ARTÍCULO

A new genus for Habrothrix angustidens and Akodon serrensis (Rodentia, Cricetidae): again paleontology meets neontology in the legacy of lund

 

Ulyses F. J. Pardiñas¹, Lena Geise², Karen Ventura³, and Gisele Lessa⁴

¹ Instituto de Diversidad y Evolución Austral (IDEAus, CONICET), CC 128, 9120 Puerto Madryn, Chubut, Argentina. [Correspondence: <ulyses@cenpat-conicet.gob.ar>].
² Laboratório de Mastozoologia, Departamento de Zoologia, Instituto de Biologia, Universidade do Estado do Rio de Janeiro, Rua São Francisco Xavier, 524, Maracanã, 20550-900 Rio de Janeiro, Brazil.
³ Departamento de Genética e Biologia Evolutiva, Instituto de Biociências, Universidade de São Paulo, Rua do Matão 277, 05508-090 São Paulo-SP and Instituto de Recursos Naturais, Universidade Federal de Itajubá, Avenida BPS 1303, 37500-903 Itajubá, Minas Gerais, Brazil.
⁴ Museu de Zoologia, Departamento de Biologia Animal, Universidade Federal de Viçosa, 36571-000 Viçosa, Minas Gerais, Brazil.

Recibido 30 setiembre 2015.
Aceptado 13 abril 2016.
Editor asociado: B Patterson

---

ABSTRACT.

We describe a new genus of sigmodontine rodent to allocate a fossil species from Lagoa Santa cave deposits, Habrothrix angustidens Winge, and a living species from southeastern Atlantic Forest described at the start of the past century, Akodon serrensis Thos. Tentatively, both forms are considered as synonyms pending a detailed revision of the genus. The new genus belongs to the Akodon Division of the diverse tribe Akodontini where, based on molecular evidence, it is sister to Thaptomys, another forest-dwelling akodontine. The new genus is characterized by a unique combination of morphological traits including tail shorter than head-and-body; manual and pedal digits with short claws; skull robust with pointed rostrum, broad and some­what flat interorbital region with frontal borders divergent posteriorly, and enlarged braincase without crests; large incisive foramina with expanded palatal process of premaxillary; broad mesopterygoid fossa with anterior margin rounded; alisphenoid strut present; carotid arterial circulation pattern primitive; molars noticeably large in relation to the skull and moderately hypsodont; main molar cusps arranged in opposite pairs; first upper molar without anteromedian flexus; length of third lower molar subequal to those of second lower molar; mandible without distinct capsular process; gall bladder present; stomach unilocular and hemiglandular; 2n = 46 (FN = 46). The new genus is an Atlantic Forest endemic, indicating that the diversity of Akodontini has been overlooked outside of the Andes.

RESUMEN.

Un nuevo género para Habrothrix angustidens y Akodon serrensis (Rodentia, Cricetidae): de nuevo palentología y neontología se encuentran en el legado de Lund.

Describimos un nuevo género de roedor sigmodontino para ubicar una especie fósil de los depósitos de Lagoa Santa, Habrothrix angustidens Winge y una especie viviente del sudeste de la Selva Atlántica descripta en los comienzos de la centuria pasada, Akodon serrensis Thos. Tentativamente, ambas formas son consideradas sinónimos a la espera de una revisión detallada de la diversidad del género. El nuevo género pertenece a la División Akodonde la diversa tribu Akodontini donde, sobre la base de evidencia molecular, resulta hermano de Thaptomys, otro akodontino especialista de selva. El nuevo género se caracteriza por una combinación única de rasgos morfológicos incluyendo una cola más corta que el largo cabeza-cuerpo, garras de los dedos de la mano y el pie cortas, cráneo robusto con rostro puntiagudo, región interorbitaria ancha y algo chata con los bordes frontales divergentes hacia atrás y caja craneana agrandada y sin crestas, forámenes incisivos expandidos con proceso palatal del premaxilar bien desarrollado, fosa mesopterigoidea ancha con el borde anterior redondeado, barra alisfenoidea presente, patrón de circulación carotídea primitivo, molares notablemente grandes en relación a las proporciones del cráneo y moderadamente hipsodontes, principales cúspides de los molares dispuestas en pares opuestos, primer molar superior carente de flexo anteromediano, largo del tercer molar inferior subigual al del segundo molar inferior, mandíbula sin proyección capsular evidente, vesícula biliar presente, estómago unilocular-hemiglandular, 2n = 46 (FN = 46). El nuevo género es endémico de la Selva Atlántica indicando una insospechada diversidad de los Akodontini en regiones extraandinas.

Key words: Akodontini; Atlantic Forest; Deltamys; Lagoa Santa; Muroidea.

Palabras clave: Akodontini; Deltamys; Lagoa Santa; Muroidea; Selva Atlántica.

---

INTRODUCTION

During the last 2 decades, several species of sigmodontine rodents named by Winge (1887) based on material from the rich paleontological deposits of Lagoa Santa area (Minas Gerais, Brazil) were recognized as senior synonyms of living entities (e.g., Voss, 1993; Voss and Carleton, 1993; Musser et al., 1998; Pardiñas and Teta, 2011, 2013). These findings, partly based on earlier observations (e.g., Ávila-Pires, 1960; Massoia and Fornes, 1965; Hershkovitz, 1966) helped identify redundancies in sigmo­dontine classification. In addition, these studies provided new data for understanding the tempo and mode of evolution of this rodent group and its past diversity (Voss, 1988; Voss and Myers, 1991). In fact, of 25 species recorded by Winge (1887) in Quaternary deposits (summarized in Voss and Myers, 1991:table 4), only Habrothrix angustidens, Habrothrix clivigenis, Oxymycterus cosmodus, Oxymycterus talpinus, and Calomys anoblepas, still remain as presumptively extinct forms (Pardiñas and Teta, 2013).

During the course of our studies of the col­lections of Lund housed in the Danish city of Copenhagen, once again paleontology meets neontology. Habrothrix angustidens Winge, 1887, a putative fossil species recorded from the area of Lagoa Santa, Minas Gerais, Brazil, appears morphologically undistinguishable from Akodon serrensis Thomas, 1902, a living form from the Atlantic Forest of Argentina and Brazil. The present paper documents this synonymy-first found and generously commu­nicated to us by Robert Voss-and proposes a new genus for both forms. The need for a new genus was uncovered by the phylogenetic analy­ses of D'Elía (2003) and D'Elía et al. (2003), who employed mitochondrial and nuclear DNA sequences and broad taxonomic coverage and recovered A. serrensis as sister to Thaptomys and not as part of the genus Akodon. That hypoth­esis is here morphologically and karyologically supported and expanded through descriptions and comparisons against related genera that conform the Akodon Division (sensu D'Elía, 2003) of the tribe Akodontini (see also D'Elía and Pardiñas, 2015).

MATERIALS AND METHODS

In this report we studied fossil and recent specimens (see Appendix 1) housed at the following institu­tions: American Museum of Natural History, New York, USA (AMNH); The Natural History Museum, London, UK (BMNH); Centro Nacional Patagónico, Puerto Madryn, Argentina (CNP); Field Museum of Natural History, Chicago, USA (FMNH); Colección Nacional de Mastozoología, Museo Argentino de Ciencias Naturales "Bernardino Rivadavia," Buenos Aires, Argentina (MACN); Museu Nacional, Rio de Janeiro, Brazil (MN); Museu de Zoologia, Departamento de Biologia Animal, Universidade Federal de Viçosa, Viçosa, Minas Gerais, Brazil (MZUFV); Universitets Zoologisk Museum, Copenhagen, Denmark (ZMK); and voucher specimens CIT and BIO that will be deposited at Museu de Zoologia, Universidade de São Paulo, São Paulo, Brazil. Fossil specimens are represented by craniodental fragments; recent specimens are preserved as dry skins, skulls, skeletons and individuals stored in fluid. Most of the studied material came from several localities in southeastern Brazil, while a few are from Misiones Province in northeastern Argentina (see an annotated gazetteer of recording localities in Supplementary Material S1).

Anatomical descriptions and terminology follow Reig (1977), Carleton (1980), Voss (1988), and Carleton and Musser (1989). Standard external mea­surements (length of head and body; length of tail; length of hind foot with and without claw; length of ear from notch; all in millimeters) and weight (in grams) were transcribed from the specimen labels or museum catalogs as originally recorded by collectors. Craniodental dimensions were taken using a digital caliper and recorded to the nearest 0.01 mm and follow definitions provided by Wek­sler and Geise (1996) and Voss et al. (2002). They were, arranged in alphabetical order: breadth across M1s, breadth across occipital condyles, breadth of braincase, breadth of incisive foramina, breadth of M1, breadth of rostrum, breadth of zygomatic plate, condylo-incisive length, depth of incisor, height of skull, height of mandible, least interorbital breadth, length of bulla, length of diastema, length of inci­sive foramina, length of mandible, length of orbital fossa, length of palatal bridge, length of rostrum, length of maxillary molar row, and zygomatic breadth. Measured specimens were sorted into age categories following toothwear classes defined for Zygodontomys by Voss (1991). Because our aim is to provide a general characterization of a new genus, we adopt a univariate statistics approach based on mean values for large samples of adult individuals (toothwear classes 3-5).

Karyotypes from selected specimens (see Appendix 1) were prepared in the field. Chromosomes at metaphases were obtained via in vitro culture (culture of bone marrow grown in Dulbecco's MEM with 10% fetal bovine serum and colchicine 10-6 M), following incubation in KCl 0.075M solution at 37 ºC by 40 minutes, centrifuged, fixed in Carnoy solution (methanol: acetic acid, 3:1), dropped onto clean slides and air-dried (Geise, 2014). Conventional coloration with Giemsa 5% was used to observe diploid (2n) and fundamental (FN, excluding sexual chromosomes) numbers and chromosome morphology variation; 50 metaphases of each specimen were analyzed with a Nikon Eclipse 50i photomicroscope under 1000 x magnification. Karyotypes were compared with literature reports. For intergeneric chromosomal comparisons, Deltamys sp. (CIT 944, male, Esmeralda, Rio Grande do Sul, Brazil), Necromys lasiurus (CIT 952, male, Rio Formoso, Pernambuco, Brazil), Thalpomys lasiotis (BIO 288, male, Brasília, Distrito Federal, Brazil), and Thaptomys nigrita (CIT 331, male, Iguape, São Paulo, Brazil), metaphases were also obtained from fibroblast cell cultures derived from ear biopsies grown in Dulbecco's modified Eagle's medium supplemented with 20% fetal calf serum. For all species, GTG-banding was based on routine cytogenetic procedures, and FISH with whole chromosome probes was conducted following Ventura et al. (2009) using the Akodon paranaensis chromosome set as paints. FISH with telomeric probes was also performed only on T. lasiotis using Telomere PNA FISH Kit/FITC (K 5325-Dako) fol­lowing the manufacturers' protocol for the purpose of cytogenetic comparisons. After FISH, the slides were counterstained with DAPI diluted in Vecta­shield (Vector) and analyzed using a fluorescence microscope (Zeiss Axiophot) equipped with software for image capture (MetaSystems).

