Revision of the attaphilous genus Phoxonotus ( Coleoptera : Histeridae : Saprininae )

In this paper the strictly attaphilous Centraland South American genus Phoxonotus Marseul, 1862 is revised. Based on the structure of their antennal cavities, the species of Phoxonotus are newly split into two subgenera: Phoxonotus Marseul, 1862 and Alienosternus subgen. n. (type species Saprinus venustus Erichson, 1834). The subgenus Phoxonotus contains three species: Phoxonotus (P.) tuberculatus Marseul, 1862 (French Guyana, Suriname, Brazil: Pará, Mato Grosso), P. (P.) lectus Lewis, 1902 (eastern Peru) and the newly described P. parvotuberculatus sp. n. (Guatemala, Costa Rica). The subgenus Alienosternus subgen. n. contains two species: Phoxonotus (Alienosternus) venustus (Erichson, 1834) (Brazil: Bahia) and P. (Alienosternus) fryi Lewis, 1879 (Brazil: Rio de Janeiro and Bahia). P. tuberculatus Marseul, 1862 is synonymized with P. suturalis Lewis, 1907, syn. n. Lectotype of Phoxonotus fryi Lewis, 1879 is designated. Genitalia of the males for the species for which they are available are illustrated, most type specimens are imaged and scanning electron micrographs of P. tuberculatus provided. Mouthparts and sensory structures of the antennal club of P. tuberculatus are depicted. ZooBank Article LSID: 4C3BB497-274C-4E0E-B397-FCCD44AF8086 * Present address: Bavarian State Collection of Zoology, Münchhausenstraße 21, 81247 Munich, Germany. INTRODUCTION Genus Phoxonotus was erected by Marseul (1862), most likely based on a single specimen of a newly described species P. tuberculatus. Lewis (1879) briefl y described another species, P. fryi, based on several specimens; later he (Lewis, 1902) described another species, P. lectus, and redescribed P. fryi in more detail. Based on the description of Saprinus venustus Erichson, 1834, Lewis (1886: 280) transferred this species into the genus Phoxonotus most likely without ever examining the type specimen. Finally, Lewis (1908) added another species to the genus, P. sutu ralis, which he described based on a single specimen. From its establishment, the genus Phoxonotus was placed in the subfamily Dendrophilinae, based on the structure of its antennal cavity and prosternum, (see, e.g., Bickhardt, 1914, 1916 or Mazur, 1984). Although Lewis (1915) was the fi rst to question Bickhardt’s (1914) placement, it was Wenzel (unpubl.) who moved Phoxonotus into the Saprininae and this was adopted by Mazur in his last two catalogues (1997, 2011). The present paper critically examines all of the type speci mens of Phoxonotus (except for that of P. tuberculatus, Eur. J. Entomol. 113: 240–258, 2016 doi: 10.14411/eje.2016.029


INTRODUCTION
Genus Phoxonotus was erected by Marseul (1862), most likely based on a single specimen of a newly described species P. tuberculatus.Lewis (1879) briefl y described another species, P. fryi, based on several specimens; later he (Lewis, 1902) described another species, P. lectus, and redescribed P. fryi in more detail.Based on the description of Saprinus venustus Erichson, 1834, Lewis (1886: 280) transferred this species into the genus Phoxonotus most likely without ever examining the type specimen.Finally, Lewis (1908) added another species to the genus, P. sutu ralis, which he described based on a single specimen.From its establishment, the genus Phoxonotus was placed in the subfamily Dendrophilinae, based on the structure of its antennal cavity and prosternum, (see, e.g., Bickhardt, 1914, 1916or Mazur, 1984).Although Lewis (1915) was the fi rst to question Bickhardt's (1914) placement, it was Wenzel (unpubl.)who moved Phoxonotus into the Saprininae and this was adopted by Mazur in his last two catalogues (1997,2011).
