Lonchophylla orienticollina, Dávalos & Corthals, 2008

Dávalos, Liliana M. & Corthals, Angelique, 2008, A new species of Lonchophylla (Chiroptera: Phyllostomidae) from the eastern Andes of northwestern South America, American Museum Novitates 3635 (1), pp. 1-16 : 5-12

publication ID

https://doi.org/ 10.1206/464.1

persistent identifier


treatment provided by


scientific name

Lonchophylla orienticollina

new species

Lonchophylla orienticollina   , new species

Eastern Cordilleran Nectar Bat

Figures 1 View Fig , 2

Lonchophylla robusta: Handley 1976: 21   part, not Lonchophylla robusta Miller, 1912   .

Lonchophylla robusta: Alberico, Cadena   , Hernández- Camacho, and Muñoz-Saba 2000: 153 part, not Lonchophylla robusta Miller, 1912   .

Lonchophylla robusta: Dávalos 2004: 14   part, not Lonchophylla robusta Miller, 1912   .

Lonchophylla robusta: Woodman and Timm 2006: 476   part, not Lonchophylla robusta Miller, 1912   .

Lonchophylla robusta: Woodman 2006: 356   part, not Lonchophylla robusta Miller, 1912   .

HOLOTYPE: Skin and skull of a female specimen ICN 10280 View Materials collected by M.P. Rivas (original number RST409) on 9 July 1988 at the intersection of caño Guamalito and caño la Curía, northern part of the Serranía de la Macarena, San Juan de Arama , 500 m elevation, Departamento del Meta, Colombia.  

REFERRED SPECIMENS: (16) One male ( ICN 10278 View Materials ) and one female ( ICN 10279 View Materials ) collected by María del Pilar Rivas (original numbers RST041 and RST076) on 23 and 26 April 1988 at caño la Curía, northern part of the Serranía de la Macarena , San Juan de Arama, 500 m elevation, Departamento del Meta, Colombia   . One female ( ICN 9702 View Materials ) collected by the students of Alberto Cadena (original number ACG1747) on 5 December 1985 at Colegio Departamental Agropecuario , vereda San Jose´, Acacías, 625 m elevation, Departamento del Meta, Colombia   . One female ( ICN 10114 View Materials ) collected by Alberto Cadena (original number ACG2405) on 5 May 1988 at bocatoma Caney Alto , Restrepo, Departamento del Meta, Colombia   . One female ( ICN 13839 View Materials ) collected by A. Cadena (original number ACG2784) on 24 September 1995 at the forest, vereda Brisas del Guayuriba , Acacías, 480 m elevation, Departamento del Meta, Colombia   . One male ( ICN 13839 View Materials ) collected by Alberto Cadena (original number ACG2782) on 26 September 1995 at the forest surrounding the Guayuriba river , vereda Brisas del Guayuriba, Acacías, 480 m elevation, Departamento del Meta, Colombia   . Two males ( ICN 14399 View Materials and 14400) collected by students of group 5 of the Introducción Sistemática Animal course (original numbers 11 and 15) on 5 June 1996 at finca La Estrella, vereda El Vergel, Cubarral , 750 m elevation, Departamento del Meta, Colombia   . One male ( IAvH 6679 ) collected by Yaneth Muñoz Saba (original number YMS702)   on 23 September 1999 at P.N.N. Tamá, Finca ‘‘ San Isidro’ ’ de Pablo Contreras (07 ° 07 9 22 0 N 72 ° 14 9 42 0 W), 1000 m elevation, Rio Negro , Mpio. Toledo, Norte de Santander, Colombia GoogleMaps   . One male and one female ( USNM 419409 View Materials and 419410) collected by Norman E. Peterson, Fred P. Brown, John O. Matson, and C. E. Yunker (original numbers 22129 and 22600) on 17 April 1968 at Kasmera 9 ° 59 9 N 72 ° 43 9 W, 21 Km SW of Machiques, Zulia, Venezuela GoogleMaps   . One male and one female ( USNM 419425 View Materials and 419426) collected by Arden L. Tuttle, Benjamin Inquilla, and Ernest L. Stromeyer (original numbers 34095, and 34274) in January 1968 at Altamira (8 ° 50 9 N 70 ° 30 9 W), Barinas, Venezuela GoogleMaps   . Three females (BM-NH 78-1354, 78-1356, and 78-1359) collected by Liam Hutson and R.E. Stebbins on 30 July 1976 at Yaupi (2 ° 93 9 S 77 ° 54 9 W), Morona Santiago, Ecuador   .

