Description of a new shrew of the genus Cryptotis (Mammalia: Soricomorpha: Soricidae) from the Sierra de Aroa, an isolated mountain range in northwestern Venezuela, with remarks on biogeography and conservation
MARCIAL QUIROGA-CARMONA1 & JESÚS MOLINARI2
1Departamento de Biología, Facultad Experimental de Ciencias y Tecnología, Universidad de Carabobo, Valencia, Venezuela. E-mail: [email protected]. 2Departamento de Biología, Facultad de Ciencias, Universidad de Los Andes, Mérida, Venezuela. E-mail:[email protected] (corresponding author).
Abstract
In South America, shrews of the genus Cryptotis have a primarily Andean distribution. Based on specimens from the non-Andean Sierra de Aroa in Venezuela, we name Cryptotis aroensis sp. nov., which we assign to the C. thomasi group of thegenus owing to its possession of characters that include a relatively large body size, luxuriant fur, moderately enlargedforefeet with elongated and narrow claws, unicuspid teeth relatively narrow and concave on the posteroventral margin,ectoloph of first upper molar with the anterior element reduced relatively to the posterior element, and mandible with thearticular process not robust, high, and broad, and with the coronoid process joining the ramus at a low angle. The newspecies can be differentiated from other members of the C. thomasi group on the basis of its possession of a uniquecombination of characters, that include a rich grayish brown pelage, a narrow palate at the level of the second upper molars(no overlap in this measurement observed with 146 specimens of 10 other species of the group), nasal cavity and ethmo-turbinals partially visible in occlusal view of palate, lacrimal foramina wide and deep, tympanic process of petromastoidsshowing a minute foramen, unicuspid teeth with posterolingual cuspules, fourth unicuspid tooth labially placed, thirdupper molar complex and nearly as wide as the second upper molar, and bicuspulate lower incisors. The new species isknown from only 3 specimens obtained at the type locality, which is in a pristine patch of cloud forest at elevation 1730m, and represents the first mammalian taxon known to be endemic to the Sierra de Aroa, a small and isolated mountainrange that was previously reported to possess numerous biotic elements that are either exclusive, or shared with the muchlarger Cordillera de la Costa to the east. The clear morphological differentiation of the new species with respect to itsAndean relatives suggests that its ancestors colonized the Sierra de Aroa several glacial maxima ago, when montanevegetation belts were lower than today. The presumably optimal habitat of the new species (cloud forests of the Sierra deAroa above 1500 m) covers less than 40 km2, and is threatened by deforestation and global climate change. Therefore, thenew species can be categorized as endangered by application of criteria B2a and E of the current “Red List Categories andCriteria” of the International Union for Conservation of Nature.
The genus Cryptotis Pomel (Mammalia: Soricomorpha: Soricidae) is composed of 33 extant species of small tomedium-sized shrews with reduced eyes and pinnae. Only 1 species, Cryptotis parvus (Say), with an ampledistribution in México, occurs in the eastern half of the United States and adjacent Canada, 21 species occur only inMéxico and Central America, and 11 species occur only in South America (Hutterer 2005; Woodman & Péfaur2008; Woodman 2010; Quiroga-Carmona, 2011, in press). Based on morphological characters, extant species ofCryptotis are divided into 4 informal, but likely monophyletic, species groups, namely the C. mexicanus (Coues),C. nigrescens (J. A. Allen), C. parvus, and C. thomasi (Merriam) groups (Choate 1970; Woodman et al. 2003). Twoof these species groups occur in South America: C. nigrescens, with 2 species (we do not count as South AmericanC. merus Goldman, a member of the C. nigrescens group occurring in the Serranías de Darién and Pirre, along the
Panamanian-Colombian border; Woodman & Timm 1993); and C. thomasi (an exclusively South Americangroup), with 9 species. In South America, species of Cryptotis occur from the Venezuelan Cordillera de la Costaand the Venezuelan and Colombian Andes in the north, to the Ecuadorian and Peruvian Andes in the south, atelevations ranging from 1300 to 4055 m (Tate 1932; Ojasti & Mondolfi 1968; Vivar et al. 1997; Hutterer 2005;Woodman & Péfaur 2008; Quiroga-Carmona, 2011, in press).
Cryptotis was first recorded from Venezuela by Thomas (1898) on the basis of the specimens that he used todescribe Blarina meridensis, for which, according to Woodman (2002), the correct type locality is “Montes delValle Merida 2165 m” (ca. 08°41'N, 71° 05'W), in the Cordillera de Mérida. Osgood (1912) reported Blarinameridensis from the Andes south of the Cordillera de Mérida, namely “Paramo de Tama, head of Tachira River,Venezuela and Colombia”. Thomas (1912) transferred meridensis from Blarina Gray to Cryptotis. Cabrera (1958)recognized meridensis as a subspecies of C. thomasi. Ojasti and Mondolfi (1968) reported the first specimen ofCryptotis from the central part of the Cordillera de la Costa (10°28'N, 67°05'W) as C. thomasi. Handley (1976)reported Cryptotis to occur in both north (Cordillera de Mérida) and south (near Páramo del Tamá) of theDepresión del Táchira, and assigned all the specimens to C. thomasi. Hutterer (1986) recognized C. meridensis as afull species distinct from C. thomasi. Duarte and Viloria (1992) reported the first specimen of Cryptotis from theSierra de Perijá (10°24'N, 72°53'W), which they assigned to C. thomasi. The specimen was obtained in Colombianterritory a short distance from the northwestern border of Venezuela and Colombia, thus they concluded that itcame from a population distributed in both countries. Hutterer (1993) recognized 2 species of Cryptotis forVenezuela, C. thomasi (Páramo del Tamá) and C. meridensis (Cordillera de Mérida). Linares (1998) assigned allknown Venezuelan populations of Cryptotis (Páramo del Tamá, Cordillera de Mérida, Sierra de Perijá, Cordillerade la Costa) to C. m. meridensis. In the taxonomic arrangement that persists today (Hutterer 2005; Woodman &Péfaur 2008), Woodman (2002) assigned all the specimens from Páramo del Tamá and 3 Colombian localities inthe Departamento de Santander to a new species, C. tamensis (Type locality: 07°27'N, 72°26'W), which herecognized as distinct from C. thomasi, and C. meridensis. Woodman (2002) concluded that the Sierra de Perijáspecimen (Duarte & Viloria 1992) was not likely to be C. thomasi, and might possibly represent C. tamensis.However, Corredor-Carrillo and Muñoz-Saba (2007) still assigned 1 specimen from the Colombian side of theSierra de Perijá (10°15'N, 72°58'W) to C. thomasi. Finally, based on the original specimen reported by Ojasti andMondolfi (1968), and on 2 additional specimens from a nearby locality (10°25'N, 67°13'W), Quiroga-Carmona(2011, in press) described Cryptotis sp. A, which he considered likely to be endemic to the Cordillera de la Costa.
Recently, we obtained 3 specimens of Cryptotis from the Sierra de Aroa, an isolated mountain range innorthwestern Venezuela in which the presence of shrews was not previously known. Comparisons with specimensof Cryptotis from the Cordillera de la Costa, the Cordillera de Mérida, the Sierra de Perijá, and the Páramo delTamá, and comparisons with published descriptions of the species of Cryptotis found in Colombia, Ecuador, andPerú, indicate that the population from the Sierra de Aroa belongs to a new species referable to the C. thomasigroup. The purpose of this paper is to describe this new species.
Materials and methods
Specimens examined (Appendix 1) are housed in the following institutions: Colección de Vertebrados de laUniversidad de Los Andes (CVULA); Museo de Biología de la Universidad Central de Venezuela (MBUCV);Museo de Biología de La Universidad del Zulia (MBLUZ); Museo de Historia Natural La Salle (MHNLS); andMuseo de la Estación Biológica de Rancho Grande (EBRG).
Cranial terminology follows Gaughran (1954). Dental terminology follows Reppening (1967), Choate (1970),and Reumer (1984). We used in our analyses 5 external and 21 craniomandibular measurements (see Table 1 fornames and abbreviations), which are described and used in previous studies in Cryptotis systematics (Woodman &Timm 1993, 1999, 2000; Woodman 1996; Vivar et al. 1997). We recorded most cranial and mandibularmeasurements to the nearest 0.1 mm by using an ocular micrometer mounted in one of the oculars of a binocularstereomicroscope (Leica MZ 8 with a 1x planapochromatic objective). We also used this stereomicroscope to counttail scales per cm, and to examine external soft anatomy. Exceptions are condylobasal length (CBL) and cranialbreadth (CB), which we recorded using a Mitutoyo® handheld digital caliper with pointed jaws (model 573-221-20). External measurements are those recorded by the collector. All measurements are in millimeters.
We restrict comparisons to members of the C. thomasi group of Cryptotis, to which the new species belongs(see description). We compare the measurements of the specimens of the new species with those of our ownsamples of Cryptotis meridensis and C. tamensis (Table 1), and with those available for Cryptotis sp. A (Quiroga-Carmona, 2011, in press), C. meridensis, C. tamensis, C. thomasi, and C. medellinius (Woodman 2002); C.equatoris equatoris, C. e. osgoodi, and C. peruviensis (Vivar et al. 1997); C. montivagus (Barnett 1992; Vivar et al.1997), and C. squamipes (Allen, 1912). For C. meridensis and C. tamensis, we aggregated our own measurements(Table 1) to those of Woodman (2002). For 2 external measurements (HB, TL) of C. montivagus, we aggregated themeasurements of Barnett (1992) and Vivar et al. (1997) We used the resulting aggregated data for statistical tests.The aggregations were done by adding the sample sizes, by pooling the variances, and by pooling the means ofsamples without shared specimens. We pooled variances as explained by Rudmin (2010). To obtain, for a givenmeasurement, the pooled mean (or weighted average) of 2 samples of the same species, we used the methodexplained, with varying mathematical notations, in many statistics textbooks (e.g., Sokal & Rohlf, 1981: 41): first,the values N, v1, and v2 were obtained, as follows, N = n1 + n2, v1 = n1 × mean1, and v2 = n2 × mean2; then, the value
V was obtained, as follows, V = v1 + v2; finally, the pooled mean was obtained, as follows, pooled mean = V / N;where n1 and mean1 are the sample size and the mean of the first sample, and n2 and mean2 are the sample size andthe mean of the second sample.
