ANDEAN MOUNTAIN CAT, OREAILURUS JACOBITA: MORPHOLOGICAL DESCRIPTION AND COMPARISON WITH OTHER FELINES FROM THE ALTIPLANO

Transcription

1 Journal of Mammalogy, 8(1): , 2002 ANDEAN MOUNTAIN CAT, OREAILURUS JACOBITA: MORPHOLOGICAL DESCRIPTION AND COMPARISON WITH OTHER FELINES FROM THE ALTIPLANO ROSA GARCíA-PEREA* Museo Nacional de Ciencias Naturales, CL. J. Gutierrez Abascal 2, Madrid 28006, Spain Recent field surveys searching for the rare Andean mountain cat, Oreailurus jacobita, have had difficulty in identifying this species from sightings and ski. This is caused by the paucity of museum specime (only skulls available to date) and the lack of criteria to differentiate this species from other small sympatric felines (e.g., Lynchailurus pajeros). In a study to solve these problems, new skulls of were identified and a total of 5 skulls and 41 ski examined. Skulls of average 12 % larger than skulls of, showing a large anterior chamber in the bulla. Andean mountain cats have vertical series of yellowish-brown blotches on the sides, and a long, bushy tail with wide dark rings. Two keys to differentiate these felines, based on these and other characters, are offered. Key words: Andean altiplano, Andean mountain cat, feral cat, Lynchailurus pajeros, morphological keys, Oreailurus jacobita, pampas cat, South America Two species of small wild cat occur in and around the Andean altiplano of Peru, Bolivia, Chile, and Argentina: Andean mountain cat (Oreailurus jacobita Cornalia, 1865) and pampas cat (Lynchailurus pajeros (Desmarest, 1816)). The 1st species is a poorly known felid, whose morphology and biometry have been poorly described on the basis of skulls (Kuhn 7; Pearson 57; Philippi 1870), ski (Burmeister 1879; Cabrera 61; Cornalia 1865; Greer 65; Matschie 12; Philippi 1870, 187; Pine et al. 79; Pocock 41; Scrocchi and Halloy 86; Yepes 29), and photographs of individuals in their natural habitat (Sanderson 99; Scrocchi and Halloy 86; Ziesler 92). Moreover, the skulls were subadult specime, and 1 of them was lost at the beginning of the 20th century. Based on such a small sample, no attempt at a detailed description of the morphology and biometry of this species has * Correspondent: mcng10@mncn.csic.es been made. Also, age and sex variatio are unknown. The pampas cat has traditionally been coidered a unique, polymorphic species (Lynchailurus pajeros, after Allen, or Oncifelis colocolo, after Wozencraft 9), although it has recently been split into species (García-Perea 94): Lynchailurus pajeros, L. colocolo, and L. braccatus. The amount and nature of morphological variation within this species group have been described in some detail by García-Perea (94). No museum specime of the pampas cat from the altiplano were found by García- Perea (94), but recent reports indicate that altiplano farmers keep ski of both kinds of cat in their homes (A. Iriarte, L. Villalba, and N. Bernal, in litt.). Although recent molecular studies (Johon et al. 98) suggest that Oreailurus and Lynchailurus are not closely related genera, specime of both genera can be confused in altiplanic areas because of the external 110

