We report the discovery of a new species of doglike bat (Peropteryx) from the lowland Amazonian forests of Ecuador and Peru. It has transparent wing membranes that are faintly tinged brown with pale-brown arms and digits; ears that are separated on the forehead; and a skull with small, shallow pterygoid pits that are anterolateral to an undivided basisphenoid pit and that are separated by a mesopterygoid extension. These characters distinguish the new species from morphologically similar species with which it was previously confused (P. leucoptera and P. macrotis). A molecular phylogenetic analysis of unlinked loci from each of the four genetic transmission systems of mammals (mitochondrial, nuclear-autosomal, X, and Y chromosomes) independently corroborated the placement of the new species as the sister taxon to a clade that includes P. kappleri, P. macrotis, and P. trinitatis; the basal lineage for the genus is P. leucoptera. This phylogeny suggests that transparent wings (sometimes described as “white” but actually lacking pigment), the traditional character used to diagnose Peronymus, is not a unique synapomorphy. Furthermore, based on a molecular dating analysis, the depth of divergence of Peropteryx is equivalent to that of another New World emballonurid genus (Balantiopteryx). Therefore, Peronymus does not warrant higher-level recognition as a subgenus or genus.
INTRODUCTION
There are four species of doglike bats currently recognized in the genus Peropteryx Peters, including P. kappleri Peters, P. leucoptera Peters, P. macrotis (Wagner), and P. trinitatis Miller (Simmons, 2005). This genus belongs to the tribe Diclidurini Robbins and Sarich, which is the only group of sheath-tailed bats (family Emballonuridae Gervais) that is endemic to the Neotropics (Robbins and Sarich, 1988; Lim et al., 2008). Peropteryx is one of four genera of emballonurids (the others being Balantiopteryx Peters, Cormura Peters, and Saccopteryx Illiger) that have a distinctive saclike structure in the wing membrane between the leading edge of the upper and lower arm (propatagium). Each genus has a different-sized sac that is uniquely positioned in the propatagium, and these structures are typically better developed in adult males than in females. The sac is comparatively large and located adjacent to the forearm in Saccopteryx, whereas it is medium sized and located in the middle of the propatagium in Balantiopteryx. In contrast, the sac is located near the leading edge of the propatagium in Cormura and Peropteryx. In Cormura the sac is small but moderately well developed, whereas in Peropteryx it is small and poorly developed (Sanborn, 1937; Lim and Dunlop, 2008). Peropteryx is also differentiated from Cormura, the genus to which it is most morphologically similar, by a combination of other characters including having a paler medium-brown fur color (darker in Cormura), a naked face, wing attachment at the ankle (attachment at base of toe in Cormura), and less robust body proportions (Emmons and Feer, 1997; Reid, 1997; Lim and Engstrom, 2001).
New World emballonurids roost in a variety of habitats in lowland forests including tree trunks, tree hollows, rotted logs, palm fronds, caves, and buildings (Emmons and Feer, 1997; Reid, 1997; Simmons and Voss, 1998). However, most species are seldom documented during faunal surveys, and they are relatively understudied because of their typically agile, high-flying behavior and a sensitive echolocation system that lessens the chance of capture in mist nets. There is a better possibility of encountering them by actively searching for roosts (Simmons and Voss, 1998). Acoustic monitoring is another approach for detecting emballonurids by comparing species-specific echolocation search-call parameters, such as peak frequency, call duration, and pulse interval (Kalko, 1995; Jung et al., 2007). Although the widely distributed Peropteryx macrotis can be locally common in limestone caves of the Yucatan Peninsula of Mexico, they are sporadically documented and poorly known across the vast expanse of the Amazon basin. Similarly, P. trinitatis sometimes roosts in large colonies among granite outcrops in the Venezuelan Llanos, but its occurrence throughout the Guiana Shield of northern South America is rarely reported. Our understanding of the relative abundance, distribution, and general biology of the other species of Peropteryx is even sparser. During biodiversity studies in Amazonian Ecuador and Peru, we discovered another species in this genus that is new to science. Herein we describe its diagnostic characters, analyze its phylogenetic relationships, and summarize what little is known about its natural history.
