INTRODUCTION

Glyphonycteris^{Thomas, 1896} is a poorly known genus of New World leaf-nosed bats that inhabit lowland and mountain forests of Central and South America (^{Williams & Genoways 2007}; ^{Zortea et al. 2008}; ^{Tirira et al. 2016}; ^{Solari 2018}). This genus comprises gleaning foraging bats that are less frequently collected in studies using mist net or harp trap and are therefore poorly represented in museums (^{Simmons & Voss 1998}; ^{Pedro et al. 2001}; ^{Sampaio et al. 2003}; ^{Gregorin & Rossi 2005}). Glyphonycteris and other related genera, such as Lampronycteris, Neonycteris, and Trinycteris, were traditionally treated as subgenera of Micronycteris after ^Sanborn’s (1949) revision. However, this taxonomic arrangement was frequently questioned due to the paraphyly of Micronycteris (see ^{Simmons 1996}; ^{Simmons & Voss 1998}; ^{Wetterer et al. 2000}). In a taxonomic revision of the mammals of Paracou, Simmons & Voss (1998) included Barticonycteris daviesi (Hill 1965) in Glyphonycteris; the authors recognized it as a full genus and provided an emended diagnosis. More recently, several molecular phylogenetic analyses recovered Glyphonycteris + Trinycteris as a sister clade of Carollia (^{Baker et al. 2002}; ²⁰⁰³; ²⁰¹²; ²⁰¹⁶; ^{Amador et al. 2018}). In this phylogenetic context, Baker et al. (²⁰¹²; ²⁰¹⁶) proposed and diagnosed the subfamily Glyphonycterinae in which the authors included Glyphonycteris, Trinycteris, and Neonycteris,and suggested that the observed morphological similarities with Micronycteris and Macrotus represent plesiomorphic features.

Currently, Glyphonycteris comprises three nominal species: Glyphonycteris behnii (Peters 1865), G. daviesi (Hill 1965), and G. sylvestris (^{Thomas 1896}) (^{Simmons 2005}). Glyphonycteris behnii is a rare species that has been suggested to be a senior synonym of G. sylvestris (^{Simmons & Voss 1998}) and it is only known by two specimens from Cuzco in Peru, one from Mato Grosso and one from Minas Gerais in Brazil (^{Andersen 1906}; ^{Peracchi & Albuquerque 1985}). Glyphonycteris daviesi, the largest of the genus (forearm 53.8–58.1 mm and skull over 25.0 mm; ^{Williams & Genoways 2007}), is distributed in Central America (Honduras, Costa Rica and Panama), island of Trinidad and South America (Venezuela, Colombia, Guiana, Surinam, French Guiana, Brazil, Ecuador, Peru, and Bolivia; ^{Williams & Genoways 2007}; ^{Morales-Martínez & Suárez-Castro 2014}). Lastly, G. sylvestris, the typespecies of the genus (^{Thomas 1896}, b:301), is also the smallest species (forearm 38.0–44.0 mm and the skull less than 22.0 mm; ^{Williams & Genoways 2007}), and inhabits Central America (Mexico, Panama and Costa Rica) and northern South America (Colombia, Venezuela, Trinidad and Tobago, Surinam, French Guiana, Ecuador, and Peru; ^{Williams & Genoways 2007}; ^{Datzman et al. 2010}; ^{Morales-Martínez & Suárez-Castro 2014}; ^{Tirira et al. 2016}). Additional populations were reported from Atlantic Forests Ecoregion Complex in southeastern Brazil, with a few records from Minas Gerais, Rio de Janeiro, São Paulo, and Paraná States (^{Trajano 1982}; ^{Sekiama et al. 2001}; ^{Pedro et al. 2001}; ^{Dias et al. 2003}; ^{Tavares et al. 2010}), including a new record from the Cerrado in the Tocantins state (^{Felix et al. 2016}).

