Mygalodelphys, Pavan & Voss, 2016

Pavan, Silvia E. & Voss, Robert S., 2016, A Revised Subgeneric Classification of Short-tailed Opossums (Didelphidae: Monodelphis), American Museum Novitates 2016 (3868), pp. 1-44 : 19-22

Mygalodelphys , new subgenus

TYPE SPECIES: Monodelphis adusta ( Thomas, 1897) .

CONTENTS: adusta Thomas, 1897 (including melanops Goldman, 1912); peruviana Osgood, 1913 ; osgoodi Doutt, 1938 ; kunsi Pine, 1975 ; reigi Lew and Pérez-Hernández, 2004 ; ronaldi Solari, 2004 ; handleyi Solari, 2007 ; and pinocchio Pavan, 2015 .

DIAGNOSIS: Dorsal body pelage unpatterned; ventral pelage uniformly colored or with self-whitish median markings. 5 Mammae 2–0–2 = 4 (e.g., in M. peruviana ; AMNH 264562), 3–0–3 = 6 (e.g., in M. adusta ; AMNH 202650), or 3–1–3 = 7 (e.g., in M. pinocchio ; MZUSP MTR15815), all abdominal-inguinal. Thenar and first interdigital pad of pes separate, not fused; hypothenar pad of pes present (but unknown for M. reigi , M. peruviana , and M. ronaldi ). Body pelage extends onto tail farther ventrally than dorsally; tail scales arranged in annular or spiral series. Infraorbital foramen dorsal to M1; frontal process of jugal absent or indistinct; parietal usually (> 90% of examined specimens) not in contact with mastoid; length of incisive foramina variable; length of maxillopalatine fenestra variable; sphenorbital fissure small (basisphenoid laterally concealed); infratemporal crest of alisphenoid distinct or indistinct; secondary foramen ovale usually absent 6; tympanic wing of alisphenoid small; tip of anterior process of malleus exposed on external bullar surface between ectotympanic and alisphenoid; rostral tympanic process of petrosal narrow and triangular, not concealing fenestra cochleae in ventral view; stapes columelliform, imperforate or microperforate; subsquamosal foramen large. Anterior cingulids of m2 and m3 narrow; entoconids of m1–m3 very small, indistinct; dp3 small, with incomplete trigonid and indistinct anterior cingulid in some species (e.g., M. adusta , M. reigi ), but dp3 large, with complete trigonid and distinct anterior cingulid in other species (e.g., M. handleyi ; the morphology of dp3 is unknown for M. peruviana , M. osgoodi , M. ronaldi , M. pinocchio , and M. kunsi ).

COMPARISONS: Members of the subgenus Mygalodelphys differ from currently recognized species in other subgenera of Monodelphis by several unique external and craniodental traits, including: (1) soπ body pelage that extends onto the tail farther ventrally than dorsally; (2) frontal process of jugal absent or indistinct; (3) parietal-mastoid contact absent; (4) a small sphenorbital fissure that does not expose the basisphenoid to lateral view; (5) narrow lower molar anterior cingulids; and (6) indistinct entoconids on m1–m3. Self-whitish midventral pelage markings are also unique to Mygalodelphys , although they are oπen polymorphic and are not present in all member species.

Among other diagnostic comparisons ( table 2 ), Mygalodelphys additionally differs from Pyrodelphys by its unpatterned dorsal pelage, separate thenar and first interdigital pads on the hind foot, small alisphenoid tympanic wing, exposure of the anterior process of the malleus on the external surface of the bulla, narrow-triangular rostral tympanic process of the petrosal, and a large subsquamosal foramen. Mygalodelphys additionally differs from the usual morphology seen in the nominotypical subgenus by possessing a distinct hypothenar pad on the hindfoot, an infraorbital foramen that is dorsal to M1, and a columelliform stapes. Mygalodelphys additionally differs from Microdelphys by its consistently unpatterned dorsal pelage, small alisphenoid tympanic wing, exposure of the anterior process of the malleus on the external surface of the bulla, and narrow-triangular rostral tympanic process of the petrosal. Mygalodelphys additionally differs from Monodelphiops by its unpatterned dorsal pelage, lack of pectoral mammae, and possession of a hypothenar pad of the hind foot.