We used 2-tailed two-sample t tests to compare, separately for each external or craniomandibularmeasurement, the mean of the new species with the corresponding mean of other species of the C. thomasi group.Because raw data were not available for most species, we used sample sizes, the means, the standard deviations,and statistical tables to carry out the t tests manually. We assumed equality of variances because the Levene’s testfor homogeneity of variances, carried out using SPSS version 12 for Windows, seldom detected significantdifferences between species in preliminary comparisons involving our raw data for the new species (n = 3, externaland craniomandibular measurements) and C. meridensis (n ≥ 22, external, and n = 32, craniomandibularmeasurements). To avoid an increase in the proportion of false positives (Type I errors) as a result of themultiplicity of measurements being compared in the t tests without disproportionally increasing the proportion offalse negatives (Type II errors), we used the sequential correction proposed by Benjamini and Hochberg (1995).
The qualitative differences that we enumerate in the description and comparison sections are based on thespecimens listed in Appendix 1, and on the comprehensive accounts of South American Cryptotis provided byWoodman (1996, 2002), Vivar et al. (1997), and Woodman and Péfaur (2008). In the qualitative comparisons of thenew species with C. meridensis and Venezuelan C. tamensis, we were able to consider both external (based on freshand ethanol-preserved specimens) and craniomandibular characters, whereas in the comparisons with othermembers of the C. thomasi group we were able to consider only craniomandibular characters. For qualitativecomparisons, we also took into account the original descriptions of all species of the C. thomasi group, andexamined good quality photographs of the skins, crania, and mandibles of the holotypes of C. squamipes and C.montivagus (American Museum of Natural History, AMNH M–32378, M–47200).
Results
Cryptotis aroensis, new speciesSierra de Aroa shrew
Holotype. CVULA I–8548, an adult male (Fig. 1A, 1B), with complete skull and mandibles (Figs. 2 and 3) plusethanol-preserved body, collected on March 12, 2011. Measurements in Table 1.
Type locality. Las Cumaraguas Sector, Sierra de Aroa, Municipio Cocorote, Estado Yaracuy, Venezuela(10°22'02.6''N, 68°49'20.4'W), elevation 1730 m. This site is a pristine patch of cloud forest near a dirt road andaround a small creek (Fig. 4). See Map (Fig. 5).
Paratypes. CVULA I–8546, an adult female, collected on March 8, 2011 (right hindfoot shown in Fig. 1;dentition in Fig. 3). CVULA I–8547, an adult male, collected on March 12, 2011 (dentition shown in Fig. 3). Bothobtained in the type locality, and prepared as skulls and mandibles (which are in good conditions, except for the leftangular process which is broken in CVULA I–8546) plus ethanol-preserved bodies. Measurements in Table 1.
Etymology. The species name aroensis [Aro(a) + ensis] is a toponym meaning “from Aroa.”
FIGURE 1. External appearance of Cryptotis aroensis. A, lateral view of the holotype (CVULA I–8548), which had a head-and-body length of 81mm, and a tail length of 40 mm (Table 1); B, ventral view of the holotype; C, ventral view of the righthindfoot of a paratype (CVULA I–8546), which had a length of 13 mm (Table 1). Photographs by J. Molinari.
Diagnosis. A member of the C. thomasi group (for characters defining this group, see Choate 1970; Woodmanet al. 2003; and below in description). Pelage rich grayish brown. Pinnae almost as broad as high, and thereforevisible frontally. Anterior border of zygomatic plate in line and above the metastyle of M1, and posterior border inline and above the paracone of M3. Palate at the level of the second molars narrower than that of any other knownmember of the C. thomasi group (M2B averaging 5.5 mm; Table 1). In occlusal view of the posterior palate, a smallpart of the nasal cavity and the ethmo-turbinals can be observed at each side of the vomer, which is the only bone ofthe roof of the mesopterygoid fossa penetrating the choanae. Lacrimal foramina wide and deep. A minute foramenon the posterior edge of the tympanic process of each petromastoid. U1, U2, and U3 with posterolingual cuspules.U4 small to medium-sized for a member of the C. thomasi group. U4 labially placed, thus U4 visible in lateralview. Upper molars conspicuously pigmented. M3 complex, and nearly as wide as M2. Bicuspulate lower incisors,each with a poorly developed, almost imperceptible, anterior cusp separated from the posterior cusp by a shallowanterior notch.
Description. External and craniodental measurements, and body masses of the holotype, and the 2 paratypes,are shown in Table 1.
External characters (Fig. 1).—A medium-sized and long-tailed member of the C. thomasi group (mean HB >72 mm is a diagnostic character for members of the C. thomasi group; Woodman et al. 2003), HB averaging 79.0mm, and tail averaging 46% of HB (Table 1). General coloration rich grayish brown. Long (about 6–7 mm),luxuriant fur (a diagnostic character for members of the C. thomasi group; Woodman et al. 2003). Hairs bicolored,with gray bases and brown tips. Dorsal pelage darker and brighter than ventral pelage, but lacking a cleardorsoventral demarcation. Tail unicolored. Muzzle with 8 or 9 pairs of postnasal warts, and with mystacialvibrissae reaching the ears. Upper half of philtrum with a smooth, long and narrow wart. Pinnae moderately long(standard ear length, which measure height, averaging 6.8 mm; Table 1) and broad (projecting about 6 mmlaterally), each with a laterally well-developed helix, antihelix and antitragus, and deep scapha. Flanks of body
without noticeable bare patches marking the location of lateral glands (a diagnostic character for members of the C.thomasi group; Woodman et al. 2003). Forefeet somewhat enlarged (a diagnostic character for members of the C.thomasi group; Woodman et al. 2003), robust, covered by abundant short hair, and having long central digits.Foreclaws, not broadened, relatively straight, and elongated (a diagnostic character for members of the C. thomasigroup; Woodman et al. 2003). Forefeet and hindfeet with moderately developed and circular thenar and hypothenareminences, and palmar pads. Thenar eminences slightly more distal than hypothenar eminences (Fig. 1C). Digitswith moderately developed, granular, and juxtaposed scales arranged in 2 longitudinal rows (Fig. 1C). Tail coveredwith short and coarse hairs, and with small scales (averaging 32 per cm) lacking sharp edges.
Cranial and mandibular characters (Fig. 2).—Nasal openings narrow (rectangular-shaped in occlusal view),fully encased laterally by the premaxillary bones. In occlusal view, zygomatic plate with its anterior border in lineand above the metastyle of M1, and its posterior border in line and above the paracone of M3. Palate short (PLaveraging 9.1 mm; Table 1), and very narrow at the level of M2 (M2B averaging 5.5 mm; Table 1). Posteriorbranch of maxillary process aligned over the mesostyle of M3, and not in touch with the posterior border of palate.Lacrimal foramina wide and deep. Apertures formed by the combined foramen rotundum and inferior orbitalfissure of each side (Gaughran 1954) wide and ovoid, located posteriorly on alisphenoids. Ovale foramina wide.The vomer is the only portion of the roof of the mesopterygoid fossa penetrating the choanae, thus, in occlusalview, a small part of the nasal cavity and the ethmo-turbinals can be observed at each side of the vomer (Fig. 3).Petromastoids each with a short, thin, and low anterior process. Posterior border of tympanic processes ofpetromastoids each with a minute foramen that is posterior in position to the paraoccipital process. Mandibles long(averaging 7.3 mm; Table 1), short behind m3 (AC3 averaging 5.1 mm; Table 1). Coronoid processes low (adiagnostic character for members of the C. thomasi group; Woodman et al. 2003), wide, and straight. Articularprocesses high, broad, and not robust (a diagnostic character for members of the C. thomasi group; Woodman et al.2003). Upper sigmoid notches angled. Lower sigmoid notches shallow to very shallow (a diagnostic character formembers of the C. thomasi group; Woodman et al. 2003).Angular processes straight, short (not surpassingposteriorly the condyle), broad at base, and blunt-tipped.
Dental characters (Figs. 2 and 3).—Teeth robust (nonbulbous). U1, U2, and U3 relatively narrow, and concaveto very concave on the posteroventral margin (a diagnostic character for members of the C. thomasi group;Woodman et al. 2003). U1, U2, and U3 with well-developed cuspules. U1, U2, U3, and U4 with low and broadcingula, and dark red cusps. U4 small to medium-sized for a member of the C. thomasi group, of approximatelycircular contour in occlusal view, and labially placed, thus, if cranium is rotated sideways as necessary, U4 visiblein lateral view. P4 with an anteriorly well developed and well pigmented parastyle, and a well pigmentedprotocone. M1 with the anterior element of ectoloph shorter than the posterior element (a diagnostic character formembers of the C. thomasi group; Woodman et al. 2003). M1 and M2 having well-developed and intenselypigmented protocones, little developed and unpigmented hypocones, and precentrocristae and postcentrocristaewith fully developed and very high labial sides. M3 complex (mesostyle, postcentrocrista, and metaconeconspicuous), and nearly as wide as M2. Bicuspulate lower incisors with dark red terminal halves, each with apoorly developed, almost imperceptible, anterior cusp separated from posterior cusp by a shallow anterior notch.Lower molars as follows: m1 and m2 with well-developed and pigmented entoconids and hypoconids; m3 havingan elongated talonid, a high and pigmented hypoconid, and a high and posteriorly well-developed hypoconulid.
Distribution. Known only from the type locality (Fig. 5). Cryptotis aroensis is most likely an endemic speciesto the highlands of the Sierra de Aroa, in northwestern Venezuela.
Comparisons. The morphometric values to which we refer in the comparisons are taken from Table 1, andfrom the sources mentioned above (Materials and methods). In the comparisons, we enumerate as morphometricdifferences only measurements whose means are significantly different (in each case we indicate p-values) betweenthe new species and other members of the C. thomasi group.
