Background and study regions

In mid-April 1999 we undertook a mammalian expedition to eastern Bolivia (Brooks et al. 2002). On the evening of 17 April 1999 a single mist net was erected in drizzling weather in the middle of a small patch of savanna surrounded by patches of forest at Estancia Patuju (Fig. 1; 17°37′04.9″S, 59°32′9.5″W; 220 m). A single bat (voucher CBF 6154) was netted—a specimen of the genus Micronycteris that could not be assigned to any known species previously reported for the region (e.g., Anderson 1997). The regional landscape was Cerrado, a semideciduous savanna composed of a mosaic of open grassland and dry forest (Redford and da Fonseca 1986) found predominately in central Brazil (Ratter et al. 1997). Cerrado is typified by low cover composed of grasses and small shrubs, and tree species that are at low densities and often have adaptations to water stress such as twisted trunks, deep roots, and hard, leathery leaves (Redford and da Fonseca 1986; Ibisch et al. 2002).

The specimen was caught in 1 of the 4 subecoregions of Bolivian Cerrado known as Cerrado Chiquitano, which is found in central and eastern Santa Cruz Department and is a continuation of the Brazilian Cerrado (Ibisch et al. 2003). Cerrado Chiquitano is 23,500 km² in size, ranges from 120 to 1,000 m above sea level, has an average annual temperature range of 21–27°C, and has a landscape dominated by plains and inselbergs with open and wooded savanna and fire-resistant forests that are short in height (Ibisch et al. 2003).

Once the specimen (CBF 6154) was determined to be something potentially unique, we put the call out to other chiropterologists working in Bolivia to be on the lookout for this Micronycteris, yet none were found. Then we caught what was then unbeknownst to be an additional specimen on 14 November 2005 (voucher MHNC-M 141) in a private reserve called Refugio Los Volcanes (Fig. 1; 18°06′42.5″S, 63°36′07.8″W; 1,060 m). This reserve is approximately 300 ha in size and adjacent to the southern border of Amboró National Park, directly on the transition from the humid inner tropics to the seasonally dry subtropics (Linares-Palomino et al. 2008). Although the site is located in the Yungas ecoregion, Los Volcanes is adjacent to the Tucuman–Bolivian Forest and Chaco Serrano ecoregions (Ibisch et al. 2003). The substrate in this area consists primarily of red sandstone that forms cliffs several hundred meters high that are intersected by narrow valleys (Linares-Palomino et al. 2008). Annual precipitation is approximately 120–150 cm with most of the rainfall from October–November through March–April, with high temporal variability (Linares-Palomino et al. 2008).

A 3rd specimen was collected on 18 February 2007 (voucher MHNC-M 157) from a house in the small town of Zurima (Fig. 1; 18°46′40.2″S, 65°07′40.3″W; 1,800 m). Zurima is located on the main road between the cities of Cochabamba and Sucre. The area is located in Inter-Andean Dry Forest, an ecoregion of 44,805 km² with an altitudinal range of 500–3,300 m above sea level, an average annual temperature range of 12–16°C, and a landscape dominated by moderately dissected valleys and small plains with largely disturbed or destroyed deciduous dry forest ranging 10–20 m in height (Ibisch et al. 2003). All specimens collected for this study were captured in accordance with animal welfare guidelines established by the American Society of Mammalogists (Sikes et al. 2011).

Specimens examined

The 3 specimens of the new species of Micronycteris were deposited in Bolivian natural history collections: Colección Boliviana de Fauna (CBF) in La Paz and Museo de Historia Natural Alcide d'Orbigny (MHNC) in Cochabamba. For comparison, 84 specimens of 10 species of Micronycteris were used for phylogenetic and morphometric analyses. Specimens examined (see Supporting Information S1, DOI:  10.1644/12-MAMM-A-259.S1 (10.1644_12-MAMM-A-259.S1.docx)) were deposited in the following natural history collections: American Museum of Natural History (AMNH), Carnegie Museum of Natural History (CMNH), Colección Boliviana de Fauna (CBF), Museo de Historia Natural Alcide d'Orbigny (MHNC), Museum of Southwestern Biology (MSB—catalog number, NK—tissue number), Museum of Texas Tech University (TTU—voucher number, TK—tissue number), Pontificia Universidad Católica del Ecuador (QCAZ), Royal Ontario Museum (ROM), and United States National Museum of Natural History (NMNH).

