Phylogeny of the genus Yumtaax Boucher (Coleoptera, Passalidae, Proculini): Taxonomic and evolutionary implications with descriptions of three new species

Abstract Yumtaax Boucher (Coleoptera: Passalidae) is an endemic genus from the temperate sierras of Mexico and includes six narrowly distributed species. Yumtaax species have been assigned to several genera of Passalidae throughout history, and a phylogenetic approach is necessary to understand species delimitation and interspecific relationships. This study reconstructed the molecular phylogeny of six Yumtaax morphotypes using parsimony and Bayesian analysis of DNA sequence data from the ribosomal nuclear gene region 28S and the mitochondrial gene regions 12S and cytochrome oxidase I (COI) in addition to morphological characters. Analyses recovered two well-supported Yumtaax clades (the Yumtaax laticornis and Yumtaax imbellis clades) that are possible sister lineages. One synapomorphic morphological character state and the geographic isolation of the group provide corroborative evidence for monophyly. Molecular phylogenetic analyses and traditional morphological examinations also resulted in the discovery of two undescribed Yumtaax species and the discovery of two separate evolutionary lineages (cryptic species) within Yumtaax recticornis. As a result we describe three new species (Yumtaax veracrucensis Beza-Beza, Reyes-Castillo & Jameson, sp. n., Yumtaax cameliae Beza-Beza, Reyes-Castillo & Jameson, sp. n., and Yumtaax jimenezi Beza-Beza, Reyes-Castillo & Jameson, sp. n.), redescribe two species (Yumtaax recticornis [Burmeister 1847] and Yumtaax laticornis [Truqui 1857]), and provide a key to all nine Yumtaax species. This study is one of two studies to use molecular data to evaluate the evolutionary relationships of a genus of Bess Beetles (Passalidae), an ecologically important insect group exhibiting low morphological variability and heretofore lacking molecular phylogenetic study.


Introduction
Yumtaax Boucher (Coleoptera: Passalidae: Proculini) is an endemic genus of the southern and eastern Sierra Madre (Boucher 2006). As other members of the family Passalidae, these beetles feed on rotten wood and are important in the process of nutrient cycling in forests (Cano and Schuster 2012). Due to competition for this food resource with other Passalidae and resulting resource partitioning, Yumtaax species specialize on feeding in the periphery of large logs or on twigs and branches with a diameter less than 15 cm (Castillo and Reyes-Castillo 1984). Species of Yumtaax are associated with high altitude habitats such as cloud and pine-oak forests Reyes-Castillo 1984, Boucher 2006).
Yumtaax was described by Boucher (2006) for six species previously considered by Castillo and Reyes-Castillo (1984) as part of the genus Petrejoides Kuwert: Yumtaax recticornis (Burmeister, 1847), Yumtaax laticornis (Truqui, 1857), Yumtaax imbellis (Casey, 1897), Yumtaax nebulosus (Castillo & Reyes-Castillo, 1984) (Fig. 1A), Yumtaax mazatecus (Castillo & Reyes-Castillo, 1984), and Yumtaax olmecae (Castillo & Reyes-Castillo, 1984). Boucher (2006) considered this a morphologically and biogeographically cohesive group that deserved generic status based on the dorsal mesotibial ridge that is elevated at the middle and setose on its dorsal edge (Fig. 2). Yumtaax species exhibit low morphological variability (Castillo and Reyes-Castillo 1984), rendering a traditional morphological phylogenetic approach of limited utility. A traditional morphological approach in combination with molecular data are needed to define species and reconstruct the phylogeny of the genus. Molecular data have historically proven useful in the family Passalidae (Villatoro 1997, Archila 2009, Beza-Beza et al. 2011, Jiménez-Ferbans et al. 2016, and these data are essential for species delimitation and phylogeny reconstruction in the absence of strong morphological data. Although passalids are a potentially informative group for understanding the dynamics of New World cloud forests (Beza-Beza et al. 2011, Schuster andCano 2006), a strong phylogenetic hypothesis is needed for such applications. The aims of this study are to: (1) test the monophyly of Yumtaax and (2) reconstruct the phylogenetic relationships among Yumtaax species.
The morphological characters of the Y. laticornis species group include a short frontal area (Fig. 1B), dorsal mesotibial ridge elevated at the middle (described as "quilla dorsal de la tibia II corta" by Castillo and Reyes-Castillo 1984) (Fig. 2), presence of the infraocular ridge (Fig. 3), and the striatopunctatus-type mesofrontal structure (MFS) (Castillo and Reyes-Castillo 1984). The morphological characters of the Y. recticornis species group include the short frontal area (Fig. 1A), dorsal mesotibial ridge that is elevated at the middle (Fig. 2), and the central tooth of MFS short ( Fig. 1A) (long in P. tenuis) (Castillo and Reyes-Castillo 1984). Morphological character states shared by Yumtaax species (P. laticornis species group + P. recticornis species group [-P. tenuis]) include the short frontal area (Fig. 1A, B) and the dorsal mesotibial ridge elevated at the middle (Boucher 2006).