2 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 111 similarity in their coat patter. Native farmers even give them the same name in many localities (L. Villalba and N. Bernal, in litt.). Field researchers have found it difficult to distinguish the 2 species in sightings, ski stuffed by native people, or skulls found occasionally while conducting fieldwork. For example, some of the Bolivian records of included in Yeen et al. (94) belong to (see Conclusio ). This situation is caused by the paucity of museum specime of available for comparative purposes and the lack of published information providing field researchers with effective tools to identify specime. Awareness of these difficulties led me to conduct a study focused on providing a key for differentiating small wild cats living in the area. The goals of my study were characterization of external and skull morphologies of and elaboration of a key for differentiating altiplanic individuals of the Andean mountain cat and the pampas cat. I have included, wherever possible, some compariso with the domestic cat (Felis catus), because feral specime of the domestic form of the African wild cat (Felis silvestris lybica group, after Nowell and Jackson 96) seem to be widespread in the altiplano and other remote places of South America. However, the high amount of variation in skull and pelage characteristics of these feral specime makes it impossible to treat these differences in depth now. A recent revision of the information available on, based on literature, has been released (Yeen and Seymour 2000). In this article, I am providing firsthand data from the largest sample ever studied of the rare Andean mountain cat. MATERIALS AND METHODS A sample of 44 specime ( skulls, 9 ski, and 2 skulls plus ski) of, 50 specime (29 skulls and 8 ski) of, and 16 specime (15 skulls and 6 ski) of kept in American and European ititutio was studied (Appendix I). Specime were from FIG. 1. A) Dorsal view of subadult skull of Oreailurus jacobita (based on MVZ 11617) showing an unossified interparietal bone (IB) and a developing sagittal crest (SC). B) Ventral view of a juvenile skull of Lynchailurus pajeros (based on CBF 2960) showing upper milk dentition (dc, dp, dp4) and unossified sphenooccipital synchondrosis (SS). Peru, Bolivia, Chile, and Argentina. Specime were identified to species a priori by skull and skin characters, based on the information provided by the few descriptive papers available (e.g., Cabrera 61; García-Perea 94) and my own data. Only specime of pampas cats from the altiplano and nearby areas were coidered, in order to remove geographic variation from the sample (García-Perea 94). For morphological characterization of and for purposes of comparison among species, a relative age was assigned to each individual. Age was estimated on skulls based on the developmental stages of several morphological structures (Barone 76; García-Perea 91, 96; García-Perea et al. 96; Gaunt 59). Juveniles (1st year) can be identified as those having deciduous dentition either erupting, fully erupted, or being replaced by permanent teeth (Fig. 1B), and as those lacking sagittal crest or showing only an incipient structure. Subadults (2nd year) can be identified as those having permanent dentition, unossified sphenooccipital synchondrosis, and interparietal bone, and a developing sagittal crest (Fig. 1A). Adults are specime more than 2 years old, showing totally ossified sphenooccipital synchondrosis and interparietal bone. The word adult is used here with an ontogenetic meaning, indicating that specime have ended their growth and have