Diagnosis.Ovoid, dorsally convex, light to dark brown non-metallic beetles with (prominent) tubercles on pronotum and elytra.Dorsum of body usually wholly punctate; apical pronotal angles prominent, pronotal hypomeron or body venter asetose.Frontal stria always interrupted and prolonged onto clypeus.Pronotal depressions absent, marginal pronotal stria single, carinate.Marginal elytral and inner subhumeral striae usually complete, carinate; elytral disc with striae 1-3 usually marked, stria 4 often marked by only a row of tubercles; in subgenus Alienosternus there is another row of tubercles between 4 th stria and elytral suture; in P. tuberculatus a single tubercle is also present on elytral suture near elytral base.Antennal cavity either a small depression in the ventral part of the composite plate (subgenus Phoxonotus; Fig. 30) or formed by the (Bratislava, Slovakia); scanning electron micrographs (SEM) were taken at the laboratory of the Faculty of Agriculture, Hokkaido University, Sapporo, Japan.All specimens were measured using an ocular micrometer.Beetle terminology follows that of Ôhara (1994) and Lackner (2010).
Abbreviations.Abbreviations of morphological measurements follow Ôhara (1994) and are used throughout the text as follows: APW -distance between anterior angles of pronotum; ELlength of an elytron measured along elytral suture; EW -distance between outer margins of elytra; PEL -distance between anterior angle of pronotum and apex of elytra; PPW -distance between posterior angles of pronotum.prosternal process and outer lateral costa of the antennal groove (subgenus Alienosternus; Fig. 64).Anterior third of prosternal process always elevated and broadened; carinal prosternal striae always divergent anteriorly, in the nominotypical subgenus forming a round loop; surface between carinal prosternal striae in one species with faint longitudinal tubercle.Apical part of mesoventrite protruding and fi tting into base of prosternal process.Apex of eighth sternite of male genitalia with an accessory sclerite (Figs 17,31,42,65); concealed under setae in P. venustus and P. fryi.Differential diagnosis.An unmistakable genus whose external morphology does not resemble any other taxon of currently known Saprininae.Among the phylogenetically meaningful characters are the deeply incised base of the prosternum, formed to accommodate the acutely angulate apex of the mesoventrite (no other currently known Saprininae taxon exhibits this feature); unique tubercles on pronotum and/or elytra (likewise absent in all other currently known Saprininae taxa); accessory sclerite of eighth sternite of male genitalia (absent in most Saprininae taxa); and absence of the sutural elytral stria (this stria is, interestingly, also absent in several other, apparently unrelated myrmecophilous or termitophilous forms, e.g., Myrmetes paykulli Kanaar, 1979 and an undescribed myrmecophilous genus from Australia: Lackner & Leschen, forth coming).

Key to the subgenera and species of Phoxonotus
Biology.Reichensperger (1935) was the fi rst to state that members of Phoxonotus are attaphilous based on a series of specimens of P. fryi collected from the nests of Atta ants.This was confi rmed by Kanaar (1997) who published a paper on attaphilous Histeridae from Suriname, which includes scores of Phoxonotus tuberculatus collected by D.C. Geijskes in detritus chambers of both Atta sexdens and A. cephalotes.I was able to examine specimens of P. tuberculatus mentioned by Kanaar (1997); one specimen of P. tuberculatus was collected using a fl ight interception trap.The two Guatemalan specimens of the newly described P. parvotuberculatus were collected in a large 6-8 feet (1.8-2.4 m) deep detritus cavity of a species of Atta.Biological data for other species of Phoxonotus are lacking.
Distribution.Central-and South America: known from Costa Rica, Brazil, Peru, French Guyana and Suriname (Fig. 74).
Note.The type specimen(s) of P. tuberculatus Marseul, 1862 were not located in the collection of S.A. Marseul, which is now part of the general collection of MNHN.According to the collection curator (A. Taghavian, pers. comm., 2014), it is possible that the type is elsewhere in the collection and therefore I am hesitant to designate a neotype pending the discovery of the original type specimen(s) of Marseul.Furthermore, the existence of a single specimen from "Cayenne" in the general collection of MNHN is mentioned by Dégallier & Kanaar (2001: 206).The following redescription is therefore based on the available material from French Guyana (which is the type locality) and Suriname, which is a neighbouring country, and carefully compared with Marseul's description (1862).Redescription.Body (Fig. 1) PEL = 3.25-3.65mm; APW = 1.25-1.50mm; PPW = 2.50-2.70mm; EW = 2.75-3.10mm; EL = 2.00-2.25 mm; oval, dorsally convex, ventrally fl attened, cuticle light to dark brown, without metallic lustre, wholly punctuate; punctures more prominent on elytra and metaventrite than on rest of body.Body appendages similarly coloured; antennal club lighter, ambercoloured.