RESULTS AND DISCUSSION

Taxonomic history

Habrothrix angustidens was described by Winge (1887:28-29) based on fossil remains recovered by the Danish naturalist Peter Lund in Lagoa Santa cave deposits in the first half of the XIX century. Winge (1887) also recognized or named 5 other forms in Habrothrix. One of these, H. clivigenis was referred to Akodon by Trouessart (1898) and remains a putative fossil form. The other 4, H. cursor, H. orycter, H. lasiotis, and H. lasiurus are now allocated, respectively, to the recent genera Akodon, Thaptomys, Thalpomys, and Necromys. Inter­estingly, all these genera belong to the tribe Akodontini. This highlights the accurate systematic perception of Winge, despite working mostly with fragmentary fossil material repre­senting a large number of taxonomic groups.

Winge (1887:28) emphasized the similarities between H. angustidens with H. cursor but also was impressed by the absence of a capsular projection of the incisors and the smaller size of the former. He specifically highlighted the peculiar condition of the dentition of H. angustidens, and most remarkably, mentioned the difficulty of distinguishing the enamel from the dentine on the occlusal surfaces. Winge (1887:pl. II, figure 9) illustrated 2 remains belonging to H. angustidens, both from Lapa da Serra das Abelhas; an anterior skull fragment and a right mandible. Shortly after, Trouessart (1898:536) listed H. angustidens under Akodon; this action must be viewed as a taxonomic act without major implications because this author trans­ferred to this genus all the Habrothrix species of Winge plus a wide array of sigmodontine forms.

In 1902 Oldfield Thomas received from Brazil a large mammal collection made by Alphonse Robert in the Serra do Mar of the Brazilian state of Paraná. Thomas (1902:61) recognized one sigmodontine as new and described it as Akodon serrensis with type locality at "Roça Nova... at an altitude of about 1000 meters" (Thomas, 1902:59). Thomas also received in the same collection specimens of A. cursor, and he directly compared both species. Akodon serrensis was characterized basically on color­ation traits and said to be intermediate in size between A. azarae (= "arenicola" in the original work) and A. cursor. Thomas (1902:61) also noted "molars very large in proportion, as large as those of the much bigger species A. cursor;" however, he did not make any mention about their morphology.

Miranda Ribeiro (1905:188) added a new name to the nominal forms allied to A. serrensis. Based on a single female from Retiro de Ramos (also known as Retiro do Ramos), Rio de Janeiro, Brazil, he erected [Hesperomys] (A.) serrensis var. leucogula, a subspecies char­acterized by a more olive dorsal coloration, a yellowish-ochraceous belly, and a narrow pure white band in the lower part between the chin and the anterior limbs.

The XXth century passed without major systematic or nomenclatural modifications. While H. angustidens was basically forgot­ten, A. serrensis-including H. (A.) serrensis leucogula-was maintained under Akodon from the time of the catalog of Gyldenstolpe (1932) up to the more recent treatments (e.g., Musser and Carleton, 2005; Patton et al., 2015). Hershkovitz (1990a:14) ranked leucogula as a valid and different species from serrensis, in­cluding both forms in an informal group that he named Akodon mollis size class. Other research­ers addressed morphological and chromosomal considerations of A. serrensis populations, especially in southern Brazil. In part, these studies were triggered by a misunderstanding involving several species of the A. cursor group (sensu Smith and Patton, 2007) and A. serren­sis. This contorted history, which started when Liascovich and Reig (1989) erroneously referred one Argentinean Akodon specimen to A. serren­sis and reported for it a diploid complement of 44, was fully resolved by Christoff et al. (2000; see also Pardiñas et al., 2003; Gonçalves et al., 2007). It is of interest to transcribe a paragraph of Liascovich and Reig (1989:394) in order to reflect the understanding of A. serrensis at that time: "Akodon serrensis is similar to A. cursor in size and color pattern, and the two species are difficult to distinguish based on external characters. Fresh skins of A. serrensis have a conspicuously more ochraceous venter, especially evident in the anal region; however, even this distinction must be confirmed by examination of a large series of specimens. The skulls of both species also are similar in having long, narrow rostra, moderately in­flated frontal regions, short zygomatic plates, a reduced interparietal, and elongated palatal foramina that reaches the level of the protocone of M1. The main differences between the two species are: the nasals are longer and project farther behind the maxillo-frontal suture in A. cursor; the zygomatic plate is much deeper in A. serrensis; the palate is shorter in A. cursor, its posterior border terminating anterior to the level of the alveoli of M3; the molars are narrow, elongated, and larger in A. serrensis; and the mandible in A. cursor is slender and elongated with a more posteriorly projecting condylar process and a more inflated capsular projection (root of incisor)." It is difficult to know how influential it was that one of the specimens used by Liascovich and Reig (1989) was, in reality, an Akodon paranaensis (see Pardiñas et al., 2003; Gonçalves et al., 2007). Despite having the holotype of A. serrensis in hand, these authorities failed to distinguish these species. This necessitates that we call attention to the similarity of A. serrensis and other species of the cursor group of Akodon (e.g., cursor, montensis, paranaensis).

However, a new scenario emerged from mo­lecular studies. D'Elía et al. (2003:358) explicitly called attention to the generic position of A. serrensis as follows: "A novel finding of our study with respect to akodontine systematics is the placement of Akodon serrensis well out­side the Akodon clade... A. serrensis appears as sister to the monotypic genus Thaptomys, and this clade is sister to the genus Necromys... The phylogenetic position of Akodon serrensis is surprising because to the best of our knowl­edge nobody has cast doubt upon the inclu­sion of A. serrensis within the genus Akodon... We will not make any taxonomic judgment about this topology until further studies... are carried out. However, a combined analysis of nuclear and mitochondrial DNA sequences points in the same direction." In fact, D'Elía (2003:319) revisited this topic in a combined cytb and IRBP analysis "Ten extra steps are needed to force A. serrensis within the Akodon clade. Akodon serrensis appears to be sister to the monotypic genus Thaptomys (node 18), although this relationship is weakly supported (JK<50%, BS = 1). This clade is sister to the clade Thalpomys-Necromys. Further studies, including morphological comparisons and the sequencing of additional specimens, are needed to determine if A. serrensis should be placed on its own genus or in Thaptomys." Despite the relevance, these findings were basically ignored in the chromosomal research developed with Akodon species in southeastern Brazil (e.g., Ventura et al., 2006; Hass et al., 2008; Bonvicino, 2011). More recently, analyzing cytb over a taxonomically diverse sample of Akodon, Smith and Patton (2007:848) corroborated the separate position of serrensis although they highlighted its alternative relationship to Akodon + Deltamys rather than Thaptomys.

The taxonomic hypothesis that H. angustidens and A. serrensis were congeners was achieved by R. Voss after a restudy of the type mate­rial of both forms. According to him "The type series of Habrothrix angustidens consists of subfossil cranial fragments and disassoci­ated hemimandibles from several caves in the vicinity of Lagoa Santa… All of this material qualifies as syntypes because Winge did not designate a holotype… Altogether, this mate­rial… preserves the entire upper and lower dentition, the left zygomatic plate, diastema, palate, interorbital region, and mandibles. We compared this material side-by-side with a complete topotypical cranium of Akodon serrensis (ZMUC 1257) and found these two taxa to be phenotypically indistinguishable. In addition to qualitative resemblances, the few measurements that we were able to take from the lectotype of angustidens all fall within the range of variation seen in the topotypic series of serrensis… The only possible taxonomic conclusion is that angustidens and serrensis are synonyms" (R. Voss, in litt. November 4, 2008; Supplementary Material S2).

SYSTEMATICS

Family Cricetidae Fischer, 1817 Subfamily Sigmodontinae Wagner, 1843 Tribe Akodontini Vorontsov, 1959 Castoria, gen. nov. Figs. 2-5; Tables 1-3

Fig. 2. Dorsal and plantar view of left hind foot of Castoria gen. nov., based on an adult female from RP 2, 6 km NE Arroyo Paraíso, Misiones, Argentina (CNP 449). Abbreviations are: I-V, digits; 1-4, interdigital tubercles; h, hypothenar; t, thenar.

Fig. 3. Castoria gen. nov., based on an adult male from Fazenda da Neblina, Parque Estadual da Serra do Brigadeiro, Brazil (MZUFV 1635): skull in right lateral (top), dorsal (left) and ventral (right) views, and right dentary in labial view. Scale = 5 mm.

Fig. 4. Selected anatomical features of the skull and dentition of Castoria gen. nov., based on adult individuals MZUFV 1630, 1655, 2555, and CNP 449: A) rostrum in dorsal view; B) maxillary-frontal area in dorsal view; C) left zygomatic arch in lateral view; D) left upper toothrow (note the hypsodonty); E) incisive foramina; F) mesopterygoid and parapterygoid fossae (the dashed line highlights the short condition of the palate); G) posterior portion of the skull in lateral view; H) alisphenoid-squamosal area; I) angular process of the mandible in lingual view. Abbreviations are: a, angular process of the mandible; ab, auditory bulla; as, alisphenoid strut; bmt: buccinatory-masticatory thorough; f, frontal; fs, frontal sinus; if, incisive foramen; j: jugal; zn, zygomatic notch; l, lachrymal; lpp, lateral projection of the parietal; mf, mesopterygoid fossa; n, nasal; nc, nasolachrymal capsules; p, parietal; pgf, postglenoid foramen; pmp, premaxillary process; spv: sphenopalatine vacuities; ssf, subsquamosal fenestra; zpm, zygomatic process of the maxillary; zps, zygomatic process of the squamosal.

Fig. 5. Occlusal view of upper left (above; MZUFV 1234, 1242, 1630, 1076) and lower left (below; MZUFV 1234, 1242, MN 69813, MZUFV 1076) toothrows in Castoria gen. nov., arranged from left to right by increased wear. Scale = 1 mm.

Table 1 Selected measurements (in mm) of the lectotype of Habrothrix angustidens (ZMK 1/1845:13246), the type series of Akodon serrensis (BMNH 3.7.1.66-BMNH 3.7.1.73, ZMK 1257; mean, minimum - maximun and sample size in parenthesis), and the holotype of Akodon serrensis leucogula (MN 2513).

Table 2 Descriptive statistics (in mm; mean ± standard deviation, range, and sample size in parenthesis) for 2 samples of adult individuals of Castoria angustidens nov. comb., from Brazil (see Supplementary Material S1 for additional data about these localities); PE = Parque Estadual; PN = Parque Nacional.

Table 3 Morphological comparisons of selected traits among Castoria gen. nov. and other related sigmodontines of the Akodon Division.

Synonymy.-

Habrothrix sensu Winge (1887:28), part, not Habrothrix Wagner, 1843

Akodon sensu Thomas (1902:61), part, not Akodon Meyen, 1833

Hesperomys sensu Miranda Ribeiro (1905:188), part, not Hesperomys Waterhouse, 1839

Type species.- Habrothrix angustidens Winge, 1887.