3 112 JOURNAL OF MAMMALOGY Vol. 8, No. 1 reached their definitive physical characteristics (i.e., not related to reproductive condition). Skulls were classified by age as follows:, 2 subadults and adults;, 5 juveniles, 5 subadults, and adults; and, 15 adults. No criteria have been established for estimating age from felid ski except relative size, to use which we need to know already the size of adult specime, and this was not the case for the Andean mountain cat. However, the large sample (n 20) of ski studied from Catamarca, Argentina, offered a unique opportunity to analyze age variation. Coidering the fact that these specime came from a relatively small area, I assumed that they represented a single population. All ski were preserved with the same tanning method, so I assumed that size differences I observed reflected real differences in size between individuals. Five external measurements were recorded wherever available, either from the collection labels or from measurement of specime of : head and body length, tail length, hind foot length, ear length, and body weight. Based on these premises, I used the variables head and body length and tail length on the ski, searching for differences in global size that could be attributed to age. A comparison of values obtained allowed me to establish size classes. Because males are likely larger than females, as in most felids (García-Perea 94, 96, in litt.; García-Perea et al. 85), some differences in size could be attributed to sexual dimorphism. However, the size groups identified showed noticeable differences in spotting pattern and color, something unusual among males and females of the same species of wild felid but linked to growth in many species such as the lion, the cougar, or the lynx (Ewer 7; R. García-Perea, in litt.). I assumed therefore that the size groups represented age classes, covering age periods perhaps not strictly equivalent to those described based on skull (values of head and body length and tail length in millimeters): juveniles (n 1, head and body length 500, tail length about 0); subadults (n 8, head and body length , tail length 0 420); and adults (n 11, head and body length , tail length ). Twenty skull and teeth variables commonly used for carnivores were measured on each specimen of and (Fig. 2), using a digital caliper accurate to 0.02 mm. Descriptive statistics (mean, standard error of mean, range) were calculated for each variable. Because I had only 2 subadult and adult specime of, no statistical tests were applied to test differences observed between the 2 age groups. However, significant differences have been found between subadults and adults of other felid species (Andersen and Wiig 84; García-Perea et al. 96); so differences among species were tested only on adult specime. Thus, the sample used for that purpose coisted of,, and (craniometric data of were recorded from literature Pocock 51 because only qualitative characters were available from the 15 skulls mentioned earlier). For dental measurements, 5 skulls of were coidered, because carnassials erupt with their definitive dimeio. Differences among pairs of species were tested by Mann Whitney U-test and Student s t-test, using SPSS software (SPSS Inc. 9). Further, 1 morphological characters were checked on each skull to record the character states present in the sample, but only 10 of them showed variation when specime of Oreailurus and Lynchailurus were compared (Figs. 5): palate position of anterior palatine foramina relative to palatine-maxilla suture; rostrum shape of nasal bones; zygomatic arches size of dorsal postorbital process of jugal; basicranium shape of presphenoid body; bullae position of vagina processus hyoideus relative to stylomastoid foramen; bullae shape of paroccipital processes; teeth 1 ridge between 2 grooves in lingual side of upper canine; teeth size of parastyle related to protocone of P4; teeth metaconid on m1; and bullae relative size of ecto- and entotympanic chambers. Seven characters related to body proportio, external anatomy, and coat pattern were examined on each skin wherever possible (Figs. 6 and 7): legs black, copicuous rings around forearms; ears color of dorsal region; body spinal crest (band of long, erectile fur from shoulders to base of tail); body arrangement and color of blotches or rosettes on sides; face color of rhinarium; tail number, size, and color of rings; and tail length percentage of head and body length represented. A Spanish summary of the results is available at the web site of Cat Action Treasury (CAT),

4 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 11 FIG. 2. Variables measured on each skull. P4, m1: upper and lower carnassials, respectively, in occlusal views. GLS greatest length of skull, CBL condylobasal length, BAL basilar axis length, ZW zygomatic width, PL palatal length, FW facial width (distance between tips of postorbital processes), IOW interorbital width, POW postorbital width, CH neurocranial height, RL rostral length, IBD interbullae distance, MW mastoid width, RWC rostral width across the canines, SCL sagittal crest length, ML mandible length, MRH height of mandible ramus, P4L length of upper P4, P4W maximum width of P4, Lm1 length of lower m1, Wm1 maximum width of m1. Spanish tralatio of figure legends and titles of tables are available from the author. RESULTS AND DISCUSSION Age-related variation in skulls of. The youngest specime of this species studied were 2 subadults (MVZ 11617, UG D27). Besides the small size compared with adults (Table 1) and an unossified sphenooccipital synchondrosis (Fig. 1A), subadults have different skull proportio: postorbital processes are less developed, zygomatic arches are narrower, and sagittal crest is shorter (i.e., lower values of facial width, zygomatic width, and sagittal crest length). Also, postorbital cotrictio of adults are narrower (lower values of postorbital width). These results are coistent with those found for other felid species, such as the Lynx (Andersen and Wiig 84; García-Perea 91) and other carnivores (e.g., Genetta genetta Crawford-Cabral 81). These ontogenetic changes must be coidered carefully, because several characters described by Pocock (41) and Cabrera (40, 61) to differentiate Oreailurus (Colocolo after Pocock) from Lynchailurus and other South American felids showed age-related variation, and the skull of they based their results on was a subadult (Philippi 187). For example, Cabrera (61:20) described the sagittal crest of