Head: frons (Fig. 2) feebly convex, frontal disc covered with microscopic punctures, frontal stria widely interrupted, prolonged onto clypeus; supraorbital stria indistinct; clypeus slightly depressed, subtrapezoidal, margined anteriorly, disc on clypeus with elongate fi ne, almost confl uent punctures.Eyes rather small, inconspicuous from above.Antennal scape with several thin setae dorsally, rather  long, elbowed, slender, fi tting into deep furrow situated below eye laterad to mandibular base when head is retracted.Pedicel approximately twice as long as each 3 rd to 7 th antennal segment; antennal club (Fig. 3) oval, without apparent intersegmental sutures (although Marseul in his original description mentions them; see Remarks below), fl attened dorsoventrally, completely covered with dense short sensilla intermingled with sparse much longer erect sensilla; sensory structures on the antenna (Fig. 16) in form of a single pear-shaped vesicle situated on internal distal margin of antennal club under small inconspicuous round sensory area; another, smaller pacifi er-like vesicle situated on external proximal margin of antennal club without a complementary sensory area.Mandibles (Fig. 14) large, robust, microscopically wrinkled laterally, pointed apically; subapical tooth of left mandible blunt.Labrum (Fig. 15) wholly punctate, subtrapezoidal; basal half of labral disc distinctly elevated, apical half situated lower than basal; a pair of large labral pits situated near the lateral margin of labral disc, pits furnished each with two unusually long, slender labral setae.Setae of epipharynx rather long and dense, distinctly protruding.Labral process absent.Maxilla: cardo subrectangular, apical margin with two long thin setae, basal margin with numerous shorter setae; stipes rather small, triangular, with two thin rather long setae.Lacinia approximately three times smaller than galea, lacinial hook (= uncus) absent; galea with two distinct sclerites and large brush of dense long setae.Terminal maxillary palpomere slender, elongated, approximately four times longer than penultimate, narrowing apically, palpal organ inconspicuous (= absent?)Mentum (Fig. 4) subquadrate, slightly longer than wide, apical margin with prominent notch, disc of mentum smooth, laterally furnished with two rows of irregular short setae, anterior margin in anterolateral angles each with two longer setae.
Pronotum distinctly convex, moderately narrowing anteriorly, anterior angles prominent, anterior pronotal margin with deep incision for head.Marginal pronotal stria present, thin, slightly carinate laterally, anteriorly very thin, almost inconspicuous.Pronotal disc with rather dense microscopic punctuation, punctures separated by several times their diameter, punctures do not become coarser or denser laterally.Pronotal base with a row of four equallyspaced prominent tubercles: two outer ones almost round, globular, while two inner ones larger than two outer ones, elongate.Another two, distinctly smaller and lower tubercles each situated anterad of outermost tubercles of fi rst row; tubercles of both rows form together an imaginary trapezoid.Pronotal depressions absent, prescutellar depression faint; pronotal hypomeron asetose, glabrous.Scutellum small, triangular.
Elytra widest across humeri, wholly punctuate, punctures separated by twice to several times their diameter, more prominent along elytral suture.Elytral epipleuron almost smooth, marginal epipleural stria fi ne, complete; marginal elytral stria carinate and complete; apical elytral stria absent.Inner subhumeral stria long, carinate and complete, subparallel to marginal elytral stria, for short distance continuous along elytral apex.Oblique humeral stria faint, occasionally doubled.Elytral disc with three distinct dorsal elytral striae (in several specimens from Suriname the third stria is indistinct): 1 st the longest, almost complete; 2 nd stria present on basal third (roughly); 3 rd stria of approximately identical length as 2 nd , along its course running over two distinct tubercles: a smaller, round one basally, and a larger, elongate one apically; another, even smaller than basalmost round tubercle situated under large elongate median one.Fourth dorsal elytral stria absent; in its stead there are three approximately evenly-spaced prominent tubercles: basalmost one the smallest, median and apicalmost larger, approximately of the same size.Sutural elytral stria absent; surface between elytral suture and tubercles distinctly concave, forming a shallow furrow.Elytral suture elevated, elevation forming before scutellum an indistinct tubercle.
