Taxonomic re-evaluation of New World Eptesicus and Histiotus (Chiroptera: Vespertilionidae), with the description of a new genus

ABSTRACT Eptesicus Rafinesque, 1820 is widely distributed in the Old and New World (26 species), and Histiotus Gervais, 1856 is a South American endemic (11 species). Molecular phylogenies have recovered Eptesicus (sensu lato) as polyphyletic, with New World Eptesicus and the sister genus Histiotus in a paraphyletic American clade sister to Old World Eptesicus. Based on these phylogenetic reconstructions, authors have treated Histiotus as either a subgenus of Eptesicus or restricted Eptesicus to the New World species, treating Histiotus as a full genus, and using the name Cnephaeus Kaup, 1829 at the generic rank to comprise Old World Eptesicus. Based on recently published molecular studies, and on novel qualitative and quantitative morphological comparisons of representatives of Histiotus and New and Old World Eptesicus, we provide evidence for restricting the name Eptesicus to the species E. fuscus (Palisot de Beauvois, 1796) and E. guadeloupensis Genoways & Baker, 1975, allocating the remaining New World species under a new genus, keeping Histiotus as a full genus, and raising Cnephaeus to generic rank to comprise all Old World taxa currently under Eptesicus. This arrangement resolves the paraphyly of New World Eptesicus, and promotes taxonomic stability for Histiotus, which is a well-established genus of easily recognizable Neotropical bats and treated separate from Eptesicus by most authorities.

Initial taxonomic arrangements for the family were based mostly on cranial and dental morphology, later refined by bacular morphology and karyotypic and molecular data (Tate 1942, Hill and Harrison 1987, Volleth et al. 2006).Recently, molecular analyses have revealed that similar phenotypes evolved independently (Volleth et al. 2006, Roehrs et al. 2011, Amador et al. 2018, Ruedi et al. 2017).Phylogenetic reconstructions based mainly on molecular evidence have resulted in several taxa formerly treated as part of Eptesicus being rearranged in different genera, including the following: Vespadelus Troughton, 1943, a former synonym of Eptesicus recognized as a valid genus based on karyotypic and molecular evidence that place it within Vespertilionini, closely related to Pipistrellini (Volleth and Tidemann 1991, Volleth and Heller 1994, Amador et al. 2018); Neoromicia Roberts, 1926, formerly a subgenus of Eptesicus, and now elevated to generic level and placed within Vespertilionini based on karyotypic evidence (Volleth et al. 2001, Simmons 2005, Amador et al. 2018); Arielulus Hill and Harrison, 1987, a former synonym of Eptesicus later recognized as a valid genus and placed within Eptesicini based on karyotypic, molecular, and morphological evidence (Volleth and Heller 1994, Csorba and Lee 1999, Simmons 2005, Amador et al. 2018); Rhyneptesicus Bianchi, 1917, formerly a subgenus of Eptesicus, now elevated to generic level and placed as sister to Pipistrellini based on morphological and molecular evidence (Horáček et al. 2000, Juste et al. 2013, Amador et al. 2018); and Cassistrellus Ruedi et al., 2017, a genus recently described based on molecular evidence, which was previously related to Eptesicus based on morphology (Ruedi et al. 2017).
In addition to the former subgenera of Eptesicus mentioned above, additional subgenera, such as Rhinopterus Miller, 1916, andAmblyotus Kolenati, 1858, are still recognized by some authors despite the lack of molecular support (Tate 1942, Hill and Harrison 1987, Simmons 2005, Artyushin et al. 2018).Rhinopterus includes only Eptesicus floweri (de Winton, 1901) from northern Africa, and has not been assessed phylogenetically (Hill and Harrison 1987, Simmons 2005, Artyushin et al. 2018).Amblyotus includes the E. nilssonii species group and has been treated as either a subgenus or as a full genus.However, recent molecular studies by Artyushin et al. (2009Artyushin et al. ( , 2018) ) have brought into question the species composition of Amblyotus because their phylogenetic reconstruction does not recover all the species currently assigned to this genus as monophyletic.Also, E. nilssonii and E. serotinus mtDNA haplotypes are remarkably similar, which can be explained by some extent of introgression (Artyushin et al. 2009).
Qualitative and quantitative morphological external (N = 3) and skull (N = 16) characters (Tables 1 and 2) were selected following Davis (1966), Barquez et al. (1999), Miranda et al. (2006), Davis and Gardner (2008), Handley and Gardner (2008), Moratelli et al. (2013), Feijó et al. (2015), and Sánchez et al. (2019).We recorded qualitative and quantitative data from adults only, classified as such based on the closed epiphyses (q.v., Brunet-Rossini and Wilkinson 2009).Cranial measurements were taken under binocular microscopes with low magnification (usually 6×), to the nearest 0.01 mm, and their abbreviations are described in Table 1.A Multivariate Analysis of Variance (MANOVA) and subsequent pairwise Hotelling's tests were run in the software PAST 3.3 (Hammer et al. 2001) to test whether specimens representing the species we sampled of Eptesicus and Histiotus differ in skull dimensions.In all tests we calculated p-values with Bonferroni correction and considered them statistically significant for α ≤ 0.001.