There is 1 unique distinguishing character of C. aroensis that, for the sake of brevity, we do not repeat in thecomparisons that follow. Namely, the species has a very narrow palate at the level of molars, i.e., the mean value ofM2B of C. aroensis is significantly (p < 0.01, or p << 0.01, depending on the species) smaller than the M2B valuesof all other species of the C. thomasi group, with the exception of C. peruviensis, of which only 2 specimens areknown (M2B is smaller but p is not significant). Moreover, the individual M2B values of the 3 specimens of C.aroensis (Table 1) fall below the range of the individual M2B values of the 146 specimens of the 10 species(including C. peruviensis, and C. squamipes) of the C. thomasi group for which cranial measurements are available(for the M2B value of the holotype of C. squamipes, see Allen 1912).
FIGURE 2. From top to bottom, dorsal, ventral, and lateral views of crania, and lateral view of mandible of the holotype(CVULA I–8548) of Cryptotis aroensis.
FIGURE 3. Occlusal view of the upper (right) and lower (left) dentitions showing variation in 3 specimens of Cryptotisaroensis. A, views of the holotype (CVULA I–8548); B, views of the paratype CVULA I–8546; C, views of the paratypeCVULA I–8547.
Comparison with Cryptotis sp. A Quiroga-Carmona, in press. –Cryptotis aroensis differs from this species,which occurs in the Venezuelan Cordillera de la Costa (Fig. 5), in having: 1) pelage rich grayish brown (as opposedto dark gray with a dark brown luster); 2) better developed pinnae that are nearly as broad as high, thus visiblefrontally, each with a laterally well-developed helix, antihelix and antitragus, and deep scapha (as opposed topinnae that are higher than broad, thus hidden in pelage, each with scarcely developed helix, antihelix, andantitragus, and a superficial scapha); 3) muzzle with a long and narrow wart restricted to the upper half of philtrum(as opposed to a long and narrow wart extending from the edge of the upper lip to the base of the rhinarium); 4)muzzle with 8 or 9 pairs of postnasal warts (as opposed to 4 pairs); 5) digits robust, with poorly developed, semi-square, and juxtaposed scales, arranged in 2 longitudinal rows (as opposed to digits delicate, with well-developed,granular, and juxtaposed scales, arranged in 3 longitudinal rows); 6) tail scales smaller, averaging 32 per cm (asopposed to larger, averaging 26 per cm); 7) in occlusal view, anterior margin of the mesopterygoid fossa coveredby the palatine extension of the hard palate only centrally (as opposed to anterior margin of the mesopterygoidfossa completely covered by the palatine extension of the hard palate); 8) U4 visible in lateral view if cranium is
rotated sideways as necessary (as opposed to U4 not visible in lateral view, even if rotated sideways); 9) M3 aboutas wide as M2 (as opposed to narrower than M2); 10) lower incisors with a well defined, albeit shallow, anteriornotch (as opposed to with an almost imperceptible anterior notch); 11) coronoid processes slightly higher thancondyles (as opposed to much higher than condyles); 12) angular processes narrow-tipped (as opposed to broad-tipped).
Comparison with Cryptotis meridensis (Thomas, 1898), and Cryptotis tamensis Woodman, 2002.–Cryptotisaroensis differs from either or both species (C. meridensis occurs in the Venezuelan Cordillera de Mérida, and C.tamensis occurs in the Colombian Cordillera Oriental, and adjacent Venezuelan Andes, namely in the Páramo delTamá; Fig. 5) in having: 1) head-and-body shorter (p << 0.01), HB averaging 79.0 mm (as opposed to head-and-body longer, HB averaging 86.0 mm, in C. tamensis); 2) pelage rich grayish brown (as opposed to dark gray orgray, with a brownish luster, in C. meridensis; and to chocolate-brown, in C. tamensis); 3) better developed pinnaethat are nearly as broad as high, thus visible frontally, each with a laterally well-developed helix, antihelix andantitragus, and deep scapha (as opposed to pinnae that are higher than broad, thus hidden in pelage, each withscarcely developed helix, antihelix, and antitragus, and a superficial scapha); 4) muzzle with a long and narrowwart restricted to the upper half of philtrum (as opposed to upper half of philtrum with a very short wart in contactwith the rhinarium); 5) muzzle with 8 or 9 pairs of postnasal warts (as opposed to 4 or 5 pairs); 6) digits robust,with poorly developed, semi-square, and juxtaposed scales (as opposed to delicate, with well-developed, granular,and juxtaposed scales); 7) tail scales smaller, averaging 32 per cm (as opposed to larger, averaging 30 per cm); 8)lacrimal foramina wide and deep (as opposed to very narrow and shallow); 9) in occlusal view, anterior margin ofthe mesopterygoid fossa covered by the palatine extension of the hard palate only centrally (as opposed to anteriormargin of the mesopterygoid fossa completely covered by the palatine extension of the hard palate); 10) palateshorter (p < 0.03), PL averaging 9.1 mm (as opposed to longer, PL averaging 9.5 mm, in C. tamensis); 11) palatenarrower at the level of the third unicuspid (p < 0.03), U3B averaging 3.2 mm (as opposed to broader, U3Baveraging 3.5–3.7 mm); 12) U4 small to medium-sized and visible in lateral view if cranium is rotated sideways asnecessary (as opposed to extremely reduced or absent, and not or little visible in lateral view when present, even ifrotated sideways, in C. meridensis); 13) M3 complex, and as wide as M2 (as opposed to simple in most specimens,and narrower than M2, in C. meridensis; and simple or complex, and narrower than M3, in C. tamensis); 14) lowerincisors bicuspulate, each with an almost imperceptible anterior cusp separated from the posterior cusp by ashallow anterior notch (as opposed to a well-developed anterior cusp separated from the posterior cusp by a deepanterior notch); 15) m1 and m2 with well-developed and pigmented entoconids (as opposed to m1 and m2 with lowentoconids showing pigmentation only on tips); 16) m3 having an elongated talonid, a high hypoconid, and aconspicuous hypoconulid (as opposed to a short talonid, and a medium hypoconid in C. meridensis and C.tamensis; no hypoconulid in C. meridensis; a reduced hypoconulid in C. tamensis); 17) angular processes short (notsurpassing posteriorly the condyle), broad at base, straight, and blunt tipped (as opposed to long and narrow).
Comparison with Cryptotis thomasi (Merriam, 1897).—Cryptotis aroensis differs from this species, whichoccurs near Bogotá, in the central part of the Colombian Cordillera Oriental, in having: 1) tail longer (p << 0.01),TL averaging 36.3 mm (as opposed to shorter, TL averaging 24.0 mm); 2) pelage rich grayish brown (as opposed tobrown); 3) a minute foramen on the posterior edge of the tympanic process of each petromastoid (as opposed to alarge foramen on the same position); 4) posterior border of zygomatic plate in line and above the paracone of M3(as opposed to above or slightly behind the M2–M3 border); 5) mandible shorter behind m3 (p << 0.01), AC3averaging 5.1 mm (as opposed to longer, AC3 averaging 5.7 mm); 6) upper sigmoid notch of each mandiblesemicircular, and much deeper than coronoid process and condyle (as opposed to upper sigmoid notch irregularly-shaped, and slightly deeper than coronoid process and condyle); 7) angular processes short (not surpassingposteriorly the condyle), and broad at base (as opposed to long and narrow).
Comparison with Cryptotis medellinius Thomas, 1921.—Cryptotis aroensis differs from this species, whichoccurs in the northern parts of the Colombian Cordillera Central and Cordillera Occidental, in having: 1) bodymass lower (p < 0.02), BM averaging 11.0 g (as opposed to higher, BM averaging 16.2 g); 2) pelage rich grayishbrown (as opposed to light brown); 3) a minute foramen on the posterior edge of the tympanic process of eachpetromastoid (as opposed to a huge foramen on the same position); 4) anterior border of zygomatic plate in line andabove the metastyle of M1, and posterior border in line and above the paracone of M3 (as opposed to anteriorborder of zygomatic plate at the level of the metacristae of M1, and posterior border above the metacristae ormetastyle of M2); 5) forehead narrower (p < 0.04), IO averaging 5.0 mm (as opposed to broader, IO averaging 5.4
mm); 6) palate shorter (p < 0.04), PL averaging 9.1 mm (as opposed to longer, PL averaging 9.8 mm); 7) unicuspidtoothrow shorter (p < 0.01), UTR averaging 2.7 mm (as opposed to longer, UTR averaging 3.0 mm; 8) U4 small tomedium-sized, and labially placed, thus U4 visible in lateral view (as opposed to U4 reduced and linguallydisplaced, with U3 and M1 almost in touch labially, thus U4 barely visible in lateral view).
Comparison with Cryptotis squamipes (J. A. Allen, 1912).—Cryptotis aroensis differs from this species, whichoccurs in Colombia (southern parts of the Cordillera Central and Cordillera Occidental, and Cordillera del Sur), inhaving: 1) external and cranial measurements (Table 1) apparently lesser (comparable measurements of theholotype of C. squamipes are: HB, 86; TL, 42; HL, 18, PL, 9.8; MTR, 6.0; M2B, 6.0; Allen 1912); 2) pelage richgrayish brown (as opposed to blackish or dark grayish brown; Allen 1912); 3) U1, U2, and U3 with posterolingualcuspules (as opposed to U1, U2, and U3 usually lacking posterolingual cuspules; Woodman & Péfaur 2008); 4) U4labially placed (as opposed to U4 lingually displaced; photograph of the holotype); 5) coronoid process lower,HCP/ML averaging 64% (as opposed to higher, HPC/ML averaging 71%; Table 1, Woodman & Péfaur 2008).