Propygidium (Fig. 5) half-covered by elytra, broad, covered with sparse fi ne punctures; pygidium (Fig. 5) triangular, covered with regular round punctures separated by approximately several times their diameter becoming smaller, fi ner and sparser apically.
Prosternum (Fig. 6): base of prosternum acutely inwardly emarginated to accommodate anteriorly projecting mesoventrite; prosternal process on basal 2/3 evenly to only slightly elevated, abruptly elevated and broadened on ante-  rior third.Carinal prosternal striae carinate, subparallel to slightly divergent on basal 2/3, on apical third widely divergent and united anteriorly under large round loop; surface between carinal prosternal striae fi nely punctuate.Lateral prosternal striae carinate, slightly convergent anteriorly, attaining carinal prosternal striae approximately at the point where the latter widely diverge anteriorly.Lateral costa of antennal groove terminates anterior to procoxa and does not reach prosternal process; antennal cavity for accommodating antennal club (while head is retracted) situated on lateral portion of anterior margin of composite ventral plate between tergopleural suture and procoxa, forming a well-defi ned, albeit shallow semiglobular impression; its base marked by a distinct ridge.
Metaventrite: disc of metaventrite with median longitudinal depression that is more prominent in male, surface around median longitudinal line covered with large and deep dense punctures separated by approximately their diameter; punctures become smaller and sparser anterolaterad; marginal metaventral stria almost straight, carinate, stopping short of metacoxa.Lateral disc on metaventrite (Fig. 8) with shallow confl uent ellipsoid punctures; anepisternum and metepisternum covered with identical punctation.
First visible abdominal sternite completely striate laterally, apices of striae slightly curved outwardly; disc with scattered microscopic punctation more prominent along base and lateral margins.
Legs: protibia (Fig. 9) fl attened and widened, on outer margin four distinct triangular teeth diminishing in size in proximal direction topped by prominent denticle followed by a row of 8-9 short denticles growing out from extremely low teeth diminishing in size in proximal direction; protarsal groove wide but shallow; setae of outer row very small, thin; setae of median row even shorter, microscopic; protibial stria vestigial, present only basally.Protibial spur long, prominent, growing out from near protarsal insertion.Apical margin of protibia with 3-4 short denticles; outer part of posterior surface of protibia (Fig. 10) with a single row of evenly-spaced short denticles; otherwise glabrous.Posterior protibial stria complete, terminating in four inner posterior denticles; setae of inner margin of protibia double.Mesotibia (Fig. 11) on outer margin with a row of 15+ denticles growing in size in proximal direction outer row of setae regular, shorter than denticles; posterior margin of mesotibia covered with short sparse setae not forming distinct rows; mesotibial spur long, prominent.Anterior margin of mesotibia with another row of tiny denticles; anterior mesotibial stria complete, terminating in two inner anterior denticles; setae of inner row strongly sclerotized, present only on proximal half of mesotibia (Fig. 12).Metatibia (Fig. 13) with on outer margin three short thin denticles separated from another two, longer denticles situated near apex of metatibia.Setae of outer row thin, regularly-spaced, growing in size apically; setae of median row much shorter.Metatibial spur long, prominent; anterior margin of metatibia with a single row of tiny denticles; anterior metatibial stria complete, terminating in three inner anterior denticles.Male genitalia: eighth sternite (Fig. 17) separated medially; apically with additional subquadrate sclerite furnished with several inconspicuous setae; eighth tergite widely inwardly arcuate basally, only slightly inwardly arcuate apically; eighth sternite and tergite not fused laterally (Fig. 19).Ninth and tenth tergites (Figs 20-21) typical for the subfamily; spiculum gastrale (Fig. 22): basal end ("tail") spoon-like, outwardly arcuate, apical end ("head") gradually dilated from approximately halfway along its length; laterally with downwardly turned "horns".Aedeagus (Fig. 24) tube-like, parameres fused along their basal half, apex of aedeagus pointed, basal piece dilated basally, rather short, its length : length of parameres approximately 1 : 4; aedeagus strongly curved ventrad (Fig. 25).