Morphological methods

Morphological characters and measurements were used to describe the new taxon and make comparisons with other Micronycteris specimens. Morphological terminology was based on Lira et al. (1994), Czaplewski and Morgan (2003), and Giannini et al. (2006). Measurements of study skins were taken with digital calipers to the nearest 0.01 mm following Lira et al. (1994) and Simmons and Voss (1998). Measurements of only adult specimens were taken by 1 person to ensure consistency. External measurements included total body length (TL), tail length (Tail), length of hind foot (HF), ear length (Ear), forearm length (FA), 3rd metacarpal length (Mc3), 4th metacarpal length (Mc4), thumb length (Thu), tibia length (Tib), calcar length (Cal), and dorsal hair length (DHL) at the shoulder region. Cranial measurements included greatest skull length (GSL), condylobasal length (CBL), zygomatic breadth (ZB), postorbital constriction width (POC), braincase breadth (BB), mastoid breadth (MB), maxillary toothrow length (MTL), greatest breadth across upper molars (BAM), and braincase height (BH).

Cranial measurements also were taken from other species of pale-bellied Micronycteris (sensu Simmons et al. [2002]—M. brosseti, M. minuta [includes M. homezi—Ochoa and Sánchez 2005], M. sanborni, and M. schmidtorum) that were available (see Supporting Information S1, DOI:  10.1644/12-MAMM-A-259.S1 (10.1644_12-MAMM-A-259.S1.docx)). Measurements of M. brosseti presented by Simmons and Voss (1998) and of a marginal record of M. sanborni from western Brazil presented by Ferreira Santos et al. (2010) also were used to make comparisons. Measurements presented in Simmons and Voss (1998), Simmons et al. (2002), Fonseca et al. (2007), and Larsen et al. (2011) were used to compare dark-bellied Micronycteris (sensu Simmons et al. 2002—M. buriri, M. giovanniae, M. hirsuta, M. matses, M. megalotis, and M. microtis) with the new taxon. Because of the small sample size of most taxa, comparisons were made using descriptive statistics and measurement ranges.

We performed a principal component analysis of 8 cranial characters (ZB was excluded) using PAST version 2.14 (Hammer et al. 2001). Only pale-bellied specimens (n = 43) of Micronycteris were compared, including M. brosseti (n = 1), M. minuta (n = 27), M. sanborni (n = 3), M. schmidtorum (n = 9), and Micronycteris sp. nov. (n = 3). Principal components (PCs) were calculated using the covariance matrix to preserve the information about relative scale among variables.

Molecular methods

Genomic DNA was extracted and sequenced for 5 specimens: Micronycteris sp. nov. (n = 3) and paratypes of M. sanborni (n = 2). DNA extractions were conducted using standard phenol–chloroform methods (Sambrook et al. 1989) on biopsies taken from the plagiopatagium of 3 skin vouchers (CBF 6154 and M. sanborni [CMNH 98915 and CMNH 98916]) and muscle preserved in ethanol (MHNC-M 141 and MNHC-M 157). Amplification of cytochrome-b was performed using primers LGL765 and LGL766 (Bickham et al. 2004). Reaction volumes of 25 μl included approximately 200 ng of DNA, 0.12 μM of each primer, 1.5 mM of MgCl₂, 0.012 mM of deoxynucleoside triphosphates, 1X reaction buffer, 0.32 mg/ml of bovine serum albumin, and 0.625 U of Taq DNA polymerase (Promega Corporation, Madison, Wisconsin). Thermal cycling conditions were 94°C for 2 min followed by 30–34 cycles of denaturation at 94°C for 45 s, annealing at 47°C for 1 min, extension at 72°C for 1 min 15 s, followed by a final extension of 72°C for 10 min. When necessary, further optimization was performed with alternate DNA template quantities. Polymerase chain reation amplicons were purified using QIAquick PCR Purification Kits (Qiagen Inc., Chatsworth, California) and sequenced using ABI Big Dye chemistry version 3.1 and an ABI 3100–Avant Genetic Analyzer (PE Applied Biosystems, Foster City, California). For all 3 specimens of Micronycteris sp. nov., the entire cytochrome-b gene of 1,140 base pairs (bp) was sequenced by using amplification primers in addition to 4 internal primers: ART16 (Larsen et al. 2007), G1L and G7L (Hoffmann and Baker 2001), and MVZ04 (Smith and Patton 1993). Sequences were manually checked and aligned using Sequencher version 4.9 (Gene Codes Corporation, Ann Arbor, Michigan).