Morphological comparisons
New World Eptesicus (NWE) and Old World Eptesicus (OWE) are morphologically similar to each other when compared to Histiotus, which can be easily distinguished from NWE and OWE based on both external and skull qualitative and quantitative characters (Figs 1, 2, Table 2).Also, Histiotus, NWE, and OWE can be distinguished based on quantitative characters (Wilk's lambda = 3.10E-194; F = 5.65E97, p < 0.001).In NWE and OWE, ears are small, narrow, separated, and have a total length ranging from 7.0 to 24.0 mm, usually less than 20 mm (N = 398; Fig. 1, Table 2).Comparatively, the ears of Histiotus are greatly enlarged, wide, have a total length ranging from 21.0 to 39.1 mm, usually longer than 22 mm (N = 143), and are fully or partially connected across the head by a band of variable development (Fig. 1, Table 2).The tragus in NWE and OWE is small, blunt, and straight to strongly curved inward, with total length ranging from 4.8 to 12.0 mm (N = 129); in Histiotus, the tragus is long and slightly to strongly curved outward, with total length ranging from 8.5 to 20.0 mm (N = 123).The eyes in NWE and OWE are reduced, whereas in Histiotus they are larger-for comparative purposes, eyes are smaller or about the same size as nostrils and lower canines in the first two, and larger in Histiotus.The muzzle is strongly inflated in NWE and OWE, but not conspicuously inflated in Histiotus (Fig. 1).NWE and OWE have a robust and broad skull (mean value of GLS/BAM = 2.4; N = 389), and the jugal is relatively slender in comparison with Histiotus.Histiotus has a narrower skull (mean value of GLS/BAM = 2.7; N = 94), with a well-developed postorbital process of the jugal.The auditory bullae are much larger in Histiotus than in NWE and OWE.The slope of the frontals is steeper than the nasals in NWE and OWE, but the nasal-frontal profile is relatively straight in Histiotus (Fig. 2), except in H. humboldti.NWE can be separated into two distinct morphological groups; one includes the large species E. fuscus and E. guadeloupensis and can be differentiated from smaller congeners by a set of morphological characters: fuscus and guadeloupensis are large, with forearm length usually close to 50 mm and not overlapping with most congeners (usually less than 45 mm in other congeners); length of dorsal fur usually larger than 10 mm (most congeners have dorsal fur length less than 9 mm); tragus broader and rounded at the distal extremity when compared to congeners; robust skull, with a straight lateral profile of the braincase in lateral view (comparatively delicate in congeners, with a more domed shape); and well-developed sagittal and lambdoidal crests (crests absent to well developed in congeners, but comparatively smaller).A similar situation can be found within OWE, with distinct morphological groups in which E. gobiensis and E. nilssonii are smaller, with GLS < 16.5 mm and forearm length < 44 mm; E. isabellinus, E. pachyomus, and E. serotinus are larger, with GLS > 18 mm and forearm length > 44 mm; and E. bottae is intermediate in size, with differences between subspecies.

Morphometric analyses
Morphological differences between Eptesicus and Histiotus are summarized in the principal component and canonical variate analyses (Figs 3-6).In the PCA, the first principal component (PC1) accounts for 91% of the total variation (Fig. 3A) and is strongly influenced by size, as observed in the loadings of all variables (Fig. 3B).The plot of PC1 scores also reflects the differences in size variation in each genus, with NWE including a group of small forms (E.diminutus, E. innoxius, E. furinalis) that overlap only with Histiotus humboldti.Eptesicus brasiliensis, E. ulapesensis, E. andinus, and E. taddeii are intermediate in size and overlap with most species of Histiotus.The large species Eptesicus fuscus and E. guadeloupensis do not overlap with Histiotus along PC1, except for a slight overlap between E. fuscus and H. macrotus.The plot of PC1 scores of OWE species also reflects size differences, with the small species E. nilssonii and E. gobiensis overlapping only with H. humboldti.The intermediate species E. bottae extensively overlaps with all the species of Histiotus except H. humboldti, and the large species E. isabellinus, E. serotinus, and E. pachyomus do not overlap with any of the Histiotus species along PC1.