Biology.Beetles of the "Suriname series" were collected in detritus-rooms of Atta cephalotes L. and A. sexdens L. by D.C. Geijskes in 1938 in northern Suriname.According to Kanaar (1997) the detritus-rooms of the nests of both Atta species are situated on the periphery of the system of galleries in the nests.In addition to exhausted material from the fungus gardens of the ants, these rooms contain numerous dead ants (Kanaar, 1997).It is likely that both the larvae and adults of Phoxonotus feed on some arthropods associated with the detritus-rooms of Atta ants.One specimen (from Mato Grosso, Brazil) was collected using a fl ight intercept trap.
Differential diagnosis.This species differs from the very similar P. lectus by the denser punctuation of the meso ventrite and a thinner aedeagus.See also differential diagnosis of P. lectus.From the newly described P. parvotuberculatus, P. tuberculatus differs in the presence of a second row of tubercles on the pronotum, and more prominent and numerous elytral tubercles.
Remarks.Marseul's original description of the genus and species (1862: 35-38) in general agrees with the present redescription based on specimens from French Guyana and Suriname, with the following discrepancies: (a) Marseul writes that the mandibles lack a subapical tooth ("sans dent interne").This was not observed as there is a short and blunt subapical tooth present on the mandibles of the specimens studied (Fig. 14); (b) antennal club formed by four distinct segments (Marseul, 1862: 36;Fig. 1b;Fig. 26 in the present paper; "massue…de quatre articles peu distincts"…); this again was not observed, there were no traces of intersegmental sutures on the antennal clubs of specimens studied (Fig. 3); (c) feebly acute anterior pronotal angles of Marseul (1862: 36 "…les angles arrondis, peu saillants…"), whereas, the anterior angles of the pronotum are rather prominent and acute in the specimens studied (Fig. 1); (d) Marseul's four dorsal elytral striae (1862: 36: "…quatre stries dorsales fi nes, obliques…"; in his Fig. 1, however, this fourth stria is not fi gured), whereas, the fourth stria is always lacking and represented only by a row of tubercles in the specimens studied (Fig. 1); (e) Marseul's misinterpretation of the subhumeral striae.Marseul (1862: 36) writes "…deux subhumérales parallèles sur le bord lateral…", whereas, in my opinion one of the two striae is the true inner subhumeral and the second the marginal elytral stria; (f) Marseul (1862: 36, 38, Fig. 1a) mentions carinal prosternal striae being almost parallel and then widely divergent anteriorly.In the specimens studied, the carinal prosternal striae are never completely parallel (e.g., Fig. 6), they are more or less divergent along their course before they become widely divergent and form a rounded loop apically.Lewis (1907) described P. suturalis from Obidos River, Amazonia (actually situated in Pará State of modern Brazil).He distinguished this newly described species from all other congeners by the presence of "the sutural tubercle on the dorsum".However, all specimens of P. tuberculatus examined from French Guyana and Suriname also have this sutural tubercle.Marseul (1862) in his description of P. tuberculatus does not mention this "sutural tubercle", which he probably overlooked.As the original Marseul's type specimen(s) of P. tuberculatus cannot be located in the collection of MNHN, it is assumed that the specimens examined here from French Guyana and Suriname are conspecifi c with P. tuberculatus (as they agree with the description in most morphological characters) and thus this "sutural tubercle" is present in P. tuberculatus.The type locality of P. suturalis, Obidos city on the Amazon River, lies south of the "Guyanas", in the Amazon River basin; both French Guyana and Suriname are likewise part of this vast ecosystem.Type material examined.Holotype, ♂, mounted on tip of a triangular mounting point, right antenna reduced to scape and pedicel, other segments missing, right protibia and mesotibia missing, left mesotibia broken off, glued to the same mounting point as the specimen (devoid of tarsal segments 2-5), segments 3-5 of left metatarsus missing, with extracted and dismembered male genitalia glued onto the same triangular point as the specimen with the following labels: "♂" (printed); followed by: "E.Peru" (printed); followed by: "Type" (red-margined round label); followed by: "Phoxonotus / lectus / Type / Lewis" (written) (BMNH).