5 1 JOURNAL OF MAMMALOGY Vol. 8, No. 1 FIG. 4. Qualitative characters and proportio differing in skulls of A) Felis catus, B) Oreailurus jacobita, and C) Lynchailurus pajeros. EC: ectotympanic chamber of bulla, EN: entotympanic chamber of bulla, PPJ: postorbital process of jugal, PP: paroccipital process, P2: upper 2nd premolar. Black arrows highlight certain anatomical differences mentioned in text. FIG.. Qualitative characters and proportio differing in skulls of A) and B) Oreailurus jacobita and C) and D) Lynchailurus pajeros. NB: nasal bones, SC: sagittal crest, PB: presphenoid body, EC: ectotympanic chamber of bulla, EN: entotympanic chamber of bulla, APF: anterior palatine foramina, C: upper canine, P: rd upper premolar, P4: upper carnassial, M1: 1st upper molar. Oreailurus as only well developed in its occipital tip, and this is true for all subadults and 1 adult of small skull size (MUSM 6015), but not for 2 other adults (CBF 445, CGECM 027), which have long sagittal crests. It must be noted that values of sagittal crest length increase with age in other felids, sometimes remarkably so, and that sagittal crest length is positively correlated with skull size (García-Perea 96). Age-related variation in ski of. The only juvenile examined (MACN 7.4) had a higher number of blotches, smaller than those of adults and of markedly darker color, especially those from the ventral parts, which are chocolate over a creamy background. Markings from the sides are rosette-like, reddish-brown iide with darker borders. Tail rings are narrower, closer to each other, and darker brown than those of adults. Head is gray, as in adults, but also darker. Black markings on front legs are more copicuous in the kitten, but not forming rings. In subadults, the general color becomes lighter. Compared with adults, the ground color is creamy itead of gray, and blotches look reddish. As in the juvenile, blotches are smaller and more numerous in subadults than in adults and are distributed irregularly. There is a blackish band along the back (nonerectile), lacking the reddish tinge of that in adults. Some subadults show copicuous black markings on the front legs, but these never form rings. Belly spots are darker than those of adults. Tail rings are more like those of adults than those of the juvenile, but still narrower. From the previous descriptio, data previously published (García-Perea 94), and information provided subsequently (under Coat characteristics of ), it follows that subadults of could be mistaken for adult Lynchailurus from the

6 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 115 FIG. 5. Qualitative differences between upper carnassials (P4) of A) Lynchailurus pajeros and B) Oreailurus jacobita. Pr: protocone, Pa: parastyle, Ec: ectostyle. FIG. 7. Main external differences observed between A) Oreailurus jacobita and B) Lynchailurus pajeros from the altiplano. Black arrows indicate characters differentiating the 2 species. altiplano (especially in sightings), unless the details of tail rings, front leg stripes, and spinal crest are identified. Skull and body measurements of. Values for skull and teeth variables measured on specime of are summarized in Table 1. Because the only skull of known sex studied was a male (Table 1), it is risky to say anything about sexual dimorphism of this species. However, the relatively small size of specimen MUSM 6015, an old adult with totally ossified sutures and rough braincase surface (Table 1), suggests that MUSM 6015 could be a female. This pattern of females being smaller than males occurs in other felid species (e.g., Lynx pardinus García-Perea et al. 85). Data suggest that, although some large FIG. 6. External appearance of adult Oreailurus jacobita. Based on pictures of a living individual and on museum specime (see Appendix I).

