COI mtDNA barcoding and morphology for the description of a new species of ricinuleid of the genus Pseudocellus (Arachnida: Ricinulei: Ricinoididae) from El Triunfo Biosphere Reserve, Chiapas, Mexico

A new species of epigean ricinuleid of the genus Pseudocellus Platnick, 1980 from El Triunfo Biosphere Reserve, Chiapas, Mexico is described. DNA barcoding utilizing mitochondrial cytochrome c oxidase subunit 1 (CO1) and morphology were used for species delimitation. Molecular analyses and species delimitation included four methods: 1) General Mixed Yule Coalescent model (GMYC), 2) Automatic Barcode Gap Discovery (ABGD), 3) Bayesian Poisson Tree Process (bPTP), and 4) Assemble Species by Automatic Partitioning (ASAP). All molecular methods and morphology were consistent in delimiting and recognizing the new species described herein. The average interspecific genetic distance (p-distance) among analyzed species of Pseudocellus was 11.6%. The species is described based on adult males and females: Pseudocellus giribeti sp. nov. This is the seventh species described from Chiapas, which holds the highest number of ricinuleids species for the country. The total number of described species of Pseudocellus from Mexico increases to 21, having the highest species diversity of known ricinuleids worldwide.

In modern taxonomy, it is generally accepted that the use of multiple lines of evidence, such as morphology, ecology, reproduction, and biogeography (integrative taxonomy), are more effective in the diagnosis, delimitation, and even the description of species (Valdez-Mondragón 2020). Several recently developed molecular delimitation methods have highlighted the extensive inconsistency in classical morphological taxonomy used on different groups of arachnids (Hamilton et al. 2011(Hamilton et al. , 2014(Hamilton et al. , 2016Ortiz & Francke 2016;Tahami et al. 2017;Cruz-López et al. 2019;Valdez-Mondragón et al. 2019;Navarro-Rodríguez & Valdez-Mondragón 2020). However, the molecular delimitation methods using mitochondrial markers VALDEZ-MONDRAGÓN A. & CORTEZ-ROLDÁN M.R., New Mexican species of Pseudocellus 3 have never been used in ricinuleids so far. Molecular methods have provided a new way to resolve species delimitation problems by using the infraspecifi c genealogical information in DNA markers which provides objective implementation of modern species concepts (e.g., biological, phylogenetic, genotypic cluster) (Valdez-Mondragón et al. 2019). For species delimitation research, the appropriate way is to analyze the data with a wide variety of methods and different lines of evidence to delimit lineages, which must be consistent across the results, to understand the behavior of the molecular species delimitation methods and thus contribute to integrative taxonomy (Carstens et al. 2013;Luo et al., 2018).
In this study we describe a new epigean species of the genus Pseudocellus from a cloud forest of the state of Chiapas, Mexico, based on morphological evidence and DNA barcoding (CO1). This study currently represents the fi rst one using different molecular species delimitation methods including morphology with ricinuleids.

Biological material
The collected specimens were preserved in 80% ethanol. The type material are deposited in the Invertebrate Zoology collection in the Museum of Comparative Zoology, Harvard University, USA (MCZ), and in the Colección Nacional de Arácnidos (CNAN), Institute of Biology, Universidad Nacional Autónoma de México (IBUNAM), Mexico City. The descriptions and observations of the specimens were done using a Zeiss Discovery V8 stereo microscope. A Zeiss Axiocam 506 color camera attached to a Zeiss AXIO Zoom V16 stereo microscope was used to photograph the different structures of specimens. All measurements are in millimeters (mm). Male copulatory apparatus and female spermathecae were dissected previous to observation. Photography was conducted with specimens and structures submerged in commercial gel alcohol (to hold them in the appropriate position), and the preparation completely covered with 80% ethanol. Digital images were edited in Adobe Photoshop CS6. The distribution map was produced using QGIS 3.18.3 "Zürich" (QGIS.org). The terminology used for leg segments follows Gertsch (1971), whereas for the copulatory structures we follow Pittard & Mitchell (1972) and Valdez-Mondragón et al. (2018, 2020. The length/diameter (l/d) ratio of femur II of males was calculated in prolateral view. The abbreviations below are used for some copulatory structures following Botero-Trujillo & Valdez-Mondragón (2016).