Etymology.- In homage to Castor Cartelle, a con­temporany paleontologist (born in Ourense, Spain in 1938) studying Quaternary Brazilian mamma­lian fauna who effectively resurrected in Brazil the genius and figure of Peter W. Lund (see Cartelle, 2002). Continuous efforts of Cartelle over the last 4 decades have enlarged our knowledge of biotic evolution in tropical areas through the excavation, cleaning, sorting, and study of a huge quantity of osteological material recovered from karstic caves in northeastern Brazil.

Geographic distribution.- Known from southeast­ern Brazil and the northeastern tip of Argentina, roughly between 20º to 30º S and mostly restricted to an elevational range above 800 m (Fig. 1; Supplementary Material S1).

Fig. 1. Locality records for Castoria gen. nov.; type localities of available nominal forms are indicated. See Supplementary Material S1 for a geographic gazetteer.

Chronological distribution.- Late Pleistocene? -Recent. Fossils were found in 5 cave deposits - Lapa do Capão Secco, da Cerca Grande, a chamber near Escrivania, Quebra Chavelha, and Serra das Abelhas- in the area of Lagoa Santa, Minas Gerais, Brazil (Winge, 1887). Absolute radiometric dates performed on fossils and breccias excavated by Lund from Lagoa Santa indicate an array of ages covering at least the last 0.5 MA (e.g., Auler et al., 2006).

Contents.- Tentatively, the genus is considered to be composed by one species, Cas­toria angustidens (Winge, 1887), nov. comb. Neither morphology nor metrics (Table 1) allow us to distinguish, at the specific level, the anterior skull portion that represents H. angustidens from those of living popula­tions traditionally ascribed to A. serrensis. In fact, an exploratory multivariate analysis based on principal components placed the lectotype of H. angustidens within the morphospace of recent Brazilian specimens of A. serrensis (Supplementary Material S3). Minor differences between the fossil H. angustidens fragment-particularly the narrow condition of the molar toothrow-and recent A. serrensis individuals are here attributed to age variation. Given the broad geographic range of this genus and modest morphometric segregation (results not shown; see also Kosloski, 1997), the variation of Castoria is a topic that deserves further study. The names serrensis Thomas, 1902 and leucogula Miranda Ribeiro, 1905 are available.

Morphological diagnosis.- Rodents of the subfam­ily Sigmodontinae (sensu Reig, 1980) with the fol­lowing combination of characters: small-medium size (head-and-body length ~ 95 mm; condylo-incisive length ~ 25 mm); tail shorter than head-and-body (~ 85% of head-and-body length); manual and pedal digits with short claws covered by long and sparse ungual tufts; pelage soft and long; skull robust with pointed rostrum, broad and somewhat flat interor­bital region with frontal borders moderately divergent backwards, and enlarged braincase without patent crests; broad, anteriorly rounded, and slanting zygomatic plate; zygomatic arches with their lower border parallel with respect to the alveolar plane; long incisive foramina, anteriorly constricted, with expanded palatal process of premaxillary; short, broad, and flat bony palate; broad mesopterygoid fossa with anterior margin rounded; alisphenoid strut present; tegmen tympani overlapping squamosal with the participation of a distinct posterior suspensory process of the latter bone; carotid arterial circulation pattern primitive; bullae short and rounded; upper incisors opisthodont; molars noticeable large in relation to the cranium proportions, terraced, enlarged, and moderately hypsodont; main molar cusps arranged in opposite pairs; M1 without anteromedian flexus with large procingulum and marked trilophodont pattern; M3 with cylindrical outline; m1 with anteromedian flexid weakly present and procingulum anterior-posteriorly compressed; m1-2 with small ectolophids present and mesolophids probably fussed to metaconids; length of m2 slightly shorter than those of m1; length of m3 subequal than those of m2; lower incisor alveolus without distinct capsu­lar process on lateral dentary surface; gall bladder present; stomach unilocular and hemiglandular; diploid number (2n) = 46 and fundamental number (FN) = 46.

Morphological description.-The following de­scription is based on representative samples from localities in Minas Gerais and Rio de Janeiro states in Brazil and 2 individuals from Argentina. It aims to provide a general characterization of the genus. Additional morphological details and measurements can be found in Hershkovitz (1998:215-218).

Castoria includes small-medium sized sigmodontine rodents (Tables 1 and 2) characterized by a darker general coloration above and below, short mystacial vibrissae, short claws, and a moderately long tail although shorter than the combined length of head and body. The dorsal and ventral colors are subtly delimited with weak countershading (Supplementary Material S4). The skins exhibit thick, soft fur (guard hairs ~ 9 to 12 mm); the cuticle pattern of guard hairs was described as narrow and leaf-shaped; medullar pattern of these same hairs were a mixture of alveolar and striped types, with three to four layers of cells of irregular shape (Silveira et al., 2013). The dorsal fur is dark yellowish- to orangish-brown; the ventral surface is washed with bright buff or tawny over dark-gray hair bases. The eyes are small. The mystacial vibrissae are short, not extending posteriorly much beyond the base of the pinnae; genal vibrissae are present. The ears are medium in size, rounded, and internally and externally covered by brown hairs. The tail is dark brown, unicolored, and sparsely haired, covered by sets of 3 somewhat hispid hairs per scale, each dorsal hair covering 2-3 scale-rows; apical tuft absent. The manus and pes in Castoria are covered with short hairs, with their basal portion brown and apical portion whitish. The claws, basally open on the pes, are inconspicuous and dorsally covered by several whitish or darker ungual hairs slightly surpassing their apices. The plantar surface is naked, pigmented, and slightly squamated (sensu Pacheco, 2003). The pes has 6 pads, 4 interdigital and 2 plantar, the thenar and hypothenar; the thenar pad is enlarged and narrow; the hypothenar pad is subequal in size to the 1^st interdigital pad; dIII and dIV are the largest, subequal in size and showing a slightly syndactylous condition (Fig. 2).

The skull in dorsal view is characterized by a flat, broad appearance with a pointed rostral region and a rounded braincase (Fig. 3). The nasals are pointed and usually slightly projecting beyond the anterior margin of the premaxillary (Fig. 4A). The rostrum is flanked by shallow but distinct zygomatic notches (Fig. 4B); the frontal sinuses are inflated and the interorbital region is broad, with rounded supraor­bital margins in young and adult specimens (older animals develop small, sharp, convergent postorbital edges). The zygomatic arches are delicate, widest across the squamosal roots; they converge anteriorly and their midportions run parallel to the alveolar plane; the jugal is well developed (Fig. 4C). The braincase is large, smooth, and globose; the lambdoid crests are developed only in older individuals. The interparietal varies in size and form, being typically rhomboidal and compressed. The zygomatic plate in lateral view has a straight anterior margin that aligns vertically or slopes slightly backward; the origin of the superficial masseter is marked by a scar (not a tubercle) that is just below and slightly behind the anteroventral margin of the zygomatic plate. The foramina of the carotid circulation consist of pattern 1 (sensu Voss, 1988); a thick alisphenoid strut is present, and the subsquamosal fenestra is large (Figs. 4G and 4H). The incisive foramina are long (extending between the M1 anterocones) and characterized by a constricted anterior portion (at or near the premaxillary-maxillary sutures) and a markedly inflated premaxillary process (Fig. 4E); the palate is broad and uncomplicated by keels or deep lateral grooves; and the mesopterygoid fossa extends anteriorly to or between the M3s (Fig. 4F). There are no large posterolateral palatal pits, just tiny foramina flanking the mesopterygoid fossa. The mesopterygoid roof is perforated by very narrow but long sphenopalatine vacuities (Fig. 4F). The auditory bullae are small and flask-shaped, with broad bony Eustachian tubes.

The mandible is relatively robust, moderately low and elongated; the lower border of the ramus bends upward and backward behind the level of the m2 and descends again behind the m3. The mental foramen opens on the dorsolateral surface of the diastema. The alveolus of the lower incisor is contained in a ridge, not a distinct process, on the lateral surface of the mandible behind the base of the coronoid process. The angular process is pointed and bent upwards (Fig. 4I).

The upper incisors are opisthodont in orientation, with yellow-orange enamel bands, straight or very slightly inclined in frontal view and sharp, not inclined medially in ventral view (cf. Voss, 1988:376- 377). The exterior edge of the upper incisors shows a thick enamel band; the dentine lake is straight (sensu Steppan, 1995:figure 10). The molars are very narrow in proportion to their length, with a marked unilateral tubercular hypsodonty (sensu Hershkovitz, 1962:89; Fig. 4D). They are also characterized by the opposite pattern of the main cusps, strongly bi-level (subadults) to terraced (adults) coronal surface (sensu Hershkovitz, 1962:86), simplification, and a tendency to retain little occlusal detail even in young animals; the cusps and lophs wear rapidly away, exposing large dentine basins (Fig. 5). The latter feature, shared with other akodontines (e.g., Akodon, Oxymycterus), makes it necessary to study a series of individuals of different ages in order to adequately assess dental variation. The upper molar tooth-rows are parallel or slightly convergent backwards in adult and old-adult specimens. The M1 is typically composed by 3 subequal "lobules" characterized by central basins and abrupt posterior walls inclined backwards. The procingulum lacks any trace of an anteromedian flexus and shows a well-developed surface occlusal area. The paracone "lobule" is the largest because it likely involves the mesoloph plus the paralophule; a trace of the latter is present as an acute posterior edge. The proto- and hypoflexus are subequals and scarcely developed, while para- and metaflexus are moderately directed backwards, with their internal margins reaching the middle line of the molar; the mures flanked by these flexus are particularly enlarged, lingually disposed, and longitudinally oriented. No evidence of a free posteroloph was detected in any studied specimen. The M2 shows a tetralophodont pattern with anteroloph and posteroloph obsolescent in young individuals to bilophodont in adults. The metacone "lobule" is smaller than the paracone one. The M3 is certainly complex in juveniles, with an internal enamel ring of equivocal identity (me­sofossette?) and reduced hypoflexus and posterior lobe. In subadult to adult individuals, M3 acquires a cylindrical outline without internal accidents. We documented a deformation of the occlusal figure of the upper molars by the differential wear of labial and lingual sides as a consequence of age; whereas the labial portion is partially eroded to the coronal base, the lingual remains with persistent striae and thicker enamel. Lower molars are characterized by occlusal simplicity in subadult and adult individu­als, a compressed procingulum mostly fused with the remainder portion of the m1, and a large m3 subequal to or even exceeding the m2 in length.

On the basis of 9 individuals, Castoria has 13 tho­racic ribs, 13 thoracic vertebrae, 6 lumbar vertebrae, and 29-30 caudal vertebrae. The sacrum involves 3 subequal and squared sacral vertebrae. The neural spine of the axis is enlarged anteriorly-posteriorly and is fan-shaped. The neural spine of the 2^nd thoracic vertebra is enlarged, twice or more as long as nearby spines. The caudal vertebrae lack hemal arches but hemal processes are present starting between the 3^rd and 4^th caudal vertebrae, becoming most pronounced between the 5^th and 6^th, and then decreasing in size.