7 116 JOURNAL OF MAMMALOGY Vol. 8, No. 1 TABLE 1. Values (in millimeters) of 20 skull and teeth variables measured on 2 subadult and adult specime of. Sex is indicated where known. Additional data from 1 skull from the literature are included. Specimen numbers refer to museums listed in Appendix I. Variable MVZ UG D27 MUSM 6015 CBF 445 CGECM 027 From literature a Age class Sex Greatest length of skull Condylobasal length Rostral length Rostral width across the canines Mastoid width Palatal length Interorbital width Facial width Postorbital width Zygomatic width Basilar axis length Neurocranial height Interbullae distance Sagittal crest length Length of P4 Maximum width of P4 Mandible length Height of mandible ramus Length of ml Maximum width of ml Subadult Male Subadult Adult Adult Adult Subadult a Skull illustrated by Philippi (187) and measured by Cabrera (61). specime of the pampas cat have skull dimeio similar to those of, the latter species is significantly larger (greatest length of skull 12% larger, condylobasal length % larger) than altiplanic specime of and (greatest length of skull % larger, condylobasal length 21% larger; Table 2). Carnassial teeth (P4 and m1) are also significantly more massive in than in and in (P4 length, m1 length; Table 2). Note that average values and ranges of skull measurements of adult given in Table 2 are larger that those offered by Yeen and Seymour (2000:2), because their sample included subadult specime. Observation of skulls also shows other differences: dorsal profile of skull is more flat and elongated in than in L. pajeros; dorsal profile of rostrum is lower and more flat in than in ; adult specime of usually show a long sagittal crest, which is never so well developed in adult specime of from the altiplano; and auditory bullae are relatively smaller and more separate in (Figs. and 4) than in (interbullae distance; Table 2). An elongated muzzle places postorbital processes of closer to the middle of the skull (Figs. A and 4B), as Yeen and Seymour (2000) mention, but not as much as Pocock (41) indicates. As I noted above, Pocock and Cabrera based their descriptio on the subadult specimen illustrated by Philippi (187). Values of external measurements obtained on tanned ski from Catamarca were useful for establishing age classes, but they could not be used for descriptive purposes because they were likely overesti-

8 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 117 mated. In fact, specimen MVZ (Pearson 57), a subadult male, shows lower values of body measurements (on the flesh) than those measured on subadult Catamarca specime (Table ). These differences are likely a coequence of change in size due to the tanning method. Average values given by Yeen and Seymour (2000), based on the five specime recorded in the literature (also given in Table ), are just suggestive, because age of specime is unknown, except for 1 subadult, and subadult felids are significantly smaller than adults (e.g., García-Perea 96; García-Perea et al. 96). Regarding the 10 qualitative traits showing variation in skulls of and L. pajeros, character states observed on the sample are shown in Table 4. These characters confirm that, as Kuhn (7) mentioned, the relative sizes of ecto- and entotympanic chambers of auditory bulla are diagnostic for in the altiplano, not the presence of external groove, or sulcus, separating both chambers (Pocock 41), which is not always obvious, as the sulcus does not appear better developed in all old individuals, as Seymour (99) claims. In, the ectotympanic chamber is equal to or larger than the entotympanic (Table 4), representing more than 50% of bullar volume. However, this character is not unique to, because L. colocolo from central Chile, and some species of Felis (e.g., F. manul, F. margarita) also have an ectotympanic chamber equal to or larger than the entotympanic chamber (García-Perea 94; Pocock 16). As Cabrera (61) suspected, the 5 skulls of lack P2 (similar to ). Its dental formula is then I /, C 1/ 1, P 2/2, M 1/1, total 28. Feral cats show a high amount of variation in qualitative characters of skull. For this reason, I am providing in Table 5 some typical, although not exclusive, skull characters that, if restricted to the altiplano, can help to identify (Fig. 4A). Coat characteristics of. A general view of an adult showing the most copicuous features characterizing the coat pattern of this species is shown in Fig. 6 based on photographs of living individuals and on museum study ski. Former illustratio of shown by Cabrera and Yepes (40:plate 29), by Seideticker and Lumpkin (91: 48), and by Redford and Eisenberg (92: plate 9h) fail to reflect accurately external characters such as shape and distribution of body markings, tail proportio, and tail rings of the species. The ground color of an adult is ashy-gray, with yellowish-brown, irregular blotches arranged in vertical series on the flanks (traverse lines, when observed from the top). The head and face are gray, with cheeks and areas around the lips white. As in many other felids, 2 dark brown lines run across the cheeks, converging laterally. In some specime, 2 dark gray lines start above the eyes, running up to the space between the ears. From there, 2 wide yellowish-brown bars run laterally down to the base of the neck, sometimes with 1 or 2 bars of the same color between them. Ears are gray, with darker borders. There is a rusty-black band (nonerectile fur) along the back. Some black spots occur in the ulnar region of forelegs, but never forming complete rings like those in. However, hind limbs may have 1 or 2 narrow, dark rings, dorsally black and ventrally reddish. Belly is white or creamy with light brown spots laterally and black spots medially. One or 2 dark brown, traverse stripes occur in the ventral part of the neck, sometimes incomplete. Tail is long (around 66 75% of head and body length in fresh specime), bushy, and cylindrical, showing 6 9 rings varying from black to dark brown. The or 4 distal rings are remarkably wider (up to 60 mm wide) than those of the proximal half. Two adult females had well-developed mammae with no fur around them, suggesting that they were rearing kitte when captured. Both females had 2 pairs of mammae located in abdominal position, and