Comparison with Cryptotis equatoris equatoris (Thomas, 1912), and C. e. osgoodi (Stone, 1914).–Cryptotisaroensis differs from either or both taxa, which occur in the central Ecuadorian Andes, in having: 1) pelage richgrayish brown (as opposed to dark chocolate brown); 2) a minute foramen on the posterior edge of the tympanicprocess of each petromastoid (as opposed to absence of such foramen); 3) anterior border of zygomatic plate in lineand above the metastyle of M1, and posterior border in line and above the paracone of M3 (as opposed to anteriorborder of zygomatic plate at metacristae and slightly anterior to metastyle of M1, and posterior border behind theM2–M3 border); 4) longer cranium (p < 0.01), CBL averaging 21.3 mm (as opposed to shorter, CBL averaging20.7 in C. e. osgoodi); 5) cranium broader (p < 0.04), CB averaging 10.6 mm (as opposed to narrower, CBaveraging 9.9 mm, in C. e. osgoodi); 6) zygomatic plate narrower (p < 0.03), ZP averaging 1.9 mm (as opposed tobroader, CB averaging 2.2 in C. e. equatoris); 7) palate anteriorly broader (p < 0.01, both for U1B and U3B), U1Baveraging 2.8 mm, and U3B averaging 3.2 mm (as opposed to narrower, U1B averaging 2.4 mm, and U3Baveraging 2.9 mm, in C. e. osgoodi); 8) upper molars conspicuously pigmented (as opposed to almost entirelywhite owing to absence of red pigment on hypoconal basins); 9) upper toothrow longer (p < 0.03), TR averaging8.5 mm (as opposed to shorter, TR averaging 7.9–8.0 mm); 10) lower incisors bicuspulate, with an almostimperceptible anterior cusp (as opposed to tricuspulate); 11) coronoid process higher (p << 0.01), HCP averaging4.7 mm (as opposed to lower, HCP averaging 4.2 mm in C. e. osgoodi); 12) coronoid valley higher (p << 0.01),HCV averaging 3.2 mm (as opposed to lower, HCV averaging 2.8 mm); 13) articular condyle much higher (p <<0.01), HAC averaging 4.2 mm (as opposed to much lower, HAC averaging 2.6–2.7 mm); 14) lower sigmoid notchdeep (as opposed to shallow, or moderately deep); 15) m3 with both hypoconid and hypoconulid on talonid (asopposed to with only hypoconid).
Comparison with Cryptotis montivagus (Anthony, 1921).—Cryptotis aroensis differs from this species, whichoccurs in the southern Ecuadorian Andes, in having: 1) pelage rich grayish brown (as opposed to medium gray,with a speckled appearance); 2) a minute foramen on the posterior edge of the tympanic process of eachpetromastoid (as opposed to absence of such foramen); 3) anterior border of zygomatic plate in line and above themetastyle of M1, and posterior border in line and above the paracone of M3 (as opposed to anterior border ofzygomatic plate vertically oriented and above metacristae of M1, and posterior border above the M2–M3 contact);4) upper molars conspicuously pigmented (as opposed to almost entirely white owing to absence of red pigment onhypoconal basins); 5) U4 labially placed (as opposed to in line with other unicuspids); 6) articular condyle muchhigher (p << 0.01), HAC averaging 4.2 mm (as opposed to much lower, HAC averaging 3.0 mm); 7) lower incisorsbicuspulate, with an almost imperceptible anterior cusp (as opposed to conspicuously tricuspulate); 8) m3 withboth hypoconid and hypoconulid on talonid (as opposed to only hypoconid).
Comparison with Cryptotis peruviensis Vivar, Pacheco and Valqui, 1997.—Cryptotis aroensis differs from thisspecies, which occurs in the northern Peruvian Andes, in having: 1) pelage rich grayish brown (as opposed to darkgrayish brown); 2) nasal openings narrow, fully encased laterally by the premaxillary bones (as opposed to broad,partially encased laterally by the premaxillary bones); 3) a minute foramen on the posterior edge of the tympanicprocess of each petromastoid (as opposed to absence of such foramen); 4) posterior border of zygomatic plate inline and above the paracone of M3 (as opposed to posterior border of zygomatic plate above M3); 5) U4approximately circular in occlusal view, and labially placed, thus U4 visible in lateral view (as opposed to U4triangular in occlusal view, and lingually displaced, with U3 and M1 in touch labially, thus U4 usually not visible inlateral view); 6) lower incisors bicuspulate, with an almost imperceptible anterior cusp (as opposed to
conspicuously tricuspulate); 7) coronoid process relatively straight (as opposed to flared to labial side ofmandible); 8) articular condyle much higher (p < 0.01), HAC averaging 4.2 mm (as opposed to much lower, HACaveraging 2.9 mm); 9) angular process short and broad at base (as opposed to long and slender).
Remarks. Sexual dimorphism is not apparent in our small series of Cryptotis aroensis. This is consistent withobservations made on larger samples of C. meridensis and C. tamensis (Durant & Péfaur 1984; Woodman 2002).
Discussion
Biogeography. Based on the fossil record, present distribution, and morphological diversity of Cryptotis, Choate(1970: 297) suggested that the species groups of the genus, including the endemic South American C. thomasigroup, might have diverged during the late Pliocene (3.58 to 1.77 Ma), and that southern México was the center oforigin and dispersal of most recent taxa. Although the time at which Cryptotis arrived to South America cannot beestablished at present, the endemicity of the C. thomasi group, the presence of at least 12 well-differentiated (e.g.,comparisons with C. aroensis presented above) species of the genus in South America, and the fact that membersof the genus have been able to colonize mountain ranges beyond the Andes (Sierra de Aroa and Cordillera de laCosta: Ojasti & Modolfi 1968; Quiroga-Carmona 2011, in press; Fig. 5), indicates that the history of Cryptotis inSouth America may be longer and more complex than previously thought (e.g., Simpson 1945; Hershkovitz 1969).
Because tropical Cryptotis are fundamentally highland forms (the only exception is C. mayensis, which occursat low elevations <200 m in xeric habitats in the Yucatán Península; Choate 1970; Woodman 2010), overlandarrival of the genus to South American after the completion of the Panamanian isthmus, about 2.8 Ma, may haveoccurred during 1 or more of the about 10 glacial maxima (Woodburne 2010) of the last 2 million years, whenmuch cooler conditions (8 to 9°C less than present during the last glacial maximum) led to the altitudinal descent ofthe cloud forest and páramo vegetation belts in the mountains of Central and South America (Hooghiemstra &Cleef 1995; Islebe & Hooghiemstra 1997; Lachniet & Seltzer 2002; Hooghiemstra & Van Der Hammen 2004), thusmaking Cryptotis-inhabited habitats of southern Central America (Serranías de Darién and Pirre) geographicallyclose to equivalent habitats in the northern extreme of the Colombian Cordillera Occidental. Ancestors of theendemic South American C. thomasi group may have arrived at an earlier glacial maximum than ancestors of the 2South American species (C. brachyonyx, C. colombianus) of the C. nigrescens group because the latter still isprimarily Central American (Woodman 2003). The high degree of morphological differentiation among species ofthe C. thomasi group (e.g., comparisons with C. aroensis presented above) indicates that, as implied by Choate(1970: 297), even if speciation in the new Andean ranges was fast, the ancestor of the group most likely arrived toSouth America several glacial maxima ago.
The altitudinal descent of montane vegetation belts, concomitant with glacial maxima, were likewise necessaryfor the colonization by Cryptotis of the Cordillera de Mérida (from the Cordillera Oriental de Colombia), and of theSierra de Aroa and the Cordillera de la Costa (directly or indirectly from the Cordillera de Mérida). These mountainranges are separated by significant lowland barriers, namely the Depresiones del Táchira, Lara, and Yaracuy (Fig.5). At present, the gap of habitats (<1500 m elevation and lacking cloud forests) unsuitable or suboptimal forCryptotis created by the Depresión del Táchira is ∼20 km (Cordillera Oriental to Cordillera de Mérida), whereasthe gap created by the Depresión de Lara is ∼50 km (Cordillera de Mérida to Sierra de Aroa), the gap created by theDepresión de Yaracuy is ∼40 km (Sierra de Aroa to Cordillera de la Costa), and the gap created in combination bythe Depresiones de Lara and Yaracuy is ∼70 km (Cordillera de Mérida to Cordillera de la Costa). Colonization byCryptotis of Venezuelan mountain ranges is here hypothesized to have occurred during glacial maxima in which adescent of the montane vegetation belts caused these gaps to be greatly reduced. Given the greater magnitude (∼70km) of the gap separating the Cordillera de la Costa from the Cordillera de Mérida (Fig. 5), it is likely thatCryptotis arrived to the former (Ojasti & Mondolfi 1968; Quiroga-Carmona, 2011, in press) indirectly, from theSierra de Aroa. It should be noted that because the drier climates characteristic of glacial periods causes aridregions to expand (Hooghiemstra & Cleef 1995; Hooghiemstra & Van Der Hammen 2004), and because at thepresent interglacial the Depresiones del Táchira and Lara are partly arid, these lowland barriers may have createddifficulties for the dispersal of Cryptotis even during glacial maxima owing to an increase of xeric vegetation.
Topography of mountains and intermontane depressions is determined by the complex and rapidly changinginterplay between tectonics and climate. Although the Andes may have nearly reached their present appearance by
the time of the completion of the Panamanian isthmus (Díaz de Gamero 1996; Gregory-Wodzicki 2000), tectonicand erosional activity have continued at fast paces during the last 2 Ma. For example, current uplift rates of thecentral part of the Cordillera de Mérida range from 0.4 to 1.7 km/Ma (Bermúdez et al. 2011), erosion owing totopographic uplift and increasing aridity in the Peruvian Andes are capable of incising the terrain at rates of 0.5 to2.0 km/Ma (Schildgen et al. 2007; Thouret et al. 2007), and sediment accumulation rates in the SubandeanBolivian Chaco may reach 0.63 km/Ma (Uba et al. 2007). Therefore, it is likely that during the glacial maxima ofthe last 2 Ma the topographies of the lowland barriers (the Depresión de Atrato between the Serranías de Dariénand Pirre, and the Cordillera Occidental; Depresiones del Táchira, Lara, and Yaracuy) for the dispersal of Cryptotisin northern South America were variable, and, as a consequence, Cryptotis may have been able to disperse acrosseach barrier when a glacial maxima coincided with a favorable characteristic of the barrier, such as presence of arelatively high pass with a humid windward slope. Because of the topographical, climatic, and ecologicalpeculiarities of each lowland barrier and surrounding mountains, dispersal across different barriers most likely wasnot simultaneous. In this context, the fact that Cryptotis aroensis is morphologically more differentiated withrespect to C. meridensis than the latter with respect to C. tamensis would seem to indicate that, of the 3 species, C.aroensis is the one that has been geographically isolated for the longest time.