Gross stomach morphology in Castoria conforms with the unilocular-hemiglandular type (sensu Carleton, 1973:figure 2), a widespread condition in sigmodontine rodents. The bordering fold bisects the stomach on a line from the incisura angularis to a point opposite it on the greater curvature (Supplementary Material S5). Animals from Rio de Janeiro, São Paulo, and Misiones indicate that a gall bladder is present (cf. Geise et al., 2004; this paper).

Cytogenetics.- The specimens of Castoria angustidens cytogenetically analyzed (Appendix 1) presented the same karyotype, with 2n = 46 and FN = 46 (Supplementary Material S6); the auto­somes comprising 21 acrocentric pairs and 1 small metacentric pair (the smallest of the complement). The X chromosome is a medium-sized acrocentric, and the Y one of the smallest of the complement, also acrocentric. This karyotype was previously recorded for specimens collected in states of Rio de Janeiro (Geise et al., 1998; Olifiers et al., 2007), Paraná (Barros-Battesti et al., 1998; Hass et al., 2008), Santa Catarina (Testoni et al., 2012), Minas Gerais (Moreira et al., 2009), and Espírito Santo (Bonvicino et al., 2002). Cytogenetic information is lacking for Argentinean populations; the animal referred to A. serrensis by Liascovich and Reig (1989) and characterized as having 2n = 44 (FN = 44) was correctly reidentified as Akodon paranaensis by Christoff et al. (2000).

Phylogeny.- Phylogenetic analyses of combined molecular markers (e.g., D'Elía, 2003; D'Elía et al., 2003; Smith and Patton, 2007) showed that Castoria belongs to the Akodon Division of the tribe Akodon­tini, a well-supported clade composed by the recent genera Akodon, Deltamys, Necromys, Podoxymys, Thalpomys, and Thaptomys (D'Elía, 2003; Leite et al., 2015). Within the Akodon Division, its posi­tion varies depending upon the molecular markers used, taxonomic sampling, and the type of analysis conducted. In fact, in those studies with moderate taxonomic coverage (e.g., excluding Vilela et al., 2014), Castoria was recovered as sister to a clade including Akodon + Deltamys (Smith and Patton, 2007; Ventura et al., 2013; Abreu et al., 2014; Pardiñas et al., 2014b), but elsewhere, it appears as sister to Thaptomys (e.g. D'Elía, 2003; D'Elía et al., 2003), or as sister to Deltamys (Fabre et al., 2012; Leite et al., 2015). Despite its uncertain position, no study has recovered Castoria within or as sister to the genus Akodon. This finding supports the hypothesis that this lineage deserves generic recognition.

Comparisons.- The fact that A. serrensis remained allocated as a species of Akodon for more than a century is a clear signal of the strong morphologi­cal similarity between these 2 genera. In addition, some character traits that were used as evidence of the distinctiveness of some genera of the Akodon Division, such as the morphology of the male accessory glands (see Reig, 1987), were-and re­main-unknown for A. serrensis. Castoria is smaller than Akodon cursor or Akodon montensis and can be externally distinguished by its shorter manual claws and overall darker coloration. The relative size of molar toothrows, noticeably larger in Castoria than in Akodon (as Gyldenstolpe, 1932:104 high­lighted), the general morphology of the interorbital region-flat and divergent posteriorly in Castoria and smooth and hourglass-shaped in Akodon, and the marked differences in the development of the premaxillary process as seen through the openings of the incisive foramina-more inflated and extended in Castoria than in Akodon, are the main cranial features separating these genera (Fig. 6). However, there are more trenchant differences be­tween Castoria and Akodon in molar morphology. Castoria displays a conspicuous pattern of opposite main cusps in contrast to Akodon where these ele­ments are moderately alternating; in addition, the former lacks the anteromedian flexus that typically characterizes the M1 procingulum in Akodon species (Table 3; Fig. 7). An additional important trait to distinguish Castoria from species of the cursor group of Akodon, which are sympatric with Castoria, is the morphology of the incisive foramina. In Castoria, the foramina are constricted anteriorly, anterior to the premaxillary-maxillary suture. In Akodon, the margins are not constricted, and the lateral margins of the foramina are convex and wider medially, resembling parentheses and showing no changes in curvature (Fig. 6; see also Supplementary Material S7). However, this degree of morphological differen­tiation is considered as interspecific variation within the genera Handleyomys (cf. Voss et al., 2002:22) and Drymoreomys (Percequillo et al., 2011:371).

Fig. 6. Selected aspects of qualitative anatomy contrasted in the skulls of Akodon, Thaptomys, Deltamys, and Castoria gen. nov., scaled to the same length. The figure portrays contrasts between the char­acteristics of these genera and Castoria as follows: Akodon has rounded in­terorbital margins on fron­tals (fr), smaller toothrow (m), and shorter palate (p); Thaptomys has a shorter rostrum (ro), less devel­oped zygomatic notches (zn), rounded interorbital margins, shorter incisive foramina (if ), and more ossified lacerate foramen (lf); Deltamys has a shorter rostrum, less developed zygomatic notches, rounded interorbital margins, and an almost completely ossified lacerate foramen.

Fig. 7. Upper right (above) and lower left (below) molar toothrows in occlusal view in Akodon, Deltamys, Thaptomys, and Castoria gen. nov.: A, E, Akodon montensis (CNP 4112); B, F, Deltamys kempi (CNP 4141); C, G; Thaptomys nigrita (MZUFV 1112); D, H, Castoria (MZUFV 1234 [upper], MN 69804 [lower]). Not to scale.

Thaptomys appears as the sister taxon for Castoria in some molecule-based phylogenies (e.g., D'Elía, 2003). Both share weak pelage countershaping and inflation of the premaxillary process. However, morphological differences are strong. Castoria has a moderately long tail and rostrum, contrasting with the shorter condition of both in Thaptomys. Castoria has a well-developed upper free margin of the zygomatic plate whereas Thaptomys does not. Castoria lacks the defined capsular projection of the lower incisor root characterizing the mandible of Thaptomys. A gall bladder is present in Castoria and absent in Thaptomys. Molar morphology in Thaptomys is very similar to that in Akodon, so that Castoria differs from Thaptomys in all the respects noted earlier regarding Akodon.

Although Castoria differs in numerous aspects with respect to Deltamys, both share weak countershading, a moderately long tail, and the general "soricine" external aspect. Their skulls are easily distinguished by the pointed rostrum and broad interorbital re­gion of Castoria versus the comparatively shortened rostrum and narrow interorbital region of Deltamys. In addition, Castoria has a broad zygomatic plate and an alisphenoid strut whereas Deltamys has a narrow plate without free upper margin and lacks the alisphenoid strut. In keeping with the molecu­lar phylogenetic hypoth­esis of Smith and Patton (2007), which recovered Castoria close to Delta­mys, these genera show many similarities in mo­lar occlusal morphology (Fig. 7). In particular, both have the main cusps arranged in opposite pairs, the tendency to a trilophodont M1 in adults, and an anteriorly-posteriorly compressed procingulum that coupled with its coalescence to the metaconid produce on the m1 a nearly bilophodont pattern.

The GTG-banding patterns of Castoria chromo­somes allowed the identification and pairing of all homologues (Supplementary Material S6), and therefore, comparisons with the GTG-banded chromosomes of Akodon paranaensis (2n = 44 and FN = 44; reported in Ventura et al., 2009). Chromo­some painting using A. paranaensis (APA) revealed that a tandem rearrangement, in which chromosome 7 of A. paranaensis is homologous to two chromo­some pairs (16 and 21) of Castoria angustidens, explains their differences in 2n and FN. Meanwhile, analyzed species of Akodon share the sequences of APA 7 conserved as one chromosome (see Ventura et al., 2009). Comparisons of karyo­types among C. angustidens and other closely related akodontines (Deltamys, Necromys, Thalpomys and Thaptomys), showed that two groups besides Akodon can also be delimited with respect to APA 7. The first group unites Castoria and Thaptomys, both with homologous sequences of chromosome 7 of A. paranaensis reorganized into two acro­centric chromosomes (data not shown). The second group comprises Deltamys sp., Necromys lasiurus and Thalpomys lasiotis, which present the 2 chromo­some pairs homologous to APA 7 (as Castoria and Thaptomys), but also pres­ents an exclusive association of chro­mosomes APA 7 and APA 8. Castoria presents a distinct pattern and nature of heterochromatin, distributed in large and positive pericentromeric blocks after telomeric FISH with (TTAGGG) n sequences as probes (Ventura et al., 2006, as Akodon serrensis). Thalpomys lasiotis also exhibits a pericentromeric region rich in (TTAGGG)n-like se­quences (for CBG-banding patterns see Yonenaga-Yassuda et al., 1987, as Akodon reinhardti, and Andrade et al., 2004); this homoplasic feature in light of currently understood phylogenetic relationships, suggests that accumulation of the (TTAGGG) n-like repetitive sequences could lead to generic-level differentiation in the Akodontini.

Natural History.- Castoria has been recorded in at least 44 localities, almost all in the Atlantic Forest of Brazil (41) and Argentina (3), encompassing 7 ecoregions (Supplementary Material S1). Little is known about its natural history. The principal vegetation where Castoria is recorded is humid mountain forest, including those with Araucaria augustifolia (Supplementary Material S8). The elevational range for Castoria collection localities is from 300 to 2100 m (correlation between latitude and elevation was r = 0.5; Supplementary Material S9). Usually Castoria is interpreted as an element of montane fauna (Gonçalves et al., 2007; Abreu et al., 2014). Paglia et al. (2012) reported the diet of this rodent as insectivorous/omnivorous, although the basis for this assessment is unknown. Morphometric analysis of the appendicular skeleton suggests that this species uses terrestrial locomotion (Coutinho et al., 2013). Vieira and Monteiro-Filho (2003) captured this mouse exclusively on the ground. Miranda Ribeiro (1905) indicates that Castoria inhabits "galerias sub­terraneas," that according to Voss (1993:35) possibly correspond to "shallow tunnels under matted roots and moss" and piles of litter in thickets of Chusquea. The species was the most frequently trapped small mammal in the Parque Estadual do Desengano, a disturbed montane forest (~1000-1425 m elevation) in the northern portion of Rio de Janeiro (Modesto et al., 2008). Castoria angustidens was also the dominant species among recorded sigmodontines in an elevational transect of Parque Estadual da Serra do Brigadeiro from 1200 to 1800 m, particularly at middle to upper elevations (1450-1800 m; Moreira et al., 2009). In Parque Nacional de Itatiaia, it was trapped in montane forests and "campos de altitude" up to ~2100 m (Geise et al., 2004). Captive animals fed mainly on pulp of fruits and also on small to medium-sized seeds of a wide array of plant species from the Atlantic Forest (Vieira et al., 2003). A rich assemblage of parasites, including Acari, Anoplura, and Siphonaptera has been recorded from Brazilian individuals (Barros-Battesti et al., 1998; Linardi and Guimarães, 2000; Moraes et al., 2003). Teixeira et al. (2014) recorded natural spillover infection between this species and the Jaborá strain of hantavirus, whose primary host is Akodon montensis. According to these authors, spillover infection in secondary hosts, including C. angustidens and A. paranaensis, may play an important role in maintaining the hantavirus sylvatic cycle during periods of low prevalence in primary hosts.