MATERIALS AND METHODS
Mammals were surveyed in 1995 and 1996 as part of a larger faunal and floral inventory of the Block 16 oil concession in Orellana province, Ecuador, by MAXUS Inc. (Reid et al., 2000). Sampling was concentrated along the main access road through predominantly terra firma primary forest ranging between 200 and 350 m above sea level west of Yasuni National Park. The main method of capturing bats was by mist net, either 6 m or 12 m long, with a mesh diameter of 36 mm set at a height of 3 m above the ground in different situations (such as across trails, over streams, in clearings, and near roosts). Nets were opened at dusk and checked on a regular basis, typically until midnight.
Surveys were also conducted at the Matses Indian village of Nuevo San Juan (5°14′50″S, 73°9′50″W, ca. 150 m above sea level) on the Río Gálvez, a left-bank tributary of the Río Yavarí, in the Peruvian department of Loreto. Mist-netting was carried out by Voss on 21 nights from May to July of 1998, when an average of 40.9 m of nets were deployed at ground level (0–3 m) for 2.6 hr per night, for a total of 2309 net/meter/hour. Nets (in 6 m and 12 m lengths, both with 36 mm mesh) were usually opened just before dusk and were tended continuously until they were closed, usually before midnight. Local habitats sampled by mist-netting included gardens and clearings around Matses houses, secondary growth (abandoned swiddens), well-drained primary forest, palm swamps, mineral licks, and river beaches. Bats were also surveyed at Nuevo San Juan by searching for roosts (Voss and Emmons, 1996: 20–22). Voss searched for roosts from May to June of 1998, and several Matses men were trained to search for roosts by Fleck from September to November of 1999 (for details of this operation, see Fleck et al., 2002). Bats observed at roosts were collected by hand or by shooting.
Specimens examined (appendix 1) are deposited at the American Museum of Natural History (AMNH; New York, USA); the Museo de Historia Natural de la Universidad Nacional Mayor de San Marcos (MUSM; Lima, Peru); and the Royal Ontario Museum (ROM; Toronto, Canada). External measurements (mm) and mass (g) were taken from specimen labels including total length (TL), length of tail (TV), length of hind foot (HF), length of ear (EAR), length of tragus (TR), and length of forearm (FA). Cranial measurements were recorded with digital calipers to the nearest 0.1 mm including greatest length of skull (GLS), condylo-incisive length (CIL), breadth across zygomatic arches (ZB), breadth across mastoids (MB), breadth of braincase (BBC), breadth across postorbital constriction (POC), length of maxillary tooth row (CM3), and breadth across upper molars (M3M3; Reid et al., 2000). One-tailed Student's t-tests were used to evaluate sexual size dimorphism in our measurement data.
RESULTS
Fig. 1
The contrasting color patterns of the wings spread on the left and closed on the right for three species of Peropteryx: top, P. leucoptera (ROM 112531) with white wing tips, brown toward the elbow and brown digits; middle, P. pallidoptera (ROM 104396) with uniform white from the wing tips to the elbow and light brown digits; and bottom, P. macrotis (ROM 108523) with uniform brown from the wing tips to the elbow and brown digits.
Fig. 2
Differences in basisphenoid pit and lateral pterygoid pits at the anterior end (toward the top) for three species of Peropteryx: top, P. leucoptera (ROM 112531) with large pterygoid pits separated by a mesopterygoid extension; middle, P. pallidoptera (ROM 104396) with small pterygoid pits separated by a mesopterygoid extension; and bottom, P. macrotis (ROM 108467) with small pterygoid pits separated by the basisphenoid pit.
Fig. 3
Skull of the holotype of Peropteryx pallidoptera (ROM 104396) with dorsal, ventral, and lateral views of the cranium, and lateral view of the left mandible from top to bottom. The greatest length of the skull is 14.1 mm.
Holotype
An adult male (ROM 104396) collected 66 km S of Pompeya Sur (00°48′S, 76°24′W), Orellana Province, Ecuador by F.A. Reid, and F. Sornoza (field number F37227) on 8 May 1995. The holotype was prepared as a skin, skull, and partial skeleton.