The Sierra de Aroa has not been fully inventoried faunistically or floristically, especially at higher elevations.However, it is known to share faunal and floristic elements with the Cordillera de la Costa (Meier 1998, 2005;Barrio 1999ab; Smith & Field 2001; Hilty 2003) with which it has a common geological origin (Meier 1998). Forexample, 8 birds [the apodiform Coeligena coeligena coeligena (Lesson), and the passeriforms Anabacerthiastriaticollis venezuelana (Hellmayr), Chamaeza turdina chionogaster (Hellmayr), Grallaricula loricata (Sclater),Lochmias nematura sororia (Sclater & Salvin), Pipreola formosa (Hartlaub), Pseudocolaptes boissonneautistriaticeps (Hellmayr & Seilern), and Syndactyla guttulata (Sclater)], 1 polichrotid lizard (Anolis squamulatusPeters), 1 plethodontid salamander (Bolitoglossa borburata Trapido), 1 hemiphractid frog (Gastrotheca walkeriDuellman), and 1 bufonid toad [Atelopus cruciger (Lichtenstein & Martens)], occur only in both mountain ranges(Barrio 1999ab; Hilty 2003; Rodríguez-Contreras et al. 2008; Ugueto et al. 2009). Equivalent montane forest beltstend to occur at lower elevations in coastal than in inland mountains (Bruijnzeel & Hamilton 2000). The Sierra deAroa and the Cordillera de la Costa are closer to the Caribbean than the Cordillera de Mérida (Fig. 5). Therefore,isolation of mountain biotas by elevation may have been less pervasive for the Sierra de Aroa/Cordillera de laCosta than for the Sierra de Aroa/Cordillera de Mérida.
Cryptotis aroensis is the first mammal species known to be endemic to the Sierra de Aroa. Previously endemicanimal species known from this mountain range include 1 hylid frog, Dendropsophus yaracuyanus (Mijares-Urrutia & Rivero), and 1 dragonfly, Philogenia polyxena Calvert (Mijares-Urrutia & Rivero 2000; De Marmels2008). In addition, 10 fish species are endemic to the Sierra de Aroa and the two small (∼4700 km2) northeastward-draining basins that surround it (Aroa Basin in the northwestern slope, Yaracuy Basin in the southeastern slope),namely 3 characiforms (Creagrutus lassoi Vari & Harold, Creagrutus lepidus Vari, Harold, Lasso and Machado-Allison, Hyphessobrycon fernandezi Fernández-Yépez),1 gymnotiform (Brachyhypopomus diazi Fernández-Yépez), and 6 siluriforms [Ancistrus nationi (Fernández-Yépez), Chaetostoma yurubiense Ceas and Page,Farlowella martini Fernández-Yépez, Hypostomus pagei Armbruster, Pseudopimelodus mathisoni (Fernández-Yépez), Trichomycterus arleoi Fernández-Yépez] (Rodríguez-Olarte et al. 2006). The cloud forests of the Sierra deAroa have been said to contain “a significative number” of endemic plant species (Huber 1986, 1997).
Natural history and conservation. The Sierra de Aroa is halfway between the farthest northeast portion of theCordillera de Mérida and the non-Andean Cordillera de la Costa (Fig. 5). It is an isolated, and modest-sized (about70 km long and 20–25 km wide) mountain range (Fig. 5), with elevations ranging from approximately 400 m, to1833 m at the Cerro Negro (10°25'N, 68°47'W), and to 1945 m at the Cerro Tigre (10°24'N, 68°48'W). As in allrelatively high tropical wet mountain ranges, cloud forests are present at higher elevations at which sufficient levelsof persistent cloud condensation are reached (Huber 1986, 1997; Bruijnzeel & Hamilton 2000).
At Parque Nacional Henri Pittier, in the Cordillera de la Costa, approximately 120 km to the east of the Sierrade Aroa, montane forest belts include semi-deciduous forest below 800–1000 m, transitional cloud forest below1300–1400 m, lower cloud forest below 1450–1550 m, and upper cloud forest below 1850–2400 m (Huber 1986).A description of altitudinal forest zonation in the Sierra de Aroa is not available. However, montane forest beltsthere appear to be equivalent to those described for the Parque Nacional Henri Pittier (Huber 1986). Given thegreater proximity to the Caribbean (Fig. 5), the upper elevation limits of montane forest belts in the Cordillera de la
Costa and the Sierra de Aroa tend to be, as expected (Bruijnzeel & Hamilton 2000), lower than in the more inlandCordillera de Mérida (Huber 1986, 1997). Throughout neotropical mountains, from México to Bolivia, the conifer,Podocarpus oleifolius D. Don, is an indicator of cloud and elfin forests (Eckenwalder 2009). This large tree (up to25 m high and 1 m diameter) is very common in the Sierra de Aroa at elevations above 1500 m (Veillon 1962;personal observation). In the Cordillera de la Costa, it occurs in upper cloud forests at 1700–2200 m (Veillon 1962;Meier 1998). In the Cordillera de Mérida, though sometimes it is found at 2000 m in lower cloud forests, it occursmainly at 2200–2500 m to 3300 m in upper cloud and elfin forests (Veillon 1962; Little 1986). Therefore, based onthe observed elevation range of P. oleifolius in the Sierra de Aroa, we assume that upper cloud forests there occur atelevations of 1500–1600 m to 1945 m. In addition to having dense stands of P. oleifolius, the upper cloud forests ofthe Sierra de Aroa are very rich in small palms (2 to 10 m high) of the genera Geonoma and Wettinia, and tree fernsof the genus Cyathea (Fig. 4). As in all cloud forests, epiphytes such as mosses, aroids, bromeliads, and orchidsgrow abundantly on tree branches.
FIGURE 4. Cloud forest at the type locality of Cryptotis aroensis; left, forest edge viewed from road; right, understoryvegetation. Photographed by J. Molinari on March 12, 2011.
Known elevation ranges for members of the Cryptotis thomasi group are: C. squamipes, 1500–3375 m; C.equatoris, 1600–4300 m; C. meridensis, 1670–3950 m; C. aroensis, 1730 m; C. medellinius, 2000–3800 m; C.peruviensis, 2050–3150 m; Cryptotis sp. A at the Cordillera de la Costa, 2100–2283 m; Cryptotis sp. at the Sierrade Perijá, 2100–2850 m; C. montivagus, 2300–4000 m; C. tamensis, 2385–3330 m; and C. thomasi, 2800–3500 m(Ojasti & Mondolfi 1968, Duarte & Viloria 1992, Corredor-Carrillo & Muñoz-Saba 2007, Tirira 2007, Woodman& Pefaur 2008, Quiroga-Carmona, 2011, in press; Appendix 1). There are very few records for members of thegroup at elevations below 2000 m, at even these seem to involve cloud forest localities. Woodman & Díaz dePascual (2004) erroneously claimed that C. meridensis occurs in seasonal forest (800–1700 m elevation), withscarce epiphytes and lianas, and a pronounced 1–3 months dry season. This claim was based on specimen CVULA
I–0821 (Appendix 1) obtained at an elevation of 1670 m (reported as 1640 m by Woodman & Díaz de Pascual2004), and on the generalization that, in the Cordillera de Mérida, seasonal forests occur below 1700 m, and cloudforests above 1700 (Soriano et al. 1999). However, the locality at which specimen CVULA I–0821 was obtained isin the very humid Llanos-facing southern slopes of the Cordillera de Mérida, where cloud forests typically descendto elevations of around 1600 m (Gómez & Molina 2007). We have visited this locality and verified that it is atypical Andean cloud forest, with abundant epiphytes, tree ferns, and white-leaved yagrumos (Cecropia telenitidaCuatrec., a conspicuous cloud forest tree). Therefore, available evidence indicates that members of the C. thomasigroup primarily are cloud forest, elfin forest, and páramo inhabitants.
FIGURE 5. Map of Western Venezuela and adjacent Colombia showing collection localities of specimens of small-earedshrews examined in the present study. Areas shaded in gray indicate elevations from 1000 to 2000 m. Areas shaded in greenindicate elevations above 2000 m. Species symbols are: Cryptotis aroensis (star), Cryptotis sp. A (inverted triangles), C.meridensis (solid circles), C. tamensis (triangle), and Cryptotis sp. from Sierra de Perijá (solid squares).
The only study of the diet of a member of the C. thomasi group, involving C. meridensis in a cloud forestenvironment, found that the bulk (nearly 70%) of the diet of the species is composed of earthworms and othersubterranean invertebrates (Díaz de Pascual & De Ascenção 2000). The robust forefeet of Crytotis aroensis (Fig.1A, 1B) seem to indicate that the species is also a mainly hypogeal feeder. We suggest that members of the C.thomasi group are restricted to cloud forests, elfin forests, and páramos for 4 reasons. First, as a result ofphylogenetic inertia, they retain the physiology of their Euroasian and North American ancestors, and still are bestadapted to colder environments. Second, as elevation increases, the decrease in abundance of soil litterinvertebrates (Olson 1994) is compensated by an increase in the abundance of earthworms (Zinck 1986; Rickart etal. 1991; Cavelier & Peñuela 1990; González et al. 2007), which are the most important prey type for C. meridensis
(Díaz de Pascual & De Ascenção 2000). Third, the soils of cloud forests, elfin forests, and páramos, unlike those ofhabitats found at lower elevations, have an open and porous mineral structure, and contain high proportions of softhumus and peat in their upper layers (Zinck 1986; Cavelier & Peñuela 1990; Buytaert et al. 2006), thus facilitatingthe excavation in search of prey by Cryptotis shrews. Fourth, most importantly, owing to such factors as frequentrains, daily presence of cloud cover and fog, sloped topography, and high water storage capacity of terrain, cloudforests, elfin forests, and páramos contain microhabitats in which soils are simultaneously and permanently moistand aerated (Zinck 1986; Bohlman et al. 1995; Bruijnzeel & Hamilton 2000; Buytaert et al. 2006), thus assuring ayear-round supply of invertebrate prey that Cryptotis shrews would not easily get in the more seasonal forests (inwhich soils typically fluctuate from waterlogged in the wet season, to desiccated in the dry season) found at lowerelevations.