DISTRIBUTION: Currently known from the southeastern versant of the Cordillera de Mérida, the eastern versant of the Serranía del Perijá in Venezuela, the eastern and western versants of Cordillera Oriental and the northern portion of the Serranía de la Macarena of Colombia, and the eastern versant of the Cordillera Oriental of Ecuador (fig. 3).

ETYMOLOGY: From the Latin oriens (‘‘eastern’’) and collis (‘‘hill’’), to summarize the known distribution of the species in northwestern South America.

DIAGNOSIS: A medium-sized species of Lonchophylla   (forearm 40–47 mm; weight 11–16 g) with dorsal fur ranging from intense ochraceous orange to buckthorn brown to snuff brown and light tawny olive ventral fur

Fig. 2. Dorsal view of Lonchophylla orienticollina USNM   419409 ♀ (A), and dorsal view of Lonchophylla robusta USNM   419415 ♀ (B). Arrows indicate fixed differences helpful in distinguishing these species.

( Ridgway, 1912); ventral hairs sometimes bicolored, the banding is almost imperceptible and disappears towards the abdomen; pinnae short with rounded tips; no furry fringe along uropatagium; calcar shorter than foot; and long feet and thumbs relative to body size. The shape of the skull is distinctive: the braincase is tall; the rostrum is inflated and short, appearing thick in profile; the palate is wide, with postpalatine torus. The height of the braincase makes the slope to the rostrum have a relatively high angle, visible in profile. The spatial arrangement of teeth in the palate is diagnostic: P4 is angled outward about 15 ° with respect to P3, making the palate appear wide and almost round. There are narrow gaps between I1 and I2, outer upper incisors I2 point ventromedially; wide gaps are present between C and P3; narrower to no gaps between P3 and P4; height of P3 less than P4; lingual cusp on P4 well defined; M1 longer (anteroposterior axis) than M2; M3 is the smallest of the molars. Lower incisors are small, with crowns taller than they are wide, bilobed or trilobed.

Cleaned skulls are necessary to diagnose this species from its sympatric congeners robusta   and handleyi   . Five characters are particularly useful in this regard: in dorsal view the rostrum appears short and wide, inflated at the center; in profile the rostrum appears thick, particularly in the postorbital region; the braincase appears tall and forms a marked angle with respect to the rostrum; in ventral view P4 is at about a 15 ° angle outward from P3; and this considerable widening makes the palate seem wide and almost round. Lonchophylla orienticollina   is the only species of its size in the genus to display this combination of traits.

MEASUREMENTS: A summary of all known specimens of Lonchophylla orienticollina   is provided in table 1. Comparisons with a representative series of other similarly sized congeners are also compiled in table 1.

DESCRIPTION AND COMPARISONS: Lonchophylla orienticollina   is a medium-sized member of the genus, larger than mordax   , concava   , fornicata, thomasi, pattoni, cadenai, and dekeyseri, and smaller than handleyi   , chocoana   , orcesi and most, but not all, robusta   ; see table 1 (Albuja V. and Gardner, 2005; Dávalos, 2004; Taddei et al., 1983; Woodman, 2007; Woodman and Timm, 2006). Lonchophylla orienticollina   can be unambiguously distinguished from the smaller Lonchophylla species   on the basis of forearm length (. 40 mm) and greatest length of skull (. 24.5 mm). Additionally, orienticollina   can be distinguished from mordax   and thomasi based on dorsal fur coloration. The latter are dark brown bister or sepia of Ridgway (1914), whereas orienticollina   exhibits the orangebrown dorsal fur coloration common in robusta   .

There is overlap in measurements between orienticollina   specimens and smaller robusta   individuals (table 1). Qualitative characters are diagnostic for this comparison. In dorsal view the skull of orienticollina   appears blunter and shorter, with all features shortened when compared to robusta   of similar size; the rostrum clearly inflated above P4 (fig. 2). In profile, the blunt and shortened appearance persists, with the braincase taller and the rostrum shorter and thicker than in robusta   ; the braincase slopes at a higher angle from the horizontal plane than in robusta   , and in robusta   the postorbital area appears inflated relative to orienticollina   . In ventral view, the palate of orienticollina   is wide, appearing shorter than that of robusta   of similar size although palatal length is not diagnostic; P4 forms an angle about 15 ° with respect to P3, widening the palate and making it look almost round compared to the rectangular appearance of the palate of robusta   .