The contrast between PC2 correlations, when the subset of measurements POB, BCB, WFH, GLS, CIL, CCL, MAB, and BAL is contrasted to those of ZYG, MAL, COH, BAM, BAC, MAN, and MAB suggests a differentiation pattern between most NWE, OWE, and Histiotus species when the effect of general size represented by PC1 is excluded.That contrast, except for the mandibular measurements MAL and COH, highlight differences related to cranial length and width that also characterize species of each genus.Exceptions to this pattern are E. gobiensis and E. nilssonii, which show similar relationships between these two subsets of measurements with those of most Histiotus species, despite the general size differences expressed along PC1; Eptesicus andinus, which extensively overlaps with H. velatus along PC2; and E. fuscus and E. guadeloupensis, which slightly overlap with Histiotus species along PC2, despite the general size differences expressed along PC1 (Fig. 3).
A similar pattern of differentiation between the two genera is revealed by the first CVA analysis (Fig. 4A).Representatives of Histiotus species are recovered as morphometrically distinct from representatives of NWE and OWE along the second axis (CV2), with scores for the last two widely distributed (Fig. 4A).CV1 accounts for 29% of the total variation and is clearly related to skull size, with the extremes of the CV1 continuum (Fig. 4) similar to those revealed along PC1 (Fig. 3A).Along the axis for CV2, which accounts for 18% of the variation between species, scores for representatives of all species of NWE and OWE, except E. fuscus, E. guadeloupensis, E. nilssonii, and E. gobiensis, do not overlap with those of Histiotus species.Nevertheless, despite the partial overlapping of scores in relation to either CV1 or CV2 axes, different trajectories of size-related differentiation along these two axes can be identified between the species samples of the genera.Because size alone does not permit clear characterization of each genus, we opted to conduct a "size-free" CVA in search of cranial patterns characterizing species distinctions between genera.In this analysis (Fig. 4B) almost all the species of NWE were retrieved in a single cluster, separate from the cluster for species of Histiotus, except Table 2. Summary of diagnostic quantitative characters of external and skull morphology between New World Eptesicus (NWE), Old World Eptesicus (OWE), and Histiotus species.Mean values are presented in millimeters for "forearm length", "ear length", "tragus length", and the ratio between "greatest length of skull" (GLS) and "breadth across molars" (BAM).The number of specimens analyzed (N) for each measurement is provided in parentheses.Vector correlations of original variables with the size-free CV1 (Fig. 5), which account for 23% of the size-independent variation among species, are like those revealed by the original variables and PC2, and summarize major craniometric differentiation patterns between NWE, OWE, and Histiotus species.Here again, the correlations of skull length measurements GLS, CCL, CIL, and BAL contrast with BAM, ZYG, and MAN to separate content in the genera.In parallel to the scores of PC2, the exceptions are represented by H. magellanicus, H. macrotus, H. velatus, E. gobiensis, E. nilssonii, E. diminutus, and E. innoxius, species that overlap in this size-free canonical function.Also as noted with respect to the PC2 distribution of scores, these species are clearly distinguishable otherwise based on skull size, with E. gobiensis, E. nilssonii, E. diminutus, and E. innoxius being the smallest species of Eptesicus, and H. magellanicus, H. macrotus, and H. velatus being the largest Histiotus.Previously mentioned qualitative and quantitative traits (Table 2) show these species to be morphologically distinct and easily assignable to their respective genus.
Patterns of distribution in the morphospace in the CVA that includes only New World species (Fig. 6A) are similar to that observed in Fig. 4A, with Histiotus species clustering together and not overlapping with any of the Eptesicus species included in the analysis.CV1 accounts for 34.1% of total variation, and, again, is strongly influenced by skull size, with species clusters located at the extremes of the axis reflecting the size variation within samples.Along CV1, Histiotus as a group overlaps only with E. andinus, E. brasiliensis, E. chiriquinus, E. fuscus, and E. taddeii.And the large species, E. fuscus and E. guadeloupensis, do not overlap with any of the remaining Eptesicus.Along CV2, which accounts for 19.1% of the total variation, Histiotus spp.partially overlaps only with the cluster that includes small Eptesicus, and the large species E. fuscus and E. guadeloupensis also partially overlap with the cluster of small Eptesicus.Vector correlations (Fig. 6B) also show a strong influence of size, with a small contrast between the set of measurements POB, GLS, BCB, CIL, WFH, CCL, BAL, and MAB in relation to COH, ZYG, BAM, BAC, MAN, MTL, MAL, and M1M3, which also suggests a differentiation pattern between taxa included in the analysis.Here, again, the clusters of species clearly show different trajectories over the morphospace.

DISCUSSION
Taxonomic decisions at the generic level should consider phylogenetic relationships, synapomorphic character status, phenotypic distinctiveness, and ecological factors in the formal recognition of distinctiveness among clades (Isler et al. 2013, Garbino 2015).Also, the zoological nomenclature "should convey evolutionary relationships, diversity, divergence, and the potential to clarify conservation priorities" (Baird et al. 2021: 285).