9 118 JOURNAL OF MAMMALOGY Vol. 8, No. 1 TABLE 2. Descriptive statistics for 20 cranial variables of adult and and of 8 variables of adult (after Pocock 51); n sample size. Level of significance is indicated for Mann Whitney U-test (comparing with each of the other species) and Student s t- test (comparing with ). Symbols: *, P 0.05; **, P 0.005; and ***, P 0.001;, not significant (P 0.05). Skull size (mm) n X SE Range Greatest length of skull * ** ** Condylobasal length ** ** ** Rostral length Rostral width across the canines Mastoid width * Palatal length ** Interorbital width ** ** * Facial width * Postorbital width *** Zygomatic width * * Basilar axis length Neurocranical height

10 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 1 TABLE 2. Continued. Skull size (mm) n X SE Range Inter-bullar distance ** Sagittal crest length Length of P *** *** *** Maximum width of P *** Mandible length ** * Height of ramus of mandible Length of ml *** *** *** Maximum width of ml *** both were captured in December. Taking into account that weaning usually occurs in small felids about 2 months after birth (Ewer 7; Kitchener 91), birth of both litters could have occurred around October, during austral spring. Pelage of both and differ copicuously in 7 external characters, illustrated in Fig. 7 and described in Table 4. Characteristics described are coistent with those observed by L. Villalba and N. Bernal (in litt.) Tail of is shorter than that of. It is necessary to be careful when checking tail proportio, because measurements taken on fresh specime differ from those taken on collection ski. Regarding, data recorded from literature (fresh specime, n 5), indicate that its tail length is about 66 75% of head and body length, whereas data recorded from collection specime (n 11) give a figure of 52 65%. Similarly, data on fresh specime of recorded from specimen labels (n 9) indicate that tail length is about 40 50% of head and body length, whereas these variables measured on collection specime yield values that are 4 8% of the length (n 9). My data on fresh specime of agree with those given by Yeen and Seymour (2000) 60 75% of length but data on L. pajeros given by these authors (ca. 0%) seem to refer to collection specime.

11 120 JOURNAL OF MAMMALOGY Vol. 8, No. 1 TABLE. Values of 5 external measurements for, either found in the literature or measured on tanned collection ski for this study. n sample size. Body weight (g) Reference Length of hind foot (mm) Ear length (mm) Tail length (mm) Head and body length (mm) Sex Age n 4000 Pearson 57 a Cornalia 1865 Matschie 12 Pine et al Male Subadult Cabrera 61 This study This study Adult Subadult Adult Male Males and females Males and females a MVZ A great amount of variation is observed in coat pattern of feral cats, too large to be treated here. However, I include in Table 5 the most striking characters of feral cats to be coidered when comparing with the 2 other wild genera, Oreailurus and Lynchailurus, occurring in the altiplanic area. CONCLUSIONS Felids are especially difficult to assess morphologically. Usually, individual variation is moderate, and often odd character states are found at very low frequencies. For this reason, finding diagnostic characters for closely related species is very difficult, and it is safer to look for sets of traits characteristic of a species, with a high probability of finding them all together. Based on this premise, a dichotomous key alone would not be enough for my purposes; so I am providing sets of characters (Tables 4 and 5) that are likely to occur together in a species, and if one fails, we can confirm our identification by looking for a match with the others. From my experience, 1 or 2 characters are expected to be missing together. For, only typical characters are provided, but we must remember that they are not exclusive to and may not be present in some specime. The following keys are only useful within the limits of the altiplanic region, because shows a certain amount of geographic variation. Key for skulls. It is necessary to determine first whether the skull is from an adult (with sphenooccipital synchondrosis ossified). If it is not, the key may not be useful. After using the following keys, confirm identification using characters in Tables 4 and 5, especially for individuals with condylobasal length close to 94 mm (as certain domestic cats could also have similar condylobasal length). DICHOTOMOUS KEY FOR SKULLS 1a. Condylobasal length 94 mm, lower m1 length 9.8 mm, dorsal profile of skull flat and low, anterior chamber of

