Two new species of Passalus Fabricius (Coleoptera: Passalidae) from the western Brazilian Amazon with comments on the taxonomic limits of the subgenera

Passalus Fabricius, 1792 is the largest genus of Passalidae and presents great diversity in South America, mainly in the Amazon region. Currently this complex and heterogeneous genus is subdivided into three subgenera: P. (Pertinax) Kaup, 1869, P. (Mitrorhinus) Kaup, 1871, and P. (Passalus). Herein, two new species of Passalus from the western region of the Brazilian Amazon, a diverse but poorly studied area, are described and illustrated. The new species Passalus (Pertinax) deuterocerus sp. nov. and Passalus (Passalus) cleidecostae sp. nov. are compared with similar species occurring in nearby regions. Also, the diagnostic characters used to define the subgenera and sections of Passalus are summarized.


INTRODUCTION
Passalus Fabricius, 1792, the largest of the 64 genera of Passalidae, is exclusively Neotropical with about 139 valid species (Schuster & Cano, 2008), being dominant in the bess beetle fauna of South America. Currently, the genus is divided into three subgenera: P. (Pertinax) Kaup, 1869, P. (Mitrorhinus) Kaup, 1871, and P. (Passalus) which, in turn, is subdivided into three species groups, the "sections": Phoroneus, Petrejus and Neleus (Luederwaldt, 1931;Hincks & Dibb, 1935, 1958. The three subgenera differ mainly by the number of secondary mediofrontal tubercles on the anterior frontal edge (Luederwaldt, 1931). Boucher (2015a) proposed the revalidation of seven genera previously synonymized with Passalus, including Pertinax, without use of cladistic methods. However, we prefer to maintain the classification previously proposed by Luederwaldt (1931) and enhanced by Hincks & Dibb (1935, 1958, since it is the most used by Passalidae researchers. According to Fonseca & Reyes-Castillo (2004), 26 species of Passalus occur in the Brazilian Amazon. Recently, Bevilaqua & Fonseca (2019) published a knowledge synthesis of the Passalidae for the west-most region of the Brazilian Amazon, and established five new records for Passalus species, increasing the total number to 31. In a new evaluation of the material coming from this rich area, it was verified the existence of two new species, here described and illustrated.

MATERIAL AND METHODS
The studied material belongs to the invertebrate collection of the Instituto Nacional de Pesquisas da Amazônia, Manaus, Brazil (INPA) and Coleção Zoológica Professor Paulo Bührnheim of the Universidade Federal do Amazonas, Manaus, Brazil (CZPB).
The images were obtained in Leica M165C stereomicroscope with a coupled DFC295 camera and processed in the LAS version 4.2 software to perform the morphometric measurements. Then, they were stacked in layers by the software Helicon Focus version 5.3 in order to generate a single image of combined focus. For uniform and more effective illumination, we used the geodesic dome of lighting according to Kawada & Buffington (2016). The images were edited in Adobe Lightroom® Software for light and contrast correction, and the plates were made using Adobe Photoshop®. The labels data of the exam-ISSN On-Line: 1807-0205 ISSN Printed: 0031-1049 ISNI: 0000-0004-0384-1825 ined material were transcribed in verbatim. A slash (/) indicates a new line on the label, additional information are given in square brackets ([ ]).

Aedeagus (Fig. 2F):
Median lobe: narrower than parameres, almost as long as parameres and basal piece together; ventrally with two lateral sclerotized plates lacking at median region. Basal piece separated from parameres by distinct suture. Parameres ventrally, with median anterior region with central emargination; laterally, with rounded projections almost reaching half of median lobe length; dorsally, with unjoined projections. Basal piece ventrally, short, with convexity in posteromedian region; dorsally lateral projections separate.

Etymology:
The specific epithet derives from the union of two Greek words, "deúteros" ("secondary" or "double") and "keros" ("horn" or "any hornlike projection"), in reference to the two secondary mediofrontal tubercles, the most remarkable feature of the species.