Relationships of the genus Yumtaax
The tumultuous nomenclatural history of Yumtaax species is due in part to the lack of molecular phylogenetic study of generic relationships in Passalidae. Most phylogenetic studies in the family have concentrated on the resolution of deeper relationships (subfamily and tribal level) (e.g., Fonseca 1987, Gillogly 2005, Boucher 2006, Fonseca et al. 2011 or have addressed the phylogeny of genera using morphology alone (e.g., Marshall 2000, Schuster et al. 2003, Boucher 2015, Jiménez-Ferbans and Reyes-Castillo 2015. The most complete generic-level phylogenetic analysis of Passalidae is that of Boucher (2006), who conducted a phylogenetic analysis of the tribe Proculini based on 51 morphological characters. Based on this analysis, Boucher placed Yumtaax within the tribe Proculini and hypothesized that Yumtaax is sister to Spurius Kaup (Yumtaax + Spurius), that the Yumtaax + Spurius clade is sister to Popilius sensu Boucher (2006), and that this clade (Yumtaax + Spurius + Popilius) is sister to Petrejoides sensu Boucher (2006).

Specimen acquisition, DNA extraction, and amplification
Both freshly collected and museum specimens were used (Suppl. material 1). Adults and larvae were field-collected by opening rotting logs with an axe and actively searching for tunnels, adults, and larvae. Specimens were stored in 95% ethanol and kept at a cool temperature. Muscle tissue was obtained from the right hind legs of specimens. DNA was extracted using the protocol detailed in Tagliavia et al. (2011) with two modifications. In order to more fully macerate tissue, legs were ground to a fine powder using the modified reciprocating saw approach described in Alexander et al. (2007). Additionally, 240 µl of lysis buffer with detergent was used in the first step instead of 80 µl of lysis buffer per mg of ground leg.
Steps 2-3 were repeated for 25 cycles. All amplicons were sequenced at the University of Chicago Comprehensive Cancer Center DNA Sequencing and Genotyping Facility. Sequence contigs were constructed using CLC MAIN WORKBENCH (CLC bio, Aarhus, Denmark). If samples required amplification with internal primers (see above) sequences were assembled using GENEIOUS R9.1 (Kearse et al. 2012). Uninterpretable sequences were cropped or discarded using the GENEIOUS default trimming tool. Sequences were then aligned with MEGA version 5 (Tamura et al. 2011). The Muscle algorithm with default settings (Edgar 2004) was used first, followed by Clustal W (Higgins et al. 1994) alignment with default settings. Alignments were then checked by eye and manually modified if necessary. Sections of missing data were replaced with Ns.

Phylogenetic analysis
Phylogenetic inference using maximum parsimony and Bayesian optimality criteria was conducted for each locus independently (COI, 12S, and 28S) and the total combined dataset (COI+12S+28S). Maximum parsimony bootstrap analyses were conducted using PAUP 4.0 (Swofford 2002) and included 1000 bootstrap replicates, each involving a heuristic search with 100 random additions. Clades with bootstrap support higher than 80% were considered well supported (Baum and Smith 2013). Bayesian analyses were performed using MR. BAYES 3.2 (Ronquist and Huelsenbeck 2003). Each analysis included 4 independent runs of one million generations, with trees sampled every 1,000 trees generations. For the COI and 28S datasets, 500,000 extra generations were run after the first one million generations until the split frequency reached less than 0.00. A concatenated alignment of the three loci was assessed with PARTITION FINDER v.1.1.0 (Lanfear et al. 2012), which suggested treating each locus as a separate partition with the GTR+I+G model applied to the C01 and 28S partitions and the GTR+G model applied to the 12S partition. For the mitochondrial datasets (12S and COI), the genetic code Bayes function (lset code) was set to invertebrate mitochondrial (lset code=invertmt). The first 100 trees of each run were discarded the remaining 901 trees for each run of the 12S and combine dataset were then used to create a 50% majorityrule consensus tree of posterior probability values. The remaining 1,401 trees for each run of the 28S and COI dataset were then used to create a 50% majority-rule consensus tree of posterior probability values Clades with Bayesian posterior probabilities equal to or higher than 0.95 were considered well supported (Baum and Smith 2013).

Species delimitation and species descriptions
The species status of each OTU was evaluated using two criteria. In order to be considered a species, an OTU must (1) be morphologically distinctive and (2) the molecular phylogeny must provide either evidence of its status as an evolutionary lineage or not provide contrary evidence. Species are segments of evolutionary lineages which can be diagnosed by a variety of criteria ("The General Lineage Concept"; de Queiroz 1998, Hey 2006), among them morphological distinctiveness. We view morphological distinctiveness alone as a sufficient criterion for species diagnosis, with the supporting phylogenetic data (when present) as confirmation. The taxa we diagnose represent working hypotheses and future workers should test these hypotheses with additional criteria (Carstens et al. 2013). Type specimens for the six described Yumtaax species were examined in order to properly associate species names. Species descriptions used the morphological terminology of Reyes-Castillo (1970) and Castillo and Reyes-Castillo (1984) with the following modifications: total body length was measured from the anterior apex of the left mandible to the posterior apex of the left elytrum. Head width was defined as the distance between the posterior tubercles of the supraorbital ridge. Eyes were considered large if the distal edge of the eye projected beyond the distal edge of the canthus (e.g., Fig. 1A; Y. nebulosus), moderately reduced if the distal edge of the eye was subequal to the canthus (e.g., Figs 1B, 9C; Y. cameliae), and greatly reduced if the distal edge of the eye did not surpass the distal edge of the canthus (e.g., Fig. 7C; Y. laticornis). Using the terminology of Reyes- Castillo (1970), borders or edges of structures and sutures are described as concave (curved posteriorly), straight, or convex (curved anteriorly).