Treating Histiotus as a subgenus to resolve the non-monophyly of Eptesicus is based mainly on the argument that it would "bring less turmoil to the taxonomy of the Palearctic forms" (Juste et al. 2013: 448).Earlier, Hoofer and Van Den Bussche (2003: 35) argued that this approach would "underscore cranial and dental similarities between Histiotus and Eptesicus (sensu stricto) and de-emphasize the assumption that large ears were gained secondarily in Histiotus after the divergence between New and Old World Eptesicus."Although this arrangement can be useful to avoid controversy in the taxonomy of Old World Eptesicus, it does not fully reflect other convergences observed in the phenotypes within Vespertilionidae (e.g., long, and wide ears in Histiotus and Laephotis; Hoofer and Van Den Bussche 2003), and the systematic history of the lineages involved.A similar phylogenetic situation to that found within Eptesicus and Histiotus can be seen in the African genera Laephotis Thomas, 1901, andNeoromicia Roberts, 1926.Neoromicia was previously treated as a subgenus of Eptesicus but was raised to generic rank by Volleth et al. (2001).Subsequently, Kearney et al. (2002) suggested that Laephotis might nest within Neoromicia based on bacular structure.Laephotis is currently considered a full genus, with two sister groups: one that includes a long-eared specialized offshoot from an eptesicoid ancestor, closely related to Neoromicia based on molecular and morphological evidence, with topologies like those retrieved for American Eptesicus and Histiotus (Hill and Harrison 1987, Hoofer and Van Den Bussche 2003, Roehrs et al. 2011, Monadjem et al. 2021); and another group including short-eared species previously considered as Neoromicia (Monadjem et al. 2021).The inclusion of short-eared species within Laephotis, according to the authors, suggests that the size of ears may not be a good diagnostic character at genus level, however, the integration of morphological, and molecular evidences still led to the split of Neoromicia into four genera rather than the treatment of Laephotis as a subgenus of Neoromicia (see Monadjem et al. 2021).
Also paralleling the case of Laephotis versus Neoromicia, morphological distinction in ear size and shape between Histiotus and Eptesicus likely reflect ecological and behavioral traits that can play a key role in niche differentiation, with ecomorphs adaptive to different prey or feeding strategies (Giménez andGiannini 2017, Yi andLatch 2022).According to Yi and Latch (2022), these differences in ear shape and size could also have allowed Histiotus to avoid competition and co-occur with Eptesicus.
These adaptations are reflected in the echolocation patterns of these taxa, which are variable within Eptesicus (sensu lato) and Histiotus.Most species of both genera have frequency modulated calls that often include quasi constant frequency components, and call structures are as follows: (1) Old World Eptesicus have calls that sweep from ca. 52.0 to 21.0 kHz, with peak frequency of 22.7-35.8kHz, and call mean duration of 6.0-9.75 ms; (2) Histiotus have calls that sweep from ca. 53 to 15 kHz, with peak frequency of 22.4-32.1 kHz, and call mean duration of 1.3-5.0ms; (3) New World Eptesicus fuscus and Eptesicus guadeloupensis have calls that sweep from ca. 50.0 to 25.0 kHz, with peak frequency of ca. 30 kHz, and call mean duration of 3.0-10.0ms; and (4) New World small Neotropical Eptesicus have calls that sweep from ca. 80.0 to 33.0 kHz, with peak frequency of ca.32.0 to 66.0 kHz, and call duration of ca.3.0-9.4ms (Rodríguez and Mora 2006, Papadatou et al. 2008, Ossa et al. 2014, Horta et al. 2015, Arias-Aguilar et al. 2018, Cláudio 2019, Rodríguez-Posada et al. 2021) As previously discussed, and supported by our results, molecular data allow us to make two different taxonomic arrangements for Eptesicus (sensu lato) and Histiotus; but this approach alone was not sufficient to resolve the instability in the use of these names during the past two decades.The morphological data analyzed here provides robust additional evidence to resolve this issue in the light of a more integrative perspective.We considered the remarkable phenotypic discontinuity between Eptesicus (sensu lato) and Histiotus, allied to molecular data such as phylogenies, time of divergence and genetic distance between groups (see below), and echolocation data in deciding whether Histiotus and Cnephaeus should be treated at the generic or subgeneric rank (see Novaes et al. 2018); our evidence indicates that both should be treated as separate at the generic level, with Cnephaeus including Old World taxa previously treated as Eptesicus (adopted in the following paragraphs).In most phylogenetic reconstructions, E. fuscus has been recovered as sister to a clade including other American Eptesicus and Histiotus (Roehrs et al. 2010, 2011, Amador et al. 2018, Giménez et al. 2019).More recently, Yi and Latch (2022) conducted a broad phylogenetic study on Eptesicus and Histiotus based on UCEs, with extensive taxonomic and geographic sampling.Their findings indicate that E. fuscus and E. guadeloupensis are, in fact, more closely related to Histiotus than to remaining New World species of Eptesicus, which were recovered in a Neotropical clade sister to the fuscus + Histiotus clade.This suggests that the name Eptesicus should be restricted to E. fuscus and E. guadeloupensis, and the other New World species of Eptesicus allocated to a new genus.Additionally, we also recovered two distinct morphological groups within New World Eptesicus on both qualitative and morphometric analyses, which also support the split of New World species into two genera (Figs 4-6).Yi and Latch (2022) also recovered two clades within Neotropical Eptesicus (excluding fuscus and guadeloupensis), however, we did not find any morphological differences that support the split of this clade into two different genera.