Lonchophylla orienticollina   is smaller than all remaining congeners in cranial dimensions. Lonchophylla hesperia   and L. bokermanni   are absent from the range of orienticollina   , have longer skulls, with greater skull length-towidth ratios, and have shorter forearms (Taddei et al., 1983). Lonchophylla orienticollina   is smaller than chocoana   and orcesi in most dimensions (Albuja V. and Gardner, 2005; Dávalos, 2004), and their distributional ranges do not overlap. Lonchophylla orienticollina   can be distinguished from its sympatric congener handleyi   on the basis of size, the furry fringe along the uropatagium, the definition and size of lingual cusp on P4, presence of a ridge along the posterior edge of the palate postpalatine torus, and the shape of the palate, which is long and pointed at the incisors in handleyi   .

As with robusta   , handleyi   , and chocoana   , the dorsal pelage of orienticollina   is composed of bicolored hairs with cream-white bases and ochraceous orange to brown tips. The length of the dorsal fur along the upper back in chocoana   is approximately 7–9 mm, slightly longer than in robusta   , handleyi   , and orienticollina   4–8 mm. The ventral pelage of orienticollina   , as in robusta   , sometimes shows bicolored hairs around the neck and in the abdominal region.

The cranial morphology of orienticollina   is similar to that of other members of the genus. L. orienticollina   has long rostrum relative to non–nectar-feeding phyllostomids, a small but noticeable anteorbital inflation, and a large, relatively tall braincase (see Woodman and Timm, 2006, for diagnosis of the genus). Zygomatic arches are almost never preserved after preparation, as in all other lonchophyllines. As all Lonchophylla   , orienticollina   has a dental formula I2/2, C1/1, P2/3, M3/3 X 2 5 34. As found in smaller members of the genus (Woodman and Timm, 2006), L. orienticollina   may have supernumerary premolars. The inner upper incisors are large compared to the outer incisors and are separated by a gap from each other and from the canine. In this orienticollina   resembles all other species of Lonchophylla   .

The dentition of orienticollina   resembles that of robusta   , even in close detail. Both orienticollina   and robusta   show a tall gap between the inner upper incisors, meeting at the distal quarter of the crown. The upper canines of orienticollina   are similar in absolute size to those of handleyi   , robusta   , and thomasi whose canines are large for their body size, larger than those of mordax   and concava   , and smaller than those of chocoana   , which are exceptionally large. The posterior cusp of the upper canines orienticollina   is sharp, similar to that of robusta   , thomasi, and mordax   . In orienticollina   P4 is slightly longer than P3 anteroposterior dimension, as it is in chocoana   , robusta   , handleyi   , mordax   , and thomasi. P3 is shorter than P4 dorsoventral dimension in orienticollina   , as it is in robusta   , handleyi   , and less so in chocoana   . In mordax   , concava   and thomasi P3 is taller than P4.

Beginning with the description of Lonchophylla handleyi ( Hill, 1980)   , the degree of development of the basal lingual cusp of P4 has been used as a character to distinguish among species in this genus. After close examination of a large series of robusta   , orienticollina   , and specimens of chocoana   , handleyi   , mordax   , concava   and thomasi; differences in this character deserve further comment. The basal lingual cusp on P4 can be seen as a small protuberance in mordax   , and it has been described as ‘‘well developed’’ (Taddei et al., 1983: 629). This cusp is absent or imperceptible in concava   . In handleyi   this cusp has been described as ‘‘small and undeveloped, occasionally very small and insignificant’’ (Hill, 1981: 235), and in chocoana   as ‘‘well developed’’ (Dávalos, 2004: 8). All the handleyi   and chocoana   specimens examined here had a basal lingual protuberance on P4, but the degree of development varied from a fully formed cusp projecting upward (fig. 4B), to a small shapeless knob (fig. 4D). In both robusta   and orienticollina   the basal lingual cusp is invariably visible and distinct. There are individual differences in the sharpness and height of the cusp, probably related to dental wear. A few robusta   individuals had rounded cusps, comparable to those of some chocoana   and handleyi   specimens, though never as blunt or small (fig. 4). Individual variation might thus confound some pairwise comparisons aimed at identifying large Lonchophylla   congeners based on this character alone. In addition to the features of P4, other characters, e.g., fringe of the uropatagium, size of canines, or length and shape of the palate should be used for species identification.

As in robusta   and handleyi   , the M1 and M2 of orienticollina   have similar widths (lateral dimension), with M1 longer (anteroposterior dimension) than M2, and an overall smaller M3. In contrast, in thomasi, mordax   , and concava   the first two molars are similar in length and height.