Paratypes
An adult male (AMNH 74107) prepared as a skin and skull collected at Orosa (03°26′S, 72°08′W), Amazon River, Loreto, Peru, by Alfonso and Ramón Olalla on 28 November 1926. Seventeen individuals prepared as fluid-preserved specimens collected at Nuevo San Juan (05°14′50″S, 73°09′50″W, 150 m above sea level), Río Gálvez, Loreto, Peru, by Robert S. Voss and David W. Fleck in 1998 and 1999, including 1 adult male (MUSM 13226), 14 adult females (AMNH 272671, AMNH 272726, AMNH 272854, AMNH 272855, AMNH 272827, AMNH 273042, AMNH 273116, AMNH 273185, MUSM 13228, MUSM 13230, MUSM 15246, MUSM 15249, MUSM 15250, MUSM 15252) and 2 subadult females (MUSM 13227, MUSM 13229).
Distribution
The new species is currently known from just three localities in primary lowland forests of eastern Ecuador and northern Peru at elevations <400 m above sea level. It is found sympatrically with P. kappleri and P. leucoptera at Nuevo San Juan. The new species and P. macrotis have overlapping distributions in the western Amazon but their closest points of documented occurrence are approximately 50 km apart (between Orosa and Puerto Indiana on the Amazon River of Peru).
Etymology
The specific epithet refers to the pale wings and is compounded from the Latin words pallidus (“pale”) and ptera (“wing”).
Measurements
External measurements are presented in table 1 and cranial measurements are presented in table 2 for Peropteryx pallidoptera and other congeneric species.
TABLE 1
External measurements (in mm) and weight (in g) for adult specimens of Peropteryx The average and range of measurements for each sex of the five species of Peropteryx and the probability (* < 0.05) of sexual dimorphism within each species with sample sizes >2 are summarized in boldface. Abbreviations of measurements are listed in the Materials and Methods section, and localities of specimens examined are listed in appendix 1.
TABLE 2
Cranial measurements (in mm) for adult specimens of Peropteryx The average and range of measurements for each sex of the five species of Peropteryx and the probability (* < 0.05) of sexual dimorphism within P. macrotis (other species have small sample sizes) are summarized in boldface. Abbreviations of measurements are listed in the Materials and Methods section, and localities of specimens examined are listed in appendix 1.
Diagnosis
Superficially, Peropteryx pallidoptera appears most similar to P. leucoptera because they both have transparent wings. However, the new species has paler brown arms and digits (dark brown in P. leucoptera), the skin from the wing tip to the elbow is translucent and evenly colored with a tinge of brown (gradually darkens in color from wing tip to body in P. leucoptera; fig. 1), the ears are not connected by a band of skin, the rostrum of the skull is not broad, and the lateral pterygoid pits are not as large and deep as they are in P. leucoptera (fig. 2). In contrast, P. pallidoptera is morphologically similar to P. kappleri, P. macrotis, and P. trinitatis with the ears separated, rostrum of the skull not broad, and the lateral pterygoid pits small and shallow (fig. 2). However, the new species does not have dark brown wings and digits, the rostrum is not inflated dorsally, and the posterior border of the mesopterygoid extension is not level with the anterior border of the basisphenoid and lateral pterygoid pits, all of which are characters seen in P. kappleri, P. macrotis, and P. trinitatis. Furthermore, P. pallidoptera has a tiny peglike upper anterior premolar similar to that of P. trinitatis but different from those of P. kappleri and P. macrotis; the latter two species have a slightly larger tooth, bearing a distinct posterior cusp. Peropteryx kappleri is the largest species in the genus with no overlap in forearm length and cranial measurements (except postorbital constriction) with the other species (tables 1 and 2). In general, P. pallidoptera is smaller in cranial size than P. leucoptera, P. macrotis, and P. kappleri, but it is larger than P. trinitatis. This size gradation is most evident in same-sex comparisons of condyloincisive length (table 3).
TABLE 3
Morphological comparison among the five species of Peropteryx
Description
The new species is a medium-sized Peropteryx (table 2). The dorsal fur is relatively long (approximately 8 mm) and is a uniform medium brown with slightly paler hair bases. The ventral fur is similar but slightly lighter in color. The ears are brown and are not in contact with each other, although they are close together on the forehead. The interfemoral membrane is brown, as is the wing membrane from the body to the elbow, but the distal portion is translucent with a tinge of brown (fig. 1). The arms and digits of the wings are medium brown. There is a poorly developed wing sac situated on the leading edge of the propatagium with an outward lateral opening.