Additional supporting evidence for this taxonomic arrangement is the time of divergence within Eptesicus (sensu lato) and Histiotus clades found by Yi and Latch (2022).According to Yi and Latch (2022), the split between Old and New World clades occurred around ca. 17 million years ago (mya), while the clade of E. fuscus and Histiotus diverged from the clade of Neotropical Eptesicus around ca. 14 mya (Fig. 7), which is similar to the time of divergence of other bat genera (see Amador et al. 2018, Baird et al. 2015, 2017, 2021).The intergeneric genetic distances between genera proposed here also support this arrangement; distances reported between the pairs of genera, based on cyt-b sequences, are: ( 1 Lastly, we recovered Cnephaeus nilssonii and C. gobiensis in a cluster separate from C. isabellinus, C. pachyomus, C. serotinus, and C. bottae in our morphometric analyses.This suggests robust divergence in skull morphology in both size and form, which could also be observed in our qualitative and quantitative results.C. nilssonii and C. gobiensis are included in the controversial subgenus Amblyotus, whose validity was questioned by recent molecular studies (see Artyushin et al. 2009Artyushin et al. , 2018)).Moreover, Yi and Latch (2022) also recovered Cnephaeus as two paraphyletic clades that roughly resemble the divergence between Eurasian and South African species.Their sampling of Cnephaeus species was, however, limited and no further taxonomic decisions can be made now.Here we resolved the paraphyly of New World Eptesicus, which includes the type species and has no-menclatural priority, and promoted the stability of Histiotus.Future studies including larger geographical and taxonomic samples of both morphological and molecular analyses are needed to solve the paraphyly within Cnephaeus, and to better evaluate the validity of Amblyotus.
Distribution: Neoeptesicus is a Neotropical genus, widely distributed across South America, and Central America.In Central America it is restricted to the southern mainland, with only the species Neoeptesicus furinalis and N. brasiliensis ranging northward into Mexico.Diagnosis: Neoeptesicus can be distinguished by the following characters: ears short (<20 mm) and not connected; postorbital process of the jugal weakly developed; eyes comparatively small, with diameter smaller or about the same size as height of lower canines and diameter of nostrils (Fig. 1); auditory bullae small, basioccipital region is larger than the width of each auditory bullae; length of dorsal fur usually short (averaging 8 mm); length of forearm ranging from 31.4-48.8mm (averaging 40.1 mm); tragus short and rounded, smooth and slender; GLS 12.9-18.0mm (averaging 15.8 mm); and MTL 4.5-7.0mm (averaging 5.8 mm).
Description: Neoeptesicus includes small-sized to medium species, with forearm length ranging from 31.4 to 48.8 mm.Dorsal fur usually short (LDF 4.5-13.1 mm), less than 10 mm for most species; hairs ranging from blackish and unicolored to strongly bicolored; dark basal color extending from 1/2 to almost the entire length of hairs; the contrast between bands ranges from faint to strong; tips of hairs ranging from blackish to golden orangish or yellowish.Ventral fur usually short (LVF 4.3-11.8mm), less than 8 mm for most species; hairs bicolored with dark brown bases from 1/2 to about 4/5 of hairs length, tips of hairs ranging from dark brown to nearly white.Wing membranes naked, usually dark brown.Plagiopatagium attached to the base of the toe.Dorsal surface of the uropatagium somewhat paler than wing membranes, nearly naked with short sparse hairs that do not extend beyond the knees.Ventral surface of the uropatagium brown and sparsely haired near the base of the tail.Ears well separated, medium sized, usually dark brown, and with rounded tips; tragus wider at the base, straight to slightly curved, long and rounded.Muzzle broad and inflated.