Note.The original description does not specify the number of type specimens, but does not indicate that more than one existed.Only one type specimen was found in BMNH, which is thus assumed to be the holotype.
Additional material examined.None.

Diagnostic description.
Because of the overall similarity of Phoxonotus lectus with P. tuberculatus, only a very brief diagnostic description the former species is provided, which outlines the differences between the two taxa.Body (Figs 27, 28) PEL = 3.25 mm; APW = 1.25 mm; PPW = Differential diagnosis.P. lectus differs from P. tuberculatus mainly in its more dilated aedeagus and fi ner and sparser punctation on the mesoventrite, and from P. parvotuberculatus in the same characters.
Head: fronto-clypeal region faintly depressed, frontal disc with scattered microscopic punctures, frontal stria widely interrupted, prolonged onto clypeus; supraorbital stria indistinct; clypeus not margined anteriorly, disc of clypeus with very faint confl uent punctures.Eyes rather small, inconspicuous from above.Antenna generally similar to that of P. tuberculatus; sensory structures on antenna not examined.Mandibles in general like those of P. tuberculatus.Labrum glabrous, subtrapezoidal; with large longitudinal median convexity followed by strongly depressed anterior third; a pair of large labral pits each with two unusually long, slender labral setae situated on boundary between the anterior third and longitudinal median convexity.Rest of mouthparts not examined.
Pronotum (Fig. 40) in general similar to that of P. tuberculatus, but marginal pronotal stria prolonged for a short distance along pronotal base; pronotal disc with a single row of faint tubercles situated medially on basal pronotal third: two inner tubercles more prominent than the two vague outer ones.Scutellum small and triangular.
Elytra widest across humeri, wholly punctuate, punctures separated by about their diameter.Elytral epipleuron fi nely microscopically punctuate, marginal epipleural stria fi ne, complete and carinate; marginal elytral stria carinate and complete; apical elytral stria absent; inner subhumeral stria long, carinate and complete, subparallel to marginal elytral stria, for short distance continuous along elytral apex.Oblique humeral stria faint; elytral disc with three distinct dorsal elytral striae: 1 st the longest, almost complete, costate on apical two-thirds; 2 nd and 3 rd dorsal elytral striae shortened basally, stopping short of elytral half.Base of third elytral stria marked by a small vague tubercle; fourth dorsal elytral stria absent, its imaginary base marked by a prominent round tubercle; another, very faint elongate tubercle present at approximately half way along elytra; surface between these two tubercles and elytral suture depressed; elytral suture elevated.
Propygidium almost entirely covered by elytra; pygidium triangular, covered with regular round punctures separated by approximately their or several times their diameter becoming smaller, fi ner and sparser apically.
Prosternum (Fig. 41): structure of prosternum overall similar to that of P. tuberculatus; lateral prosternal striae carinate; structure of antennal cavity is similar to that of P. tuberculatus.
Mesoventrite (Fig. 41) in general similar to that of P. tuberculatus, but disc of mesoventrite with faint scattered tiny punctuation, punctures separated by their own diameter; meso-metaventral sutural stria faint, thin, in form of interconnected punctures.
Metaventrite (Fig. 41): disc of metaventrite in general overall similar to that of P. tuberculatus; marginal metaventral stria almost straight, carinate, stopping short of metacoxa; lateral disc on metaventrite as in P. tuberculatus.
First visible abdominal sternite as in P. tuberculatus.Legs: very similar to those of P. tuberculatus.Male genitalia .Generally quite similar to the type of the genus, P. tuberculatus, differing from it chiefl y in having an apically notched spiculum gastrale (compare Figs 22 and 47) and an aedeagus that is slightly broader apically (compare Figs 24 and 49).
Biology.The three Guatemalan specimens were collected from a large detritus cavity of Atta sp., 6-8 feet (= 1.8-2.4m) deep; biological data on Costa Rican specimens are lacking.
Distribution.Known only from Guatemala and Costa Rica (Fig. 74).
Differential diagnosis.P. parvotuberculatus is in most external morphological characters similar to both P. tuberculatus and P. lectus, but differs from both of them in the absence of a second row of tubercles on the pronotum and much less developed elytral tubercles.