Cytogenetic studies of bats from the family Glyphonycterinae are scarce, and the karyotype of only three species has been previously studied with the use of conventional chromosome staining techniques (^{Baker & Hsu 1970}; ^{Patton 1976}; ^{Honeycutt et al. 1980}; ^{Baker et al. 1982}). Karyological data are restricted to just one specimen of G. daviesi from Surinam with 2n = 28 chromosomes and autosomal fundamental number FN = 52 described from one specimen (^{Honeycutt et al. 1980}; ^{Baker et al. 2016}), whereas 2n = 22 chromosomes and FN = 36 was described from seven female specimens of G. sylvestris (^{Honeycutt et al. 1980}). Other cytogenetic methods such as staining with different base-specific fluorochromes to determine DNA sequence composition (banding technique) have not been employed for species of Glyphonycteris, evidencing the lack of knowledge of this genus of bats.

Fig. 1 Location of records of Glyphonycteris sylvestris in the Atlantic forests from Brazil (1-6) and a new record for Argentina (7). All numbered localities are listed in Table 2.  

Here, we report the first record of G. sylvestris in Argentina and in the Araucaria Moist Forest ecoregion extending its latitudinal distribution. In addition, the goals of this study were to examine dental features, external and skull measurements of our specimen, and compare them with information of bats of this species from other tropical sites. We described for the first time the karyotype for G. sylvestris from the Atlantic Forest populations and compared our results with the available information.

MATERIALS AND METHODS

The study site was located in Parque Provincial El Piñalito, 19.5 km NE of Cruce Caballero along National Route 14, Department of San Pedro (26° 25’ 41.28"S, 53° 50’ 39.6" W, 762 m a.s.l.; Fig. 1), northeast of Misiones Province, Argentina, in the Araucaria Moist Forests. This forest is one of the 15 terrestrial ecoregions that compose The Atlantic Forests Global 200 Ecoregion (hereafter “Atlantic Forest”) (^{Dinerstein et al. 1995}; ^{Olson et al. 2001}). In Argentina, the Araucaria Moist Forest forms a narrow belt along the Sierra of Misiones in the center and northeast of Misiones province (^{Giraudo et al. 2003}). The forest canopy is characterized by the presence of conifer trees (Araucaria angustifolia; Araucariaceae). The subcanopy is characterized by Nectandra megapotamica (Lauraceae), Balfourodendron riedelianum (Rutaceae), Cordia americana (Boraginaceae), Diatenopteryx sorbifolia (Sapindaceae), Parapiptadenia rigida and Myrocarpus frondosus (Fabaceae). Understory is characterized by Alsophila setosa (Cyatheaceae), Piper (Piperaceae), Sorocea bonplandii (Moraceae), Trichila (Meliaceae), Psichotria, Ixora (Rubiaceae) and ferns were dominant in the herbaceous stratum. Vascular epiphytes such as ferns, cacti (Lepismium and Rhipsalis), and bromeliads are abundant (^{Cabrera 1976}). The climate is hot and wet, with mean maximum temperature of 17.7 °C, and only mildly seasonal, since the annual rainfall (1921 mm) is distributed rather uniformly throughout the year (data estimated from Worldclim using DIVA-GIS, version 4.2; ^{Hijmans et al. 2005}).

We deployed 10 mist-nets/night at the same sampling site from 6 to 9 March 2016. Nets were placed at the ground level up to the subcanopy level (6–8 m high) inside the forest, in flight pathways, riparian forest and forest edge. Mist nets were left opened for approximately 6 h from sunset and were checked every 30 min. Each captured bat was removed from the net and placed in cloth bags for data collection and identification; taxonomic treatment follows ^{Gardner (2007)}. Bats were marked on the back by trimming the hair to avoid overestimation of the catches and were released near the capture site. Fecal samples were obtained from the bags and the bats handled in the net, and each sample was preserved as an individually labelled dry pellet. The plant dietary items were identified using a reference collection of seeds that was initiated by ^{Giannini (1999)} and continued by ^{Sánchez et al. (2012)} and ^{Sánchez & Dos Santos (2015)}.