Skull delicate to robust; rostrum short, wide, and flattened, weakly sloping upwards to the braincase; braincase slightly wider than the rostrum.Posterior region of the braincase ranging from rounded, regular to straight, slightly projected upwards.Nasal opening V-to U-shaped.Frontal expanded laterally towards the orbit.Sagittal and lambdoidal crests weakly to well developed, connected or not; occipital helmet weakly-to well developed.Triangular, flattened bony plate where the sagittal and lambdoidal crests connect ranging from absent to well developed, the region where crests connect may be enlarged.Zygomatic arches thin and slightly widened medially.Basisphenoid pits absent.Palate extends well beyond molars, ending in a concave posterior edge, with a weakly-developed to small medial spine.
Dental formula is I 2/3, C 1/1, P 1/2, M 3/3 (×2) = 32.I 1 separated, ranging from slender and weakly bilobed to spatulate and strongly bilobed; wide and short to long and nar-row, with inner and outer cusps weakly-to well-developed.I 1 usually about three to four times the size of I 2 .I 1 usually not aligned to I 2 on a transversal axis of the skull.I 2 and C 1 usually separated by a small gap, C 1 with two slightly concave faces on the lingual region, and one slightly concave face on the labial region.P1 well-developed, reaching 1/2 of C 1 in height; P 1 in contact with C 1 and molars.M 1 and M 2 about the same size, almost square shaped, with W-shaped cusps.M 3 reduced, triangular, with only 3 cusps.I 1 to I 3 reduced, trilobed, and fully occupying the space between canines.The morphologically similar genera Neoeptesicus and Eptesicus can be distinguished using the following set of characters: length of dorsal fur, usually > 10 mm in Eptesicus and < 10 mm in Neoeptesicus; the only exceptions are N. andinus and N. chiriquinus, which can be differed from Eptesicus by the dark and nearly unicolored fur, with only the tips of dorsal hairs washed with lighter color.Length of ventral fur, usually > 8 mm in Eptesicus and < 8 mm in Neoeptesicus; the only exceptions, again, are N. andinus and N. chiriquinus.Color pattern, with strongly bicolored dorsal fur in Eptesicus, with the basal 1/2 of hairs dark brown and distal 1/2 brownish to golden brown; and ventral hairs also bicolored with the basal 2/3 of hairs dark brown, and distal 1/3 grayish yellow.In Neoeptesicus the pattern of dorsal and ventral fur color is highly variable, with most species present-ing a different color combination than found in Eptesicus; dorsal fur can vary from unicolored to strongly bicolored and ventral fur can vary from dark brown to whitish, the few species that have the same pattern of color as Eptesicus can be easily differentiated based on other characters such as length of fur, length of forearm, and cranial characters.Length of forearm usually > 42 mm in Eptesicus, and usually < 45 mm in Neoeptesicus; species that overlap in size can be separated on other diagnostic characters; the only species of Neoeptesicus that overlap in both size and distribution with Eptesicus are N. brasiliensis and N. chiriquinus, two species that can be easily distinguished from Eptesicus based on length and color of dorsal fur.The skull in Eptesicus is robust, with well-developed crests and helmet, the crests are always connected and the triangular plate where lambdoidal and sagittal crests converge is always absent; in Neoeptesicus the shape of skull and development of these cranial features is highly variable, and only some of the largest species show the pattern seen in Eptesicus; again these species can be easily distinguished from Eptesicus by the combination of the characters described above.The length of skull is > 17 mm in Eptesicus and usually < 18 mm in Neoeptesicus, with no overlap when compared together with forearm length.Eptesicus is mainly distributed across North and Central Americas, overlapping in distribution only with N. andinus, N. brasiliensis, N. chiriquinus, and N. furinalis.
See Table 3 for additional comparisons.

Eptesicus Rafinesque, 1820
Vespertilio: Diagnosis: Eptesicus can be distinguished by the following characters: ears short (<21 mm) and not connected; skull very robust, wide, and short; sagittal and lambdoidal crests well developed; postorbital process of the jugal weakly developed; eyes reduced, smaller or about the same size as lower canines and nostrils; auditory bullae small, the space between auditory bullae is larger than the width of each bullae; length of dorsal fur usually long (averaging 10.1 mm); length of forearm long, ranging from 40.1-54.0mm (averaging 46.6 mm); tragus short and rounded, smooth and broad; GLS 16.9-20.8mm (averaging 19.1 mm); and MTL 6.2-7.7 mm (averaging 7.1 mm).