Etymology.The specifi c epithet of this species refers to the small tubercles on the dorsum.It is composed of the Latin words "parvus", meaning "small", and "tuberculatus", meaning "tuberculate", or furnished with bolts, tubercles.Diagnosis.Antennal cavity typically Saprininae-like (see, e.g., Lackner, 2010: Figs 148-149) formed by the prosternal process and outer lateral costa of antennal groove (Fig. 64); carinal prosternal striae angulate apically, not forming a rounded loop (Fig. 64).Between the row of tubercles indicating the course of the fourth dorsal elytral stria and elytral suture there is another row of four tubercles.
Differential diagnosis.See Differential diagnosis for Phoxonotus s. str.
Biology.See biology of Phoxonotus s. l.Distribution.Brazil: Bahia and Rio de Janeiro states (Fig. 74).
Note.The original description does not specify the number of type specimens, but does not indicate that more than one existed.Only one type specimen was found in ZMHB, which is thus assumed to be the holotype.
Additional material examined.None.
Note.This species was described from "three or four specimens in Mr. Fry's collection all from Rio Janeiro district" (Lewis, 1879: 61).There are two specimens in BMNH with "Type" labels from "RioJan" [= Rio de Janeiro], designated here as a lectotype and paralectotype, respectively, and another specimen from Bahia originating also from Fry's collection that was labelled as P. fryi.Since this last specimen was not labelled as "type" and originates from a different locality than the two syntypes (Bahia vs Rio de Janeiro), only the two syntypes were labelled as the lectotype and paralectotype, respectively.Another specimen, found in the collection of FMNH, although labelled as "Paratype" by Lewis, and originating from Fry's collection, is from a locality that is different from the one given by Lewis (Espirito Santo vs Rio de Janeiro) and therefore was not labelled as a paralectotype.
Diagnostic description.Because of the overall similarity of Phoxonotus fryi with P. venustus, the former species is provided only with diagnostic description outlining the differences between the two taxa.Body (Figs 61, 62) PEL = 3.00-3.10mm; APW = 1.00 mm; PPW = 2.20-2.50mm; EW = 2.50-2.60 mm; EL = 2.00-2.10mm.This species is externally very similar to P. venustus, differing from it chiefl y in the shape of the spiculum gastrale, which is slightly dilated approximately half way along its length, forming a distinct "head" with characteristic, downwardly  curved lateral "arms" (Fig. 70).In all other external morphological characters it resembles P. venustus Er.
Differential diagnosis.It differs from P. venustus mainly in the shape of its spiculum gastrale (compare Figs 57 and 70).
Biology.Reichensperger (1936) reports that the 9 specimens from Mendes (State of Rio de Janeiro, Brazil) are clearly attaphilous, as they were collected in nests of both A. cephalotes (L., 1758) and A. sexdens (L., 1758).The whereabouts of these 9 specimens are unknown.
Remarks.Lewis (1902: 271) mentions a "straight crenulated line at the suture of the mesosternum" (= mesoventrite) as a "marked character of this species…".According to my observations, this "straight crenulated line at the suture of the mesosternum" (= meso-metaventral sutural stria) is present in all species of the genus, albeit variously "crenulated".Based on the external morphology and most of the characters of its genitalia it is possible that this species is synonymous with Erichson's P. venustus, which Lewis combined with Phoxonotus in his 1886 paper, citing "ten [sic!; he probably meant "the" instead of "ten"] disposition of the tubercles on the thorax and the arrangement of the elytral striae" as the characters of P. venustus, which would surely differentiate it from both P. tuberculatus and P. fryi (Lewis 1886: 280).Lewis most likely never examined the type specimen of P. venustus Erichson and it is uncertain what he exactly meant by the disposition of the tubercles on the thorax and the arrangement of elytral striae.