28S nuclear data partition
The 28S data partition consisted of 1083 aligned characters of which 184 (16%) were variable. Of the variable characters, 74 (40%) were parsimony-informative. Because the parsimony and Bayesian analyses provided concordant tree topologies, bootstrap support (BS) and Bayesian posterior probabilities (PP) are shown on a Bayesian 50% majority-rule consensus tree (Suppl. material 4). Analysis of the 28S dataset did not provide support for a monophyletic Yumtaax. However, most Yumtaax OTUs (6 out of 7) were placed within a poorly supported clade (0.71 PP, <50 BS); Y. CM was placed sister to P. orizabae with maximum Bayesian and parsimony support. This relationship was also the only strongly supported conflict between the mitochondrial and nuclear 28S datasets (Suppl. material 4).

Total combined data
The Bayesian 50% majority rule consensus phylogram resulting from analysis of the combined mitochondrial (COI, 12S) and nuclear (28S) dataset (Fig. 4)

Monophyly of Yumtaax
The monophyly of Yumtaax was supported by the total combined molecular data set examined in this study; however monophyly of Yumtaax was not supported when genes were analyzed individually. The 12S and COI data strongly support two Yumtaax clades and do not provide strong evidence for the non-monophyly of Yumtaax. The 28S dataset does not support a monophyletic Yumtaax whereas the combined data suggest a monophyletic Yumtaax (Fig. 4). However, a morphological character state and geographic distribution (endemic to the temperate sierras of Mexico) both suggest that Yumtaax species form a monophyletic group. Evaluation of all Yumtaax OTUs revealed that a dorsal mesotibial ridge elevated at the middle (Fig. 2) is a character state unique to this group. This contrasts with a mesotibial ridge that is elevated for the entire length of the tibia and that is observed in the remaining 16 species of this study. Boucher (2006) also considered this dorsal mesotibial ridge elevated at the middle a synapomorphy for Yumtaax, and thus far, the dorsal mesotibial ridge elevated at the middle has not been observed in any other species of Passalidae (approximately 930 spp.) (Boucher 2006).
A combined consideration of the molecular phylogenetic, morphological, and geographic data suggests that the best working hypothesis is of Yumtaax as a valid, monophyletic genus. Further study should include both broader taxon sampling (including the genera Petrejoides sensu Boucher, Popilius sensu Boucher, and the excluded species of Yumtaax [Y. nebulosus, Y. olmecae]) and data from additional gene regions (particularly nuclear). For the remainder of this work, Yumtaax is treated as a monophyletic genus.

Yumtaax species delimitation
Based on combined morphological, molecular, and geographic data, we provide evidence that the Yumtaax OTUs analyzed in this study include seven distinctive species: (Fig. 4, Suppl. material 2, 3). Analysis of the combined molecular dataset recovered clear evidence for three independent evolutionary lineages (= species) corresponding to Y. imbellis, Y. recticornis OM, and Y. recticornis VM (Fig. 4, Suppl. materials 2, 3, 4). Although the lineage status of Y. LCM, Y. CM, Y. mazatecus, and Y. LM could not be established due to the lack of multiple samples per OTU, the phylogeny does not provide evidence that these are not lineages. For instance, each individual sample of these OTUs is genetically distinguishable from other Yumtaax lineages (has a nonzero branch length) and does not render Y. imbellis, Y. recticornis OM, or Y. recticornis VM paraphyletic. We treat each of these OTUs as species (see species diagnoses below).
Molecular and morphological data both suggest that Y. recticornis s. l. is composed of two independent lineages. First, eye size in Y. recticornis s. l. is geographically dependent; populations from Veracruz have reduced eyes whereas those from Oaxaca have large eyes (Castillo and Reyes-Castillo 1984). Molecular data reveal that these morphotypes form separate lineages: Y. recticornis VM is part of the Y. laticornis clade, and Y. recticornis OM is part of the Y. imbellis clade (Fig. 4). Based on examination of the type specimen of Y. recticornis, the name should be assigned to Y. recticornis OM; this species is distributed exclusively in Sierra Madre del Sur in Mexico. Yumtaax recticornis VM is an unnamed species that it is currently known exclusively from the transverse neo-volcanic system in Mexico. Within the Y. imbellis clade, molecular (Fig. 4) and morphological (see diagnosis below) data suggest that Y. LCM is a distinct species currently known only in the transverse neo-volcanic system in Mexico. Based on examination of type specimens in the genus Yumtaax, this morphotype also represents an undescribed species. Close examination of the type specimen of Y. laticornis indicated that this name should be applied to the Y. CM morphotype rather than the Y. LM morphotype (Yumtaax laticornis sensu Castillo and Reyes-Castillo 1984). Based on the type specimen, we re-circumscribe Y. laticornis and describe a new species for Y. LM.
Due to geographic isolation and morphological differences we follow Castillo and Reyes-Castillo (1984) and consider Y. nebulosus and Y. olmecae as independent, evolutionary lineages (= species) within the Yumtaax genus.