Polymorphisms:
The secondary mediofrontal tubercles may vary in size and distance from each other; the anterofrontal ridges may be almost absent in some specimens.
Remarks: Due to the convex body, well-developed inner tubercles, barely delimited metasternal disc, and metasternal pubescence absent, Passalus deuterocerus sp. nov. is assigned to the subgenus Passalus (Pertinax). It is very similar to P. latifrons Percheron, 1841, P. morio Percheron, 1835, and especially to P. epiphanoides (Kuwert, 1891). The differences between P. deuterocerus sp. nov. and P. latifrons are more pronounced because the new species has small-er size, two secondary mediofrontal tubercles, mesosternal scars, and absence of pubescence on posterior region of the pronotum ventrally, while P. latifrons is larger, has no secondary mediofrontal tubercles or mesosternal scars, and has pubescence on the posterior region of the pronotum ventrally (Table 1). Passalus deuterocerus sp. nov. differs from P. morio by having acute latero + mediofrontal tubercles (obtuse in P. morio), anterior frontal edge of the head thinner (which is thicker in P. morio), presence of secondary mediofrontal tubercles (absent in P. morio), and by the general shape of aedeagus (Table 1). When comparing the new species with P. epiphanoides, the differences are more subtle and require attention. Reyes-Castillo (1973) redescribed P. epiphanoides adding some characters as interspecific variations, such as protruding edges of the median notch of the anterior frontal edge of the head forming two secondary mediofrontal tubercles and mesosternal scars, which may be presented as two triangular-shaped matte areas covered with submicroscopic spicules. However, a reexamination the specimens studied by Bevilaqua & Fonseca (2019) showed that these features are constant in series of specimens. A better analysis showed that these so-called polymorphic characters are diagnostic features that distinguish the two species. Thus, P. deuterocerus sp. nov. can be differentiated from P. epiphanoides by the narrower mediofrontal area (Fig. 3A); two secondary mediofrontal tubercles (Fig. 3A); mediobasal area of the mentum with rounded anterior region, without punctures or bristles (Fig. 3B); shallow mesosternal scars with slightly triangular shape with matte surface and presence of micro bristles (Fig. 3C); metasternal disc with a group of punctures on the lateroposterior region (Fig. 3C); narrower aedeagus, with median lobe with narrower sclerotized plates, parameres with median anterior emargination and basal piece with concavity at posterior edge. Passalus epiphanoides presents wider mediofrontal area (Fig. 3D), absence of secondary mediofrontal tubercles ( Fig. 3D), mediobasal area of the mentum may have punctures or bristles and a notch in the anterior region ( Fig. 3E), deep and oval or drop shape mesosternal scars with matt surface and presence of micro bristles (Fig. 3F), metasternal disc with a small group of punctures in the lateroposterior region (Fig. 3F), wider aedeagus, with median lobe with wider sclerotized plates, parameres with median convexity on the anterior edge and basal piece with convexity on the posterior edge. The main difference between P. deuterocerus sp. nov. and other species currently allocated in P. (Pertinax) is the presence of secondary mediofrontal tubercles, a characteristic that apparently belongs to the Rhodocanthopus group. However, other characters such as eyes reduction, presence of secondary inner tubercles and presence of large spines on the external face of the meso-and metatibia, besides the general shape of the aedeagus (Jiménez-Ferbans et al., 2016a), do not allow the new species to be allocate in this group. Furthermore, it seems to be more related to species belonging to P. (Pertinax) s. str. such as P. morio, P. latifrons, P. epiphanoides and P. convexus Dalman, 1817. However, phylogenetic analysis that is in process may have clarified this issue further.  Prothorax: Pronotum (Figs. 4A, 5A, C, 6B): anterior edge slightly straight; anterior angles slightly acute and not protruding; marginal groove well-marked, deep, narrow and punctate throughout, with apex dilated, not reaching ⅓ of pronotal width; lateral fossae large, well-marked, deep and rounded; with group of coarse punctures only in the region near the lateral fossae. Prosternum (Fig. 4B): prepisternum with a few long, sparse setae on outer face; prepimerum with a few sparse bristled punctures; prosternelum rhomboidal, with acute base.
Elytra (Fig. 4A): Approximately 2.5× longer and 1.1× wider than pronotum; striae: narrower than interstriae, marked by circular punctures, on dorsal striae smaller than that on lateral striae; epipleura with setae on region closest to the humeri which have a tuft of long setae. Abdomen (Fig. 4B): sternite VII with complete and wellmarked groove; sides rough.
Aedeagus (Fig. 5F): Median lobe: narrower than parameres and basal piece together; ventrally with two sclerotized plates lacking at median region. Basal piece attached to the parameres forming a tegmen. Tegmen ventrally with median anterior region with V-shaped median notch; laterally, with slightly straight projections almost reaching half-length of median lobe; dorsally, with projections attached.

Etymology:
We dedicate this new species to Dr. Cleide Costa, a researcher with extensive career in Coleoptera systematics, working especially with Neotropical immature insects and for her dedication on training taxonomists.

Polymorphisms:
The size of the inner tubercles, and the density of punctures on the lateropostfrontal areas may vary.
Remarks: Passalus cleidecostae sp. nov. can be assigned to subgenus Passalus (Passalus), precisely in the "Neleus" section by the following characters: mediofrontal area short and transverse; presence of two well-developed secondary mediofrontal tubercles; mediofrontal and laterofrontal tubercles not fused, but close together; inner tubercles small; metasternal disc well-delimited, and humeri, epipleura, lateroanterior region of the metasternum and metasternal groove pubescent. It is more similar to species whose apex of the central tubercle is not free, such as P. interstitialis Eschscholtz, 1829 and P. bucki Luederwaldt, 1931. The new species differs from P. interstitialis by presenting smaller and more obtuse secondary mediofrontal tubercles; smaller mediofrontal tubercles and not overlapping laterofrontal tubercles; less developed inner tubercles; less straight frontal ridges; shorter and thicker antennal lamellae, not turned inwards; weaker carina on metasternal disc, and sparser pubescence on the humeri, epipleura, and metasternal anterolateral region ( Table 2). The similarities with P. bucki are more pronounced, since both species have secondary mediofrontal tubercles smaller than laterofrontal tubercles, and with obtuse apices, presence of punctures on anterior mediofrontal and lateropostfrontal areas. However, they can be differentiated since P. cleidecostae sp. nov. presents secondary mediofrontal tubercles with more acute apices ( (Table 2). Although P. cleidecostae sp. nov. and P. bucki are similar species, they have enough distinctive characteristics, which was possible to see after comparison with the P. bucki type series deposited at the Museu de Zoologia da Universidade de São Paulo (MZUSP), as well as specimens from other localities and species that also belong to the Neleus section. Both species may be more similar, since the presence of flattened central tubercle flatted, and absence of free apex are not common characteristics in this group. It is possible that within the Neleus section these species form a group which the main characteristic is related to the central tubercle.