Amplification of cytochrome-b from paratypes of M. sanborni required the design of primers encompassing approximately 250-bp regions. These primers were Micro_First_F (5′ ACC-CTC-AAG-CCT-GAC-ATC-CT 3′), Micro_First_R (5′ TAC-AGG-CCT-CGG-CCT-ACA-T 3′), Micro_Mid_F (5′ CCA-CCG-CAT-TTA-TGG-GTT-AC 3′), and Micro_Mid_R (5′ CGT-AGG-GTT-GTT-GGA-TCC-TG 3′). Primers were designed to be exact matches for Micronycteris. We tested multiple manufacturers' polymerases for amplification success because DNA yield was very low for these samples. Phire polymerase (Fisher Scientific, Pittsburg, Pennsylvania) was subsequently used based on these comparisons. Reaction conditions were 12.5 μM of each deoxynucleoside triphosphate, 1 μl of Phire polymerase, manufacturer buffer, and 25 pM of each primer in a total volume of 50 μl. Thermal conditions were 98°C for 1 min 45 s, followed by 40 cycles of 98°C for 20 s, 55°C for 20 s, 72°C for 40 s, followed by a final extension of 72°C for 10 min. Amplicons were sized on 1.5% agarose gels and purified and sequenced as described above using the amplification primers. Two partial cytochrome-b fragments were obtained for CMNH 98915 (241 and 212 bp) and CMNH 98916 (226 and 254 bp).

Additionally, 45 complete cytochrome-b sequences from 9 species of Micronycteris were compiled from GenBank (Porter et al. 2007; Larsen et al. 2011) and used for phylogenetic comparisons (list of specimens in Supporting Information S1, DOI:  10.1644/12-MAMM-A-259.S1 (10.1644_12-MAMM-A-259.S1.docx)). Maximum-likelihood and maximum-parsimony analyses were performed using MEGA version 5 (Tamura et al. 2011). The HKY+G+I model was selected using the Akaike information criterion in MEGA 5 (Tamura et al. 2011). Genetic distance values were estimated using the Kimura 2-parameter model (Kimura 1980). Lampronycteris and Macrotus were used as outgroups for all phylogenetic analyses, because previous studies indicate these genera are outgroups with respect to Micronycteris (Simmons 1996; Simmons and Voss 1998; Simmons et al. 2002; Baker et al. 2003). The partial sequences obtained from the 2 specimens of M. sanborni were analyzed independently to verify their position in the phylogenetic tree. Once confirmed, we compiled the 4 fragments into 1 sequence of 495 bp (see Supporting Information S2, DOI:  10.1644/12-MAMM-A-259.S2) to reduce the number of missing data and to increase the power of the analyses. A Bayesian analysis was performed on the complete data set using MrBayes version 3.2 software (Ronquist et al. 2012), running 5 × 10⁶ generations with 1 cold and 3 incrementally heated Markov chains, random starting trees for each chain, and trees sampled every 100 generations.

Principal component analysis

Principal components 1 and 2 accounted for 92.7% of the total variation (PC 1 = 57.3% and PC 2 = 35.4%) and were correlated with maxillary toothrow length and condylobasal length (Fig. 2; Table 1). The new species formed a cluster separate from M. schmidtorum, M. minuta, and M. brosseti along PC 1 and also separate from M. sanborni along PC 2 (Fig. 2). These results indicate that the new species of Micronycteris is morphometrically distinct from its congeners (Fig. 2).