Species descriptions of Yumtaax
As a result of this work, we describe three new species: Y. veracrucensis sp. n. (= Y. LCM), Y. cameliae sp. n. (= Y. LM), and Y. jimenezi sp. n. (= Y. recticornis VM). In order to stabilize nomenclature, we re-circunscribe two species (Y. recticornis and Y. laticornis). Yumtaax recticornis s. l. is composed of two unrelated and heretofore cryptic species (Y. recticornis OM and Y. recticornis VM). Close examination of the Y. recticornis lectotype and one paralectotype indicates that the name Y. recticornis corresponds to our Y. recticornis OM. Redescription of this species is necessary to re-circumscribe Y. recticornis sensu stricto (Y. recticornis s. s.) and distinguish it from the Y. recticornis VM lineage (= Y. jimenezi). Close examination of the Y. laticornis holotype indicates that this name corresponds to Y. CM rather than to Y. laticornis sensu Castillo and Reyes-Castillo (1984) (= Y. LM). A redescription of the holotype of Y. laticornis is provided to clarify this finding. As such, the genus Yumtaax includes nine species that can be separated morphologically using the following key.
Redescription of lectotype (Fig. 5). Head (Fig. 5C). Labrum: anterior border concave, dorsal surface smooth and glabrous medially, punctate and setose apicolaterally, apically, and basally; anterior edge excavated. Clypeus: inclined, rectangular, concave, shiny, and smooth. Frontoclypeal suture: concave and opaque; external tubercles rounded, directed anteriorly and laterally. Frontal area: inclined, concave, smooth and shiny; frontal ridges present without inner tubercles. Frontal fossae: punctate and setose. Mesofrontal structure (MFS): of the "falsus" type (see Reyes-Castillo 1970); base subparallel and as wide as the lateral ridge of MFS; center horn short with apex rounded, largely free and directed anteriorly and dorsally (Fig. 5D), not reaching the Head width (between posterior tubercles of the supraorbital ridge) = 3.0 mm. Ratio of sums of both eye widths/total head width = 0.36; postocular area punctate and setose. Ligula: tridentate, central tooth surpassing apex of lateral teeth, lateral teeth rounded; setose punctures present in discal area; posterior border convex. Mentum: lateral lobes rounded and wide, with setose punctures. Basomedial portion protruding ventrally; anterior border at middle convex; basal fossae present and rugose. Hypostomal process: with lateral depression; separated from mentum by a distance shorter than the width of the anterior width of hypostomal process. Infraocular ridge (e.g., Fig. 3): short and narrow anteriorly. Mandible: with 3 apical teeth; internal tooth of left mandible bidentate (teeth 1 and 2 of internal tooth fused); dorsal tooth occupies less than half length of the mandible. Pronotum: anterior angles rounded. Anterior fossae of marginal sulcus punctate. Lateral fossae with scattered punctures. Marginal groove lacking punctures. Prosternum: opaque; prosternellum with anterior half and lateral edges opaque and posterior half and middle area shiny. Scutellum: smooth and glabrous. Mesosternum: with lateral areas opaque. Metasternum: with setae anterolaterally, lacking punctures in lateral margins of metasternal disc. Lateral fossae wide across metathorax, with setose punctures. Elytra: anterior border straight. Meeting point of striae 1-10 (see Reyes-Castillo 1970) with one line of punctures. Wings: well developed. Legs (Fig. 5E): femur I with longitudinal anteroventral groove weakly developed, not reaching distal end of femur, posteroventral half pubescent; setae long, dense, reddish. Abdomen: last sternite with marginal groove complete (Fig. 5B). Aedeagus (Fig. 5F, G, and H) (description based on non-type material): dorsal view phallus elongated (longer than wider). In ventral view distal edges of the phallus more or less at the same level of distal edges of parameres.
Variation. The paralectotype and other specimens vary from the lectotype by the following characteristics: frontoclypeal suture varies from opaque to shiny, from concave to almost straight; frons and clypeus inclined to nearly vertical (always concave); concavity of frons and clypeus vary from strongly developed to nearly flat. Internal tubercles strongly developed or absent; frontal ridges always present, but not always terminating in internal tubercles; ratio of eyes and head width varies from 0.32-0.57; supraocular ridge weakly developed or absent. Occipital sulcus biconcave to concave in a few individuals. Small portion of individuals with frontal ridges fused to the base of the central horn of MFS (apex of the central horn always free).
Distribution. The lectotype and paralectotype are labeled "Mexico" (Burmeister 1847). Castillo and Reyes-Castillo (1984) suggested the Sierra de Juarez in Oaxaca, Mexico, as the possible type locality. The species is known only from cloud forest (1424-2986 m elevation) in Oaxaca, Mexico.
Remarks. Originally, this species was thought to be widely distributed across the Sierra Madre Oriental, the Mexican Transvolcanic Belt, and Sierra Madre del Sur (Reyes- Castillo 1970, Castillo and Reyes-Castillo 1984, Boucher 2006). Our phylogenetic analysis (Fig. 4), as well as close examination of morphological characters, provide evidence that Y. recticornis s. l. is composed of at least two cryptic species: Y. recticornis ("Y. recticornis OM" in the Y. imbellis clade, Fig. 4 and Suppl. materials 2-4) and Y. jimenezi sp. n.