Taxon sampling
The molecular analyses presented here are based on a total of 30 samples of 12 species, including the new species described herein and one single outgroup: Ricinoides aff. feae (Hansen, 1921) used only to root the trees. The ingroup includes 29 samples of Pseudocellus. The samples used in the molecular analyses are listed in Table 1. All the CO1 sequences were downloaded from GenBank (Table 1). The total process of DNA extraction, sequencing and sequence-editing was described by Murienne et al. (2013) and Fernández & Giribet (2015). DNA voucher numbers correspond to the database of the Museum of Comparative Zoology (MCZ) (mczbase.mcz.harvard.edu), Harvard University (Table 1).

DNA sequence alignment and editing
Sequences were aligned using the default Gap opening penalty of 1.53 in MAFFT (Multiple sequences Alignment based on Fast Fourier Transform), ver. 7 (Katoh & Toh 2008), available online, using the following alignment strategy: Auto (FFT-NS-2, FFTNS-i or L-INS-i; depending on data size). Inspection and editing of sequences and alignments were done using GENEIOUS ver. 10.2.6. The matrix obtained from the multiple sequence alignments was then used in the different analyses.

Starting trees under Maximum Likelihood (ML)
The approaches for DNA barcoding tree-based delimitation explicitly use the phylogenetic and cladistic species concepts. A starting tree was inputted with ML using MEGA ver. 7.0. The analysis recognizes a monophyletic cluster by searching differential intra-and interspecifi c branching patterns (Ortiz & Francke 2016; Valdez-Mondragón 2020). The ML analysis was calculated with the following parameters: Number of replicates = 1000, Bootstrap support values = 1000 (signifi cant values ≥ 50%), Models of sequence evolution selected using jModelTest = GTR, Rates among sites = G+I, No. of discrete Gamma Categories = 6, Gaps Data Treatment = Complete deletion, Select Codon Position = 1 st +2 nd and 3 rd , ML Heuristic Method = Subtree-Pruning-Regrafting -Extensive (SPR level 5), Initial Tree of ML = Make initial tree automatically (Default -NJ/BioNJ). The models of sequence evolution were selected using jModelTest ver. 2.1.10 with the Akaike information criterion (AIC) (Posada & Buckley 2004). The models selected for CO1 for each partition block were GTR+G+I (1 st and 2 nd codon positions) and GTR+G (3 rd codon position).

P-distances under neighbor joining (NJ)
The Bootstrap values in NJ analysis were calculated with the following commands: Number of replicates = 1000, Bootstrap support values = 1000 (signifi cant values ≥ 50%), Substitution type = nucleotide, Method = p-distance, Substitution to Include = d: Trasitions+Transeversions, Rates among Sites = Gamma distributed with Invariant sites (G+I), Missing Data Treatment = Pairwise deletion, Select Codon Position = 1 st +2 nd and 3 rd .

General mixed yule coalescent model (GMYC)
The GMYC species delimitation method applies single (Pons et al. 2006) or multiple (Monaghan et al. 2009) time thresholds to delimit species in a Maximum Likelihood context, using ultrametric trees (Pons et al. 2006;Ortiz & Francke 2016). Phylogenetic analyses were run in BEAST ver. 2.6.0 ) using a coalescent (constant population) tree prior. As independent lognormal relaxed clock was applied to each partition, and analyses of 20 × 106 generations were run. Convergence was assessed with TRACER ver. 1.6 (Rambaut & Drummond 2014). TreeAnnotator ver. 2.6.0 (BEAST package) (Rambaut & Drummond 2014) was used to build maximum clade credibility trees, after discarding the fi rst 25% of generations as burn-in. Following gene tree inference, GMYC was implemented in the web interface for single and multiple threshold GMYC (https://species.h-its.org/gmyc/). The backend European Journal of Taxonomy 778: 1-25 (2021) 6 of this web server runs the original R implementation of the GMYC model authored by Fujisawa & Barraclough (2013).