ETYMOLOGY: From mygale, ancient Greek for “shrew,” which members of this clade strikingly resemble in general aspect.

REMARKS: Mygalodelphys corresponds to “clade E” or the “Adusta Group” ( Pavan et al., 2014; Pavan et al., 2016), which was recovered with consistently robust support in our previous phylogenetic analyses. Although taxon-dense phylogenetic analyses incorporating morphological characters have yet to be done, it seems likely that several features unique to this subgenus (e.g., body pelage extending onto the tail farther ventrally than dorsally; frontal process of the jugal absent or indistinct; no parietal-mastoid contact; narrow lower molar anterior cingulids) will eventually be found to optimize as subgeneric synapomorphies.

Phylogenetic analyses based on mitochondrial and nuclear gene sequences ( Pavan et al., 2014; Vilela et al., 2015; Pavan et al., 2016) have consistently recovered a basal dichotomy among the species that we refer to Mygalodelphys : one clade including Monodelphis kunsi and M. pinocchio (M. “species 1” of Pavan et al., 2014; Vilela et al., 2015), and another including M. adusta , M. reigi , M. peruviana , M. osgoodi , M. handleyi , and a still-undescribed form (M. “species 2”). Although these clades are robustly supported by sequence data, morphological data does not support their formal taxonomic recognition. Despite being sister taxa, M. pinocchio and M. kunsi are externally and cranially dissimilar ( Pavan, 2015), and we are not aware of any phenotypic trait shared by these two species that consistently distinguish them from the remaining species of Mygalodelphys .

Although Monodelphis ronaldi has not been included in any phylogenetic analysis to date, we allocate this species to the subgenus Mygalodelphys based on its close phenetic similarity to M. handleyi (previously noted by Solari, 2007) and to its shared possession of morphological traits that seem likely to optimize as subgeneric synapomorphies, including (1) lack of a distinct frontal process of the jugal, (2) a small sphenorbital fissure within which the basisphenoid is not laterally exposed, (3) lack of parietal-mastoid contact, and (4) narrow anterior cingulids on m2 and m3. Including M. ronaldi in future phylogenetic analyses will effectively test the hypothesis that it is a member of Mygalodelphys .

NOTES ON DISTRIBUTION AND SYMPATRY: Species of the subgenus Mygalodelphys are known from eastern Panama; the humid tropical and subtropical Andes (to ca. 3000 m) of Colombia, Ecuador, Peru, and Bolivia; the Guiana Highlands of southern Venezuela and western Guyana; western and southeastern Amazonia 7; the Atlantic Forest of southeastern Brazil; the Cerrado landscapes of central Brazil; and the Cerrado, Chaco, and adjacent dry-forested biomes of Bolivia, Paraguay, and northeastern Argentina ( table 3 ). Species of Mygalodelphys are sympatric with Pyrodelphys in southwestern and southeastern Amazonia (e.g., in the lower Urubamba region of eastern Peru; Solari et al., 2001), with species of the subgenus Monodelphis in Amazonia and the Cerrado (e.g., at Bosque Mbaracayú in eastern Paraguay; de la Sancha et al., 2007), with species of the subgenus Microdelphys in the Andes and the Atlantic Forest (e.g., at Riacho Grande, São Paulo, southeastern Brazil; Pavan, 2015), and with species of Monodelphiops in the Atlantic Forest (e.g., at Parque Nacional do Itatiaia, southeastern Brazil; Pavan, 2015).

Given this wide distribution and extensive sympatry, the absence of Mygalodelphys throughout most of northeastern Amazonia (north of the Amazon and east of the Rio Negro), where only species of the nominotypical subgenus are known to occur in lowland habitats, is noteworthy. It is also worth noting that Mygalodelphys is the only subgenus known to occur in the northern Andes (north of the Huancabamba Deflection), and in northwestern Amazonia (north of the upper Amazon and west of the Rio Negro). Whether historical or ecological factors account for such distributional phenomena is unknown.