The original extent of cloud forests worldwide was thought to be around 0.5 million km2 in the early 1970s,which is much less than lowland tropical forest. For example, Andean cloud forests originally occupied about 5%the area of Amazonian forests, in spite of which they are home to a similar number of plant species (Henderson etal. 1991; Bruijnzeel & Hamilton 2000; Brown & Kappelle 2001). The rate at which cloud forests are being lost isthe highest of all tropical forest biomes. As a result, up to 90% of the cloud forest in the Colombian Andes havealready been cleared, mostly for pastures and agriculture (Bruijnzeel & Hamilton 2000). In Venezuela, the extent ofcloud forests was about 9300 km2 in the late 1980s. Since then, up to 32% of these forests have been cleared,mostly for pastures that are used for extensive cattle-raising with low productivities (Gómez & Molina 2007;Oliveira-Miranda et al. 2010). In the late 1980’s, nearly 100% of Venezuelan cloud forests had low levels of humanintervention, whereas at present, up to 87% of these forests have medium to very high levels of humanintervention, and as a consequence now most Venezuelan cloud forests are categorized as either endangered (EN)or critically endangered (CR) ecosystems (Oliveira-Miranda et al. 2010). It should be noted that thesecategorizations ignore global climate change, to which cloud forests are likely to be extremely vulnerable, even inthe conservative scenario of a global surface temperature increase of 1–2°C (which might be higher in the Andes)within the 21st century. Among other potentially catastrophic consequences, this temperature increase wouldrequire cloud forest organisms to migrate ∼800 m upslope to avoid extinction (Still et al. 1999; Foster 2001;Nadkarni & Solano 2002; Bush et al. 2004; Sekercioglu et al. 2008; Hillyer & Silman 2010; Herzog et al. 2011).Indeed, several studies have already documented substantial upslope shifts of elevation ranges for a diversity oftropical montane animals as a likely response to observed global warming during the last few decades (LaVal 2004;Peh 2007; Seimon et al. 2007; Raxworthy et al. 2008; Chen et al. 2009).
Based on satellite data, we estimate the area of the Sierra de Aroa to be 141.8 km2 above 1300 m, 112.9 km2
above 1400 m, 37.4 km2 above 1500 m, 23.4 km2 above 1600 m, 9.6 km2 above 1700 m, and 0.81 km2 above 1800m. Assuming that the optimal habitat for C. aroensis is restricted to the cloud forest montane belt (1500–1600 to1945 m), and assuming the species has no microhabitat restrictions within the forest, and that it that can survive inthe already present pastures and agricultural fields, at present the maximum potential extent of its optimal habitat
would be 40–24 km2.In the Sierra de Aroa, large tracts of forest below 1500 m have been cleared for cattle-raising, agriculture, or by
fires, and cloud forests above 1500 m are starting to be fragmented and are therefore under threat of conversion topastures and agricultural land within the next years (ParksWatch 2001; Oliveira-Miranda et al. 2010). Part of thesecloud forests are within the limits of the Parque Nacional Yurubí, which is better preserved than other Venezuelannational parks (ParksWatch 2001). This suggests that C. aroensis is not in immediate risk of extinction unless anextraordinary event, such as a huge fire, eliminates upland forests in the Sierra de Aroa. However, the species canbe categorized as endangered (EN) by application of criteria B2a (its area of occupancy is estimated to be less than500 km2, and it is known to exist at no more than five locations) and E (probability of extinction in the wild is atleast 20% within 20 years) of the IUCN Red List Categories and Criteria, Version 3.1 (IUCN 2001). We think thatcriterion E is applicable because there is a high probability that the habitat of C. aroensis will vanish within 20years as a result of (1) deforestation for cattle-raising or agriculture, or by fires, and (2) an impossibility for cloudforests, imposed by the relatively low elevations (maxima at 1833 and 1945 m), and by the small area of the
summits (9.6 to 0.81 km2) of the Sierra de Aroa, to increase sufficiently their elevation range in response to globalwarming. We also fear that global warming will ultimately cause the disappearance of these cloud forests, and theextinction of C. aroensis, by the end of the century.
This project was carried out using field and laboratory equipment supplied by the International Foundation forScience (Stockholm, Sweden), which was provided to Jesús Molinari for the project B/2637, entitled ‘‘The aerialinsectivorous bat fauna of Venezuela: distribution, abundance, ecology, and conservation needs”. We thank thefollowing curators for allowing us to examine specimens under their care (see names and acronyms of institutionsin “Materials and methods”): Pascual Soriano, CVULA; Francisco Bisbal and Javier Sánchez, EBRG; RossanaCalchi and David Prieto, MBLUZ; Carmen Ferreira and Mercedes Salazar, MBUCV; and Arnaldo Ferrer, MHNLS.
References
Allen, J.A. (1912) Mammals from western Colombia. Bulletin of the American Museum of Natural History, 31, 71–95.Anthony, H.E. (1921) Preliminary report on Ecuadorean mammals. No. 1. American Museum Novitates, 20, 1–6.Barnett, A.A. (1992) Notes on the ecology of Cryptotis montivaga Anthony, 1921 (Insectivora, Soricidae), a high-altitude
shrew from Ecuador. Mammalia, 56, 587–592.Barrio, C.L. (1999a) Geographic distribution: Caudata: Bolitoglossa borburata. Herpetological Review, 30, 105.Barrio, C.L. (1999b) Geographic distribution: Anura: Gastrotheca walkeri. Herpetological Review, 30, 106.Benjamini, Y. & Hochberg, Y. (1995) Controlling the false discovery rate: a practical and powerful approach to multiple
testing. Journal of the Royal Statistical Society, Series B, 57: 289–300.Bermúdez, M.A., Van der Beek, P. & Bernet, M. (2011) Asynchronous Miocene–Pliocene exhumation of the central
Venezuelan Andes. Geology, 39, 139–142.Bohlman, S.A., Matelson, T.J. & Nadkarni, N.M. (1995) Moisture and temperature patterns of canopy humus and forest floor
soil of a montane cloud forest, Costa Rica. Biotropica, 27, 13–19.Brown, A.D. & Kappelle, M. (2001) Introducción a los bosques nublados del neotrópico: una síntesis regional. In: Kappele, M.,
& Brown, A.D (Eds.) Bosques nublados del neotrópico. Instituto Nacional de Biodiversidad, Heredia, Costa Rica, pp.25–40.
Bruijnzeel, L.A. & Hamilton, L.S. (2000) Decision time for cloud forests. UNESCO, Division of Water Sciences, InternationalHydrological Programme, Paris, 40 pp.
Bush, M.B., Silman, M.R. & Urrego, D.H. (2004) 48,000 Years of climate and forest change in a biodiversity hot spot. Science,3003, 827–829.
Buytaert, W., Célleri, R., De Bièvre, B., Cisneros, F., Wyseure, G., Deckers, J. & Hofstede, R. (2006) Human impact on thehydrology of the Andean páramos. Earth-Science Reviews, 79, 53–72.
Cabrera, A. (1958) Catálogo de los mamíferos de América del Sur. Revista del Museo Argentino de Ciencias Naturales"Bernardino Rivadavia", 4, 1–308.
Cavelier, J. & Peñuela, M.C. (1990) Soil respiration in the cloud forest and dry deciduous forest of Serrania de Macuira,Colombia. Biotropica, 22, 346–352.
Choate, J.R. (1970) Systematics and zoogeography of Middle American shrews of the genus Cryptotis. University of KansasPublications, Museum of Natural History, 19, 195–317.
Chen, I.C., Shiu, H.J., Benedick, S., Holloway, J.D., Cheye, V.K., Barlow, H.S., Hill, J.K. & Thomas, C.D. (2009) Elevationincreases in moth assemblages over 42 years on a tropical mountain. Proceedings of the National Academy of Sciences ofthe United States of America, 106, 1479–1483.
Corredor-Carrillo, D.A. & Muñoz-Saba, Y. (2007) Mamíferos de la alta montaña de Perijá. In: Rangel-Ch., J.O. (Ed.) Colombiabiodiversidad biótica V: la alta montaña de la Serranía de Perijá. Universidad Nacional de Colombia, Bogotá, pp.221–233.
De Marmels, J. (2008) Elfo de Aroa Philogenia polyxena Calvert 1924. In: Rodríguez, J.P., & Rojas-Suárez, F., (Eds.) Librorojo de la fauna venezolana. Provita & Shell Venezuela S.A, Caracas, p. 286.
Díaz de Gamero, M.L. (1996) The changing course of the Orinoco River during the Neogene: a review. Palaeogeography,Palaeoclimatology, Palaeoecology, 123, 385–402.
Díaz de Pascual, A. & De Ascenção, A.A. (2000) Diet of the cloud forest shrew Cryptotis meridensis (Insectivora: Soricidae) inthe Venezuelan Andes. Acta Theriologica, 45, 13–24.
Duarte, M.A. & Viloria, A.L. (1992) Nuevo hallazgo de Cryptotis thomasi (Merriam, 1897) (Mammalia: Insectivora) en laSierra de Perija, Noreste de Colombia. Acta Científica Venezolana, 43, 240–242.