Automatic Barcode Gap Discovery (ABGD)
The ABGD species delimitation method uses recursive partitioning with a range of prior intraspecifi c divergence and relative gap widths, estimating the threshold between intra-and interspecifi c genetic variation, generating species-level groupings (Ortiz & Francke 2016). ABGD analyses were conducted using both uncorrected and K2P distance matrices with default options: Pmin = 0.001, Pmax = 0.1, Steps = 10, Relative gap width (X) = 1, Nb bins = 20. The PTP species delimitation method (Zhang et al. 2013) is similar to GMYC, but uses substitution calibrated (not ultrametric) trees to avoid the potential fl aws in constructing time calibrated phylogenies (Zhang et al. 2013;Ortiz & Francke 2016).

Bayesian Poisson Tree Process (bPTP)
The Bayesian variant of the method (bPTP) was employed in the online version (https://species.h-its.org/ptp/). It was run on the Bayesian gene trees with default options: rooted tree, MCMC generations = 100 000, Thinning = 100, Burn-in = 0.1, Seed = 123. The trees generated were edited using Adobe Photoshop CS6.

Assemble Species by Automatic Partitioning (ASAP)
The ASAP species delimitation method proposes partitions of species hypotheses using genetic distances calculated between DNA sequences (Puillandre et al. 2021). ASAP analyses were conducted using Kimura (K80) distance matrices ts/tv: 2.0.

Molecular analyses and species delimitation
The analyzed matrix of CO1 includes 30 terminals (including one outgroup) and 655 aligned characters. Specimens used in this study, GenBank accession numbers and localities of the sequences used are listed in Table 1. The multiple molecular methods consistently separate the new species European Journal of Taxonomy 778: 1-25 (2021) 8 Table 2. Genetic p-distance under NJ between Pseudocellus giribeti sp. nov. and species from Central America (Honduras and Guatemala 1) used in the analyses. Average p-distance = 9.8%. (congruence among methods) and according to the morphology of sexual structures traditionally used for identifi cation and delimitation at the species level in ricinuleids (copulatory apparatus of leg III of the male, femur and tibia II of the male and female spermathecae) (20)(21)(22)(23)(24)(25)(26)(29)(30)(31). Although the morphology was not checked for the other species included in the molecular analyses (Fig. 2), the multiple molecular methods consistently separate the rest of the species (Fig. 2). The average genetic distance under NJ among analyzed species of Pseudocellus was 11.6%. Bootstrap support values were high for the new species (99%) (Fig. 1). Pseudocellus giribeti sp. nov. is closely related to the putative species composed by specimens of Guatemala 1 and Honduras ( Fig. 1; Fig. 2, blue branch) with an average p-distance of 9.8% between both species (Table 2).
Initially, GMYC was proposed by Pons et al. (2006) as a molecular method for species delimitation in a Maximum Likelihood context and based on ultrametric trees. The method was found to be robust in a range of departures from its simplifying assumptions (Fujisawa & Barraclough 2013;Talavera et al. 2013). Herein, this species delimitation method was robust and concordant for the analyzed species of Pseudocellus, recovering the new species based on traditional morphology (Fig.  2). ABGD is a molecular species delimitation method that uses cursive partitioning, with a range of prior intraspecifi c divergence and relative gap width, to estimate the threshold between intra-and interspecifi c variation and then intends to produce species-level groupings (Ortiz & Francke 2016;Valdez-Mondragón 2020). However, this species delimitation method is sensitive to sampling effect, as was demonstrated in the spider genera Aphonopelma Pocock, 1901 and Bonnetina Vol, 1999 from North America by Hamilton et al. (2014) and Ortiz & Francke (2016), respectively, tending to give moderately over-splitting solutions. In this work, the ABGD method was consistent in separating Pseudocellus giribeti sp. nov. with recursive partitions (RP) and initial partitions (IP), and the species is recovered as one single species, which corresponds and is concordant with the morphology (Fig. 2). PTP is similar to GMYC but uses substitution calibrated (not ultrametric) trees to avoid the potential fl aws in constructing time-calibrated phylogenies, as was proposed by Zhang et al. (2013). In this work, we employed the Bayesian variant of the method (bPTP), which was also concordant with morphology and the other molecular methods, recovering a total of 11 species (Fig. 2). The ASAP is a species delimitation method that proposes partitions of species hypotheses using genetic distances, which are calculated between DNA sequences (Puillandre et al. 2020). This method is similar to ABGD, using cursive partitioning with a range of prior intraspecifi c divergence and relative gap width, which is used to estimate the threshold between intra-and interspecifi c variation, producing species-level groupings. This method also recovers the new species Pseudocellus giribeti sp. nov. (Fig. 2). However, this method was discordant with the other methods, over-estimating the number of total number of species as 13 in the case of the samples from Guatemala 1 (KR180443.1-KR180445.1) and Honduras (KR180449.1), which are considered as three species under this approach (Fig. 2).