Phylogenetic analyses

The 3 Bolivian specimens and their closest relatives (M. minuta, M. schmidtorum, and M. sanborni) showed fixed nucleotide differences that resulted in 23 amino acid changes in the cytochrome-b gene (Table 2). Three of the 23 amino acid changes were unique to Micronycteris sp. nov. and are located at positions 67, 181, and 316 out of the 380 amino acids in the cytochrome-b gene. At position 67, threonine in M. minuta, M. sanborni, and M. schmidtorum was replaced by alanine in Micronycteris sp. nov. At position 181, phenylalanine in M. minuta, M. sanborni, and M. schmidtorum was replaced by leucine in Micronycteris sp. nov. At position 316, a polymorphism of methionine or threonine in M. minuta and methionine in M. schmidtorum was replaced by alanine in Micronycteris sp. nov. (Table 2).

Alignment of 50 cytochrome-b sequences was unequivocal and without internal stop codons. Of the 1,140 characters, 678 were conserved and 373 were parsimony informative across the entire data set. Maximum-likelihood analysis resulted in 1 tree with a ln-likelihood of −7,645.41 and maximum-parsimony analyses resulted in 25 most-parsimonious trees (length = 1,334; consistency index = 0.46, retention index = 0.78, composite index = 0.37). The 3 Bolivian Micronycteris formed a statistically supported clade in all phylogenetic analyses (Fig. 3). The sister relationship to M. sanborni is not supported (Fig. 3) and the maximum-likelihood tree grouped M. sanborni with M. minuta and M. schmidtorum. Genetic distance values between clades ranged from 17.89% (M. hirsuta versus M. minuta clade 1) to 2.75% (M. buriri versus M. megalotis clade 3; Table 3). The intraspecific genetic distance value for the Bolivian Micronycteris was 1.3% and the lowest genetic distance value separating the new taxon from any other species within the genus was 5.3% (M. sanborni, based on 495 bp; Table 3).

Holotype.—Voucher MHNC-M 157; adult male; standard skin and skull deposited at the Museo de Historia Natural Alcide d'Orbigny (Cochabamba, Bolivia). Collected on 14 February 2007 by L. Siles and A. Muñoz. Prepared by L. Siles, field number LSM 146. The skin and skull are well preserved, with the exception of broken zygomatic arches and separated mandibles. External measurements (mm) were: TL, 48.50; Tail, 7.00; HF, 11.71; Ear, 18.27; FA, 36.70; Mc3, 29.30; Mc4, 30.68; Thu, 8.11; Tib, 13.24; and Cal, 9.48. Cranial measurements (mm) were: GSL, 17.85; CBL, 15.92; ZB, 8.18; POC, 3.91; MB, 8.4; MTL, 6.25; BAM, 5.32; BB, 7.47; and BH, 6.52. This individual was initially identified as M. aff. sanborni (MHNC-M records, in litt.).

Type locality.—Bolivia: Chuquisaca Department, Provincia Oropeza, Zurima, 33 km northeast of Sucre (18°46′40.2″S, 65°07′40.3″W). Collected at 1,800 m above sea level.

Paratypes.—Two additional specimens were collected from Bolivia and based on morphological and genetic data are designated as paratypes. The 1st paratype is voucher CBF 6154; adult female; standard skin and skull deposited at the Colección Boliviana de Fauna (La Paz, Bolivia). Collected on 17 April 1999 by H. Aranibar, R. J. Vargas M., and J. M. Rojas from Estancia Patuju, 370 km east of Santa Cruz de la Sierra (17°37′04.9″S, 59°32′9.5″W; 220 m), Provincia Chiquitos, Santa Cruz Department. Prepared by J. M. Rojas (field number JMR 319). The skin and mandible are well preserved; the skull has broken palatine bones, mastoid bones, zygomatic arches, and auditory bullae. This specimen was originally identified as M. sanborni (Brooks et al. 2002).

The 2nd paratype is voucher MHNC-M 141; adult male; standard skin and skull deposited at the Museo de Historia Natural Alcide d'Orbigny (Cochabamba, Bolivia). Collected on 18 November 2005 by L. Siles from Refugio Los Volcanes (18°06′42.5″S, 63°36′07.8″W; 1,060 m), Provincia Florida, Santa Cruz Department. Prepared by L. Siles (field number LSM 131). The skin, skull, and mandibles are well preserved with the exception of 1 broken zygomatic arch. This specimen was initially identified as M. sanborni (MHNC-M records, included in Aguirre and Terán [2007]). Measurements of paratypes are presented in Tables 4 and 5.