Lower incisors may be trilobed or bilobed in orienticollina   , as in robusta   . In handleyi   , incisors are variably trilobed, bilobed, or neither, while both mordax   and concava   have trilobed incisors, though this can be difficult to detect in the latter. The height of the lower incisors is greater than the width of these teeth in orienticollina   , and sometimes in robusta   . Height and width of the crown of the lower incisors are roughly the same in other Lonchophylla   examined. The lower premolar dentition of orienticollina   resembles that of robusta   . The posterior cusp of p2 is distinct and hooklike in robusta   , thomasi, handleyi   , chocoana   , and orienticollina   , with some individual variation in sharpness probably caused by wear. In mordax   this cusp, if present, is not clearly distinct or hooklike. In orienticollina   p4 is taller (dorsoventral dimension) and longer (anteroposterior dimension) than p3; p3 was erroneously reported to be taller than p 4 in robusta (Dávalos 2004: 9)   . As in robusta   , the first molar of orienticollina   is the widest (lateral dimension), tallest, and longest of the molar series, followed by m2, which is in turn wider and slightly longer than m3. The molar series varies in width and length more than in height. As in robusta   , the coronoid process in orienticollina   is high and oriented at an angle of about 110 ° with respect to the tooth row.


A plot of factor scores from a principal components analysis comparing orienticollina   to robusta   shows orienticollina   at the low end of PC 1 (fig. 5), which comprises mostly cranial size, and reflects its shorter skull (table 2). In contrast, orienticollina   falls at the high end of PC 2, resulting from the longer thumb relative to robusta   . The sexes of each species do not separate along either axis, despite the longer, wider skulls of male robusta   (table 1). Specimens of orienticollina   had a narrower size distribution than the sample of robusta   , perhaps reflecting the narrower geographic range of the former. The two species overlapped considerably along each axis, but orienticollina   spans a combination of measures distinct from that of robusta   (fig. 5).

A discriminant function fitted to the variables used to generate the PC factors was highly significant, 0.0001, with a canonical correlation R* between the species assignment and the function generated of 0.78 (where R* ranges from 0, indicating no correlation, to 1, indicating perfect correlation). Skull and palatal length were the variables most relevant to the discriminant function coefficients (table 2). The resulting discriminant function based on only six metric variables successfully predicted species assignments for 21 of 23 orienticollina   (91.3%) and 30 of 34 robusta   (88.2%), in either case much better than at random (50%).


Distributional modeling using the maxent algorithm found two variables, seasonality of rainfall and the mean of monthly temperature range, as the most important determinants of presence of Lonchophylla orienticollina   . Together, these climate variables explained 63% of the variation in inferred suitability. High seasonality in precipitation characterizes the eastern slopes of the east Andes of Venezuela and Colombia, where a marked

TABLE 2 Results of principal components and discriminant function analyses Factor loadings for the first two axes of regression factors extracted from the correlation matrix of selected metric variables, and coefficients of the discriminant function fitted to those same variables.

dry season extends from December to March ( Marston, 1948). The known localities for orienticollina   suggest a continuous distribution northward from the eastern versant of the Andes in Ecuador, widening in northern Colombia and northwestern Venezuela (fig. 3). In contrast, the model inferred relatively low suitability for the northwestern versant of the Cordillera de Mérida, and most of the eastern slopes of the Cordillera Oriental in Ecuador (fig. 6), resulting in a more complex distributional pattern than expected from elevation and land cover alone. High Andean elevations, the Colombian Choco, and the Isthmus of Panama were all found to be unsuitable for orienticollina   .

The modeled distribution inferred high suitability in regions beyond the known species distribution in central Colombia and western Ecuador. The lack of records from these areas could be explained by historical events such as orogeny, failure to disperse, or ecological interactions such as competitive exclusion by a close relative ( Anderson et al., 2002), leading to a narrower range than expected from climate alone. Alternatively, poor sampling (Voss and Emmons, 1996) or failure to include one or several critical environmental variables might account for these distributional over-predictions. Only additional sampling through targeted field surveys in the relatively undersampled midelevations of the northeastern Andes has the potential to further refine our understanding of the environmental requirements of this species, and its ecological interactions.


Instituto de Ciencias Naturales, Museo de Historia Natural














Lonchophylla orienticollina

Dávalos, Liliana M. & Corthals, Angelique 2008

Lonchophylla robusta:

Handley, C. O. 1976: 21