Durant, P. & Péfaur, J. (1984) Sistemática y ecología de la musaraña de Mérida, Soricidae: Insectivora, Cryptotis thomasi.Revista de Ecologia, Conservacion y Ornitología Latinoamericana, 1(2), 3–14.
Eckenwalder, J.E. (2009) Conifers of the World: the complete reference. Timber Press, Portland, 720 pp.Foster, P. (2001) The potential negative impacts of global change on tropical montane cloud forests. Earth-Science Reviews, 55,
Gaughran, G.R.L. (1954) A comparative study of the osteology and myology of the cranial and cervical regions of the shrew,Blarina brevicauda, and the mole, Scalopus aquaticus. University of Michigan Museum of Zoology, MiscellaneousPublications, 80, 1–82.
Gómez, H.A. & Molina, M. (2007) Principales causas de la deforestación en la vertiente sur del Parque Nacional SierraNevada, Venezuela. Revista Forestal Venezolana, 51, pp. 25–34.
González, G., García, E., Cruz, V., Borges, S., Zalamea, M. & Rivera, M.M. (2007) Earthworm communities along an elevationgradient in Northeastern Puerto Rico. European Journal of Soil Biology, 43, 24–32.
Gregory-Wodzicki, K.M. (2000) Uplift history of the Central and Northern Andes: a review. Geological Society of AmericaBulletin, 112, 1091–1105.
Handley, C.O., Jr. (1976) Mammals of the Smithsonian Venezuelan Project. Brigham Young University Science Bulletin(Biological Series), 20 (5), 1–91.
Henderson, A., Churchill, S.P. & Luteyn, J. (1991) Neotropical plant diversity. Nature, 229, 44–45.Hershkovitz, P. (1969) The recent mammals of the Neotropical Region: a zoogeographic and ecological review. Quarterly
Review of Biology, 44, 1–70.Herzog, S.K., Martínez, R., Jørgensen, P.M. & Tiessen, H., Eds. (2011) Climate Change and Biodiversity in the Tropical Andes.
Inter-American Institute for Global Change Research & Scientific Committee on Problems of the Environment, São Josédos Campos, SP, Brazil, 348 pp.
Hilty, S.L. (2003) Birds of Venezuela. Princeton University Press, Princeton, 878 pp.Hooghiemstra, H. & Cleef, A.M. (1995) Pleistocene climatic change and environmental and generic dynamics in the north
Andean montane forest and páramo. In: Churchill, S.P., Balslev, H., Forero, E. & Luteyn, J.L. (Eds.) Biodiversity andconservation of neotropical montane forests. New York Botanical Garden, pp. 35–49.
Hooghiemstra, H. & Van Der Hammen, T. (2004) Quaternary Ice-Age dynamics in the Colombian Andes: developing anunderstanding of our legacy. Philosophical Transactions: Biological Sciences, 359, 173–181.
Huber, O. (1986) Las selvas nubladas de Rancho Grande: observaciones sobre su fisionomía, estructura y fenología. In: Huber,O. (Ed.) La selva nublada de Rancho Grande, Parque Nacional “Henri Pittier”: el ambiente físico, ecología vegetal yanatomía vegetal. Fondo Editorial Acta Científica, Caracas, pp. 131–170.
Huber, O. (1997) Ambientes fisiográficos y vegetales de Venezuela. In: La Marca, E. (Ed.) Vertebrados actuales y fósiles deVenezuela. Museo de Ciencia y Tecnología, Mérida, Venezuela, pp.279–298.
Hutterer, R. (1986) Südamerikanische Spitzmäuse: Cryptotis meridensis und C. thomasi als verschiedene Arten. Zeitschrift fürSaugetierkunde, 51 (Sonderheft), 33–34.
Hutterer, R. (1993) Order Insectivora. In: Wilson, D.E. & Reeder, D.M. (Eds.) Mammal species of the world, a taxonomic andgeographic reference. Smithsonian Institution Press, Washington, D.C., pp. 69–130.
Hutterer, R. (2005) Order Soricomorpha. In: Wilson, D.E. & Reeder, D.M. (Eds.) Mammal species of the world, a taxonomicand geographic reference. John Hopkins University Press, Baltimore, pp. 220–311.
Hillyer, R. & Silman, M.R. (2010) Changes in species interactions across a 2.5 km elevation gradient: effects on plantmigration in response to climate change. Global Change Biology, 2010, 1–10.
Islebe, G.A. & H. Hooghiemstra (1997) Vegetation and climate history of montane Costa Rica since the last glacial. QuaternaryScience Reviews, 16, 589–604.
IUCN (2001) IUCN Red List Categories and Criteria: Version 3.1. International Union for Conservation of Nature, Gland,Switzerland, 30 pp.
Lachniet, M.S. & Seltzer, G.O. (2002) Late Quaternary glaciation of Costa Rica. Geological Society of America Bulletin, 114,547–558.
LaVal, R.K. (2004) Impact of global warming and locally changing climate on tropical cloud forest bats. Journal ofMammalogy, 85, 237–244.
Linares, O.J. (1998) Mamíferos de Venezuela. Sociedad Conservacionista Audubon de Venezuela, Caracas, 691 pp.Little, E.L. (1986) Arboles comunes de Venezuela. Ciateáceas-Podocarpáceas-Cupresáceas-Gramíneas-Palmeras.
Universidad de Los Andes, Mérida, 47 pp.Meier, W. (1998) Flora und Vegetation des Ávila-Nationalparks (Venezuela/Küstenkordillere) unter besonderer
Berücksichtigung der Nebelwaldstufe. Dissertationes Botanicae, 296, 1–485.Meier, W. (2005) Aspectos de la flora y vegetación del Monumento Natural Cerro Platillón (Juan Germán Roscio), Estado
Guárico, Venezuela. Acta Botánica Venezuelica, 28, 39–62.Merriam, C.H. (1897) Descriptions of five new shrews from Mexico, Guatemala, and Colombia. Proceedings of the Biological
Society of Washington, 11, 227–230.Mijares-Urrutia, A. & Rivero, R. A. (2000) A new treefrog from the Sierra de Aroa, Northern Venezuela. Journal of
Herpetology, 34, 80–84.Nadkarni, N.M. & Solano, R. (2002) Potential effects of climate change on canopy communities in a tropical cloud forest: an
experimental approach. Oecologia, 131, 580–586.Ojasti, J. & Mondolfi, E. (1968) Esbozo de la fauna de mamíferos de Caracas. In: Crema, M. (Ed.) Estudio de Caracas,
Volumen 1. Ecología vegetal y fauna. Universidad Central de Venezuela, Caracas, pp. 410–461.Oliveira-Miranda, M.A., Huber, O., Rodríguez, J.P., Rojas-Suárez, F., De Oliveira-Miranda, R., Hernández-Montilla, M. &
Zambrano-Martínez, S. (2010) Riesgo de eliminación de los ecosistemas terrestres de Venezuela. In: Rodríguez, J.P.,
Rojas-Suárez, F., & Giraldo Hernández, D. (Eds.) Libro rojo de los ecosistemas terrestres de Venezuela. Provita, ShellVenezuela & Lenovo Venezuela, Caracas, pp. 107–235.
Olson, D.M. (1994) The distribution of leaf litter invertebrates along a neotropical altitudinal gradient. Journal of TropicalEcology, 10, 129–150.
Osgood, W.H. (1912) Mammals from western Venezuela and eastern Colombia. Field Museum of Natural History, ZoologicalSeries, 10, 33–66.
ParksWatch (2001) Venezuela: Parque Nacional Yurubí. http://www.parkswatch.org/parkprofiles/pdf/yunp_spa.pdfPomel, A. (1848) Études sur les carnassiers insectivores (extrait). Seconde partie, Classification des insectivores. Archive des
Sciences Physiques et Naturelles, Genève, 9, 244–251.Peh, K.S.-H. (2007) Potential effects of climate change on elevational distributions of tropical birds in southeast Asia. The
Condor, 109, 437–441.Quiroga-Carmona, M. (2011) Taxonomía del género Cryptotis Pomel 1848 (Mammalia: Soricomorpha: Soricidae) en
Venezuela. Licenciatura Thesis, Universidad de Carabobo, Valencia, Venezuela.Quiroga-Carmona, M. (in press) Description of a new species of small-eared shrew of the genus Cryptotis Pomel 1848
(Mammalia: Soricomorpha: Soricidae) from the Cordillera de la Costa, northern Venezuela. Studies on Neotropical Faunaand Environment.
Raxworthy, C.J., Pearson, R.G., Rabibisoa, N., Rakotondrazafy, A.M., Ramanamanjato, J.B., Raselimanana, A.P., Wu, S.,Nussbaum, R.A. & Stone, D.A. (2008) Extinction vulnerability of tropical montane endemism from warming and upslopedisplacement: a preliminary appraisal for the highest massif in Madagascar. Global Change Biology, 14, 1703–1720.
Reumer, J.W.F. (1984) Ruscinian and early Pleistocene Soricidae (Insectivora, Mammalia) from Tegelen (The Netherlands) andHungary. Scripta Geologica, 73, 1–173.
Reppening, C.A. (1967) Subfamilies and genera of the Soricidae – classification, historical zoogeography, and temporalcorrelation of the shrews. U.S. Geological Survey Professional Paper 565, 1–74.
Rickart, E.A., Heaney, L.R. & Utzurrum, R.C.B. (1991) Distribution and ecology of small mammals along an elevationaltransect in southeastern Luzon, Philippines. Journal of Mammalogy, 72, 458–469.
Rodríguez-Contreras, A., Señaris, J.S., Lampo, M. & Rivero, R. (2008) Rediscovery of Atelopus cruciger (Anura: Bufonidae):current status in the Cordillera de La Costa, Venezuela. Oryx, 42, 301–304.
Rodríguez-Olarte, D., Amaro, A., Coronel, J. & Taphorn, D.C. (2006) Los peces del río Aroa, cuenca del Caribe, Venezuela.Memoria de la Fundación La Salle de Ciencias Naturales, 164, 101–127.