12 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 121 TABLE 4. Character states observed on skulls of and, and external traits characterizing ski of both species. Characters Oreailurus jacobita Lynchailurs pajeros Skull Location of anterior palatine At palatine-maxilla suture Posterior to palatine-maxilla suture foramina Shape of nasal bones Slowly narrowing posteriorly Strongly narrowing posterior half Length of postorbital process of Short Long jugal Shape of body of presphenoid Widened medially No medial widening Position of vagina processus hyoideus Posterior to stylomastoid foramen Medial to stylomastoid foramen Shape of paroccipital processes Long, separate, protruding ventrally Short, not protruding, cupped around bulla Presence of lingual ridge in upper Absent Present canine Relative size of parastyle and protocone of P4 About the same size Larger than protocone or no protocone Presence of metaconid on m1 Present (80%) Absent Relative size of ectotympanic and Ectotympanic equal to or larger Ectotympanic smaller entotympanic Skin Black rings on forelegs Absent, only large spots Present, complete rings Color of ears dorsally Uniformly gray Black and cream and reddish Spinal crest of erectile fur Absent Present Pattern and color of body markings Vertical series of yellowish-brown Oblique series or rusty rosettes blotches Color of rhinarium Black Pink to coffee or dark mahogany Tail rings 6 9 dark brown-to-black rings, some of them reaching up to 60 mm wide Around 8 reddish rings, up to 20 mm wide each Tail length as percentage of head and body length Tail length about 66 75% of head and body length Tail length about 40 50% of head and body length bulla equal to or larger than posterior... 1b. Condylobasal length 94 mm, lower m1 length 9.6, dorsal profile high and convex, anterior chamber of bulla smaller than posterior a. Presence of lingual ridges on upper canines, lower m1 length 8.0 mm, anterior chamber of bulla medially expanded and inflated... 2b. Absence of ridges on upper canines, lower m1 length 8.0 mm, anterior chamber of bulla medially expanded but not inflated... Dichotomous key for ski. Because of the high amount of variation exhibited by feral cats, this key may not be as effective as the previous one for skulls; so I recommend checking characters in Tables 4 and 5 carefully. DICHOTOMOUS KEY FOR SKINS 1a. Medium-sized body; tail length around 66 75% of head and body length; no spinal crest; no complete rings in forelegs; body sides with blotches arranged in vertical series... 1b. Small-sized body; tail length 50% of head and body length; spinal crest present; complete rings in forelegs; body sides not marked or with rosettes arranged in oblique series a. Ears black and cream and reddish; body sides with rusty rosettes arranged in oblique series... 2b. Ears uniformly colored; body sides not