Biogeography.- Castoria is distributed in south­eastern Brazil, from Espírito Santo to Rio Grande do Sul, and in Misiones, northeastern Argentina. Its distributional range overlaps those of other akodontine rodents, such as Blarinomys breviceps, Brucepattersonius spp., and Thaptomys nigrita. Unlike those taxa and other Atlantic Forest sigmodontines, Castoria is much restricted to a narrow coastal fringe at relatively high elevations (500-1600 m) in well conserved forest remnants. Most of the recorded localities are located within Serra do Mar coastal forests and Araucaria Moist forests ecoregions (Olson et al., 2001; Pardiñas et al., 2015). The Serra do Mar is recognized as one of the more remarkable centers of endemism in South America, with evidence from plants (especially of the families Myrtaceae, Laura­ceae, Melastomataceae, and Orchidaceae), birds, butterflies, amphibians, and mammals (Da Silva, 2000a and references therein). Araucaria Moist forests spread from middle-level plateaus of ~500 m to the high slopes (1600 m) of the Serra da Mantiqueira. Among 352 plant species recorded in this region, 47 (13.3%) are endemic; this moist forest is also recognized as an important endemic bird area (Da Silva, 2000b). Castoria has been recorded from the Bahia Coastal and Interior Forests and at ecotones with the Alto Paraná Atlantic forests. At a regional scale, at least other 2 genera, Delomys and Juliomys, have similar distributions, with populations mostly restricted to these same ecoregions (see Pavan and Leite, 2011; Gonçalves and Oliveira, 2014). The genera Delomys and Juliomys comprise both widely distributed species as well as more elevationally restricted taxa. This pattern highlights the impor­tance of montane topographies in the evolution of sigmodontines of the Atlantic Forest (Gonçalves and Oliveira, 2014) and identifies this area as a major center of diversification of this subfamily (Smith and Patton, 1999).

Conservation.- C. angustidens was categorized as "Least Concern" by the IUCN. Several arguments were used to justify this placement, including its wide distribution, presumed large populations, its presence in a number of protected areas (e.g., Parque Estadual do Desengano, Parque Estadual da Serra do Brigadeiro, Parque Nacional de Itatiaia, Parque Nacional da Serra dos Órgãos, Parque Nacional da Bocaina), and because "it is unlikely to be declining at nearly the rate required to qualify for listing in a threatened category" (http://www.iucnredlist.org/details/full/756/0). However, it is likely that some populations of this species have been negatively affected by habitat destruction and forest fragmentation.

Remarks.- Our systematic hypothesis involves fossil (H. angustidens) and recent materials (A. serrensis). This kind of taxonomic work is obviously limited owing to the anatomical incompleteness and absence of genetic information for the fossils. However, as discussed several times revisiting the Lund collec­tions (Voss, 1993; Voss and Carleton, 1993; Pardiñas et al., 2008), fossils of Lagoa Santa often preserve enough information to make confident inferences regarding both genus and species (Supplementary Material S2 and S10).

Winge's description of H. angustidens was explicitly based on multiple specimens, none of which was designated as type (Winge, 1887:28-29). Therefore, all of those specimens must be considered syntypes in accordance with Article 72.1.1 of the Code (ICZN). We designate as lectotype of H. angustidens-ac­cording to Article 74.1-the cranial anterior frag­ment illustrated by Winge (1887:II, figure 2); this fossil was recently numbered in ZMK collections as 1/1845:13246 and is associated with Winge's hand­written label saying "Habrothrix angustidens Tegnet [figured] ?Lapa da Serra das Abelhas" (Supplementary Material S10). In addition, a photograph of this specimen was published in the catalog of Hansen (2012:58-59).

The lectotype of H. angustidens comes from "Lapa da Serra das Abelhas," and like other fossils of the same provenience (e.g., the types of H. orycter or H. clivigenis) is fragmentary and stained yellow­ish by superficial concretions. Several of the caves worked by Lund-including Serra das Abelhas- were destroyed by human activities over the last 170 years; however, a few, such as Cerca Grande located about 16 km NW Lagoa Santa city, where H. angustidens was also recorded, still remain. It is noteworthy that, despite extensive surveys of modern mammal fauna made by Lund around Lagoa Santa, Castoria was not detected (cf. Voss and Myers, 1991). In fact, the nearest record for the species is from Fazenda da Neblina in the Parque Estadual da Serra do Brigadeiro (Moreira et al., 2009), ca. 200 km SE of Lagoa Santa. The Lapa da Serra das Abelhas sigmodontine assemblage includes 8 taxa without recent records of occurrence near Lagoa Santa, including Castoria, Delomys, Euryoryzomys, Juliomys, Sooretamys, and Thaptomys (Winge, 1887; Pardiñas and Teta, 2013). The taxonomic structure of this assemblage is characterized by elements presently restricted to Atlantic Forest in mixture with Cerrado species, a non-analogue fauna probably triggered by cooling episodes during the Late Pleistocene (Voss and Carleton, 1993; Cartelle, 1999; Pardiñas and Teta, 2013).

The cursory indication by Hershkovitz (1990a:14) concerning the synonymy of A. serrensis and Akodon pervalens Thomas, 1925, a poorly known form with type locality at Caraparí (Tarija, Bolivia), must be dismissed. A. pervalens-judging from examina­tion of its holotype and paratype-looks like a true Akodon, phenotypically similar to species of the A. cursor species group as earlier suggested by Myers (1989). Although the holotype of A. pervalens is old (see Jayat et al., 2007:210) and some of its cranial and dental features (e.g., broad interorbital region, nasals tapering forward, absence of anteromedian flexus on M1) recall those of older A. serrensis, these are only general resemblances produced by aging (Supplementary Material S7).

FINAL CONSIDERATIONS

Castoria belongs to a clade of akodontine genera historically understood as a part of Akodon (cf. Reig, 1987), which has recently been reorganized into several genera (for a review, see D'Elía and Pardiñas, 2015). Morphological differences among Akodon, Deltamys, Thalpomys, Thaptomys, Necromys, and Castoria, as well as more controversial forms such as Microxus, are typically subtle (Table 3). This seems to be a signature of the Akodon Division, one that prevented the recognition of these different genera for most of the XXth century.

Since the 1970s, several distinctive lineages formerly subsumed in Akodon-directly or under subgeneric combinations-have been progressively elevated to full generic rank. Para­digmatic cases were Abrothrix (now included in a different tribe) and Thaptomys, but also more recently Thalpomys and Deltamys. Reig (1987:358) anticipated these changes when he wrote "I tentatively recognize a central genus Akodon, subdivided into five subgenera... I do not deny, however, that more detailed and com­prehensive further studies could eventually el­evate some of these taxa to generic level." These generic partitions within Akodon were possible because of the accumulation of morphological and karyological information and were then consolidated by DNA-based phylogenetic analy­ses (e.g., Hershkovitz, 1990b; Smith and Patton, 1993, 1999, 2007; Geise et al., 2001; D'Elía, 2003; D'Elía et al., 2003; Pardiñas et al., 2005; Jayat et al., 2010). It is interesting to point out that at least two putative genus-rank groupings remain subsumed under Akodon despite clear morphological indications about their singu­larities; these are Microxus and Hypsimys (see Reig, 1987:359-360; Ventura et al., 2000:497; Voss, 2003:20-21). Perhaps their retention in Akodon is more the result of historical inertia than of unequivocal phylogenetic support (see Patton et al., 1989; Smith and Patton, 2007). An example of such progress is the recent finding that Akodon bogotensis and Akodon latebricola are neither Akodon nor Akodon­tini (Alvarado-Serrano and D'Elía, 2013). The naming of Neomicroxus demonstrates that the content of Akodon, as presently understood containing about 38 living species (Pardiñas et al., 2015), requires detailed morphological and molecular scrutiny.

Castoria adds a new piece to Akodontini historical biogeography. In fact, the exclusion of the Abrotrichini (see Smith and Patton, 1999), the inclusion of Scapteromyini (see Smith and Patton, 1999; D'Elía et al., 2005), and the elevation to genera of several erstwhile subgenera or species (Deltamys, Gyldenstolpia; D'Elía et al., 2003; Pardiñas et al., 2008) have reshaped our vision about this diverse tribe. Of the 15 genera currently recognized (see Patton et al., 2015), three are Atlantic Forest endemic, Blarinomys, Brucepattersonius, and Thaptomys. Castoria offers further evidence for a potential area of endemism in southeastern Brazil, a picture also supported by several genera of non-akodontine sigmodontine rodents (Smith and Patton, 1999; Pardiñas et al., 2014a; Prado et al., 2014) and echimyid rodents (Upham and Patterson, 2015:Fig. 1).

ACKNOWLEDGMENTS

Robert Voss (AMNH) deserves credit for recognizing the congeneric status between H. angustidens and A. serrensis. He helped us not only by providing this information but also the original notes of his study and provided casts of H. angustidens that are now housed in CNP collections. Rob's selfless attitude is an example that better communica­tion among researchers is possible. J. Oliveira (MN) kindly provided on long-term loan a large sample of specimens collected at key localities in Brazil. C. Galliari, J. Notar­nicola, M. Lareschi, and R. Robles shared field efforts in order to obtain Argentinean material for study. A. Bezerra photographed for us the holotype of A. serrensis. M. Lares­chi provided parasitological data. Valuable specimens were freely accessed thanks to the curators D. Flores (MACN), J. Patton (MVZ), and B. Patterson (FMNH); K. L. Hansen (ZMK) kindly answered our request about Lund's collec­tions in Copenhagen. This paper was partially funded by Agencia Nacional de Promoción Científica y Tecnológica (PICT 2008-547 and 2014-1039 to UFJP) and Consejo Nacional de Investigaciones Científicas y Técnicas PIP 6179 (to UFJP), Centro Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ), World Wildlife Fund (WWF) and UERJ/ Prociência (to LG), and Fundação de Amparo a Pesquisa do Estado de São Paulo, FAPESP (BP.PD 2009/54300-0 to KV). Specimens from Rio de Janeiro state were collected under collecting license of IBAMA and ICMBio (license # 598633). This contribution was originally triggered by a suggestion of G. D'Elía who, after a decade of shared efforts and contributions of phylogenetic tips, declined to be an author; this manuscript also received significant input from P. Teta, who participated in several aspects of museum study and also in the masterful production of the figures. We are deeply grateful to all these individuals and institutions. Finally, it is important to mention that the manuscript benefited from three reviews, reflecting the zeal of the editors to ensure its quality; and also, the associate editor was deeply committed to ensure the linguistic correctness employed in the text. This is GEMA's (Grupo de Estudios de Mamíferos Australes) contribution #10.