Distribution.—The specimens are known from 3 localities in Bolivia spanning approximately 700 km (Fig. 1). The 3 collecting localities represent distinct ecoregions: Inter-Andean Dry Forest, Cerrado, and Yungas (sensu Ibisch et al. 2003) ranging from 220 to 1,800 m above sea level.

Etymology.—This species is named in honor of Terry Lamon Yates (1950–2007) for his pivotal contributions to the knowledge of Bolivian mammals, training Bolivian biologists, and starting collaborations that strengthened mammalian research and shaped current science and field biology in Bolivia. We suggest that the common name of this species be the Yates's big-eared bat.

Diagnosis.—Micronycteris yatesi is distinguished from its congeners by a combination of external and craniodental features. The ventral hair on the throat and sternal region in M. yatesi is white, whereas the abdominal area presents pale buff coloration (Fig. 4). The skull of M. yatesi presents palatine bones that are short and parallel with a pointed arch shape anteriorly (Fig. 5). The sutura palatomaxillaris is located between M2 and M3 (Fig. 5). Medial upper incisors are long, unilobate, and projected slightly upward.

Micronycteris yatesi also can be distinguished from its congeners based on DNA sequence data, which present nucleotide differences that result in unique amino acids in the cytochrome-b protein at positions 67, 181, and 316 (alanine, leucine, and alanine, respectively; Table 2). Phylogenetically, the cytochrome-b sequence positions M. yatesi in a clade that is divergent and statistically supported from other members of the genus studied thus far (Fig. 3).

Description.—Micronycteris yatesi is a medium-sized bat (forearm 34.54–36.7 mm). The dorsal fur is long (8–10 mm in the shoulder region) and bicolored with a white base that composes approximately one-half of each hair (white portion = 4.3–5.6 mm). The terminal portions of the dorsal fur are olive brown (Ridgway 1912:plate XL) in color in the holotype (MHNC-M 157), Prout's brown (Ridgway 1912:plate XV) in MHNC-M 141, and tawny (Ridgway 1912:plate XV) in CBF 6154. MHNC-M 157 and MHNC-M 141 have the ventral hair on the throat and sternal region completely white, whereas the abdominal area presents pale buff coloration. CBF 6154 has dorsal and ventral fur with a white base that composes one-third of each hair, and the ventral fur presents pale yellowish coloration. Fur on external surface of leading edge of the pinna (sensu Simmons 1996) is dense and of variable length; MHNC-M 157 and CBF 6154 have short hair (≤3.5 mm), and MHNC-M 141 has long hair (5.5 mm). The ears are large (15.50–18.27 mm) with rounded tips and a high, notched band connects their bases. Interauricular membrane is deeply notched with triangular flaps (observations only on dry skins). Hind feet are hairy. The calcar is shorter than the foot in MHNC-M 157 and MHNC-M 141, and similar in size to the foot in CBF 6154 (Table 4). Uropatagium and wing membranes are naked.

Rostrum is elongated; premaxillae are short and wide; maxillae are inflated over the region of the 2nd premolar and 1st molar; nasal bones are slightly inflated (Fig. 6). CBF 6154 has a less inflated rostrum than the other 2 specimens. Incisive foramina are triangular. Infraorbital canal is narrow and deep. Sagittal crest and lambdoidal crest are absent. Zygomatic arch is slender (complete arch in MHNC-M 141). Paraoccipital processes are not well developed and do not surpass the occipital condyles. Foramen magnum is oval. Basisphenoid pits are deep, separated by a well-developed septum. Basioccipital bone is narrow. There are 2 pairs of foraminae ovale; the anterior is larger than the posterior. Palatine bones are short with a pointed arch shape anteriorly (Fig. 5). The sutura palatomaxillaris is located between M2 and M3. Postpalatal extension is narrow over the mesopterygoid fossa and has a U-shaped posterior margin.