Voucher specimen was preserved as skin, skull and skeleton after recording the basic external measurements, and finally housed in the Mammal Collection of the Laboratorio de Genética Evolutiva (CM-LGE) of the Instituto de Biología Subtropical (IBS)-nodo Posadas, CONICET-UNaM.

Specimen was collected under permit obtained from the Ministerio de Ecología y Recursos Naturales Renovables of Misiones Province (Collection Permit N° 002/2016).

External and cranial measurements were taken using a digital caliper of 0.01 mm sensitivity DIGIMESS© (Buenos Aires, Argentina) and body mass to the nearest 0.5 g using a spring scale PESOLA TM (Baar, Switzerland). Measurements follow ^{Simmons & Voss (1998)}, and ^{Giménez & Giannini (2016)} (Table 1) and include: weight (g); total length; tail length; hindfoot length; ear length; forearm length; tibia length; greatest length of skull excluding incisors; condyleincisive length; postorbital breadth; zygomatic breadth; braincase breadth; mastoid breadth; length of maxillary toothrow; length of palatal; breadth across upper molars; width across upper canines; length of mandible; length of mandibular toothrow.

Mitotic chromosomes were obtained from drip of cell suspension from bone marrow of one female specimen using the cell suspension technique, after in vivo injection of colchicine, following ^{Ford & Hamerton (1956)}. Twenty metaphase spreads were selected in order to perform the karyotype characterization. This characterization was performed using different staining techniques: conventional staining with Giemsa 10% (^{Ford & Hamerton 1956}), and differential staining with fluorochromes DAPI (4,6-diamidino2-phenylindole), and CMA3 (Chromomycin A3) to identify regions rich in AT and GC base pairs, respectively (^{Schweizer 1976}; ¹⁹⁸⁰). Diploid number (2n) and fundamental number (FN) were calculated on conventionally stained metaphases using photomicrographs obtained with an Olympus BX50 epifluorescent photomicroscope equipped with Sony Exwave-Had digital camera. Subsequently, ten metaphases were used for the construction of an idiogram using Drawid V0.26 software (^{Kirov et al. 2017}).

Table 1 External and cranial measurements (in mm) of the specimen of Glyphonycteris sylvestris (CM-LGE 200) reported here, compared to specimens reported by ^{Dias et al. (2003)} for Brazil; ^{Tirira et al. (2016)} for Ecuador; ^{Simmons & Voss (1998)} for French Guiana; ^{Morales-Martínez & Suárez-Castro (2014)} for Colombia; ^{Jones & Carter (1976)} for Panama; ^{Thomas (1896)}, ^{Andersen (1906)}, ^{Goodwing & Greenhall (1964)} for Costa Rica; ^{Goodwin & Greenhall (1961)}, ^{Goodwing & Greenhall (1964)} for Trinidad and Mexico. Institutional abbreviations are: CM-LGE, Mammal Collection of the Laboratorio de Genética Evolutiva of the Instituto de Biología Subtropical, Universidad Nacional de Misiones; ALP, Collections of Adriano Lúcio Peracchi from the Instituto de Biologia da Universidade Federal do Rio do Janeiro; QCAZ, Mammal Division of Museo de Zoología de la Pontificia Universidad Católica del Ecuador; AMNH, American Museum of Natural History; ICN, Instituto de Ciencias Naturales Universidad Nacional de Colombia; USNM, Natural Museum of Natural History; BM, British Museum (Natural History). Measurements are expressed as the mean standard deviation for specimens from Trinidad and Mexico. LMnT: Length of mandibular toothrow. LMxT: Lengthof maxillary toothrow. WAUC: Width across upper canines. 