LITERATURE CITED

1. ABREU M, AU CHRISTOFF, VH VALIATI, and LR de OLIVEIRA. 2014. New distribution records of Serra do Mar Grass Mouse Akodon serrensis Thomas, 1902 (Mammalia: Rodentia: Sigmodontinae) in the southernmost Brazil. Check List 10:655-659.         [ Links ]

2. ALVARADO-SERRANO D and G D'ELÍA. 2013. A new genus for the Andean mice Akodon latebricola and A. bogotensis (Rodentia: Sigmodontinae). Journal of Mammalogy 94:995-1015.         [ Links ]

3. ANDRADE AFB, CR BONVICINO, DC BRIANI, and S KASAHARA. 2004. Karyologic diversification and phylogenetic relationships of the genus Thalpomys (Rodentia, Sigmodontinae). Acta Theriologica 49:181- 190.         [ Links ]

4. AULER AS, LB PILÓ, PL SMART, X WANG, D HOFFMANN, DA RICHARDS, RL EDWARDS, WA NEVES, and H CHENG. 2006. U-series dating and taphonomy of Quaternary vertebrates from Brazilian caves. Palaeogeography, Palaeoclimatology, Palaeoecology 240:508-522.         [ Links ]

5. ÁVILA-PIRES FD. 1960. Roedores coleccionados na região de Lagoa Santa, Minas Gerais, Brasil. Arquivos do Museu Nacional, Rio de Janeiro 50:25-45.         [ Links ]

6. BARROS-BATTESTI DM, M ARZUA, PM LINARDI, JR BOTELHO, and IJ SBALQUEIRO. 1998. Interrelationship between ectoparasites and wild rodents from Tijucas do Sul, State of Paraná, Brazil. Memorias do Instituto Oswaldo Cruz 93:719-725.         [ Links ]

7. BONVICINO CR. 2011. Diversidade cariotípica em roedores Akodontini do Brasil. Boletim da Sociedade Brasileira de Mastozoologia, Rio de Janeiro 62:7-13.         [ Links ]

8. BONVICINO CR, SM LINDBERGH, and LS MAROJA. 2002. Small non-flying mammals from conserved and altered areas of Atlantic Forest and Cerrado: Comments on their potential use for monitoring environment. Brazilian Journal of Biology 62:765-774.         [ Links ]

9. CARLETON MD. 1973. A survey of gross stomach morphology in New World Cricetinae (Rodentia, Muroidea), with comments on functional interpretations. Miscellaneous Publications, Museum of Zoology, University of Michigan 146:1-43.         [ Links ]

10. CARLETON MD. 1980. Phylogenetic relationships in neotomine-peromyscine rodents (Muroidea) and a reappraisal of the dichotomy within New World Cricetinae. Miscellaneous Publications, Museum of Zoology, University of Michigan 157:1-146.         [ Links ]

11. CARLETON MD and GG MUSSER. 1989. Systematic studies of oryzomyine rodents (Muridae, Sigmodontinae): A synopsis of Microryzomys. Bulletin of the American Museum of Natural History 191:1-83.         [ Links ]

12. CARTELLE C. 1999. Pleistocene mammals of the Cerrado and Caatinga of Brazil. Pp. 27-46, in: Mammals of the Neotropics. The Central Neotropics, Vol. 3 (JB Eisenberg and KH Redford, eds.). The University of Chicago Press, Chicago.         [ Links ]

13. CARTELLE C. 2002. Peter W. Lund, a naturalist of several sciences. Lundiana 3:83-85.         [ Links ]

14. CHRISTOFF AU, V FAGUNDES, IJ SBALQUEIRO, MS MATTEVI, and Y YONENAGA-YASSUDA. 2000. Description of a new species of Akodon (Rodentia: Sigmodontinae) from southern Brazil. Journal of Mammalogy 81:838-851.         [ Links ]

15. COUTINHO LC, JA OLIVEIRA, and LM PESSôA. 2013. Morphological variation in the appendicular skeleton of Atlantic Forest sigmodontine rodents. Journal of Morphology 274:779-792.         [ Links ]

16. DA SILVA JMC. 2000a. South America: Along the Atlantic coast of southeastern and southern Brazil. http://www.worldwildlife.org/ecoregions/nt0160. Accessed on 2 June 2014.         [ Links ]

17. DA SILVA JMC. 2000b. Araucaria moist forests. http://www.worldwildlife.org/ecoregions/nt0101. Accessed on 2 June 2014.         [ Links ]

18. D'ELÍA G. 2003. Phylogenetics of Sigmodontinae (Rodentia, Muroidea, Cricetidae), with special reference to the Akodont group, and with additional comments on historical biogeography. Cladistics 19:307-323.         [ Links ]

19. D'ELÍA G and UFJ PARDIÑAS. 2015. Tribe Akodontini Vorontsov 1959. Pp. 140-144, in: Mammals of South America, Volume 2 - Rodents (JL Patton, UFJ Pardiñas, and G D'Elía, eds.). The University of Chicago Press, Chicago.         [ Links ]

20. D'ELÍA G, EM GONZÁLEZ, and UFJ PARDIÑAS. 2003. Phylogenetic analysis of Sigmodontinae rodents (Muroidea), with special reference to the akodont genus Deltamys. Mammalian Biology 68:351-364.         [ Links ]

21. D'ELÍA G, UFJ PARDIÑAS, and P MYERS. 2005. An introduction to the genus Bibimys (Rodentia: Sigmodontinae): Phylogenetic position and alpha taxonomy. Pp. 211-246, in: Mammalian diversification: from chromosomes to phylogeography. A celebration of the career of James Patton (EA Lacey and P Myers, eds.). University of California Press, Berkeley.         [ Links ]

22. FABRE P-H, L HAUTIER, D DIMITROV, and E DOUZERY. 2012. A glimpse on the pattern of rodent diversification: A phylogenetic approach. BMC Evolutionary Biology 12:88.         [ Links ]

23. GEISE L. 2014. Procedimentos genéticos iniciais na captura e preparação de mamíferos. Pp. 221-235, in: Técnicas de estudos aplicadas aos mamíferos silvestres brasileiros (MR dos Reis, AL Peracchi, BK Rossaneis, and MN Fregonezi, org.). 2nd Edition, Technical Books Editora, Rio de Janeiro.         [ Links ]

24. GEISE L, FC CANAVEZ and HNA SEUÁNEZ. 1998. Comparative karyology in Akodon (Rodentia, Sigmodontinae) from Southeastern Brazil. Journal of Heredity 89:158-163.         [ Links ]

25. GEISE L, LG PEREIRA, DEP BOSSI, and HG BERGALLO. 2004. Pattern of elevational distribution and richness of non volant mammals in Itatiaia National Park and its surroundings, in southeastern Brazil. Brazilian Journal of Biology 64:599-612.         [ Links ]

26. GEISE L, MF SMITH and JL PATTON. 2001. Diversification in the genus Akodon (Rodentia, Sigmodontinae) in southeastern South America: mtDNA sequence analysis. Journal of Mammalogy 82:92-101.         [ Links ]

27. GEISE L, M WEKSLER, and CR BONVICINO. 2004. Presence or absence of gall bladder in some Akodontini rodents (Muridae, Sigmodontinae). Mammalian Biology 69:210-214.         [ Links ]

28. GONÇALVES PR and JA OLIVEIRA. 2014. An integrative appraisal of the diversification in the Atlantic forest genus Delomys (Rodentia: Cricetidae: Sigmodontinae) with the description of a new species. Zootaxa 3760:1-38.         [ Links ]

29. GONÇALVES PR, P MYERS, JF VILELA and JA OLIVEIRA. 2007. Systematics of species of the genus Akodon (Rodentia: Sigmodontinae) in southeastern Brazil and implications for the biogeography of the campos de altitude. Miscellaneous Publications, Museum of Zoology, University of Michigan 197:1-24.         [ Links ]

30. GYLDENSTOLPE N. 1932. A manua l of Neotropical sigmodont rodents. Kunglia Svenska Vetenskapsakademiens Handlingar, Serie 3, 11:1-164.         [ Links ]

31. HANSEN KL. 2012. E Museo Lundii - Addendum - A series of annotated catalogues based on subfossil and recent material. First Edition. Edited by the author, Copenhagen.         [ Links ]

32. HASS I, IJ SBALQUEIRO, and S MÜLLER. 2008. Chromosomal phylogeny of four Akodontini species (Rodentia, Cricetidae) from Southern Brazil established by Zoo-FISH using Mus musculus (Muridae) painting probes. Chromosome Research 16:75-88.         [ Links ]

33. HERSHKOVITZ P. 1962. Evolution of Neotropical cricetine rodents (Muridae) with special reference to the phyllotine group. Fieldiana, Zoology 46:1-524.         [ Links ]

34. HERSHKOVITZ P. 1966. South American swamp and fossorial rats of the scapteromyine group (Cricetinae, Muridae) with comments on the glans penis in murid taxonomy. Zeitschrift für Säugetierkunde 31:81-149.         [ Links ]

35. HERSHKOVITZ P. 1990a. Mice of the Akodon boliviensis size class (Sigmodontinae, Cricetidae), with the description of two new species from Brazil. Fieldiana, Zoology, New Series 57:iii + 1-35.         [ Links ]

36. HERSHKOVITZ P. 1990b. The Brazilian rodent genus Thalpomys (Sigmodontinae, Cricetidae) with a description of a new species. Journal of Natural History 24:763-783.         [ Links ]

37. HERSHKOVITZ P. 1998. Report on some sigmodontine rodents collected in southeastern Brazil with descriptions of a new genus and six new species. Bonner Zoologische Beiträge 47:193-256.         [ Links ]

38. ICZN. 1999. International Code of Zoological Nomenclature, Fourth edition. International Trust for Zoological Nomenclature, London.         [ Links ]

39. JAYAT P, PE ORTIZ, UFJ PARDIÑAS, and G D'ELÍA. 2007. Redescripción y posición filogenética del ratón selvático Akodon sylvanus (Rodentia: Cricetidae: Sigmodontinae). Mastozoología Neotropical 14:201-225.         [ Links ]

40. JAYAT P, PE ORTIZ, J SALAZAR-BRAVO, UFJ PARDIÑAS, and G D'ELÍA. 2010. The Akodon boliviensis species group (Rodentia: Cricetidae: Sigmodontinae) in Argentina: Species limits and distribution, with the description of a new entity. Zootaxa 2409:1-61.         [ Links ]

41. KOSLOSKI MA. 1997. Morfometria craniana de uma população de Akodon serrensis Thomas, 1902 (Rodentia, Cricetidae) do município de Araucária, Paraná, Brasil. Revista Estudos de Biologia 41:61-87.         [ Links ]

42. LEITE YLR, PJR KOK, and M WEKSLER. 2015. Evolutionary affinities of the ‘Lost World' mouse suggest a late Pliocene connection between the Guiana and Brazilian shields. Journal of Biogeography 42:706-715.