Rudmin, J.W. (2010) Calculating the exact pooled variance. arXiv:1007.1012v1 [physics.data-an].Schildgen, T.F., Hodges, K.V., Whipple, K.X., Reiners, P.W. & Pringle, M.S. (2007) Uplift of the western margin of the Andean
plateau revealed from canyon incision history, southern Peru. Geology, 35, 523–526.Seimon, T.A., Seimon, A., Daszak, P., Halloy, S.R.P., Schloegel, L.M., Aguilar, C.A., Sowell, P., Hyatt, A.D., Konecky, B. &
Simmons, J.E. (2007) Upward range extension of Andean anurans and chytridiomycosis to extreme elevations in responseto tropical deglaciation. Global Change Biology, 13, 288–299.
Simpson, G.G. (1945) The principles of classification and a classification of mammals. Bulletin of the American Museum ofNatural History, 85, 1–350.
Smith, R.F. & Field, A. (2001) Aspectos de la ecología de Gyranthera caribensis Pittier (Bombacaceae) y su implicación en laconservación de algunos bosques del norte de Venezuela. Acta Botánica Venezuelica, 24, 144–202.
Sokal, R.R. & Rohlf, J.F. (1981) Biometry. W. H. Freeman, New York, 859 pp.Soriano, P.J., Díaz De Pascual, A., Ochoa, J., & Aguilera, M. (1999) Biogeographic analysis of the mammal communities in the
Venezuelan Andes. Interciencia, 24, 17–25.Still, C.J., Foster, P.N. & Schneider, S.H. (1999) Simulating the effects of climate change on tropical montante cloud forests.
Nature, 398, 608–610.Stone, W. (1914) On a collection of mammals from Ecuador. Proceedings of the Academy of Natural Sciences of Philadelphia,
66, 9–19. Tate, G.H.H. (1932) Distribution of the South American shrews. Journal of Mammalogy, 13, 223–228.Thomas, O. (1898) On seven new small mammals from Ecuador and Venezuela. Annals and Magazine of Natural History,
Seventh Series, 1, 451–457.Thomas, O. (1912) Three small mammals from S. America. Annals and Magazine of Natural History, Eighth Series, 9,
408–410.Thomas, O. (1921) New Cryptotis, Thomasomys, and Oryzomys from Colombia. Annals and Magazine of Natural History,
on canyon incision and Miocene uplift of the Central Andes in Peru. Earth and Planetary Science Letters, 263, 151–166.Tirira, D. (2007) Mamíferos del Ecuador: guía de campo. Ediciones Murciélago Blanco, Quito, 576 pp.Uba, C.E., Strecker, M.R. & Schmitt, A.K. (2007) Increased sediment accumulation rates and climatic forcing in the central
Andes during the late Miocene. Geology, 35, 979–982.Ugueto, G.N., Rivas, G., Barros, T. & Smith, E.N. (2009) A revision of the Venezuelan anoles II: redescription of Anolis
squamulatus Peters 1863 and Anolis tigrinus Peters 1863 (Reptilia: Polychrotidae). Caribbean Journal of Science, 45,30–51.
Veillon, J.P. (1962) Coníferas autóctonas de Venezuela, los Podocarpus, con especial énfasis sobre las podocarpáceas de laregión central del Estado Mérida – Venezuela. Universidad de Los Andes, Mérida, 159 pp.
Vivar, E., Pacheco, V. & Valqui, M. (1997) A New Species of Cryptotis (Insectivora: Soricidae) from Northern Peru. AmericanMuseum Novitates, 3202, 1–15.
Woodburne, M.O. (2010) The Great American Biotic Interchange: dispersals, tectonics, climate, sea level and holding pens.Journal of Mammalian Evolution, 17, 245–264.
Woodman, N. (1996) Taxonomic status of the enigmatic Cryptotis avia (Mammalia: Insectivora: Soricidae), with comments onthe distribution of the Colombian small-eared shrew, Cryptotis colombiana. Proceedings of the Biological Society ofWashington, 109, 409–418.
Woodman, N. (2002) A new species of small-eared shrew from Colombia and Venezuela (Mammalia: Soricomorpha:Soricidae: Genus Cryptotis). Proceedings of the Biological Society of Washington, 115, 249–272.
Woodman, N. (2003) A new small-eared shrew of the Cryptotis nigrescens-group from Colombia (Mammalia: Soricomorpha:Soricidae). Proceedings of the Biological Society of Washington, 116, 853–872.
Woodman, N. (2010) Two new species of shrews (Soricidae) from the western highlands of Guatemala. Journal ofMammalogy, 91, 566–579.
Woodman, N., Cuartas-Calle, C.A. & Delgado-V, C.A. (2003) The humerus of Cryptotis colombiana and its bearing on thespecies’ phylogenetic relationships (Soricomorpha: Soricidae). Journal of Mammalogy, 84, 832–839.
Woodman, N., & Díaz de Pascual, A. (2004) Cryptotis meridensis. Mammalian Species, 761, 1–5.Woodman, N. & Péfaur, J. (2008) Magnorder Epitheria. Order Soricomorpha Gregory, 1910. In: Gardner, A.L. (Ed.) Mammals
of South America, Volume 1, Marsupials, Xenarthrans, Shrews, and Bats. University of Chicago Press, Chicago, pp.177–187.
Woodman, N. & Timm, R.M. (1993) Intraspecific and interspecific variation in the Cryptotis nigrescens species complex ofsmall-eared shrews (Insectivora: Soricidae), with the description of a new species from Colombia. Fieldiana, Zoology, 74:1–30.
Woodman, N. & Timm, R.M. (1999) Geographic variation and evolutionary relationships among broad-clawed shrews of theCryptotis goldmani-group (Mammalia: Insectivora: Soricidae). Fieldiana, Zoology, 91: 1–35.
Woodman, N. & Timm, R.M. (2000) Taxonomy and evolutionary relationships of Phillip's small eared shrew, Cryptotisphillipsi (Schaldach,1966), from Oaxaca, Mexico (Mammalia: Insectivora: Soricidae). Proceedings of the BiologicalSociety of Washington, 113, 339–355.
Zinck, A. (1986) Características y fragilidad de los suelos en ambiente de selva nublada: el ejemplo de Rancho Grande. In:Huber, O. (Ed.) La selva nublada de Rancho Grande, Parque Nacional “Henri Pittier”: el ambiente físico, ecologíavegetal y anatomía vegetal. Fondo Editorial Acta Científica, Caracas, pp. 31–66.
Appendix 1: Specimens examined
Specimens for which external measurements were available (Table 1) are indicated with a large solid point.Specimens whose skulls we measured (Table 1) are indicated with an asterisk. Specimens are ordered from lowerto higher elevations.
Cryptotis meridensis (n = 57).—VENEZUELA (Mérida): km 36 road Apartaderos–Barinas, 10 km E SantoDomingo, 08°52'N, 70°37'W, elevation 1670 m (CVULA I–0821•∗); El Salvajal, 8 km W Mérida, 08°36'N,71°13'W, elevation 2000 m (CVULA I–2458∗); Monte Zerpa, 6 km N Mérida, 08º38'N, 71º09'W, elevation2160 m (CVULA I–1211∗, 5888•, 6890•, 6891•∗); Escagüey, 1.5 km SSW Mucurubá, 08°42'N,71°00'W,elevation 2200 m (CVULA I–6899•∗, 6900•∗); La Carbonera, 14 km SE La Azulita, 08º37'N, 71º22'W,elevation 2250 m (CVULA I–7912); NE de Laguna Brava, Páramo de Mariño, 7 km W Tovar, 08º18'N,71º49'W, elevation 2300 m (CVULA I–2352•); Asentamiento Monterrey, 8 km NNE Mérida, 08°41'N,71°07'W, elevation 2300 m (CVULA I–3502•, 7078•); 5.5 km E and 2 km S Tabay, 08°37'N, 71°02'W,elevation 2630 m (EBRG 3639•∗, 3640•∗, 3641•∗); Páramo de San José, 4.7 km SSE San José, 08°20'N,71°18'W, elevation 3100 m (CVULA I–8537•∗, 8538∗, 8539•, 8540•∗, 8541•∗, 8542∗, 8543•∗, 8544∗,8545•); Laguna Negra, Páramo de Mucubají, 08°47'N, 70°48'W, elevation 3500 m (MBUCV I–1882–1885,2773, MHNLS 11221); 7.5 m E and 6 km S Tabay, 08°35'N, 71°00'W, elevation 3545 m (EBRG 3642•);near Laguna de Mucubají, Páramo de Mucubají, 08°47'N, 70°49'W, elevation 3600 (CVULA I–0333•∗,0334•∗, 0335•∗, 0337•∗, 0359•∗, 0421•∗, 0766∗, 0874∗, 1168•∗, 1486•∗, 1487–1496, 1650, 1762∗, 8549∗;
MHNLS 11220•∗). VENEZUELA (Táchira): Cerro Alto Duque, 2.5 km SE El Cobre, 08°01'N, 72°03'W,elevation 2500 m (MBLUZ M–0167•∗); Páramo El Zumbador, 10 km SSW El Cobre, 07°58'N, 72°06'W,elevation 2750 m (CVULA I–5744∗).
Cryptotis tamensis (n = 3).—VENEZUELA (Táchira): Buena Vista, near Páramo del Tamá, 07°27'N, 72°26'W,elevation 2380–2415 m (EBRG 3643, 3644, 11743).
Cryptotis sp. (probably C. tamensis, according to Woodman 2002; C. thomasi, according to Duarte and Viloria1992, and Corredor-Carrillo & Muñoz-Saba 2007), (n = 2).—COLOMBIA (César): Río Manaure, 1 kmFinca “El Suspiro”, Serranía de Valledúpar, 10°24'N, 72°53'W, elevation 2100 m (MBLUZ M–0105∗).VENEZUELA (Zulia): Base del Pico Tetari, Parque Nacional Sierra de Perijá, ca. 10°02'N, 72°57'W,elevation 3200 m (MHNLS 12354•).