The skull of P. pallidoptera has slender postorbital processes and a relatively narrow rostrum, but the rostrum is not dorsally inflated. The relatively shallow, undivided basisphenoid pit has two small anterolateral pterygoid pits and the mesopterygoid extension protrudes posteriorly into the basisphenoid region (fig. 2). The upper anterior premolar is tiny and peglike. The dental formula is i 1/3, c 1/1, p 2/2, m 3/3 with a total of 32 teeth.
Comments
The cranial measurements for males of P. pallidoptera are all within the observed range of variation for females (table 2), which suggests that there is little or no sexual dimorphism in skull size. However, although the sample size for male P. pallidoptera is small (N = 2), males are smaller than females with no overlap in forearm length and weight (table 1). In contrast, females of P. macrotis are significantly larger than males (p < 0.02) for all cranial measurements except breadth of braincase and postorbital constriction. Similarly, females of P. macrotis are significantly larger than males (p < 0.05) for all external measurements except length of tail and length of tragus. Brosset and Charles-Dominique (1990) did not observe sexual dimorphism for P. trinitatis but they only compared 1 male and 1 female. Although our sample size was not much better (at 2 males and 2 females), females were larger for all cranial measurements except postorbital constriction and there was no overlap in range (table 2). The male P. trinitatis reported by Brosset and Charles-Dominique (1990) was large and similar to our females except for maxillary toothrow length. However, with larger sample sizes for external measurements (10 males and 10 females), females of P. trinitatis are significantly larger (p < 0.003) than males except for length of tail and length of tragus. The sample sizes for P. kappleri and P. leucoptera were too small for making meaningful comment on cranial sexual size dimorphism. External measurements suggest that females of P. kappleri are larger than males (p < 0.005 for total length, forearm length, and weight), but there is no indication of sexual dimorphism in any of the external measurements for P. leucoptera.
Natural History
Ecuador
Peropteryx pallidoptera was one of 66 species of bats that we captured in Block 16 in 1995 and 1996 (Reid et al., 2000). The holotype was caught in a mist net set at the entrance to a small cave next to a muddy, spring-fed saladero (mineral-rich seep) that periodically flooded the surrounding forest. The cave appeared to have been used primarily as a night roost, because few bats were found to occupy it in the daytime.
Another six specimens of Peropteryx pallidoptera (easily recognized as such by their transparent wing tips) were captured and released by Engstrom, who netted them over a larger saladero (at least 1 km in diameter) located near the edge of the Tiputini River about 10 minutes by boat north of the Tiputini Biodiversity Station (00°38′S, 76°09′W). The small stream that fed this saladero tumbled over the edge of a cliff ca. 10 m high, and a shallow cave was present behind the waterfall. The bats apparently used the cave as a diurnal roost: Engstrom saw small, brown emballonurids in the cave before it got dark, and he caught the six P. pallidoptera in the net nearest the cave exit at dusk.
Peru
We recorded a total of 58 species of bats at Nuevo San Juan in 1998 and 1999 (Fleck et al., 2002). Of 503 recorded bat captures at this locality, 372 were in mist nets and 311 were made at diurnal roosts. All 17 specimens of Peropteryx pallidoptera from Nuevo San Juan were taken from roosts. In total, we found 10 roosts occupied by this species in a variety of circumstances as described below (numbers refer to entries in a field catalog of bat roosts).
Roost 2: Dark cavities among the roots and buttresses of a fallen tree in well-drained primary forest. Exactly where the bats were roosting is unknown because they had been spooked by kids who were shooting at them with arrows. A total of eight bats were collected as they perched on nearby tree trunks. Three (one adult male, one adult female, and one subadult female) were Peropteryx pallidoptera and five were Saccopteryx bilineata; an unknown number of either or both species might have escaped (22 May 1998).
Roost 6: A half-open hollow log (about 50 cm inside diameter) in old secondary growth near a stream. Two bats were found roosting here, only one of which (an adult female Peropteryx pallidoptera) was collected (11 June 1998).
Roost 15: An undercut earth bank in the side of a stream gully in well-drained primary forest. Two adult female and one subadult female Peropteryx pallidoptera occupied this roost, where they were found hanging well apart from one another (30 June 1998).
