Marmosa (Micoureus) jansae, Voss & Giarla, 2021

Voss, Robert S. & Giarla, Thomas C., 2021, A Revision of the Didelphid Marsupial Genus Marmosa Part 3. A New Species from Western Amazonia, with Redescriptions of M. perplexa Anthony, 1922, and M. germana Thomas, 1904, American Museum Novitates 2021 (3969), pp. 1-28 : 5-11

Marmosa (Micoureus) jansae , new species

Marmosa “Yasuni” Voss et al., 2020: 11 (provisional name for a putative species identified by phylogenetic analyses of molecular sequence data).

TYPE MATERIAL AND TYPE LOCALITY: The holotype consists of the skin, skull, postcranial skeleton, and frozen tissues of an adult male ( ROM 118880 ; original number F41891 ) collected 42 km south and 1 km east of Pompeya Sur in Parque Nacional Yasuní , Orellana province, Ecuador, on 29 May 2006 by M.D. Engstrom and S.L. Peters (see appendix 1 for geographic coordinates). Molecular sequence data obtained from the holotype by Voss et al. (2020) include the entire cytochrome b gene (1149 bp, GenBank accession number MN 978679 ) together with fragments of the following loci (abbreviated as in Voss et al., 2020): Anon128 (690 bp, MN 978693 ) , BRCA1 exon 10 (949 bp, MN 978705 ) , IRBP exon 1 (1180 bp, MN 978712 ) , and OGT intron 14 (645 bp, MN 978724 ) .

DISTRIBUTION AND SYMPATRY: Marmosa jansae is currently known from the Amazonian lowlands of southeastern Colombia (Putumayo), eastern Ecuador (Orellana, Pastaza), and northeastern Peru (Loreto); the three Peruvian localities from which M. jansae has been collected (see Specimens Examined, below) are all on the north (left) bank of the Amazon/Marañón ( fig. 2 ). Marmosa jansae has been collected in sympatry with several congeners, including M. (Eomarmosa) rubra , M. (Micoureus) germana , and M. (Marmosa) waterhousei (appendix 2).

DESCRIPTION: The dorsal pelage of Marmosa jansae is dull grayish brown ( fig. 3 ); in most specimens the color is close to Ridgway’s (1912) Deep Grayish Olive, but in a few the color is closer to Hair Brown. Although this drab pigmentation is unremarkable (most other members of the subgenus Micoureus also have grayish-brown dorsal pelage), the fur of M. jansae is distinctively short, ranging in middorsal length from just 8 to 10 mm. The ventral pelage is almost entirely gray-based; self-colored (pale buff) fur is only consistently present ventrally on the chin and groin. In a few specimens (e.g., ROM 40399, TTU 157974) there is a narrow, discontinuous midventral streak of self-colored fur, but the fur of the chest and the insides of the forelimbs is always gray-based. The hind feet are covered dorsally with short, pale-buff hairs, but the metacarpals and proximal phalanges of the forefeet are covered with dark (brownish) hairs, such that only the distal manual phalanges are pale. The tail is about 137% of the combined length of the head and body, on average, and only 20 mm or less of the base of the tail is covered with short (≤ 5 mm) fur; the naked, scaly part of the tail is uniformly dark dorsally, but the ventral surface is sometimes indistinctly paler.

Adult skulls ( fig. 4 ) are unremarkable in dorsal view, each with the widely flaring zygomatic arches, short, broad rostrum, and well-developed postorbital processes that are typical cranial features of the subgenus Micoureus . In ventral view, the skull is distinguished by short (3.8 ± 0.5 mm) maxillopalatine fenestrae that usually extend only from a point opposite the P3/M1 commissure to a point opposite the M2/M3 commissure, by the complete absence of palatine fenestrae, and by auditory bullae that often have well-developed anteromedial laminae. The latter are noteworthy because, in most marmosines (e.g., Marmosa murina ; Voss and Jansa, 2009: fig. 16A), the extracranial course of the maxillary nerve is unenclosed by bone; the anteromedial lamina does not completely enclose the nerve in most specimens of Marmosa jansae , but in a few specimens (e.g., ROM 106151) it forms a secondary foramen ovale resembling the morphology sometimes seen in Monodelphis ( Voss and Jansa, 2009: fig. 16C). The second upper premolar (P2) has an incomplete lingual cingulum, and the upper molar toothrow is short (LM = 7.9–8.5 mm) by comparison with other Amazonian species of Micoureus (see below). The postprotocrista of M3 terminates at or near the base of the metacone, and the lower molars lack posterior cingulids. 3