Medial upper incisors (I1) are large, unilobate, and projected slightly upward (more so in CBF 6154 than the other 2 specimens). Lateral upper incisors (I2) are unilobate, convergent, and small, less than one-half the size of the inner upper incisors. A gap is present between I2 and the canines. Upper canines are slender with cingula well developed, especially the antero- and posterolingual cingular styles. CBF 6154 has slightly smaller canines than the other 2 specimens. P4 is slightly larger than P3 in anteroposterior length. In occlusal view P3 is rectangular. The antero- and posterolingual cingula of P4 are well developed. M1 and M2 have a similar size with a square outline in occlusal view. Parastyle of M2 and M3 have a distinct accessory cusp that projects anteriorly. M2 parastyle is the most developed of all molar styles. The metacone of M2 and M3 is taller than the paracone, and the protocone is taller than the hypocone.

Lower incisors are small and bilobed with lobes parallel to each other. Lower canines are robust with cingula well developed. Crown height of lower premolars varies; p3 is much smaller than p2 and p4. Coronoid process is low; angular process is well developed and extends out to the same level of the mandibular condyle.

Taxonomic remarks

Phylogenetic analyses indicate that M. yatesi is within the subgenus Schizonycteris (sensu Porter et al. 2007). Available names for new species of the genus Micronycteris include M. typica and M. mexicana; however, Larsen et al. (2011) established that those names would be best applied to M. megalotis clades 1 or 3, and M. megalotis clade 2, respectively (Fig. 3). Another available name is M. hypoleuca Allen, 1900, a white-bellied Micronycteris that was synonymized with M. minuta by Andersen (1906) due to the lack of diagnostic characters other than a completely white belly. Because the description of M. hypoleuca is based on a single specimen from the Santa Marta region of Colombia, this name would be available for M. minuta clades 2 or 3 (Fig. 3).

Morphological comparisons

Micronycteris yatesi is distinguished from its congeners by a combination of external, cranial, and dental characters, as well as size variation. Dorsal pelage is bicolored with white bases and tawny to brown tips, a character that is shared with all the members of the genus (Simmons and Voss 1998; Simmons et al. 2002; Fonseca et al. 2007; Larsen et al. 2011). The ventral fur in M. yatesi is paler than the dorsal fur, a character that is shared with M. sanborni, M. minuta, M. schmidtorum, and M. brosseti. The color of the ventral fur in M. yatesi is completely white on the throat and sternal region, whereas the abdominal area presents pale buff coloration (Fig. 4). M. sanborni is the only pale-bellied species with a bright white ventral region, including throat, sternal, and abdominal areas (Fig. 4). All other pale-bellied Micronycteris have pale gray or pale buff underparts (i.e., M. brosseti, M. minuta, and M. schmidtorum—Simmons and Voss 1998).

Although the relative proportion of the calcar and hind foot varies among species of Micronycteris and is used as a diagnostic character (Simmons 1996; Williams and Genoways 2007), its usefulness comes into question because it relies heavily on the subjective perception of the calcar being longer or shorter than the foot (Ferreira Santos et al. 2010), and also on preparation methods (Simmons 1996), which can deform the structure. In M. yatesi the calcar is shorter than the foot in 2 specimens, and 0.2 mm longer in a 3rd specimen. Although Simmons (1996) established that M. sanborni has calcar and foot of similar size, of the 3 specimens we examined the calcar was shorter than the foot (Table 4). We report these measurements here, but based on our observations we cannot determine which category should be assigned to M. yatesi and M. sanborni.

When comparing external and cranial measurements with the dark-bellied members of the genus (Tables 4–6), M. yatesi was smaller than M. buriri, M. giovanniae, M. hirsuta, and M. matses in most measurements, with overlapping values of hind foot and ear lengths for all 4 of these species. Other overlapping values include forearm length and postorbital constriction with M. buriri. M. yatesi has a larger thumb length compared to M. giovanniae. M. yatesi was of similar size to M. megalotis and M. microtis, with most external and cranial measurements overlapping. Exceptions include greater maxillary toothrow length and braincase height in M. megalotis and M. microtis, and a larger thumb length in M. megalotis.

