Mesomys occultus, PATTON & DA SILVA & MALCOLM, 2000

PATTON, JAMES L., DA SILVA, MARIA NAZARETH F. & MALCOLM, JAY R., 2000, Mammals Of The Rio Juruá And The Evolutionary And Ecological Diversification Of Amazonia, Bulletin of the American Museum of Natural History 2000 (244), pp. 1-306 : 194-198

Mesomys occultus , new species

HOLOTYPE: INPA 2690, adult female collected by J. R. Malcolm on June 4, 1992 (original number JUR 501); body, naturally missing tail at the base, preserved in formalin and maintained in 70% ethanol with skull removed; skull and mandibles in good condition; liver sample maintained in 95% ethanol; and chromosome cell suspension maintained frozen.

TYPE LOCALITY: Colocação Vira­Volta, left bank Rio Juruá on Igarapé Arabidi, affluent of Paraná Breu, Amazonas, Brazil (3°17̍S 66°14̍W); locality 14 in the map, figure 1.

DIAGNOSIS: A spinose and moderate­sized arboreal rat. Middle band of aristiform spines of the neck and shoulder region orange. Tail especially hirsute but with narrow and largely colorless hairs associated with individual scales, and terminating in a distinct crest and elongated tuft of hairs extending 30 mm or more beyond the tip (fig. 126). Skull (fig 127) with a short, broad, and distinctly diamond­shaped incisive foramen with a short premaxillary septum, one third, or less, the length of the opening; weak grooves on anterior palate extend posteriorly from border of incisive foramen; median palatal ridge only weakly developed or absent; orthodont upper incisors; and narrow, attenuated, and angled paroccipital processes. Four flexi on PM4, M1, and M2, with the fourth small and lost with wear in older individuals; usually only three flexi on M3 or, if four, medial flexi coalesced (fig. 128). Karyotype with diploid number of 42, fundamental number of 54 (fig. 129).

REFERRED SPECIMENS: Known from four specimens in addition to the holotype, as follows: three topotypes, JUR 502 (adult fe­ male, fluid preserved body with skull removed, chromosome cell suspension, and tissues in 95% ethanol) ; JUR 567 (subadult male, fluid preserved body with skull removed, chromosome cell suspension, and tissues in 95% ethanol); and MNFS 1701 (subadult male, fluid preserved body with skull removed, chromosome cell suspension, and tissues in 95% ethanol); and MNFS 201 from the upper Rio Urucu , Pref. Tefe´, Amazonas, Brazil (adult female, fluid preserved body with skull removed and tissues in 95% ethanol) .

MEASUREMENTS OF HOLOTYPE: See table 56.

ADDITIONAL MEASUREMENTS: See table 56.

DESCRIPTION AND COMPARISONS: A small rat, equivalent in size to M. hispidus , to which it is similar in color, color pattern, and degree of aristiform spine development. Externally, these two are so similar that they were not recognized as distinct until the genetic data were examined. Subsequent comparisons of the series available to us, however, revealed the subtle, but consistent differences in external, cranial, and tooth characters enumerated above in the diagnosis. In contrast to M. occultus , M. hispidus lacks an orange midband in the aristiform spines of the neck and shoulder region. It has black, petiolate tail hairs as opposed to the color­ less, thin hairs of M. occultus . Mesomys hispidus has a longer and narrower incisive foramen with either slightly bowed or parallel sides and a longer premaxillary septum (fig 130), along with two deep groves on the anterior palate that extend to the level of M1 and flank a well­developed median ridge The paroccipital processes of M. occultus are flatter and broader, oriented more in the dorso­ventral plane. All cheekteeth have four flexi, with all flexi independent and the last small but always discernible into advanced age (fig. 128). Finally, while the length of the terminal tail tuft is highly variable in M. hispidus , it is significantly shorter than that of M. occultus . Only one of the five known specimens of M. occultus has a complete tail but the length of its tail tuft nonoverlaps with and is significantly longer than that of M. hispidus whether it is compared to the entire series of the latter or only those from the same locality (33.31 mm versus 5.04 to 21.28, critical difference = 8.521 and p <0.001 by Fisher’s PLSD, or 6.54 to 17.50, critical difference = 11.787 and p = 0.0046 respectively). The tail of M. occultus is also more hirsute and has a terminal dorsal crest in addition to the pencil, in contrast to that of M. hispidus (fig. 127). The two species also differ substantially in karyotype, as that

of M. hispidus is 2n = 60, FN = 116 (see below and fig. 129A and B).

We compared the two species morphometrically by a discriminant analysis, using log 10 ­transformed cranial variables and treat­ ed all specimens of M. hispidus as a single group. They exhibit completely nonoverlapping distributions along the single discriminant axis recovered (fig. 131). Each specimen was correctly classified to its respective taxon, with posterior probabilities of 0.96 or higher. Although no single variable differs significantly between the two species (twotailed t ­tests; p always> 0.07), overall cranial length measures contrast with the more restricted rostral, palate, and mesopterygoid fossa lengths in influencing most of the multivariate separation (table 57).

COMMENTS: Through the kindness of Louise H. Emmons, we compared the crania of our specimens of M. occultus to photographs she had taken of the holotypes of ferrugineus Günther, spicatus Thomas, leniceps Thomas and stimulax Thomas , all in the British Museum (Natural History). Although both ferrugineus and spicatus are also characterized by ‘‘tassels about an inch long on the end of their tails’’ (Thomas, 1924: 535), neither possesses the cranial characters, especially the distinctive incisive foramina, of M. occultus . Both ferrugineus and spicatus are usually listed as synonyms of M. hispidus

(e.g., Woods, 1993). It is possible that the new species described here is the same as ecaudatus Wagner, with type locality at Borba on the lower Rio Madeira in central Ama­ zonia, and another taxon considered a synonym of M. hispidus . Unfortunately, as the name implies, the type and only referred specimen of ecaudatus lacks a tail, and the skull, if extant, has not been examined or figured.

DISTRIBUTION AND HABITAT: This species is known only from two localities in central Amazonian Brazil, both south of the Rio Solimões in Estado do Amazonas: the type locality at Colocacao Vira­Volta (locality 14), left bank of Rio Jurua, and the upper Rio Urucu, Prefeitura Tefe (see map, fig. 124) We obtained all specimens in traps placed in trees in terra firme forest, one at 1.5 m above the ground, the others in canopy platforms positioned between 9.6 and 15.4 m.

ETYMOLOGY: From the Latin word meaning ‘‘unknown’’ or ‘‘hidden,’’in reference to being hidden within M. hispidus , its sympatric and morphologically similar congener.

KARYOTYPE (fig. 129A): Chromosome data are available from four individuals, including the holotype, all from Vira­Volta (locality 14) at the mouth of the Rio Jurua´. 2N = 42, FN = 54. The autosomal complement consists of 7 pairs of large and medium­large biarmed elements, the smallest of which has distinct, near­terminal secondary constric­ tions, and 13 pairs of acrocentric chromosomes varying in size from medium to small. The X­chromosome is a large submetacentric and the Y is a small subtelocentric. This karyotype contrasts sharply with the 2n = 60, FN = 116 complement of M. hispidus and M. stimulax (fig. 129B, described below). The differences between the two karyotypes are substantial, and must have involved a large number of complex rearrangements.

REPRODUCTION AND LIFE HISTORY: Of the two adult females collected in early June, one was pregnant with a single embryo, and the second had an open vagina and an enlarged uterus. Both males were nonreproductive.