("Y. recticornis VM" in the Y. laticornis clade, Fig. 4 and Suppl. materials 2-4). Based on comparison with the lectotype and one paralectotype of Y. recticornis, this name should be applied to Y. recticornis OM. The following character states provide evidence that Y. recticornis OM is conspecific with Burmeister's concept of Y. recticornis: large eye size (distal edge of the eye surpassing distal edge of the canthus), shape of the central tooth of the MFS (base subparallel and as wide as the lateral ridge of the MFS; center horn short with apex rounded, largely free and directed anteriorly and dorsally [ Fig. 5D]), and shape of the frons and clypeus concave (rather than straight as in Y. jimenezi). Castillo and Reyes-Castillo (1984) suggested that eye size variation among Y. recticornis s. l. was dependent upon the locality of the population (populations in Oaxaca possess large eyes; populations in Veracruz possess small eyes). Characters of the internal tooth of the left mandible and aedeagus are described based on the paralectotype and non-type material because the mandibles of the lectotype are closed (thus making it impossible to determine the state of this character in this specimen).
Description of holotype (Fig. 6). Head (Fig. 6C). Labrum: anterior border concave, dorsal surface smooth and glabrous medially, punctate and setose apicolaterally, apically, and basally; anterior edge excavated. Clypeus: vertical, rectangular, flat, shiny, and smooth. Frontoclypeal suture: straight, and opaque; external tubercles rounded, directed anteriorly and laterally. Frontal area: horizontal, flat, smooth and shiny, frontal ridges weak finishing in inner tubercles; inner tubercles smaller than external tubercles. Frontal fossae: punctate and setose. Mesofrontal structure (MFS): of the "falsus" type (see Reyes-Castillo 1970); base subparallel, slightly narrower than MFS' lateral ridges; center horn long with apex acute, largely free and directed anteriorly (Fig. 6D), surpass- ing posterior margin of clypeus, dorsally with sparse micro-punctures; base of the center horn wide, narrowing gradually until apex; dorsal fossa present at the base of MFS. Occipital fossa: shallow posteriorly and deeper laterally connected with the frontal fossae. Posterior occipital sulcus sinuate. Supraorbital ridge: bituberculate, tubercles of similar size; posterior half of supraorbital ridge not bifurcated. Canthus: with apex rounded, almost oblique, covering 1/3 of the eye, not expanded distally. Eyes: reduced (distal edge of the eye not reaching the distal edge of the canthus), width = 0.3 mm (each eye). Head width = 3.0 mm. Ratio of sums of both eyes widths/total head width = 0.2; postocular area punctate and setose. Ligula: tridentate, central tooth surpassing apex of lateral teeth; lateral teeth rounded; setose punctures present in discal area; posterior border convex. Mentum: lateral lobes rounded and wide, with setose punctures. Basomedial portion protruding ventrally; anterior border at the middle convex; basal fossae present with setose punctures. Hypostomal process: without lateral depression; separated from the mentum by a distance shorter than the width of the anterior width of the hypostomal process. Infraocular ridge (e.g., Fig. 3): short, weak, and wide anteriorly. Mandible: with 3 apical teeth; internal tooth in left mandible tridentate; dorsal tooth occupies at least half length of the mandible. Pronotum: anterior angles rounded. Anterior fossae of marginal sulcus punctate. Lateral fossae without punctures. Marginal groove lacking punctures. Prosternum: opaque; prosternellum with anterior and lateral edges rugose and opaque, anteriorly and posteriorly shiny. Scutellum: smooth and glabrous. Mesosternum: with anterior-lateral areas opaque. Metasternum: with setae anterolaterally, lacking punctures in lateral margins of metasternal disc. Lateral fossae wide posteriorly with setose punctuations. Elytra: anterior border straight. Meeting point of striae 1-10 (see Reyes-Castillo 1970) with one line of punctures. Wings: well developed. Legs: femur I with longitudinal anteroventral groove weakly developed, not reaching the distal part of the femur, posteroventral half pubescent; setae long, sparse, reddish. Abdomen: last sternite with marginal groove incomplete (Fig. 6B). Aedeagus (Fig. 6E, F): in dorsal view phallus elongated (longer than wider). In ventral view distal edges of the phallus more or less at the same level of distal edges of parameres.
Variation. Paratypes vary from the holotype by the following characteristics: internal tubercles weak to obsolete; frontal fossae glabrous or setose; ratio of eyes to head width vary between 0.19 and 0.22; basal fossae of mentum strong, opaque and glabrous or shiny and with setose punctures; infraocular ridge weak or absent; femur I with longitudinal antero-ventral groove weakly developed to obsolete.
Etymology. This species is named after its home state of Veracruz in Mexico. Distribution. This species is known from cloud forest between around 1900 m in the transverse neo-volcanic system, Mexico. The surrounding states and areas in which this species is distributed have been well-collected, and Y. veracrucensis has only been found at three localities in Veracruz, Mexico: La Cortadura Natural Reserve near Coatepec; Chiconquiaco (near La Parra); and the road between Las Minas and Xalapa; Chiconquiaco; Congr. La Guacamaya (19°45'51.4"N,96°48'1.7"W).