Most external measurements of M. yatesi overlapped with the pale-bellied M. schmidtorum, M. brosseti, and M. minuta (Tables 4 and 6). Exceptions included a larger thumb in M. schmidtorum, a smaller calcar in M. minuta, and a smaller forearm length in M. brosseti. External measurements between M. yatesi and M. sanborni overlap in body length, tail, metacarpals 3 and 4, tibia, and calcar. M. sanborni is smaller than M. yatesi in hind foot, forearm, and thumb length, and is only larger in ear length.

Comparing cranial measurements among pale-bellied Micronycteris (Tables 5 and 6), M. yatesi completely overlaps all measurements with M. minuta. Compared to M. brosseti most measurements overlap except greatest length of skull (smaller in M. brosseti), zygomatic breadth, and breadth across upper molars (both larger in M. yatesi). M. schmidtorum presented greater values of condylobasal length, postorbital constriction, braincase breadth, and braincase height; all other measurements overlapped. M. sanborni was smaller than M. yatesi in greatest length of skull, condylobasal length, braincase breadth, and maxillary toothrow length; with overlapping values of postorbital constriction, mastoid breadth, breadth across upper molars, and braincase height. Based on our observations it seems that maxillary toothrow and condylobasal lengths will be very important in discriminating M. yatesi from other similar species in the genus, an assessment that also is supported by the results of the principal component analysis (Fig. 2; Table 1).

Based on the phylogenetic analysis (Fig. 3), we present a more detailed comparison of the morphology between M. yatesi and specimens of the closely related M. sanborni, M. schmidtorum, and M. minuta. The rostrum is elongated in M. yatesi, M. schmidtorum, and M. minuta, whereas M. sanborni has a relatively short rostrum. Palatine bones in M. yatesi are short and parallel with a pointed arch shape anteriorly (Fig. 5). In contrast M. sanborni, M. schmidtorum, and M. minuta have longer, convergent, and fusiform-shaped palatine bones (Fig. 5). The sutura palatomaxillaris is located between M2 and M3 in M. yatesi, whereas in M. sanborni, M. schmidtorum, and M. minuta this sutura is located between M1 and M2 (Fig. 5). The shape and size of the palatine bones described for M. yatesi are unique to this taxon and not found in any other species of the genus.

Like other members of the genus, M. yatesi has a dental formula of i 2/2, c 1/1, p 2/3, m 3/3, total 34. Medial upper incisors are similar in shape and size in M. minuta, M. sanborni, and M. yatesi; all specimens examined from these taxa presented unilobate I1, with 2 exceptions (TTU 48093 and TTU 33282 from Venezuela). Lateral upper incisors are similar in size in M. minuta, M. sanborni, and M. yatesi; however, the shape and presence of lobes is variable in M. minuta. A gap between I2 and the canines is present only in M. yatesi and M. sanborni and varies in width. The upper canines, premolars, and molars are similar in the specimens examined of M. minuta, M. sanborni, and M. yatesi. Lower incisors are round in M. sanborni and M. yatesi, whereas M. minuta has more rectangular incisors. The rest of the lower dentition and mandibular characters are similar in M. minuta, M. sanborni, and M. yatesi.

A specimen of M. sanborni was reported from the western region of the Brazilian Cerrado in the state of Mato Grosso do Sul, a state adjacent to southeastern Bolivia (Ferreira Santos et al. 2010). Although some of the measurements of this specimen are within the range of M. sanborni, maxillary toothrow length and condylobasal length seem to be closer to the range of M. yatesi (Table 5). Ferreira Santos et al. (2010) also mention that the specimen has a gap between the lateral upper incisors and the canines, a character that defines both M. sanborni and M. yatesi. We recommend that all specimens from this region be examined with respect to the diagnostic characters presented here, because they may represent additional specimens of M. yatesi.