43. LIASCOVICH RC and OA REIG. 1989. Low chromosomal number in Akodon cursor montensis Thomas and karyologic confirmation of A. serrensis Thomas in Misiones, Argentina. Journal of Mammalogy 70:391- 395.         [ Links ]

44. LINARDI PM and LR GUIMARÃES. 2000. Sifonapteros do Brasil. Ed. Museu de Zoologia da Universidade de São Paulo, FAPESP, São Paulo         [ Links ]

45. MASSOIA E and A FORNES. 1965. Notas sobre el género Scapteromys (Rodentia Cricetidae). II. Fundamentos de la identidad específica de S. principalis (Lund) y S. gnambiquarae (M. Ribeiro). Neotrópica 11:1-7.         [ Links ]

46. MIRANDA-RIBEIRO A. 1905. Vertebrados do Itatiaya (Peixes, Serpentes, Sáurios, Aves e Mammíferos). Arquivos de Museu Nacional, Rio de Janeiro 13:163- 190 + 3 plates.         [ Links ]

47. MODESTO TC, FS PESSôA, MC ENRICI, N ATTIAS, T JORDÃO-NOGUEIRA, LM COSTA, HG ALBUQUERQUE, and HG BERGALLO. 2008. Mamíferos do Parque Estadual do Desengano, Rio de Janeiro, Brasil. Biota Neotropica 8:153-159.         [ Links ]

48. MORAES LB, DAP BOSSI, and AX LINHARES. 2003. Siphonaptera parasites of wild rodents and marsupials trapped in three mountain ranges of the Atlantic Forest in southeastern Brazil. Memorias do Instituto Oswaldo Cruz 98:1071-1076.         [ Links ]

49. MOREIRA JC, EG MANDUCA, PR GONÇALVES, MM MORAIS, RF PEREIRA, G LESSA, and JA DERGAM. 2009. Small mammals from Serra do Brigadeiro state park, Minas Gerais, Southeastern Brazil: Species composition and elevational distribution. Arquivos do Museu Nacional, Rio de Janeiro 67:103- 118.         [ Links ]

50. MUSSER GG and MD CARLETON. 2005. Superfamily Muroidea. Pp. 894-1531, in: Mammal species of the World: A taxonomic and geographic reference (DE Wilson and DM Reeder, eds.). 3rd edition, Johns Hopkins University Press, Baltimore.         [ Links ]

51. MUSSER GG, MD CARLETON, EM BROTHERS and AL GARDNER. 1998. Systematic studies of oryzomyine rodents (Muridae, Sigmodontinae): Diagnoses and distributions of species formerly assigned to Oryzomys "capito." Bulletin of the American Museum of Natural History 236:1-376.         [ Links ]

52. MYERS P. 1989. A preliminary revision of the varius group of Akodon (A. dayi, dolores, molinae, neocenus, simulator, toba and varius). Pp. 5-54, in: Advances in Neotropical Mammalogy (KH Redford and JF Eisenberg, eds.). Sandhill Crane Press, Inc., Gainesville.         [ Links ]

53. OLIFIERS N, A CUNHA, C GRELLE, CR BONVICINO, L GEISE, L PEREIRA, M VIEIRA, P D'ANDREA, and R CERQUEIRA. 2007. Lista de espécies de pequenos mamíferos não-voadores do Parque Nacional da Serra dos Órgãos. Pp. 183-192, in: Ciência e Conservação na Serra dos Órgãos (C Cronemberger and EB Viveiros de Castor, org.). Ibama, Brasília.         [ Links ]

54. OLSON DM, et al. 2001. Terrestrial ecoregions of the world: A new map of life on Earth. Bioscience 51:933-938.         [ Links ]

55. PACHECO V. 2003. Phylogenetic analyses of the Thomasomyini (Muroidea: Sigmodontinae) based on morphological data. Unpublished Ph. D. dissertation, City University of New York, New York.         [ Links ]

56. PAGLIA AP, et al. 2012. Lista anotada dos mamíferos do Brasil / Annotated checklist of Brazilian mammals. 2ª Edição / 2nd Edition. Occasional Papers in Conservation Biology 6:1-76.         [ Links ]

57. PARDIÑAS UFJ and P TETA. 2011. On the taxonomic status of the Brazilian mouse Calomys anoblepas Winge, 1887 (Mammalia, Rodentia, Cricetidae). Zootaxa 2788:38-44.         [ Links ]

58. PARDIÑAS UFJ and P TETA. 2013. Taxonomic status of Mus talpinus Lund (Rodentia: Sigmodontinae) from the Quaternary deposits of Lagoa Santa, Minas Gerais, Brazil and its paleoenvironmental meaning. Mammalia 77:347-355.         [ Links ]

59. PARDIÑAS UFJ, G D'ELÍA, and S CIRIGNOLI. 2003. The genus Akodon (Muroidea: Sigmodontinae) in Misiones, Argentina. Mammalian Biology 68:129-143.         [ Links ]

60. PARDIÑAS UFJ, G D'ELÍA, S CIRIGNOLI and P SUÁREZ. 2005. A new species of Akodon (Rodentia, Cricetidae) from the northern campos grasslands of Argentina. Journal of Mammalogy 86:462-474.         [ Links ]

61. PARDIÑAS UFJ, G D'ELÍA, and P TETA. 2008 [2009]. Una introducción a los mayores sigmodontinos vivientes: revisión de Kunsia Hershkovitz, 1966 y descripción de un nuevo género (Rodentia: Cricetidae). Arquivos do Museu Nacional, Rio de Janeiro 66:509-594.         [ Links ]

62. PARDIÑAS UFJ, G D'ELÍA, G LESSA, M PASSAMANI, and P TETA. 2014a. Nuevos datos morfológicos y una hipótesis filogenética para Phaenomys (Rodentia, Cricetidae). Mastozoología Neotropical 21:251-261.         [ Links ]

63. PARDIÑAS UFJ, G LESSA, P TETA, J SALAZAR-BRAVO, and EM CÂMARA. 2014b. A new genus of sigmodontine rodent from eastern Brazil and the origin of the tribe Phyllotini. Journal of Mammalogy 95:201-215.         [ Links ]

64. PARDIÑAS UFJ, P TETA, D ALVARADO-SERRANO, L GEISE, JP JAYAT, PE ORTIZ, PR GONÇALVES, and G D'ELÍA. 2015. Genus Akodon Meyen, 1833. Pp. 144-204, in: Mammals of South America, Volume 2 - Rodents (JL Patton, UFJ Pardiñas, and G D'Elía, eds.). The University of Chicago Press, Chicago.         [ Links ]

65. PATTON JL, P MYERS, and MF SMITH. 1989. Electromorphic variation in selected South American akodontine rodents (Muridae: Sigmodontinae), with comments on systematic implications. Zeitschrift für Saugetierkunde 54:347-359.         [ Links ]

66. PATTON JL, UFJ PARDIÑAS, and G D'ELÍA. 2015. Mammals of South America, Volume 2 - Rodents. University of Chicago Press, Chicago.         [ Links ]

67. PAVAN SE and YRL LEITE. 2011. Morphological diagnosis and geographic distribution of Atlantic Forest red-rumped mice of the genus Juliomys (Rodentia: Sigmodontinae). ZOOLOGIA (Curitiba) 28:663-672.         [ Links ]

68. PERCEQUILLO AR, M WEKSLER, and LP COSTA. 2011. A new genus and species of rodent from the Brazilian Atlantic Forest (Rodentia: Cricetidae: Sigmodontinae: Oryzomyini), with comments on oryzomyine biogeography. Zoological Journal of the Linnean Society 161:357-390.         [ Links ]

69. PRADO JR, PGG BRENNAND, LP GODOY, G LIBARDI, EF ABREU-JÚNIOR, P ROTH, E CHIQUITO, and AR PERCEQUILLO. 2014. Species richness and areas of endemism of oryzomyine rodents (Cricetidae, Sigmodontinae) in South America: an ndm/vndm approach. Journal of Biogeography 42:540-551.         [ Links ]

70. REIG OA. 1977. A proposed unified nomenclature for the enamelled components of the molar teeth of the Cricetidae (Rodentia). Journal of Zoology 181:227-241.         [ Links ]

71. REIG OA. 1980. A new fossil genus of South American cricetid rodents allied to Wiedomys, with an assessment of the Sigmodontinae. Journal of Zoology 192:257-281.         [ Links ]

72. REIG OA. 1987. An assessment of the systematics and evolution of the Akodontini, with the description of new fossil species of Akodon (Cricetidae: Sigmodontinae). Pp. 347-399, in: Studies in Neotropical mammalogy. Essays in honor of Philips Hershkovitz (BD Patterson and RM Timm, eds.). Fieldiana, Zoology, new series, 39.         [ Links ]

73. SILVEIRA F, IJ SBALQUEIRO, and ELA MONTEIRO-FILHO. 2013. Identificação das espécies brasileiras de Akodon (Rodentia: Cricetidae: Sigmodontinae) através da microestrutura dos pelos. Biota Neotropical 13:339-345.         [ Links ]

74. SMITH MF and JL PATTON. 1993. The diversification of South American murid rodents: Evidence from mitochondrial DNA sequence data for the akodontine tribe. Biological Journal of the Linnean Society 50:149-177.         [ Links ]

75. SMITH MF and JL PATTON. 1999. Phylogenetic relationships and the radiation of sigmodontine rodents in South America: Evidence from cytochrome b. Journal of Mammalian Evolution 6:89-128.         [ Links ]

76. SMITH MF and JL PATTON. 2007. Molecular phylogenetics and diversification of South American grass mice, genus Akodon. Pp. 827-858, in: Studies in contemporary mammalian biology. Papers honoring the remarkable career of Oliver P. Pearson, 1915-2003 (D Kelt, E Lessa, and J Salazar-Bravo, eds.). University of California Publications in Zoology, Berkeley.         [ Links ]

77. STEPPAN SJ. 1995. Revision of the tribe Phyllotini (Rodentia: Sigmodontinae), with a phylogenetic hypothesis for the Sigmodontinae. Fieldiana, Zoology, new series 80:1-112.         [ Links ]

78. TEIXEIRA BR et al. 2014. Population ecology of hantavirus rodent hosts in southern Brazil. American Journal of Tropical Medicine and Hygiene 91:249-257.         [ Links ]

79. TESTONI AF, J FUMIS, SL ALTHOFF, FR TORTATO, and JJ CHEREM. 2012. Akodon serrensis Thomas, 1902 (Mammalia: Rodentia: Sigmodontinae): Records in Santa Catarina state, southern Brazil. Check List 8:1344-1346.         [ Links ]

80. THOMAS O. 1902. On mammals from the Serra do Mar, Parana, collected by Mr. Alphonse Robert. Annals and Magazine of Natural History, series 7, 9:59-64.         [ Links ]

81. TROUESSART EL. 1898. Catalogus mammalium tam viventium quam fossilium. Tomus 2. R. Friedländer and Sohn, Berlin.         [ Links ]

82. UPHAM NS and BD PATTERSON. 2015. Evolution of caviomorph rodents: A complete phylogeny and timetree for living genera. Pp. 63-120, in: Biology of caviomorph rodents: Diversity and evolution (AI Vasallo and D Antenucci, eds.). SAREM Series A, Mammalogical Research, Volume 1, Buenos Aires.         [ Links ]

83. VENTURA J, MJ LÓPEZ-FUSTER, M SALAZAR, and R PÉREZ-HERNÁNDEZ. 2000. Morphometric analysis of some Venezuelan akodontine rodents. Netherlands Journal of Zoology 50:487-501.         [ Links ]