Remarks. Specimens of Y. veracrucensis were originally identified as P. orizabae and were collected in Reserva La Cortadura in Coatepec, Veracruz, Mexico. Based on our phylogenetic analysis, Y. veracrucensis (Y. LCM) and Y. imbellis are potential sister species (Fig. 4; PP 0.99/BS 90). Molecular distinctiveness and form of the dorsal ridge in tibia II (as in all species of Yumtaax) provide support that this is a distinct species within the genus Yumtaax.
Diagnosis. Yumtaax laticornis is a large (24.5-33.0 mm) brachypterous species and is part of the Yumtaax laticornis clade (Fig. 4). This species is diagnosed by the following character combination: the clypeus is vertical (shared with Y. veracrucensis, Y. cameliae; inclined in other members of Yumtaax) and the anterior border is straight (shared with other members of Yumtaax except for Y. recticornis and Y. olmecae with concave anterior border of clypeus); mesofrontal structure (MFS) of the "falsus" type (see Reyes-Castillo 1970) (shared with all members of Yumtaax except Y. cameliae which has the MFS of the "striatopunctatus" type), with the central tooth largely free (shared with Y. recticornis, Y. veracrucensis, Y. cameliae, Y. mazatecus; fused with frontal ridges in Y. jimenezi, Y. imbellis, Y. nebulosus, Y. olmecae), directed anteriorly (shared with Y. veracrucensis; directed dorsally in Y. recticornis, Y. imbellis, Y. nebulosus, Y. olmecae; directed dorsally and anteriorly in Y. jimenezi, Y. mazatecus; elevated in the posterior half bending abruptly forward in the anterior half in Y. cameliae), and reaching the frontoclypeal suture (shared with Y. veracrucensis, Y. cameliae; not reaching the clypeus in other members of Yumtaax); eyes are strongly reduced (shared with Y. mazatecus; eyes large in Y. recticornis, Y. imbellis; eyes moderately reduced in Y. veracrucensis, Y. cameliae, Y. jimenezi, Y. nebulosus, Y. olmecae); and the scutellum is punctate (smooth in other members of Yumtaax).
Redescription of holotype (Fig. 7). Head (Fig. 7C). Labrum: anterior border concave, dorsal surface smooth and glabrous medially, punctate and setose apicolaterally, apically, and basally; anterior edge excavated. Clypeus: vertical, rectangular, flat, shiny, and smooth. Frontoclypeal suture: straight, and shiny; external tubercles rounded weak, directed anteriorly. Frontal area: vertical, flat, smooth and shiny, frontal ridge absent. Frontal fossae: punctate and glabrous. Mesofrontal structure (MFS): of the "falsus" type (see Reyes-Castillo 1970); base subparallel and narrower than MFS' lateral ridge; center horn long with apex acute, largely free and directed anteriorly (Fig. 7D), surpassing posterior margin of clypeus, dorsally without micro-punctures; base of center horn wide narrowing gradually until apex; dorsal fossa absent at the base of MFS. Occipital fossa: shallow posteriorly and deeper laterally, not connected with the frontal fossae. Posterior occipital sulcus concave. Supraorbital ridge: bituberculate, tubercles of similar size; posterior half of supraorbital ridge bifurcated. Canthus: with apex rounded, covering more than 1/3 of the eye, expanded distally. Eyes: strongly reduced (distal edge of the eye shorter than the distal edge of the canthus), width = 0.4 mm (each eye). Head width (between posterior tubercles of the supraorbital ridge) = 5.0 mm. Ratio of sums of both eyes widths/total head width = 0.16; postocular area punctate and setose. Ligula: tridentate, central tooth surpassing apex of lateral teeth, lateral teeth rounded; setose punctures present in discal area; posterior border straight. Mentum: lateral lobes rounded and wide, with setose punctures. Basomedial portion protruding ventrally; anterior border at middle straight; basal fossae present, with setose punctures. Hypostomal process: with lateral depression; separated from mentum by a distance larger than the wide of the anterior width of hypostomal process. Infraocular ridge (e.g., Fig. 3): short and wide anteriorly, narrow posteriorly. Mandible: with 3 apical teeth; internal tooth of left mandible bidentate; dorsal tooth occupies more than half length of the mandible. Pronotum: anterior angles rounded. Anterior fossae of marginal sulcus punctate. Lateral fossae without punctures. Marginal groove with punctures. Prosternum: opaque. Prosternellum with anterior half and lateral edges opaque and posterior half and middle area shiny. Scutellum: punctate and glabrous. Mesosternum: with anterior-lateral areas opaque. Metasternum: with setae anterolaterally, lacking punctures in lateral margins of the metasternal disc. Lateral fossae wide glabrous posteriorly with setose punctures anteriorly. Elytra: anterior border straight. Meeting point of striae 1-10 (see Reyes-Castillo 1970) with one line of punctures. Wings: reduced. Legs (Fig. 7E): femur I with longitudinal anteroventral groove strongly developed, reaching the distal end of the femur, posterioventral half pubescent; setae short, dense, reddish. Abdomen: last sternite with marginal groove complete (Fig. 7B). Aedeagus (Fig. 7F, G, and H) (Description based on non-type material): dorsal view phallus globose (wider than long). Ventral view lateral edges of the phallus surpassing the laterodistal edges of the parameres.
Variation. The non-type material differs from the holotype in the following characters: internal tubercles obsolete to strongly developed; frontal ridges obsolete to strongly developed; frontal area glabrous to sparsely setose; ratio of eyes versus head width varies from 0.13-0.23; pronotum laterally with or without strong punctures, even at the individual level (right vs left side of the pronotum); prosternelum completely opaque or opaque and shiny.