Roost 20: A small dark cavity of roots and earth at the base of a large fallen tree in well-drained primary forest. Two bats were found roosting here about 50 cm above the ground, of which only one (an adult female Peropteryx pallidoptera) was collected (8 July 1998).
Roost 22: A dark horizontal chamber under a buttress of a fallen tree in well-drained primary forest. Two adult female Peropteryx pallidoptera were found roosting here, about 80 cm above the ground (8 July 1998).
Roost 31: A deep armadillo burrow in the side of a stream headwater gully in well-drained primary forest. One adult female Peropteryx pallidoptera was found roosting here in the company of at least two Micronycteris matses and three Carollia brevicauda. No bats were seen to escape this roost by flying out of it, but some might have escaped by flying deeper inside (4 September 1999).
Roost 56: The unmodified leaf of a wild banana plant in hilltop primary forest. The single bat (a pregnant female Peropteryx pallidoptera) observed and collected here was hanging from the middle of the leaf about 1 m above the ground (16 September 1999).
Roost 73: A hole in the ground in the side of a stream headwater gully on a hillside in well-drained primary forest. Two bats were observed roosting here, of which one (a pregnant female Peropteryx pallidoptera) was collected (22 September 1999).
Roost 77: A hole in the side of a stream headwater gully in hilltop primary forest. Four bats were observed using this roost, of which two were collected. One of those collected was an adult female Peropteryx pallidoptera, and the other was an adult male P. leucoptera (23 September 1999).
Roost 102: The undercut bank of a small stream in well-drained primary forest. Only a single bat (an adult female Peropteryx pallidoptera) was observed and collected here (11 October 1999).
Roost 147: The “underside of a log” [as described by the Matses collector; presumably elevated somehow above the ground] in well-drained primary forest in a valley bottom. A single lactating adult female Peropteryx pallidoptera was collected here, but it seems likely that one or more nursing young may have been present but unobserved (27 October 1999).
Roost 154: The underside of a fallen tree in primary upland forest near a small stream. A single female Peropteryx pallidoptera was in the roost (2 November 1999).
The specimen from Orosa was collected by Alfonso and Ramón Olalla, members of an Ecuadorean family of professional collectors employed by the American Museum of Natural History, who worked at or near the mouth of the Río Orosa on the right (south) bank of the Amazon from 30 August to 11 December 1926. According to Wiley (in press), the Olallas' camp at Orosa must have been in the floodplain of the Amazon, which is so wide at this point (about 15 km) that most of their specimens certainly came from floodplain habitats. Although the forest within several kilometers of the riverbank at the Olallas' presumed campsite (ca. 3°31′31″S, 73°11′22″W; Wiley, in press) is seasonally flooded varzea, the river was probably at or near its lowest annual stage in August–December, so this habitat would have been easy to traverse on foot.
DISCUSSION
Taxonomy
Several genera of New World emballonurids (including Balantiopteryx, Cormura, Peropteryx, and Rhynchonycteris) were described by Peters (1867) based primarily on the presence or absence of a wing sac and the position of this sac in the propatagium. In addition to Balantiopteryx plicata and Peropteryx kappleri, Peters (1867) also described the white-winged doglike bat (P. leucoptera) with the type locality from Suriname. The following year, Peters (1868) described a new subgenus (Peronymus) for this species. Dobson (1878) subsequently synonymized Balantiopteryx, Centronycteris, and Peropteryx with Saccopteryx (treating them as subgenera) based primarily on the shared presence of a wing sac. He considered this taxonomic grouping of species with wing sacs in Saccopteryx as analogous to the grouping of taxa with throat poaches in Taphazous and to the grouping of taxa with frontal sacs in Rhinolophus. However, Dobson's (1878) speculation that wing sacs were present in Centronycteris was a misinterpretation of Peters (1867) and further exasperated by the examination of only two females, the gender that typically has poorly developed sacs. The absence of wing sacs in Centronycteris was eventually correctly documented by Sanborn (1937) in his taxonomic revision of New World emballonurids. Dobson (1878) apparently considered Cormura part of Saccopteryx, but did not formally classify it in his synopsis of groups and genera of Emballonurinae because the presence of wing sacs could not be conclusively documented in the poorly preserved holotype.