COMPARISONS: Marmosa jansae averages larger than its trans-Andean sister species, M. perplexa , in all measured dimensions ( tables 1, 3). Despite its larger body size, however, M. jansae has shorter (8–10 mm versus 11–13 mm) dorsal fur than M. perplexa , but these species are otherwise externally similar. In dorsal view, skulls of M. jansae ( fig. 4A ) have visibly shorter, broader rostrums in proportion to cranial length than skulls of M. perplexa ( fig. 4B ), but for lack of suitable landmarks these proportions are difficult to quantify. A more conspicuous difference is that the postorbital processes are always well developed in adult skulls of M. jansae , whereas the postorbital processes are never very large in M. perplexa . In ventral cranial view,

3 For illustrations of these and contrasting molar traits of other species, see Voss et al. (2020: figs. 10, 11). M. jansae has absolutely and relatively shorter maxillopalatine fenestrae (on average) than M. perplexa , although there is some taxonomic overlap in measured fenestral lengths (2.8–4.4 mm in M. jansae versus 4.2–5.2 mm in M. perplexa ). Lastly, none of our specimens of M. jansae has palatine fenestrae, whereas palatine fenestrae are consistently present in our material of M. perplexa . Such morphological differences, together with the 8% average divergence in cytochrome b sequences reported by Voss et al. (2020: table 4) seem sufficient to support the hypothesis that these are distinct species.

The geographic range of Marmosa jansae overlaps the distributions of two other Amazonian species of the subgenus Micoureus — M. germana and M. rutteri —but these taxa are not closely related ( fig. 1 ) and they are morphologically dissimilar ( table 2). By comparison with M. jansae , M. germana is much larger in all measured dimensions ( table 4), has longer dorsal fur, has a much longer furry tail base, and lacks anteromedial bullar processes. Marmosa rutteri is also substantially larger than M. jansae (see measurement data summarized by Voss et al., 2020: table 9), but these two species have similarly short dorsal fur and short-furred tail bases. Instead, M. jansae and M. rutteri are most easily distinguished externally by ventral coloration: whereas M. jansae has almost entirely gray-based ventral fur, M. rutteri almost always has a broad midventral zone of self-colored (usually yellowish) fur that extends continuously from chin to groin, and the insides of the forelegs are also self-colored. Additionally, M. rutteri has a complete lingual cingulum on P2, a long postprotocrista on M3, and small but usually persistent posterior cingulids on m1–m3.

REMARKS: One of the specimens we refer to Marmosa jansae (BMNH 11.7.19.20) was identified by Tate (1933: 80) as M. germana germana , but later in the same publication (on p. 82) he identified two other specimens (AMNH 64025, 68142) as M. germana rutteri . 4 Such misidentifications exemplify the taxonomic confusion that has long prevailed regarding western Amazonian species of Marmosa , and they help explain why M. jansae (specimens of which were first collected>100 years ago) remained undescribed until now. More recently, specimens of Marmosa jansae were identified as M. regina by Hice and Velazco (2012), an identification consistent with the then-prevailing notion that M. germana and M. rutteri were synonyms or subspecies of M. regina . However, M. regina is not a member of the subgenus Micoureus ( Giarla and Voss, 2020) , and an application has been submitted to the International Commission on Zoological Nomenclature to suppress usage of the epithet ( Voss and Giarla, 2020).