Taxonomy
Clase Arachnida Cuvier, 1812 Order Ricinulei Thorell, 1876Suborder Neoricinulei Selden, 1992Superfamily Ricinoidoidea Ewing, 1929 Family Ricinoididae Ewing, 1929 Genus Pseudocellus Platnick, 1980 Type species is shorter and slightly oval (Fig. 1), whereas that of P. alux is slightly longer and less oval (Armas & Agreda 2016: fi g. b). The carapace in the new species is as short as in P. alux (Figs 3, 5). Males of the new species have the tibia of the pedipalp deeply curved on the prolateral part (Fig. 19), whereas in P. alux the tibia of the pedipalp is slightly curved (Armas & Agreda 2016: fi g. c). Males of P. giribeti sp. nov. and P. alux have a similar shape of femur II, but the males of the new species have a longer femur II than P. alux, with proportions of 3.45 and 2.17 times as long as wide, respectively (

Etymology
This species is dedicated to Dr Gonzalo Giribet from Harvard University in recognition of his contribution to the knowledge of ricinuleid phylogenetics and the loan of the specimens. COLORATION. Cucullus, carapace and sternal region dark reddish (Figs 3-6). Legs I, III and IV lighter dark reddish than legs II; tarsomeres paler reddish than other segments. Opisthosoma reddish dorsally, darker ventrally on sternites XI and XII (Figs 7-8).  (Fig. 5). Slightly longer than wide, trapezoidal, widest at posterior margin between coxae II and III. Tegument covered with abundant, fi ne translucent setae and granules. Anterior margin straight, lateral margins not parallel, narrowing anteriorly; posterior margin procurved. Inconspicuous translucent areas present at level of trochanters II. Carapace with seven depressions: one median along, and three lateral depressions at level of each coxae III. CUCULLUS (Fig. 11). As wide as long, widest distally; anterior margin straight, lateral margins rounded at anterior corners. Tegument covered with abundant translucent setae and granules similar to those on carapace. One slight depression each side close to anterior margin, cuticular pits absent. Distal margin with long, translucent setae. CHELICERAE (Fig. 12). Movable fi nger with eight teeth, basal one larger than others, which are equal in size. Fixed fi nger with fi ve teeth, subequal in size, distal tooth longer. Both fi ngers with long translucent setae basally.  STERNAL REGION (Fig. 6). Coxae covered with translucent setae and granules similar to those on carapace. Coxae I rhomboidal, II sub-rectangular, III and IV conical. Coxae II larger and wider than others, coxae IV smallest. Coxae II meeting tritosternum along ⅓ of its length. Coxae II anterior and posterior margins perpendicular to median axis of prosoma; coxae III slightly oblique, their posterior margins forming an obtuse angle (>90°) with each other; coxae IV oblique, their posterior margins forming an obtuse angle (little more >90°) with each other.
OPISTHOSOMA (Figs 3-4, 7-8). Longer than wide, widest at level of tergites XII and XIII. Tegument covered with abundant setae and granules similar to those on carapace; cuticular pits absent. Median plates of tergites XI-XIII with paired longitudinal depressions, those of tergite XI the smallest. Tergite X widest and shortest. Median tergites XI and XII trapezoidal, wider than long; median tergite XIII slightly longer than wide, with posterior corners curved. Lateral tergites in oblique position; X smallest and XIII largest. Lateral tergites XI trapezoidal, XII square and XIII triangular. Sternites XI-XIII with paired depressions. Sternites XI and XII dark medially, XIII paler reddish. Pygidium segments without notch (Fig. 10).