The subgenera of Saccopteryx proposed by Dobson (1878) were subsequently reinstated as full genera by Miller (1907) based on cranial differences. Miller (1907) additionally elevated Peronymus to generic rank. The primary characters for distinguishing Peronymus from Peropteryx include a broad rostrum, ears connected across the top of the head, and white wings (Miller, 1907). In total, Miller recognized eight genera of Neotropical emballonurines: Rhynchiscus ( = Rhynchonycteris), Centronycteris, Balantiopteryx, Saccopteryx, Cormura, Peropteryx, Peronymus, and Myropteryx. Myropteryx, described as new by Miller (1907), is now considered a junior synonym of Cormura.
Peronymus was considered monotypic until Thomas (1924) described a second species of white-winged doglike bat as Peronymus cyclops from Peru. Thomas (1924: 531) diagnosed cyclops as “very closely allied to the P. leucoptera of Guiana and the lower Amazon, but distinguished by its larger skull.” Although Thomas (1924) did not consider that the differences in rostral inflation and scarcely perceptible ear connection in Peronymus warranted higher-level recognition, he grudgingly followed the generic classification of Miller (1907) because of differences in the structure of the basisphenoid pits. Likewise, Sanborn (1937) retained Peronymus as a distinct genus but regarded cyclops as a larger upper Amazon subspecies of P. leucoptera. Peronymus was again demoted to a subgenus of Peropteryx by Cabrera (1958) following the classification of Peters (1868), but its taxonomic rank remained contentious. Subsequent authorities have treated Peronymus as a distinct genus (Husson, 1962, 1978; Linares, 1986; Brosset and Charles-Dominique, 1990; Corbet and Hill, 1991), as a subgenus of Peropteryx (Honaki et al., 1982; Koopman, 1982, 1984; Jones and Hood, 1993; Koopman, 1993, 1994; McKenna and Bell, 1997; Simmons and Voss, 1998; Simmons, 2005), or as a junior synonym of Peropteryx (Hood and Gardner, 2007).
Systematics
Separate molecular phylogenetic analyses representing loci from each of the four genetic transmission systems in mammals (mitochondrial, autosomal, X, and Y chromosomes) for New World emballonurids resulted in trees with congruent topologies in which P. pallidoptera and P. leucoptera appear as successively basal lineages to a clade comprised of P. kappleri, P. macrotis, and P. trinitatis (Lim et al., 2008; P. pallidoptera was identified as “Peropteryx sp.” in that publication). Although the relationships among P. kappleri, populations of P. macrotis from Central America and South America, and P. trinitatis were poorly supported, a combined analysis of the four unlinked genes (cytochrome b, chd1, usp9x, and dby) corroborated the more basal systematic relationships including the phylogenetic position of the new species (fig. 4; Lim, 2009).
Fig. 4
Topology within Peropteryx from a phylogenetic analysis of Emballonuridae (Lim, 2009). The tree is a Bayesian analysis of combined genes representing each of the four genetic transmission systems in mammals. First number along the branch is the Bayesian posterior probability percentage and the second number is the bootstrap percentage from a parsimony analysis. There are two populations of P. macrotis from Central America (CA) and South America (SA).
Parsimonious optimizations of wing pigmentation on this phylogeny suggest that transparent (“white”) wings, the principal trait used to distinguish Peronymus from Peropteryx, is not a unique synapomorphy that diagnoses that taxon. Either transparent wings evolved once in the common ancestor of Peronymus + Peropteryx and subsequently reversed on the branch that subtends the dark-winged species of Peropteryx, or transparent wings evolved independently in leucoptera and pallidoptera. Furthermore, a molecular dating study estimated a similar age of diversification for Balantiopteryx and Peropteryx (including P. leucoptera) beginning 10.8 mya and 11.7 mya, respectively (Lim, 2007). In contrast, the most recent split of genera in the subtribe Diclidurina was Cyttarops and Diclidurus, which occurred much earlier at 14.6 mya (Lim, 2007). Because Peronymus no longer appears notably distinct from all other species of Peropteryx in morphology, and because its age of divergence from other Peropteryx is similar to that seen within other embalonurid genera, we see little value in recognizing Peronymus as distinct at the generic or subgeneric level. Accordingly, we follow Hood and Gardner (2007) in treating P. leucoptera as a species of Peropteryx and in not recognizing any subgenera within that taxon.
