which are now known to be shared with Paravima, were first detected by Villarreal-M & Rodríguez-Manzanilla (2003), but not tested in a phylogenetic analysis. Our results strongly support the monophyly of Paravima, but the internal relationships within the genus are not stable enough to propose any infrageneric or supraspecific taxonomic organization at this time.
The interpretation of some states of character in some species was difficult, due to the unavailability of genitalia specimens for study (e.g. P. morritomacairensis, P. locumida) which necessitated reliance on schematic drawings from the literature, e.g. MS D and MS E, being hard to discern between if these drawings are incomplete or if certain genital structures truly do not exist. Additionally, limited information on morphological variability (Fig. 12) has led to the description of species that we have now synonymized. The case of Avima venezuelica Soares & Avram, 1981, is particularly noteworthy. Some genital and external characters, such as the shape of DS or the LP of the penis, seem to indicate a possible relationship of this species with Paravima. However, its genitalia is only known from a simplified schematic drawing that is not fully informative. The phylogenetic position of A. venezuelica remains uncertain until the genitalia can be studied. A taxonomic decision regarding this species will require expanding the generic diagnosis to include it or creating a monotypic genus, as it does not conform to the Avima core diagnosis as defined by García et al. (2022a).
A parallel analysis under equal weights offers interesting results (Fig. 17A), recovering the monophyly of the genus and even get the same internal relations. The phylogenetic relationship with other genera is perhaps the main point of discrepancy between both analyzes (Fig. 17A). This inconsistency in the trees, as well as the degree of polytomy, suggest that our understanding of the relationships in Leiosteninae is still far from optimal, however, our approach allowed for improved diagnosis of the genus, the identification of monophyletic groups with biogeographic congruence, and provides direction for future studies to refine the hypotheses of relationships.

About the geographic distribution and habitats of Paravima
Little is known about the natural history of the Paravima species, with limited information available on the biomes and microhabitats they occupy. From a biogeographical approach, most species are restricted to the biogeographic Venezuelan province in the Cordillera de la Costa in Venezuela. However, two new species have been recorded in the Magdalena province in Colombia (Fig. 18), expanding the known distribution of the genus by approximately 500 km southwest and across other biogeographical provinces. Some groups of Neotropical Opiliones have a restricted geographic distribution, and they are associated with few or even a single biogeographical region (e.g. the cosmetids of the genus Neocynorta Roewer, 1915, in the Venezuelan Andes), however, this extreme degree of endemism may also be related with the lack of sampling to cover these false gaps, what is known as the Wallacean shortfall (Whittaker et al. 2005, Bini et al. 2006. Paravima species occupy a wide range of altitudes and diverse biomes. Many of the species have been recorded from Montane Cloud Forests at altitudes between ~1200-1945 m a.s.l. (P. totoro, P. plana, P. goodnightorum) (Fig. 19F) or even higher (P. lokura and P. magistri in Andean Montane Forests, between 2170-2300 m a.s.l.) (Fig. 19H, I). Paravima quirozi is known from Subhumid tropophilous forest at 1200 m a.s.l., while P. goodnightorum is known from riparian forests at 160-900 m a.s.l. (Fig. 19C), in areas with high impact from human activities such as urban sprawl, agriculture, slash and burn (González-Sponga 1997, Villarreal-M & Rodríguez-Manzanilla 2003). Paravima morritomacairensis was described from semi-deciduous forests at ~850 m a.s.l. grown on a calcareous formation composed of three massifs, with numerous caves crossed by watercourses. Paravima propespelunca has been collected in transitional evergreen forests with marine influence (González-Sponga 1997).
Two major problems have affected the taxonomy of Paravima: (1) the use of the Roewerian taxonomic system and, (2) the incomplete original descriptions/drawings. The armature of the scutal area III was used for the creation of artificial groups, i.e. all the agoristenid without any armature on area III were classified as Avima, ignoring other characters, as the genital morphology (see García et al. 2022a). Following that criteria, a lot of unrelated species were grouped, including three that after a phylogenetic analysis have been transferred here to Paravima. Morphological homoplasy in some groups can make appearances deceiving. Additionally, the deficient/incomplete species descriptions have hindered the interpretation of characters and limited the degree of resolution in morphological analyses. Our analysis provides an important step in the diagnosis and redefinition of the genus; however, as new descriptions of the genitalia of the species are made, more robust relationship hypotheses with better resolution can be obtained and taxonomic decisions can be made more simply.

Acknowledgments
We are grateful to Quintín Arias and Vilma Savini (MIZA, Maracay, Venezuela), for the support received during the study of the material in