RESULTS AND DISCUSSION

We obtained the first record of Glyphonycteris sylvestris for Argentina, which is an adult female without evidence of reproductive activity, captured at 19:30 in a ground-level mist net placed in a forest edge. Simultaneously, we also captured individuals of Artibeus lituratus, A. fimbriatus, Pygoderma bilabiatum, and Sturnira lilium. Our specimen shares the diagnostic external characters for this species as described by ^{Thomas (1896)}, ^{Simmons & Voss (1998)}, and ^{Williams & Genoways (2007)}: short and pointed ear, with the lower part of the outer edge ending in a low rounded lobule, and with a concavity on the posterior border near the tip, short fur (2.71 mm) at the external surface of the leading edge of the ear, ventral margin of narial horseshoe grading gradually into upper lip, interauricular band absent, chin with a pair of dermal pads arranged in a "V", with no central papilla, calcar shorter than hindfoot, soft and fine fur, smoky to dark gray and tricolored dorsally, and gray fur on its ventral side. However, unlike Thomas’ description, the dorsal fur of our specimen extends to wings and limbs, including pollical metacarpals and first finger. Forearm hairs extend to approximately up to one third of this bone, and a thin tuft of hairs cover the hind limb dorsally. Tail is short but it does not perforate the interfemoral membrane on its upper surface as in the type specimen (see ^{Thomas 1896}).

Like previous descriptions (e.g., ^{Goodwin & Greenhall 1961}; ^{Simmons & Voss 1998}) the skull of our specimen has two pairs of upper incisors, outer incisors reduced and hidden by cingulum of canine, low incisors trifid with crown height approximately equal to crown width, P3 and P4 subequal in crown height, P3 molariform with well-developed lingual cingulum and cusp, P4 with lingual cingulum of convex outline, edge not raised, and lingual cusp well developed (Fig. 2), upper canine much less than twice the height of the inner upper incisor, lower premolars aligned in row on mandible, coronoid process low with little slope along dorsal margin. The skull has a domed braincase, without sagittal crest, divided into single or two raised ridges, as in specimens from Trinidad and Ecuador (see ^{Goodwin & Greenhall 1961}; ^{Tirira et al. 2016}), rostrum and anterior orbital region of skull inflated, the dorsum of the rostrum is flat, the basisphenoid pits are deep, and mastoid breadth is shorter than zygomatic breadth. All measurements of our specimen fall within the range of size variation reported for specimens from Mexico, Trinidad and Panama (^{Goodwing & Greenhall 1964}), French Guiana (^{Simmons & Voss 1998}), Colombia (^{Morales-Martínez & Suárez-Castro 2014}), Ecuador (^{Tirira et al. 2016}) and Brazil (^{Dias et al. 2003}) (Table 1).

Fig. 2 Dorsal, ventral, lateral and frontal views of the skullof the female specimen of Glyphonycteris sylvestris (CMLGE200) captured in Parque Provincial El Piñalito. Notethe skull with domed braincase without sagittal crest, and the pair of incisors reduced and hidden by canine cingulum. See Table 1 for measurements. Scale = 5 mm. 

This is the southernmost record for this species, and the first from the Araucaria Moist Forest ecoregions. Despite the scarce records throughout its distribution, G. sylvestrisis widely distributed and occurs in different ecoregions, from 45 to 1476 m a.s.l (e.g., ^{Simmons & Voss 1998}; ^{Tirira et al. 2016}). This species was recorded in 12 ecoregions from the northern populations of South America, four ecoregions from the southern populations of the Atlantic forest (Table 2), and in a moist forest relict in the Cerrado, between the northern and southern populations (^{Felix et al. 2016}). This broad distribution suggests that G. sylvestris would be flexible in habitat use and highly tolerant to different environmental conditions. Populations from Cerrado might indicate a historical connection between both tropical populations, which should be tested in the future by biogeographical analyzes.

Table 2 Occurrences of Glyphonycteris sylvestris in the Atlantic Forest ecoregion. Each locality was assigned to an ecoregion according to the World Wildlife Fund (WWF). Locality number matches geographical location on map in Fig. 1. 