Wing Sacs
The wing sac of New World emballonurids is a nonglandular structure (Scully et al., 2000) that is apparently a receptacle for various bodily secretions used in the scent-marking behavior (salting) of Saccopteryx (Voigt and von Helversen, 1999). Saccopteryx is the only genus known to exhibit the complex social behavior of salting, which is apparently associated with its harem mating system (Bradbury and Vehrencamp, 1977). Although the behavioral function of the wing sac in the other genera has not been thoroughly studied, the distinctive folding of the internal surface in Saccopteryx is lacking in Peropteryx (Scully et al., 2000). This suggests the possibility of a use other than scent holding, and possibly a different ecological role for the wing sac. Based on a study of evolutionary patterns in morphology, wing sacs evolved independently in each of the two subtribes of New World emballonurid bats: once in the lineage leading to Peropteryx, Cormura, and Balantiopteryx in Diclidurina, and independently in the ancestral lineage leading to Saccopteryx in Saccopterygina (Lim and Dunlop, 2008). Furthermore, each genus has a different-sized sac that is uniquely positioned in the propatagium. The variation of the wing sac position, degree of internal folding, and social behavior map consistently on the phylogeny suggesting a potential phylogenetic basis for these character systems. More detailed study and comprehensive taxonomic coverage of the ecology and systematics of Emballonuridae may yield additional insights about the adaptive significance and evolution of these interesting structures.
Natural History and Biogeography
Several aspects of our sparse capture data are noteworthy. First, Peropteryx pallidoptera does not appear to be a roost specialist, as it uses a variety of diurnal refugia including caves, fallen trees, holes in the ground, undercut stream banks, and leaves. Second, at least some of these roosts are shared with several other species of bats, including phyllostomids and emballonurids. Lastly, nothing about the localities where these bats were captured leads us to believe the species is rare in any ecologically meaningful sense.
The three known capture sites of Peropteryx pallidoptera (Block 16, Nuevo San Juan, and Orosa) do not appear to share any ecological characteristics that are not also present throughout much of western Amazonia. For example, all three sites are <400 m above sea level and probably receive >2800 mm of annual precipitation; lowland rainforest, including both riparian and upland formations, is the predominant vegetation type at each site (Reid et al., 2000; Fleck et al., 2002; Wiley, in press). Most of the roosts we found near Nuevo San Juan were in upland (well-drained, nonriparian) forest, which is by far the most abundant macrohabitat throughout the region. It is also relevant that P. pallidoptera occurs north and south of the Amazon, one of the most significant zoogeographic barriers in the region. Therefore, it seems probable that the species is very widely distributed and will soon turn up at other sites where appropriate survey methods are used to inventory bat diversity. It also seems likely that additional specimens may be found in museum collections where they may have been misidentified as P. leucoptera because of their similar transparent wing membranes.
The relative abundance of Peropteryx pallidoptera among the bats we collected at roosts near Nuevo San Juan (17 out of 311 specimens, about 5%) contrasts with the total absence of this species among the bats we captured in mist nets at the same locality (N = 372). Although all of the Ecuadorean specimens were taken in mist nets, the mist nets that caught P. pallidoptera were set to intercept bats flying into or out of caves. Although it is not known how the Orosa specimen was captured, the Olallas did not have mist nets, and most of the other bats they obtained at Orosa belong to species that are readily taken at diurnal roosts (e.g., Glossophaga soricina, Mesophylla macconnelli, Furipterus horrens, Thyroptera tricolor), so it seems probable that they also collected P. pallidoptera by this method. Thus, searching for roosts would appear to be the method of choice for collecting additional material of this species, which either is adept at avoiding nets or forages in situations where nets are hard to set.
Acknowledgments
We thank MAXUS Inc. for funding the fieldwork in Ecuador, Ecuambiente Inc. for logistical support, and Francisco Sornoza for field assistance. Our Peruvian fieldwork was funded by grants from the American Museum of Natural History's Center for Conservation and Biodiversity, and we thank our Matses friends for their hospitality and patience while we were their guests in Nuevo San Juan. Sergio Solari and Victor Pacheco also provided important assistance at the MUSM in Lima.