DISCUSSION
Because of the structure of its prosternum and the position and shape of the antennal cavity, Photoxonotus used to be historically (see, e.g., Bickhardt, 1914, 1916or Mazur, 1984) classifi ed among the Dendrophilinae; and even after its placement in the Saprininae (Mazur, 1997, based on Wenzel's unpublished manuscripts) it was regarded as transitional between the groups (Ślipiński & Mazur, 1999: 210).Having examined all the described and one undescribed species of this genus I conclude that the atypical form and placement of the antennal cavity (present in members of the nominotypical subgenus) is most likely derived from the "normal" condition (= plesiomorphic) in Saprininae, where the antennal cavity is formed by the prosternum and outer lateral costa of the antennal groove (sensu Lackner, 2010).It was most likely this character state, fi rst observed and described by Marseul (1862), that cast doubt on the exact sub-familial placement of Phoxonotus.The members of the newly established subgenus Alienosternus, although as the name suggests have a "weird sternum", in fact have the typical arrangement and placement of the antennal cavity for the subfamily Saprininae.I believe that only one independent modifi cation of the placement and form of the antennal cavity occurred in the genus: from the "normal" (= plesiomorphic) antennal cavity formed by the outer lateral costa of the antennal groove and prosternal process (found in Alienosternus) to the apomorphic state, in which the small antennal cavity is formed by the ventral part of the composite plate (sensu Ôhara, 1994), found in the subgenus Phoxonotus.The members of these two subgenera otherwise share several synapomorphies: (1) deeply incised base of the prosternum, formed to accommodate the acutely angulate apex of mesoventrite; (2) unique tubercles on pronotum and/or elytra; (3) accessory sclerite on eighth sternite of male genitalia; (4) widely interrupted frontal stria prolonged onto the clypeus; (5) apices of pronotum acute; (6) carinate and almost complete marginal and inner subhumeral striae; and (7) absence of sutural elytral stria.Most of these character states are possible autapomorphies of this genus.In the light of so many distinctive characters, and absence of "true" morphological modifi cations of the Dendrophilinae (e.g., presence of "deep lateral groove of the prosternum demonstrated to receive the strongly developed and hooked protibial spur in repose that forms an interlocking mechanism holding fore legs covering the antenna from below"; Ślipiński & Mazur, 1999: 218) I do not consider Phoxonotus a "basal" member of the Saprininae or a "transitional form" between groups.
Although Ślipiński & Mazur (1999: 228) state that the Reichardt's organ (sensory structures of the antennal club) is "indistinct or absent in Phoxonotus", I was able to locate two distinct vesicles inside the antennal club.According to Ślipiński & Mazur (1999: 216), who based their results on an unknown species of Phoxonotus from Costa Rica, the antennal cavity in the subgenus Phoxonotus is "weakly developed, but positioned in an impression or cavity on inner portion of prosternum adjacent to the prosternal keel".As I was unable to check their specimen, I can only guess that theirs' was the derived state found in the subgenus Phoxonotus.On the other hand, the true "basal" character states of the subfamily, as indicated by my morphological studies (Lackner, 2014), such as presence of a lacinial hook (uncus) or longitudinally divided ninth tergite in the genitalia, are absent in Phoxonotus.Perhaps in the nominotypical subgenus the position and placement of the antennal club in repose in a small cavity on the ventral part of the composite plate does give members of the subgenus Phoxonotus more exposure or sensitivity; the reasons for which remain merely speculative.Discovery and comparison of larvae of both subgenera and/or molecular characters would perhaps present an independent character system(s) for checking their inter-relationships.
ACKNOWLEDGEMENTS.I thank my wife P. Artimová for help with the Latin subgeneric name and the new specifi c epithet for Phoxonotus; I am likewise indebted for her help with Adobe Illustrator CS3.Curators of the aforementioned museums are thanked for their help with the specimens.I am thankful to A. Pütz (Eisenhüttenstadt, Germany) and P. Schüle (Herrenberg, Germany) for their help with the German language.M. Caterino (Clemson, USA) is especially thanked for providing the map of South and Central America.This research received support from the SYNTHESYS Project (http://www.synthesys.info/),which is fi nanced by the European Community Research Infrastructure Action under the FP7 Integrating Activities Program.Likewise this research was partially supported by a grant from the Faculty of Forestry and Wood Sciences, Specifi c research, grant n.B06/15.Special thanks are due to two anonymous reviewers and editor of the European Journal of Entomology who provided numerous corrections and suggestions, which increased the quality of this paper.