Description: Eptesicus includes medium-sized to large bats.Dorsal fur long (LDF 8.1-13.7 mm) and woolly; hairs usually strongly bicolored with blackish bases extending about 1/2 of hair length, distal 1/2 of hairs golden to orangish brown; the juncture between bands is well marked.Ventral fur long (LVF 7.1-12.0mm, averaging 8.8 mm) and woolly; hairs bicolored with dark brown bases extending to about 2/3 of hairs length, distal 1/3 of hairs grayish yellow.Wing membranes naked, dark brown.Plagiopatagium attached to the base of the toe.Dorsal surface of the uropatagium almost naked, or with short sparse hairs that do not extend beyond the knees.Ventral surface of the uropatagium dark brown and sparsely haired along the base of the tail.Feet densely furred with short chestnut hairs.Ears well-separated, medium sized, dark brown, and with rounded tips; tragus wider at the base, straight to slightly curved, long, rounded, and broad.Muzzle broad and inflated.Skull robust; rostrum short, wide, and flattened, and weakly sloping upwards to the braincase; braincase slightly wider than the rostrum.Posterior region of the braincase straight, slightly projected upwards.Nasal opening V-shaped.Frontal expanded laterally towards the orbit.Sagittal and lambdoidal crests well developed and connected; occipital helmet well-developed.Triangular, flattened bony plate where the sagittal and lambdoidal crests connect absent, the region where crests connect may be enlarged.Zygomatic arches thin and slightly widened medially.Basisphenoid pits absent.Palate extends well beyond molars, ending in a concave posterior edge, with a weakly-to moderately-developed medial spine.
Dental formula is I 2/3, C 1/1, P 1/2, M 3/3 (×2) = 32.I 1 separated, weakly to strongly bilobed; shape varies from narrow and long to broad and short, with outer cusps weakly to well developed.I 1 about three to four times the size of I 2 .I 2 and C 1 separated by a small gap, C 1 with two slightly concave faces on the lingual region, and one slightly concave face on the labial region.P1 well developed, reaching 1/2 of C 1 in height; P 1 in contact with C 1 and molars.M 1 and M 2 about the same size, almost square shaped, with W-shaped cusps.M 3 reduced, triangular, with only 3 cusps.I 1 to I 3 reduced, trilobed, and filling all the space between canines.P 2 about three times P 1 in height.Molars have well developed cusps and decrease in size from M 1 to M 3 .
See Neoeptesicus gen.nov.account and Table 3 for comparisons.
Diagnosis: Histiotus can be distinguished by the following characters: ears long (>21 mm) and connected by a band of skin with variable development; skull delicate, narrow, and long; sagittal and lambdoidal crests weakly developed; postorbital process of the jugal well developed; eyes enlarged, larger than lower canines and nostrils; auditory bullae large, the space between auditory bullae is smaller than the width of each bullae; length of dorsal fur usually long (averaging 11.4 mm); length of forearm ranging from 44.0-52.6 mm (averaging 47.8 mm); tragus long, pointed and notched; GLS 17.4-19.8mm (averaging 18.4 mm); and MTL 5.6-7.1 mm (averaging 6.3 mm).
Description: Histiotus includes medium-sized to large species.Dorsal fur long (LDF 8.0-14.5 mm), usually more than 10 mm; hairs ranging from almost unicolored to strongly bicolored; dark bases length from 1/2 to almost the entire length of hairs; the contrast between bands ranges from faint to very strong; distal portion of hairs ranging from dark brown to golden orangish or yellowish.Ventral fur long (LVF 6.0-12.6 mm), usually more than 8 mm; hairs bicolored with dark brown bases extending from 1/2 to about 3/4 of hairs length, distal portion of hairs ranging from dark brown to nearly white.Wing membranes naked, usually dark brown.
Plagiopatagium attached to the base of the toe.Dorsal surface of the uropatagium somewhat paler than wing membranes, almost naked, with scarce hairs that do not extend beyond the knees.Ventral surface of the uropatagium usually brown, with scarce hairs close to the base of the tail.Ears greatly enlarged, connected by a band of skin of variable development, varying from oval to triangular; tragus wider at the base, straight to slightly curved, very long, notched and pointed.Muzzle broad and slightly inflated.
Skull delicate; rostrum short and flattened, usually straight in lateral profile; braincase slightly wider than the rostrum.Posterior region of the braincase rounded, regular.Nasal opening V-to U-shaped.Frontal expanded laterally towards the orbit.Sagittal and lambdoidal crests weakly developed, not connected, occipital helmet absent.Triangular, flattened bony plate where the sagittal and lambdoidal crests connect ranging from absent to well developed.Zygomatic arches thin and greatly widened medially.Basisphenoid pits absent.Palate extends well beyond molars, ending in a concave posterior edge, with a weakly-to mid-developed medial spine.