REFERENCES
Appendices
APPENDIX 1
Specimens Examined
Locality information of specimens examined of Peropteryx. Museum acronyms are listed in the Materials and Methods section. Specimens marked with an asterisk (*) were sequenced in the molecular phylogenetic study of Lim et al. (2008).
Peropteryx pallidoptera—Ecuador: Orellana; 66 km S of Pompeya Sur, 00°48′S, 76°24′W (ROM 104396*; holotype). Peru: Loreto; Amazon River, Orosa, 03°26′S, 72°08′W (AMNH 74107); Rio Galvez, Nuevo San Juan, 05°14′50″S, 73°09′50″W, 150 m (AMNH 272671, 272726, 272827, 272854, 272855, 273042, 273116, 273185; MUSM 13226, 13227, 13228*, 13229, 13230, 15246, 15249, 15250, 15252).
Peropteryx kappleri—French Guiana: Paracou, 5°16′31″N, 52°55′25″W (AMNH 265990, 265991, 265292, 265989, 265993, 265995, 265996, 265997, 265998, 265999, 267080, 267081, 267082, 267084, 267085, 267086, 267291, 267392, 267833, 267834*, 267835, 267863). Guyana: Barima-Waini; Baramita, 07°22′N, 60°29′W (ROM 100910*, 101123*). Mexico: Tabasco; 2 km E of Teapa, 17°35′N, 92°56′W (ROM 78055). Peru: Loreto; Rio Galvez, Nuevo San Juan, 05°14′50″S, 73°09′50″W, 150 m (AMNH 272797, 272798, 272799, 273086, 273174; MUSM 15245).
Peropteryx leucoptera—French Guiana: Paracou, 5°16′31″N, 52°55′25″W (AMNH 266010, 266011, 266012, 266013, 267087, 267088, 267280, 267837, 267839*, 267838). Guyana: Demerara-Mahaica; Ceiba Biological Center, 06°30′N, 58°13′W, 50 m (ROM 112530, 112531*, 113611, 113612*). Potaro-Siparuni; Iwokrama Forest Field Station, 04°40′N, 58°51′W, 70 m (ROM 107458*). Upper Essequibo-Upper Takutu; Illiwa River, 3 days below Kuitaro River Mouth, approximately 02°40′N, 58°40′W (ROM 41530). Peru: Loreto; Rio Galvez, Nuevo San Juan, 05°14′50″S, 73°09′50″W, 150 m (AMNH 273182, 273197; MUSM 15247, 15251).
Peropteryx macrotis—French Guiana: Paracou, 5°16′31″N, 52°55′25″W (AMNH 266005, 266006, 266607, 267395, 267396). Guyana: Potaro-Siparuni; Iwokrama Forest, Cow Fly Camp, 04°20′N, 58°49′W, 100 m (ROM 108523*), Giaconda Camp, 04°38′N, 58°43′W, 75 m (ROM 107126*), 38 Mile Camp, 04°20′N, 58°51′W, 100 m (ROM 108467*). Mexico: Quintana Roo, Tulum, 20°10′N, 87°29′W (ROM FN33843*). Yucatan; Loltun, 20°15′N, 89°29′W (ROM 96442, 96443, 96444, 96445, 96446*, 96447, 96448, 96449, 96450, 96451), Ruinas Kabah, 20°15′N, 89°40′W (ROM 95945, 95946, 95947, 97765). Peru: Loreto; Amazon River, Puerto Indiana, 03°20′S, 72°40′W, 100 m (AMNH 73502, 73503). Pasco, San Pablo, 10°27′S, 74°52′W, 300 m (AMNH 230096). Suriname: Sipaliwini; Bakhuis, Area 8 Recon Fly Camp, 04°27′13″N, 56°52′00″W, 248 m (ROM 117527).
Peropteryx trinitatis—Trinidad and Tobago: Trinidad; Saint George County, Port of Spain, 10°40′N, 61°31′W (AMNH 7496; holotype). Venezuela: Amazonas; Pozon, 06°03′N, 67°25′W (ROM 107822*, 107823, 107824, 107825, 107826, 107827, 107828, 107829, 107830, 107831, 107832, 107833, 107834, 107835). Bolivar; Hato La Florida, 07°34′N, 65°52′W (ROM 107916, 107918, 107920*, 107921, 107922, 107923).