PEDIPALPS (Figs 6,(18)(19). Coxae with fi ne translucent setae and rounded granules. Trochanters 1 rounded, with sparse fi ne translucent setae and few granules restricted to ventrodistal half; trochanters 2 conical, ventrally with basal setae and granules (similar to those on 1 st ). Femora curved distally and wider proximally, without granules, with deep prolateral concavity distally close to tibial joint; tegument with abundant translucent setae, thinner and longer on prolateral and distal surfaces. Tibiae deeply curved on prolateral part, with numerous thin translucent setae longer distally. Tibiae without granules. Movable claws curved and longer than fi xed claws, which are conical. Movable claws without teeth, fi xed claws with small teeth. LEGS. Without cuticular pits but with translucent setae and rounded granules on all segments (Figs 3-4). Legs II noticeably longer and wider (Figs 3-4). Femora II almost straight, slightly curved on prolateraldistal part, wider and longer than the others; femora II ventrally with few scattered sharp-tipped granules (Fig. 17). Tibiae II ventrally with two sharp and small apophyses on basal part (Figs 3-4, 13-14). Metatarsus II with scattered sharp-tipped granules, not forming rows, ventral ones longer and more conspicuous than dorsal one, metatarsus curved ventrally (Figs 15-16). Tarsomeres of leg II dorsally with sharp-tipped granules, tarsomeres 5 with fewer granules not sharp-tipped. Femora I, III and IV ventrally with few sharp-tipped granules. Patellae I, III and IV with normal granules. Tibiae I, III and IV ventrally with few sharp-tipped granules distally. All metatarsi dorsally with V-shaped invaginations distally; metatarsi I, III and IV without granules ventrally, only few slightly sharp-tipped granules dorsodistally on metatarsi I and IV. Tarsomeres of legs I, III and IV without granules.
LEGS III AND COPULATORY APPARATUS. Metatarsi almost square-shaped in prolateral view, with numerous long translucent setae; metatarsal processes (mP) long and slightly sigmoidal in pro and retrolateral views (Figs 20-21), curved and hook-shaped in dorsal view (Fig. 22). Lamina cyathiformis (Lc) of tarsomeres 2 leaf shaped, with a small notch basally in retrolateral view (Fig. 20), Lc with long translucent setae throughout. All tarsomeres ventrally with long translucent setae, which are more numerous and longer on tarsomere 4. Tarsomeres 4 with two curved claws . Copulatory apparatus sigmoid shaped in dorsal view (Fig. 24), with tarsal process (tP) curved, arc-shaped (Figs 23,25); accessory piece (ac) of copulatory apparatus curved, with sharp tip (Fig. 23). Tip of the tP bifurcated, with a long and conical lobule 1 (L1) (Fig. 26).  Differs from male as follows: carapace paler reddish, median longitudinal depression shallower than on male. Femora II less robust than those on male. Tibiae II without ventral conical apophyses, but with some sharp-tipped granules, femora II slightly curved on prolateral-distal part. Metatarsus II with the scattered sharp-tipped granules smaller than those on male. Metatarsus II straight. Opisthosoma wider, higher and larger than in male, paler reddish (Figs 27-28). Median tergites X-XIII wider than those on male (Fig. 27). Spermathecae with two oval lobules on each side, one as long as the other, with widespread tiny pores throughout (Figs 29-31).

Natural history
The specimens were collected manually in oak forest litter.