DISCUSSION
As previously discussed by Bevilaqua & Fonseca (2017, the western region of the Brazilian Amazon, especially the westernmost part, is considered a very rich and diverse area for numerous taxa, including Passalidae, when compared to other well-studied areas of the Brazilian Amazon, and bordering countries. In this region, as well as throughout South America, the genus Passalus stands out for its great diversity, and although the literature has already pointed out that this genus may be polyphyletic (Reyes-Castillo, 1970;Boucher, 2006;Jiménez-Ferbans et al., 2016a), the researchers on Passalidae often use the classification initially proposed by Luederwaldt (1931), since so far no systematic study has been conducted to better delimit the groups present in Passalus. However, it is expected that this species-rich genus will undergo changes in its classification over the next few years, as a systematic work is underway.
The current composition was initially proposed by Gravely (1918) who, in his worldwide review, synonymized all genera included in the subfamilies Pertinacinae, Rhodocanthopinae, Pleurariinae (except Pleurarius Kaup, 1868), Neleidinae, Phoronaeinae, Petrejinae, Vatiniinae and Neleinae defined in the classification proposed by Kuwert (1898) as well as the genera Eumelus Kaup, 1871, and Mitrorhinus with Passalus (which according to Gravely (l.c.) was considered a "complex and heterogeneous genus"). The author stated that, despite proposing such synonyms, he is not sure if they are valid for some genera. Years later, his proposal was implemented by Luederwaldt (1931), who proposed subdividing Passalus into three subgenera based mainly on the presence and quantity of secondary mediofrontal tubercles, and subdividing Passalus (Passalus) into groups of species called sections. Following the proposals of Gravely (l.c.) and Luederwaldt (l.c.), Hincks (1933Hincks ( , 1934Hincks ( , 1935 and Hincks & Dibb (1935, 1958 carried out works to reassess this genus, proposing synonyms and nomenclatural corrections, which culminated in the only catalog (Hincks & Dibb (l.c.)) that lists all Passalus species, their synonyms and distributions. Reyes-Castillo (1970) redescribed the genus and commented that it is necessary to reevaluate the characters used to distinguish subgenera.
Thus, by analyzing the material from various regions of the New World, especially the Amazon region, and based on previous works, it is possible to make two comments: 1. The number of valid Passalus species is still doubtful, since several species need to be revalidated and others synonymized. However, considering the last  Avulsos Zool., 2020;v.60.special-issue: e202060(s.i.).19 9/13 species described in Jiménez-Ferbans et al. (2019) and the new status of P. occidentalis Hincks, 1950 (Bevilaqua & Fonseca in press)   ally strong and high, reaching the well-developed inner tubercles, which are located midway between the central tubercle and the latero + mediofrontal tubercles; mesosternal scars usually small and shallow, sometimes absent; metasternal grooves narrow and glabrous (except in some species, like P. mancus Burmeister, 1847, P. matilei (Boucher, 2000) and P. loici (Boucher, 2000)), and the metasternal disc is barely marked by a carina or punctures.
Passalus (Mitrorhinus) is characterized by having a single secondary mediofrontal tubercle in anterior frontal edge, which may be split or not at the apex; anterofrontal ridges which, although weak in some species, reach the latero + mediofrontal tubercles; scarce pubescence on body; and metasternal disc barely delimited by a carina or punctures, with narrow and glabrous metasternal groove.
Overall, P. (Passalus) is complex to define, but it can be characterized by having none or two secondary mediofrontal tubercles and, when present, they may be near each other and small, or large and distant from each other; concave or straight frontal edge of the labrum; and dense pubescence present or not on body. The sections  The delimitation of the subgroups and species belonging to Passalus has been a much-discussed issue in recent works (Boucher, 2006(Boucher, , 2015aJiménez-Ferbans et al., 2015, 2016a, b, 2017, 2018Bevilaqua & Fonseca, 2017 and this has been the main focus of the authors of this work in recent years, therefore, as mentioned earlier, this genus is expected will undergo changes in its classification, facilitating the delimitation of its species groups, making it better grounded. In view of this, in the future, further analysis of materials from Brazilian and foreign collections will be conducted in order to gather more information on the biodiversity of Passalus in the Amazon.  , 2020; v.60.special-issue: e202060(s.i.).19 12/13