Morphological diagnosis and geographic distribution of Atlantic Forest red-rumped mice of the genus Juliomys ( Rodentia : Sigmodontinae )

Recognition and identification of red-rumped mice of the genus Juliomys Gonzalez, 2000 has been a problem among many mammalogists, and specimens of this genus are commonly confused with other Atlantic Forest sigmodontine rodents. Herein we provide an expanded diagnosis for the genus based on the analyses of the three living species of Juliomys, and provide morphological comparisons to the small bodied and bright colored rodents Rhagomys rufescens (Thomas, 1886) and Oligoryzomys flavescens (Waterhouse, 1837), which occur in sympatry with Juliomys spp. in forested areas of southeastern Brazil. These taxa are superficially similar, and are therefore commonly misidentified in the field and museum collections. We also provide morphometric data and a key to the living species of Juliomys, and an updated distribution map of the genus and its species.

Emended diagnosis.Small-bodied sigmodontine genus (TotL: 165-238 mm in adults); dorsal pelage soft, brownish ochraceous to orange-brown, with shades of orange especially on the rump (Fig. 1); ventral fur grayish basally and white to whitish-brown distally (Fig. 1); four pairs of mammae: two inguinal, one post-axial, and one pectoral; tail equal to or slightly longer than head and body, commonly bicolored, with small scales, and a small tuft at the tip; fore and hindfeet short and broad, dorsally covered with orange to brownish hair; claws dorsally covered by silvery white hairs reaching or extending slightly beyond the claw tips; fore and hindfeet show large, bulbous plantar pads: two carpal and three interdigital on the former and two tarsal and four interdigital on the latter (Fig. 4); Skull small and delicate (ONL: 21.95-27.56 in adults), with short rostrum (Fig. 7); interorbital region hour-glass shaped with rounded or gently angled edges; zygomatic arches slightly compressed anteriorly; zygomatic plates nearly vertical, slightly inclined forward; zygomatic notch shallow and rounded; mandible small and delicate; upper incisors opisthodont; molars large (MRC: 3.33-4.24mm in adults), pentalophodont; well developed and separated procingulum cusps on M 1 ; paracone and posterior cusps of M 3 reduced.
Comparisons.Juliomys spp.specimens are superficially similar to R. rufescens and O. flavescens in size and fur color.However, differences among them are evident in both external and skull characters (Tab.I).Members of Juliomys can be distinguished from R. rufescens mainly by: a darker fur, with rump and hind legs distinctly orange, contrasting with the rest of the body; this contrast is not observed in Rhagomys Thomas, 1917, which shows a more homogeneous bright orange fur (Figs 1-3); tips of ventral hairs whitish to light-brown instead of orange; four pairs of mammae instead of three; slightly longer hindfeet, with smaller and narrower interdigital pads ; hindfeet bearing a projecting claw on the first digit, instead of a round claw, resembling a nail; skull smaller, slender and more delicate (Figs 7-9); rostrum, interorbital region and zygomatic plate narrower; parapterygoid plate wider; anterior margin of mesopterygoid fossa reaching M 3 , while in Rhagomys, this structure does not reach M 3 ; dentary smaller (Figs 10-12), and molar cusps less conspicuous.
Juliomys species differ from O. flavescens mainly by: a brighter orange fur, with rump and hind legs distinctly orange, contrasting with the rest of the body, but this contrast is not observed in Oligoryzomys Bangs, 1900, which shows a more homogeneous grayish/brownish fur (Figs 1-3); shorter tail, with smaller scales; mammae pairs in different positions: two inguinal, one post-axial, and one pectoral in Juliomys, instead of one inguinal, one abdominal, one post-axial, and one pectoral in O. flavescens; hindfeet shorter and wider, with interdigital pad larger and pads typically round (Figs 4-6); skull wider (Figs    Geographic distribution.Juliomys spp.occur in the Atlantic Forest from southeastern Brazil to northeastern Argentina and eastern Paraguay (Fig. 13).Brazilian localities of the three living species were provided in COSTA et al. (2007).Subsequently, new records of J. pictipes were reported from Argentina (PARDIÑAS et al. 2008, Fig. 13, localities 40, 43, 45), Paraguay (DE LA SANCHA et al. 2009, Fig. 13, locality 39), and the states of Espírito Santo (TONINI et al. 2010, Fig. 13, locality 3) and Rio Grande do Sul (MELO et al. 2011, Fig. 13, locality 47), in Brazil. CHEREM et al. (2004) identified some specimens from the state of Santa Catarina, southern Brazil as J. pictipes (Fig. 13, locality 46) and others as Juliomys sp. (Fig. 13,localities 47,42,41).They stated that the latter probably does not belong to J. pictipes, but did not provide any further explanations about their taxonomic assessment.CHEREM (2005) also reported Juliomys sp. from Siderópolis, Santa Catarina, southern Brazil (Fig. 13  Remarks.There is an apparent distribution gap in the south between most samples from southern Brazil and northeastern Argentina/eastern Paraguay, including one from northeastern Rio Grande do Sul, Brazil (Fig. 13).This is either due to a true historical biogeographic gap or a collecting artifact (scarcity of adequate inventories, combined with low abundance of Juliomys).The habitat in this region is dominated by Araucaria angustifolia (Bertol.)Kuntze pines (Araucariaceae), forming the Araucaria forest, and some small mammal studies failed to detect Juliomys spp. in this habitat (e.g.WALLAUER et al. 2000, CADEMARTORI et al. 2004, DALMAGRO & VIEIRA 2005).On the other hand, J. rimoforns was collected along a trapline set on vegetation varying from a forested patch with A. angustifolia to    Field inventories in this gap, combined with ecological studies on habitat preferences of Juliomys spp., will provide adequate data to answer this biogeographic question.These recent records of Juliomys, associated to the difficulties in identifying them and the existence of a potentially undescribed species call attention to our ignorance concerning the distribution and diversity of Atlantic Forest mammals in general, and small rodents in particular.Only long-term, intensive field, museum, and lab work will provide us with adequate knowledge regarding this important biodiversity hotspot.
Key to the three living species of Juliomys
; zygomatic notch shallower; interorbital region wider; anterior margin of mesopterygoid fossa reaching M 3 , while in O. flavescens this structure does not reach M 3 ; condyloid process of the dentary well developed in Juliomys spp.(Figs 10-12); molar series larger, with developed anteromedian and anterior flexi on M 1 .The basic descriptive statistics of external and cranial measurements of species of Juliomys is given in Table II, and a summary of diagnostic characters of species of Juliomys
(PARESQUE et al. 2009)NCHA  et al. (2009) included CHEREM's (2005)record as J. pictipes, but since apparently they have not examined any specimen, we kept this record under Juliomys sp.PARESQUE et al. (2009)described a new karyotype from Juliomys specimens collected at Aparados da Serra National Park, state of Rio Grande do Sul, southern Brazil (Fig.13, locality 49).Since the karyotype represents a powerful tool to diagnose Juliomys species, these authors suggested that this new karyomorph represents an undescribed species of the genus(PARESQUE et al. 2009).LIMA et al. (2010)reported the

Table II .
Descriptive statistics for measurements (mm) and weight (g) of adult (dental age classes 2-4) specimens of Juliomys, grouped by species.(N) Sample size, (S.D.) Standard Deviation, (Min) minimum, (Max) maximum.For other abbreviations see material and methods.