The fecal sample obtained from the Argentine specimen only contained plant material, including seeds of Piper aduncum. Glyphonycteris +Trinycteris has always been recovered as the sister clade of Carollia (see ^{Baker et al. 2012}; ^{Rojas et al. 2016}; ^{Amador et al. 2018}). Both Glyphonycteris and Trinycteris are insectivores that include plant material in their diets, whereas Carollia and successive branches in the phyllostomid tree (e.g., Rhinophyllinae and Stenodermatinae) are obligate frugivores. This suggests that frugivory strategies evolved from the feeding strategy of a basal insectivore (^{Baker et al. 2012}). Piper is the main resource of Carollia throughout the Neotropics (e.g., ^{Lobova et al. 2009}) and their mutualistic interaction was associated with the phylogeny and evolution of this basal linage of frugivorous bats (^{Sánchez & Giannini 2018}). Thus, the use of Piper by G. sylvestris could indicate that its diet may be related to a phylogenetic constraint due to its affinity to the Carolliinae clade (see ^{Sánchez & Giannini 2018}).

The diploid chromosome number of our Glyphonycteris sylvestris specimen is 2n = 22, and the fundamental number (including the chromosomal sex pair) FN = 42. Almost all chromosomes of the complement are biarmed, except for one pair of small acrocentric chromosomes (Fig. 3). Although this description agrees, in general, with that found in the literature for G. sylvestris, we observed a single pair of acrocentric chromosomes instead of two, as described by ^{Honeycutt et al. (1980)}.

Fig. 3 Female karyotype of Glyphonycteris sylvestris (CM-LGE 200) with 2n = 22; FN = 42. a) Giemsa-stained mitotic chromosomes; b) fluorochrome DAPI staining; c) fluorochrome Chromomicine A₃ staining. Scale = 10 µm. 

This implies a change in the fundamental number described previously for G. sylvestris, from FN = 40 to FN = 42. These discrepancies seem to be due to different criteria in the assignment of chromosomes, since we also detected one acrocentric chromosome in the figure of ^{Honeycutt et al. (1980)}. It was also considered by ^{Honeycutt et al. (1980)} that because the closely related species Glyphonycteris daviesi and Trinycteris nicefori have biarmed sexual chromosomes (^{Patton & Baker 1978}; ^{Honeycutt et al. 1980}), the sexual pair XX would be a biarmed element. Nevertheless, we cannot discuss that as we have not analyzed male specimens. The analysis of fluorochrome banding patterns in the chromosomal pericentromeric regions shows DAPI positive blocks in pairs 5 and 8, whereas pericentromeric regions stained with CMA3 are positive in pairs 1, 2, 3, 4 and 11. Moreover, the interstitial region of chromosomes shows DAPI and CMA3 positive blocks in pairs 1, 2, 3, 4 and 5 (Fig. 3 and 4).

Fig. 4 Haploid idiogramof Glyphonycteris sylvestris. Fluorochrome DAPI and Chromomycin A₃ positive bandsare shown in blue and green respectively. Grey corresponds toneutral regions. 

Among the phyllostomids, diploid chromosome numbers are highly variable, with karyotypes ranging from 2n = 14 to 2n = 46 chromosomes in 106 studied species (see ^{Sotero-Caio et al. 2017} and references therein). In Glyphonycterinae only the conventional karyotypes of Trinyicteris nicefori, Glyphonycteris davesi (both with 2n = 28, NF = 52), and G. sylvestris (2n = 22) are known (^{Patton & Baker 1978}; ^{Honeycutt et al. 1980}, this work). To date, a 2n = 22 karyotype seems to be unique within Glyphonycterinae, and although two species in the sister clade Carolliinae (Carollia benkeithi and C. castanea; ^{Solari & Baker 2006}) share this diploid number, both karyotypeshave clear differences in chromosomal morphology, and different rearrangements would be involved. The fluorochrome bands of G. sylvestris presented here are the first for Glyphonycterinae and the lack of data for related species limits comparisons at this time. Thus, it would be of high interest to develop further molecular cytogenetic studies thatcould help understand chromosomal evolution froma phylogenetic perspective in Glyphonycterinae.