HABITATS: Lowland rainforest is the natural climax vegetation throughout the known geographic range of Marmosa jansae , but only two specimens (TTU 99000, 101208) are accompanied by explicit habitat descriptions. According to Hice and Velazco (2012: 27), who summarized the results of a multiyear faunal inventory project at the Reserva Nacional Allpahuayo-Mishana in Loreto department, Peru, these specimens were trapped “in monte alto forest several hours walk from the road.” The authors explained that monte alto ( fig. 5 ) is the typical rainforest of the area, where it occurs on clayey soils in low-lying areas; among other attributes, monte alto forest has a high (30–40 m) canopy, abundant understory palms, and numerous lianas. By contrast, five sympatrically collected specimens of M. germana (TTU 98845, 98846, 98988, 101094, 101236) were trapped or shot “in varillal habitat and in older secondary growth” ( Hice and Velazco, 2012: 25). Varillal is the local name for a distinctive type of stunted forest that grows on white sand on upland terrain; among other floristic and structural differences, it has

4 Only the skin of AMNH 64025 is from Marmosa jansae ; the mismatched skull is from Marmosops noctivagus . This specimen was purchased from commercial collectors, who were probably responsible for the mismatched elements.

a much lower canopy (with a maximum height of 20 m), fewer understory palms, and fewer lianas than does adjoining monte alto habitat.

Few as they are, these observations from the Reserva Nacional Allpahuayo-Mishana suggest that, within the matrix of what is broadly considered lowland rainforest, sympatric species of the subgenus Micoureus might occupy floristically and structurally distinct habitats correlated with soil types, successional stage, or (as discussed by Voss et al., 2020), seasonal flooding regimes. Although Hice and Velazco’s (2012) results imply that Marmosa jansae prefers tall primary forest over secondary vegetation or white-sand forest, only future fieldwork carefully designed to assess faunal differences among adjacent vegetation types is likely to provide compelling evidence of such habitat associations.

ETYMOLOGY: For our friend, mentor (to T.C.G.), and colleague Sharon A. Jansa ( fig. 6 ), whose molecular-phylogenetic research has produced major advances in systematic understanding of both marsupials and rodents.

SPECIMENS EXAMINED (N = 22): Colombia — Putumayo, Mocoa (ROM 40399, 46430, 46431), Río Mecaya (FMNH 70962, 70967, 70968). Ecuador — Orellana, Cerro Guataraco (USNM 267503), 42 km S and 1 km E Pompeya Sur (ROM 118880), Río Suno Abajo (AMNH 68142), Río Suno Arriba (AMNH 64025), Tiputini Biodiversity Station (ROM 106151); Pastaza, Canelos (BMNH 11.7.19.20), Río Copotaza (FMNH 53347, 53349, 53351), Río Pastaza (BMNH 54.267). Peru — Loreto, 25 km S Iquitos (TTU 99000, 101208), 52 km S Iquitos (TTU 124792, 124793), 1.5 km N Teniente López (KU 157974, 157875).

Figures

FIG. 2. Collection localities of examined specimens of Marmosa germana, M. jansae, and M. perplexa. Numbers are keyed to entries in our gazetteer (appendix 1).

FIG. 3. Dorsal and ventral views of skins of Marmosa jansae (ROM 118880), M. perplexa (USNM 513425), and M. germana (FMNH 43179).

FIG. 4. Dorsal and ventral cranial views (×1.75) of Marmosa jansae (A, D; ROM 118880), M. perplexa (B, E; USNM 513425), and M. germana (C, F; TTU 101236).

FIG. 1. Maximum-likelihood (ML; left) and Bayesian (right) phylogenetic trees for putative species of Marmosa (Micoureus) inferred from concatenated sequences of six genes (after Voss et al., 2020: fig. 6). Numbers at nodes denote ML bootstrap support or Bayesian posterior probabilities. Outgroups are not shown. Dotted lines connect the same taxa across the two trees to highlight differences. Names of species treated in this report are in boldface.

FIG. 5. Rainforest (monte alto) habitat of Marmosa jansae at the Reserva Nacional Allpahuayo-Mishana, Loreto, Peru (photo by Christine L. Hice).

FIG. 6. Sharon A. Jansa (with Marmosa robinsoni on Tobago; photo by F. Keith Barker).