Yumtaax cameliae
Description of holotype (Fig. 8). Head (Fig. 8C). Labrum: anterior border concave, dorsal surface smooth and glabrous medially, punctate and setose apicolaterally, apically, and, basally; anterior edge excavated. Clypeus: vertical, rectangular, flat, shiny, and smooth. Frontoclypeal suture: straight, and shiny; external tubercles rounded, weak, directed dorsally and anteriorly. Frontal area: inclined, flat, smooth, and shiny; frontal ridges absent without inner tubercles. Frontal fossae: impunctate and glabrous. Mesofrontal structure (MFS): of the "striatopunctatus" type (see Reyes-Castillo 1970); base subparallel and narrower than MFS' lateral ridge; center horn long with apex acute, largely free elevated in the posterior half bending abruptly forward in the anterior half (Fig. 8D), reaching the posterior margin of clypeus, dorsally without micro- punctures; base of center horn wide, not narrowing in the posterior half and narrowing abruptly in the anterior half until apex; dorsal fossa present at base of MFS. Occipital fossa: shallow posteriorly and deeper laterally connected with the frontal fossae. Posterior occipital sulcus concave. Supraorbital ridge: bituberculate, tubercles of similar size; posterior half of supraorbital ridge not bifurcated. Canthus: with apex rounded, covering less than 1/3 of the eye, not expanded distally. Eyes moderately reduced (distal edge of the eye more or less at the distal edge of the canthus), width = 0.6 mm (each eye). Head width (between posterior tubercles of the supraorbital ridge) = 4.3 mm. Ratio of sums of both eyes widths/total head width = 0.27; postocular area punctate and setose. Ligula: tridentate, central tooth surpassing apex of lateral teeth; lateral teeth rounded; glabrous punctures present in discal area; posterior border straight. Mentum: lateral lobes rectangular and wide, with setose punctures. Basomedial portion protruding ventrally; anterior border at middle convex; basal fossae present. Hypostomal process: without lateral depression; separated from mentum by a distance shorter than the width of the anterior width of hypostomal process. Infraocular ridge (Fig. 3): short and wide anteriorly, narrow posteriorly. Mandible: with 3 apical teeth; internal tooth of left mandible bidentate; dorsal tooth occupies half of length of the mandible. Pronotum: anterior angles rounded. Anterior fossae of marginal sulcus punctate. Lateral fossae impunctate. Marginal groove lacking punctures. Prosternum: opaque; prosternellum with anterior half and lateral edges opaque and posterior half and middle area shiny. Scutellum: smooth and glabrous. Mesosternum: with anterolateral areas opaque. Metasternum: with setae anterolaterally, lacking punctures in lateral margins of metasternal disc. Lateral fossae wide posteriorly with setose punctures. Elytra: anterior border straight. Meeting point of striae 1-10 (see Reyes-Castillo 1970) with one line of punctures. Wings: well developed. Legs: femur I with longitudinal anteroventral groove strongly developed, reaching distal end of femur, posteroventral half pubescent; setae long, sparse, reddish. Abdomen: last sternite with marginal groove complete and opaque laterally (Fig. 8B). Aedeagus (Fig. 8E, F, G) (based on male paratype): in dorsal view phallus globose (wider than long). In ventral view distal edges of phallus surpassing the distal edge of the parameres.
Variation. Paratypes vary from the holotype in the following characters: internal tubercles obsolete to strongly developed; frontal ridges obsolete to strongly developed; frontal area glabrous to setose; ratio of eyes versus head width varies from 0.19-0.31; central area of the ligula always punctate, occasionally setose; pronotum with lateral fossae with or without strong punctures, even at the individual level (right vs left side of the pronotum); prosternellum shiny (one specimen of the type series) or opaque in anterior half; terminal sternite with lateral areas of the marginal groove opaque or not.
Etymology. The species is named Y. cameliae, honoring Passalidae researcher Camelia Castillo whose research (Castillo and Reyes-Castillo 1984) provided a better understanding of Yumtaax.
Distribution. This species is known only from the type locality in Veracruz, Mexico. It was collected in a small patch of oak forest (bosque mesófilo de montaña) surrounding the Puerto del Aire village at 2400 m elevation.
Description of holotype (Fig. 9). Head (Fig. 9C). Labrum: anterior border concave, dorsal surface smooth and glabrous medially, and punctate and setose in the apicolaterally, apically, and basally; anterior edge excavated. Clypeus: inclined, rectangular, shiny, and smooth. Frontoclypeal suture: straight, and shiny. External tubercles rounded and directed dorsally. Frontal area: inclined, smooth, and shiny, frontal ridges present finishing in inner tubercles. Frontal fossae: punctate and setose. Mesofrontal structure (MFS): of the "falsus" type (see Reyes-Castillo 1970); base subparallel and narrower than the MFS' lateral ridge; center horn short with apex rounded, not free (fused with frontal ridges) and directing dorsally (Fig. 9D), not reaching the posterior margin of clypeus (Fig. 9C, D), dorsally without micro-punctures; base of the center horn narrow not narrowing down along its length (central tooth tubercle like shape [ Fig. 9C, D]); dorsal fossa present at the base of MFS. Occipital fossa: shallow posteriorly and deeper laterally not connected to frontal fossae. Posterior