84. VENTURA K, PC O'BRIEN, Y YONENAGA-YASSUDA, and MA FERGUSON-SMITH. 2009. Chromosome homologies of the highly rearranged karyotypes of four Akodon species (Rodentia, Cricetidae) resolved by reciprocal chromosome painting: The evolution of the lowest diploid number in rodents. Chromosome Research 17:1063-1078.         [ Links ]

85. VENTURA K, MJJ SILVA, V FAGUNDES, and AU CHRISTOFF. 2006. Non-telomeric sites as evidence of chromosomal rearrangement and repetitive (TTAGGG)n arrays in heterochromatic and euchromatic regions in four species of Akodon (Rodentia, Muridae). Cytogenetic and Genome Research 115:169-175.         [ Links ]

86. VENTURA K, MJJ SILVA, L GEISE, YRL LEITE, UFJ PARDIÑAS, Y YONENAGA-YASSUDA, and G D'ELÍA. 2013. The phylogenetic position of the enigmatic Atlantic Forest endemic spiny-mouse Abrawayaomys (Rodentia: Sigmodontinae). Zoological Studies 52:55 doi:10.1186/1810-522X-52-55.         [ Links ]

87. VIEIRA EM and ELA MONTEIRO-FILHO. 2003. Vertical stratification of small mammals in the Atlantic rain forest of south-eastern Brazil. Journal of Tropical Ecology 19:501-507.         [ Links ]

88. VIEIRA EM, M PIZO, and P IZAR. 2003. Fruit and seed exploitation by small rodents of the Brazilian Atlantic forest. Mammalia 67:533-539.         [ Links ]

89. VILELA JF, B MELLO, CM VOLOCH, and CG SCHRAGO. 2014. Sigmodontine rodents diversified in South America prior to the complete rise of the Panamanian Isthmus. Journal of Zoological Systematics and Evolutionary Research 52:249-256.         [ Links ]

90. VOSS RS. 1988. Systematics and ecology of ichthyomyine rodents (Muroidea): patterns of morphological evolution in small adaptive radiation. Bulletin of the American Museum of Natural History 188:262-493.         [ Links ]

91. VOSS RS. 1991. An introduction to the Neotropical muroid rodent genus Zygodontomys. Bulletin of the American Museum of Natural History 210:1-113.         [ Links ]

92. VOSS RS. 1993. A revision of the Brazilian muroid rodent genus Delomys with remarks on "thomasomyine" characters. American Museum Novitates 3073:1-44.         [ Links ]

93. VOSS RS. 2003. A new species of Thomasomys (Rodentia: Muridae) from Eastern Ecuador, with remarks on mammalian diversity and biogeography in the Cordillera Oriental. American Museum Novitates 3421:1-47.         [ Links ]

94. VOSS RS and MD CARLETON. 1993. A new genus for Hesperomys molitor Winge and Holochilus magnus Hershkovitz, with comments on phylogenetic relationships and oryzomyine monophyly. American Museum Novitates 3085:1-39.         [ Links ]

95. VOSS RS and P MYERS. 1991. Pseudoryzomys simplex (Rodentia: Muridae) and the significance of Lund's collections from the caves of Lagoa Santa, Brazil. Bulletin of the American Museum of Natural History 206:414-432.         [ Links ]

96. VOSS RS, M GÓMEZ-LAVERDE, and V PACHECO. 2002. A new genus for Aepeomys fuscatus Allen, 1912, and Oryzomys intectus Thomas, 1921: Enigmatic murid rodents from Andean cloud forests. American Museum Novitates 3373:1-42.         [ Links ]

97. WEKSLER M and L GEISE. 1996. Medidas cranianas de roedores Sigmodontinae. Boletim da Sociedade Brasileira de Mastozoologia, Rio de Janeiro 30:1-2.         [ Links ]

98. WINGE H. 1887. Jordfundne og nulevende Gnavere (Rodentia) fra Lagoa Santa, Minas Geraes, Brasilien. E Museo Lundii 1(3):1-178 + 18 plates.         [ Links ]

APPENDIX 1

Studied specimens belong to the following mammal collections: AMNH, American Museum of Natural History, New York, USA; BMNH, Natural History Museum, London, UK; CNP, Colec­ción de Mamíferos del Centro Nacional Patagónico, Chubut, Argentina; FMNH, Field Museum of Natural History, Chicago, USA; MACN, Colección Nacional de Mastozoología, Museo Argentino de Ciencias Naturales "Bernardino Rivadavia," Buenos Aires, Argentina; MN, Museu Nacional, Rio de Janeiro, Brazil; MZUFV, Museu de Zoologia, Departamento de Biologia Animal, Universidade Federal de Viçosa, Viçosa, Minas Gerais, Brazil; ZMK, Universitets Zoologisk Museum, Copen­hagen, Denmark. Symbols denote specific studies carried out on each specimen (* = measured; k = karyotyped).

Akodon montensis (n = 46): Argentina, Misiones, Balneario Municipal de Aristóbulo del Valle sobre el arroyo Cuña Pirú, 10 km NN W Aristóbulo del Valle (CNP 1785, 1786, 1789, 1790, 1794, 1795, 1796, 1797, 1798, 1799, 1800, 1802, 1804, 1805, 1807, 1808, 1809, 1810, 1811, 1812, 1813, 1814, 1815, 1818, 1819, 1821, 1825, 1829, 1831, 1833, 1835, 1838, 1844, 1845, 1847), Reserva Privada de Usos Múltiples Guaraní (CNP 4243, 4244, 4245, 4247, 4248, 4249, 4250, 4251, 4252, 4253, 4254).

Castoria angustidens nov. comb. (n = 123): Argentina, Misiones, Reserva Privada de Vida Silvestre Urugua-í (MACN 22246), RP 2, 6 km NE Arroyo Paraíso (CNP 449*). Brazil, Minas Gerais, Parque Estadual da Serra do Brigadeiro, Fazenda da Neblina (MZUFV 520*, 551*, 1076, 1148*, 1150*, 1151*, 1153*, 1154*, 1155*, 1156*, 1215*, 1220*, 1234, 1242, 1266*, 1281*, 1298*, 1630, 1635, 1653, 1655, 2555; UFMG 1858*, 1859*, 1860*, 1861*, 1862*), Parque Estadual da Serra do Bri­gadeiro, Serra das Cabeças (MZUFV 1243*, 1248*, 1250*, 1267*, 1280*, 1286*, 1287*, 1294*, 1295*), Lapa da Serra das Abelhas (ZMK 1/1845:13246* [lectotype of H. angustidens]; CNP 4267 [lectotype: cast of upper molar row], CNP 4268 [paralectotype: cast of lower molar row]); Paraná, Roça Nova (BMNH 3.7.1.66*-3.7.1.73*, ZMK 1257*); Rio de Janeiro, Parque Nacional da Serra dos Órgãos, Vale das Antas (MN 69807*k, 69808*, 69809*, 69810*, 69811*, 69812* k, 69813*, 77085*, 77099*k, 77100*, 77101*, 77102*, 77104*, 77105*, 77108*, 77111*, 77112*, 77113*, 77116*, 77117*k, 77119*, 77121*, 77122*, 77123*, 77078*k, 77079*k, 77081*, 77082*, 77084*k, 81130*, 81131*, 81133*, 81134*, 81135*, 81136*, 81137*, 81138*, 77086*, 77087*k, 77088*, 77090*, 77091*, 81144*, 77096*, 81147*, 81148*, 81150*, 81151*, 81152*, 81153*, 81154*, 81155*, 81156*, 81157*, 81158*, 81159*, 81160*, 81161, 81162*, 81163*, 81164*, 81165*, 81166*, 81168*, 81169*, 81170*, 81171*, 81172*, 77097*k, 81175*), Sítio Xitaca, Debossan, Nova Friburgo (MN 35912*, 35920*, 35921*, 35923*, 35924*, 35926*), Parque Nacional do Itatiaia (MN 69803*), Retiro de Ramos (MN 2513* [holotype of H. serrensis leucogula]), Visconde de Mauá (MN 69804*); São Paulo, Fazenda Intervales (FMNH 143275).

Deltamys kempi (n = 6): Argentina, Buenos Aires, La Balandra (CNP 581, 893, 3086, 3087, 4141), General Lavalle (CNP 2377).

Necromys lactens (n = 5): Argentina, Jujuy, Bárcena, ca. 3 km S sobre RN 9 (CNP 1486), Abra de la Ciénaga Negra (CNP 1487), San Francisco (CNP 3032, 3037, 4124).

Necromys lasiurus (n = 8): Argentina, Santiago del Estero, Estación Experimental INTA "La María" (CNP 2358); Chaco, 5 km NW Puerto Las Palmas (CNP 3036, 3049); Misiones, Estancia Santa Inés (CNP 3038, 3041, 3043, 3045, 3048).

Necromys obscurus (n = 5): Argentina, Buenos Aires, Arroyo de las Brusquitas (CNP 891, 2380, 3039, 3047), Estación Experimental INTA "Balcarce" (CNP 3053).

Thaptomys nigrita (n = 23): Argentina, Misiones, Balneario Municipal de Aristóbulo del Valle sobre el arroyo Cuña Pirú, 10 km NN W Aristóbulo del Valle (CNP 1788, 1791, 1816, 1826, 1827, 1828, 1832, 1839, 1846, 1929, 1931), Refugio Moconá (CNP 1938, 3008, 3010, 3011), Asentamiento Aborigen Kaaguy Poty, 1 km NNW intersección RP 7 y arroyo Cuña Pirú (CNP 1979, 2370, 2371, 2372), Posada Las Malvinas (CNP 3526), Reserva Privada de Vida Silvestre Urugua-í (CNP 3875). Brazil: Minas Gerais, Parque Estadual da Serra do Brigadeiro, Fazenda da Neblina (MZUFV 1110, 1112)

SUPPLEMENTARY MATERIAL ONLINE

S1. Gazetteer of the recording localities for the genus Castoria.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup1.doc

S2. Study notes on Habrothrix angustidens and Akodon serrensis made by R. S. Voss in November, 2008.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup2.doc

S3. Principal component analysis of all age classes of Brazilian specimens of Castoria, including the lectotype of Habrothrix angustidens.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup3.doc

S4. External appereance of Castoria angustidens, nov. comb.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup4.doc

S5. Gastric morphology of Castoria angustidens, nov. comb.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup5.doc

S6. Chromosomes of Castoria angustidens, nov. comb.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup6.doc

S7. Comparison of the holotypes of Akodon pervalens and A. serrensis.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup7.doc

S8. Habitat of Castoria angustidens, nov. comb. at Itamonte, Brazil.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup8.doc

S9. Joint distribution of latitude and elevation for Castoria angustidens, nov. comb. collection localities.

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup9.doc

S10. Lectotype of Castoria angustidens, nov. comb. (ZMK 1/1845:13246).

https://www.sarem.org.ar/wp-content/uploads/2016/06/SAREM_MastNeotrop_23-1_Pardinas-sup10.doc