Dental formula is I 2/3, C 1/1, P 1/2, M 3/3 (×2) = 32.I 1 separated, spatulate and strongly bilobed; wide and short, with inner and outer cusps well-developed.I 1 usually about three to four times the size of I 2 .I 1 usually not aligned to I 2 on a transversal axis of the skull.I 2 and C 1 usually separated by a small gap, C 1 with two slightly concave faces on the lingual region, and one slightly concave face on the labial region.P1 well-developed, reaching 1/2 of C 1 in height; P 1 in contact with C 1 and molars.M 1 and M 2 about the same size, almost square shaped, W-shaped cusps.M 3 reduced, triangular, with only 3 cusps.I 1 to I 3 reduced, trilobed, and occupying all the space between canines.P 2 about three times P 1 in height.Molars have well developed cusps and decrease in size from M 1 to M 3 .
See Neoeptesicus gen.nov.account, and Table 3 for comparisons.

(
MNHM, Paris, France); American Museum of Natural History (AMNH, New York, New York, USA); Field Museum of Natural History (FMNH, Chicago, Illinois, USA); Louisiana State University, Museum of Natural Science (LSU, Baton Rouge, Louisiana, USA); National Museum of Natural History (USNM, Washington, D.C., USA); and Natural History Museum (BM, London, United Kingdom).See Appendix 1 for additional information about the examined specimens.

Figure 3 .
Figure 3. (A) Plot of multivariate individual scores of New World Eptesicus (Red), Old World Eptesicus (Blue), and Histiotus (Black) in the first two principal components.Analysis was performed using 16 cranial measurements.(B) Vector correlations of craniometric characters with the first two principal components eigenvectors.SeeTable 1 for abbreviations.

Figure 4 .
Figure 4. Plot of multivariate individual scores of New World Eptesicus (red), Old World Eptesicus (blue), and Histiotus (black) in the first two canonical variates: (A) canonical variates considering the effect of size; (B) "size-free" canonical variate using Mosimann shape variables.Analyses were performed using 16 cranial measurements.
Also, the decision to keep Old World forms within Eptesicus brings turmoil to the taxonomy of New World Eptesicus and Histiotus, since Eptesicus (sensu lato) must include Histiotus to keep Eptesicus monophyletic.Histiotus is morphologically distinct from Eptesicus and is a well-established genus that comprises a clade of easily recognizable Neotropical bats.Most authorities have treated Histiotus as separate from Eptesicus to avoid turmoil in the taxonomy of New World vesper bats (e.g.,Davis and Gardner 2008, Handley and Gardner 2008, Moratelli et al. 2019).Moreover, three new species were described recently within Histiotus: H. diaphanopterus, H. cadenai, and H. mochica(Feijó et al. 2015, Rodríguez-Posada et al. 2021, Velazco et al. 2021), and two new species described within Eptesicus: E. langeri and E. orinocensis(Acosta et al. 2021, Ramírez-Chaves et al. 2021).Different generic arrangements were used by the different research groups, thus contributing to the nomenclatural instability of both Eptesicus and Histiotus.

Figure 5 .
Figure 5. Vector correlations of craniometric characters with the first two eigenvectors of the canonical variates: (A) canonical variate considering the effect of size; (B) "size-free" canonical variate using Mosimann shape variables.SeeTable 1 for abbreviations.A B

Figure 6 .B
Figure 6.(A) plot of multivariate individual scores of New World Eptesicus (red and blue), and Histiotus (black) in the first two canonical variates; analysis was performed using 16 cranial measurements.(B) Vector correlations of craniometric characters with the first two canonical variates.SeeTable 1 for abbreviations.
P 2 about three times P 1 in height.Molars have well-developed cusps and decrease in size from M 1 to M 3 .Comparisons: Neoeptesicus most resembles Eptesicus, and both can be distinguished from Histiotus based on several characters.Ears in Histiotus are greatly enlarged, > 21 mm (21-39 mm, usually > 25 mm); in Neoeptesicus and Eptesicus ears are comparatively reduced, < 20 mm (8-20 mm, usually < 18 mm); the tragus in Histiotus is notched and long, with total length > 10 mm (10-20 mm); in Neoeptesicus and Eptesicus the tragus is smooth and short, > 12 mm (4.8-12 mm); there is no overlap when the length of tragus is analyzed together with ear and forearm length.The ears in Histiotus are connected by a membrane of variable development, while in Neoeptesicus and Eptesicus this membrane is absent.Eyes in Histiotus are larger than nostrils and lower canines, and smaller or about the same size as nostrils or canines in Neoeptesicus and Eptesicus.The skull in Histiotus is narrow and long (mean value of GLS/BAM = 2.7) when compared to Neoeptesicus and Eptesicus (mean value of GLS/BAM = 2.3); the auditory bullae are comparatively twice as large in Histiotus than in Neoeptesicus and Eptesicus; and Histiotus has a well-developed postorbital process of the jugal, which is weakly developed in Neoeptesicus and Eptesicus.

Table 1 .
Skull dimensions used to perform principal component and canonical variate analyses.Measurements were taken in millimeters.