Discussion
The taxonomic science has at least three distinct very important roles in biology: 1) the assignment of individual organisms into pre-existing species categories (species assignment), 2) the assignment of species to higher categories (genus, family, orders, etc), and 3) the designation of new species categories to clasify individual specimens that do not fi t into an existing species category (species delimitation) (Rannala & Yang 2020). Traditionally, all this three roles have been performed by taxonomists using morphological characters in the majority of the cases. Modern taxonomy is changing, and currently uses multiple lines of evidence for species identifi cation and delimitation ( tasks to which DNA barcodes markers are being deeply applied in modern systematics and taxonomy, fi rst distinguishing between 'sibling' species (equivalent to species identifi cation or species diagnosis), and second, the use of molecular data to discover underestimated biodiversity (equivalent to species delimitation and species description). How species boundaries are fi xed is a central debate in modern systematics, which has generated multiple molecular delimitation methods based on heuristic and parametric methods tending to solve the inconsistencies in classical taxonomy (Hamilton et al. 2011(Hamilton et al. , 2014(Hamilton et al. , 2016Ortiz & Francke 2016;Rannala & Yang 2020). The use of the mitochondrial cytochrome c oxidase subunit 1 (CO1) barcode gene has become useful as the standard animal barcoding marker in many taxonomic studies, including arachnids (Hebert et al. 2003(Hebert et al. , 2004Astrin et al. 2006;Correa-Ramírez et al. 2010;Hamilton et al. 2011Hamilton et al. , 2014Hamilton et al. , 2016Graham et al. 2015;Planas & Ribera 2015;Ortiz & Francke 2016;Tahami et al. 2017;Cruz-López et al. 2019;Valdez-Mondragón et al. 2019;Valdez-Mondragón 2020). However, the use of mitochondrial markers alone may lead to the overestimation of taxonomic diversity, as has been demonstrated in some tarantula genera (Ortiz & Francke 2019). In modern taxonomy, the use of multiple evidence is generally accepted and currently common such as morphology and molecular data (integrative taxonomy), which is more effective in the diagnosis, delimitation, and even the description of species. Recently, molecular species delimitation methods have provided a new way to resolve problems by using the infraspecifi c genealogical information, extensively using DNA barcoding markers (Hebert et al. 2003;DeSalle et al. 2005).
The CO1 barcode gene had never been tested previously for the molecular species delimitation of ricinuleids. However, this mitochondrial marker was used for the fi rst molecular phylogeny of Ricinulei by Murienne et al. (2012) and the fi rst phylogenomic study of the order by Fernández & Giribet (2015). In the present study, CO1 showed a good resolution at the species level, with an average genetic distance among the analyzed species of 11.6% and high Bootstrap support values for the new species (Figs 1-2). Also, the average intraspecifi c genetic distance is less than 1%, concordant with the morphology for most of the species delimitation methods. However, the molecular analyses are based only on a single mitochondrial molecular marker, so is not possible to establish phylogenetic relationships among the studied species that do not include all the species from Central America. The morphology of sexual structures used for identifi cation, diagnosis, and delimitation at the species level in ricinuleids is still strong evidence for the delimitation of species. Structures such as the copulatory apparatus of leg III of the male, femur and tibia II of the male, and female spermathecae are robust morphological characters for the identifi cation at the species level. In the case of the accessory piece of the copulatory apparatus (Figs 20-26), the shape and distal bifurcation seems to be specifi c for each single species, including the three genera, Pseudocellus, Cryptocellus and Ricinoides (Tuxen 1974;Botero-Trujillo & Pérez 2008Tourinho & Saturnino 2010;Pinto-da-Rocha & Andrade 2012;Botero-Trujillo & Valdez-Mondragón 2016;Botero-Trujillo & Flórez 2017;Botero-Trujillo et al. 2021a, 2021b. Additionally, sexual structures are informative at the species level, such as female spermathecae, of which the shape and the number of lobules on each side are diagnostic characters (Figs 29-31). The shape of the tibia II of the males of Pseudocellus also seems to be a diagnosis character (Figs 3-4, 13-14), although many species have similar apophyses. Perhaps a combination of sexual morphological and molecular data in a phylogenetic analysis is required to test the monophyly of the genera if the generic diversity is underestimated. Even with only a few molecular markers such as CO1, the monophyly of Cryptocellus was not recovered, which was demonstrated by Fernández & Giribet (2015).
Despite the growing knowledge of the diversity of the genus Pseudocellus in the last 10 years by Valdez-Mondragón & Francke (2011, Valdez-Mondragón et al. (2018, 2020 and Valdez-Mondragón & Juárez-Sánchez (2021), the diversity of the genus in Mexico is still poorly known. Intensive sampling is needed to know the diversity of ricinuleids. Mexico hosts the largest number of known troglobitic ricinuleids so far, which indicates that caves in Mexico could be a signifi cant source of new species, where some species of Pseudocellus are common and often collected.