occipital sulcus sinuate. Supraorbital ridge: bituberculate, tubercles of similar size; posterior half of supraorbital ridge not bifurcated. Canthus: with apex rounded covering less than 1/3 of the eye, expanded distally. Eyes: reduced (distal edge of the eye more or less at the distal edge of the canthus), width = 0.5 mm (each eye). Head width = 3.5 mm. Ratio of sums of both eyes widths/total head width = 0.24; postocular area punctate and setose. Ligula: tridentate, with central tooth surpassing apex of lateral teeth, lateral teeth rounded; setose punctures present in discal area; posterior border convex. Mentum: lateral lobes rounded and wide, with setose punctures. Basomedial portion protruding ventrally; anterior border at middle convex; basal fossae absent. Hypostomal process: without lateral depression; separated from mentum by a distance shorter than the wide of the anterior width of hypostomal process. Infraocular ridge absent. Mandible: with 3 apical teeth; internal tooth of left mandible bidentate; dorsal tooth occupies less than half length of the mandible. Pronotum: anterior angles rounded. Anterior fossae of marginal sulcus impunctate. Lateral fossae with heavy punctures. Marginal groove lacking punctures. Prosternum: opaque. Prosternellum anterior half opaque and lateral edges and posterior half shiny. Scutellum: smooth and glabrous. Mesosternum: with anterolateral areas opaque. Metasternum: with setae in anterolaterally, without punctures in lateral margins of metasternum disc. Lateral fossae wide across the metathorax, with setose punctures. Elytra: anterior border straight. Meeting point of striae 1-10 (see Reyes-Castillo 1970) with one line of punctures. Wings: well developed. Legs: femur I with longitudinal anteroventral groove weakly developed in the proximal half and strongly developed on the distal end of the femur, posteroventral half pubescent; setae long, sparse, reddish. Abdomen: last sternite with marginal groove complete (Fig.  9B). Aedeagus (Fig. 9E, F, G): in dorsal view phallus globose (wider than long). In ventral view distal edges of phallus surpassing the distal edge of the parameres.
Variation. Frontoclypeal suture can be from opaque to shiny; internal tubercles from strongly to weakly marked but always present; ratio of eyes and head with varies from 0.18-0.32; supraocular ridge from weak to absent; hypostomal process with weak lateral depression to lateral depression absent; prosternellum varies from anterior half opaque and lateral edges and posterior half shiny to anterior half and lateral edges opaque and posterior half and middle shiny to completely opaque; femur I longitudinal anterior-ventral groove from weak in the proximal half to absent; femur I longitudinal anterior-ventral groove from strongly developed in the distal half to absent.
Etymology. This species is named in honor of Passalidae worker Dr. Larry Jiménez-Ferbans who assisted in collecting trips supporting this study.
Distribution. This species is known from cloud forest (bosque mesófilo) at 2400 m elevation from the state of Veracruz, Mexico.
Remarks. Yumtaax jimenezi is a cryptic, widespread species that has been confused with Y. recticornis. Previously, Y. recticornis s. l. was thought to be broadly distributed in Mexico from the Sierra Madre Oriental in the Mexican Transvolcanic Belt and Sierra Madre del Sur (Reyes- Castillo 1970, Castillo and Reyes-Castillo 1984, Boucher 2006. Phylogenetic analysis (Fig. 4 and Suppl. materials 2-4) and close examination of morphology provide evidence that Y. recticornis s. l. comprises two cryptic species [Y. recticornis (= Y. recticornis OM) and Y. jimenezi (Y. recticornis VM)].
These species are distinguished by eye size (small in Y. jimenezi and large in Y. recticornis), shape of the central tooth of the MFS (center horn short with apex rounded, not free [fused with frontal ridges] and directed dorsally [ Fig. 9D] in Y. jimenezi; center horn short with apex rounded, largely free and directed anteriorly and dorsally [ Fig. 5D] in Y. recticornis), and the shape of the surface of the frons and clypeus (concave in Y. recticornis versus flat in Y. jimenezi). Interestingly, the reduced eye size in Y. jimenezi results in the distal expansion of the canthus. Based on seven exemplars, phylogenetic analysis ( Fig. 4) strongly supports Y. jimenezi as a unique lineage (1.0 PP/100 BS).

Conclusions
A single, unique synapomorphy (dorsal mesotibial ridge elevated at the middle), molecular phylogenetic analysis, and distributional affinities collectively support the hypothesis of Yumtaax monophyly.
Yumtaax species, as with most Passalidae, exhibit a high degree of morphological conservatism, rendering traditional systematics studies quite challenging. Cryptic species, such as those revealed in this study, are likely to be discovered by employing molecular data and careful consideration of morphological characters. Further studies, ideally those that include significant additional molecular phylogenetic data, are needed to rigorously evaluate the Passalidae species boundaries and evolutionary history.

Table S1
Authors: Cristian Fernando Beza-Beza, James Beck, Pedro Reyes-Castillo, Mary Liz Jameson Data type: specimen data Explanation note: Voucher specimens for taxa included in the molecular analysis (species, depository, preservation method, collection data, and GenBank DNA sequence accession numbers). Copyright notice: This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.

Figure S1
Authors