Molecular data

Specimens of M. sanborni, clades 1–3 of M. minuta, and M. schmidtorum are most closely related to M. yatesi (Fig. 3); therefore, we restrict discussion of molecular comparisons to these groups. The cytochrome-b gene analyses statistically support M. yatesi being genetically divergent from its closely related congeners. The genetic distance value (Table 3) separating M. yatesi from M. sanborni is 5.3%, a value that is typical for mammalian species recognized by classical morphology and is interpreted as indicative of sufficient isolation for 2 clades to speciate by an allopatric model of speciation involving the Bateson-Dobzhansky-Muller process (Bradley and Baker 2001; Baker and Bradley 2006). The genetic distance separating M. yatesi from M. schmidtorum is 9.5%. Values separating M. yatesi from members of M. minuta clades range from 9.2% (eastern Ecuador M. minuta) to 10.4% (Guyana M. minuta). Although values between 2% and 11% genetic divergence may indicate a high probability of conspecific populations (Bradley and Baker 2001), comparable values of genetic distance are found between recognized species of Micronycteris (including sister species) and of other genera of phyllostomid bats (e.g., Carollia, Glossophaga, and Mesophylla [Baker and Bradley 2006]; Artibeus [Larsen et al. 2007]; and Dermanura [Solari et al. 2009]). Furthermore, in this study we used these molecular data coupled with morphometric data and qualitative morphological characters. This combination of data sets is desirable to avoid unnecessary taxonomic splitting and makes our assessment conclusive.

Final remarks and conservation

The molecular data analyzed provide statistical support for recognizing M. yatesi as distinct from its congeners. Moreover, discrete external and cranial morphological characters discriminate this species from other congeners. Additionally we found evidence that some cranial measurements can be used to discriminate M. yatesi, but a larger sample size is needed to provide statistical support. The number of bat species recognized for Bolivia does not increase because M. yatesi replaces the record of M. sanborni. As far as we can ascertain, M. sanborni is restricted to Brazil (Simmons 1996) and M. yatesi is restricted to Bolivia.

Special attention should be given to records of pale-bellied Micronycteris from the ecoregions where M. yatesi was recorded; for example, Peñaranda and Pérez-Zubieta (2010) mention 2 records of M. minuta in the valleys of Cochabamba. M. yatesi constitutes the only species of bat thus far to be described from Bolivian specimens only. As for the specimen reported from Mato Grosso do Sul, Brazil, based on the information provided by Ferreira Santos et al. (2010) we cannot determine whether it is M. sanborni or an additional specimen of M. yatesi. We recommend that the specimen be reanalyzed under the evidence presented herein. Until more specimens are found in neighboring countries and for conservation planning purposes, we consider this species Bolivia's 1st endemic bat of the more than 130 species occurring in that country (Aguirre et al. 2010a).

In terms of conservation status, 7 species of threatened bats currently are listed in the Bolivian Red List of Vertebrates (Tarifa and Aguirre 2009) and Bat Action Plan (Aguirre et al. 2010b). An additional 5 are considered Near Threatened and 23 species are Data Deficient (Tarifa and Aguirre 2009). Only 1 species of Micronycteris (M. schmidtorum) is currently listed (but see the introduction regarding validity of this record) and its status is considered Data Deficient (Tarifa and Aguirre 2009; Aguirre et al. 2010b). We must consider M. yatesi Data Deficient until more long-term data are obtained. With only 3 known specimens captured over the last 15 years it is impossible to assess population stability, which is inherent to applying IUCN Red List criteria (International Union for Conservation of Nature and Natural Resources 2011). Nonetheless, of the thousands of net-hours and > 2,500 individuals netted in over a decade of fieldwork we were only able to obtain 3 individuals of M. yatesi, suggesting it is an extremely rare species in terms of relative abundance compared to more common taxa.

The Yungas, Cerrado, and Chiquitano Dry Forest ecoregions harbor most of the threatened species of Bolivian bats (Tarifa and Aguirre 2009; Aguirre et al. 2010b), and the paratypes were collected in Yungas (male MHNC-M 141) and Cerrado near Chiquitano (female CBF 6154). The serendipitous events leading to the initial discovery of M. yatesi are a paradox, because the site was not even targeted for sampling; inclement weather postponed our field team from visiting designated sites in Chiquitano forest and Pantanal habitats. Nonetheless, 9 (41%) of the 22 mammalian species accounted for at these sites are considered threatened enough to warrant placement on the Bolivian Red List of mammals (Brooks et al. 2002). Thus, we wish to highlight the conservation potential of this region.