13 122 JOURNAL OF MAMMALOGY Vol. 8, No. 1 TABLE 5. Skull and skin characters typical of, absent in and. Skull characters P2 present Unossified interparietal bone in adults Posterior border of palate W-shaped Ectotympanic expanded but no inflated Skin characters Plain color, no body markings Medium-sized, slender tail, 2 5 black distal rings, black tip Ears of reddish color, or another color matching body marked or with another pattern different from that described in 2a... Based on skull and skin characteristics described earlier and from a detailed comparison with and materials, specime CBF 2224, CBF 2960, and CBF 2229, previously identified as O. jacobita (Yeen et al. 94) were identified as in this study. RESUMEN Recientes expediciones en busca del desconocido gato andino, Oreailurus jacobita, se han encontrado con dificultades para identificar esta especie a partir de avistamientos y pieles. Esto se debe a la escasez de ejemplares en las colecciones (sólo se conocían cráneos hasta ahora) y a la ausencia de criterios para diferenciar esta especie de otros felinos simpátricos (por ejemplo, Lynchailurus pajeros). Con el objeto de resolver estos problemas, se ha realizado un estudio en el que se han identificado nuevos cráneos de, habiéndose examinado en total 5 cráneos y 41 pieles. Los cráneos de son un 12 % más grandes que los de, mostrando una gran cámara anterior en la bula timpánica. El gato andino presenta series verticales de manchas pardo-amarillentas en los flancos, así como una larga cola con abundante pelo y anchos anillos oscuros. Se ofrecen dos claves para diferenciar estas especies de felinos basadas en estos y otros caracteres. ACKNOWLEDGMENTS K. Nowell (Director, Cat Action Treasury) and P. Jackson (former Chairman, Cat Specialist Group, IUCN) gave me the opportunity to conduct this study. Funding was provided by Cat Action Treasury, supported by the Leonard X. Bosack and Bette M. Kruger Foundation. J. Gisbert (Museo Nacional de Ciencias Naturales, MNCN, Madrid) drew Figs. 1 5 and 7. M. Antón drew Fig. 6, using pictures generously provided by J. Sanderson. I am indebted to the people in charge of the collectio listed in Appendix I, who kindly provided access to them: E. Vivar, N. Bernal, L. Villalba, M. Lucherini, E. Luengos, J. L. Patton, H. J. Kuhn, A. Iriarte, J. Yañez, J. C. Torres-Mura, R. Kraft, M. Piantanida, O. Vaccaro, P. Jenki, D. M. Hills, L. Gordon, M. Carleton, M. Rutzmosser, S. Anderson, G. Musser, and B. Patterson. This study has also benefited from grant DGICYT PB95-01 (Spain, 99), a Large Scale Facilities grant under the Training and Mobility of Researchers Programme (European Union, 99), and from a grant under the Agreement CSIC-Deutsche Forschungsgemeichaft (Spain Germany, 95). These funds were made available thanks to J. Morales (MNCN, Madrid), P. Jenki (The Natural History Museum, London, United Kingdom), and G. Peters (Museum Alexander Koenig, Bonn, Germany). Valuable comments of an anonymous reviewer improved this article. LITERATURE CITED ALLEN, J. A.. Notes on the synonymy and nomenclature of the smaller spotted cats of tropical America. Bulletin of the American Museum of Natural History 41:41 4. ANDERSEN, T.,AND O. WIIG. 84. Growth of the skull of Norwegian lynx. Acta Theriologica 29: BARONE, R. 76. Anatomie comparée des mammifères domestiques. Tome I. Ostéologie. Fascicle 1 (texte). Vigot Frères, Paris, France. BURMEISTER, H Description physique de la Repúblique Argentine. : CABRERA, A. 40. Notas sobre carnívoros sudamericanos. Notas del Museo de La Plata (Zoologia) 5: CABRERA, A. 61. Los félidos vivientes de la República Argentina. Revista del Museo Argentino de Ciencias Naturales (Zoologia) 6: CABRERA, A., AND J. YEPES. 40. Historia natural Ediar. Mamíferos Sud-Americanos. Compañía Argentina de Editores, Buenos Aires, Argentina. CORNALIA, E Descrizione di una nuova specie del genere Felis, Felis jacobita (Corn.). Memorie della Societá Italiana di Scienze Naturali 1:1 9. CRAWFORD-CABRAL, J. 81. Análise de dados cran-

14 February 2002 GARCÍA-PEREA MORPHOLOGY OF ANDEAN MOUNTAIN CATS 12 iométricos no género Genetta G. Cuvier (Carnivora, Viverridae). Memórias da Junta de Investigações do Ultramar 66:1 29. EWER, R. F. 7. The carnivores. Cornell University Press, New York. GARCíA-PEREA, R. 91. Variabilidad morfológica del género Lynx Kerr, 1792 (Carnivora, Felidae). Ph.D. dissertation, Universidad Complutee, Madrid, Spain. GARCíA-PEREA, R. 94. The pampas cat group (Genus Lynchailurus Severtzov, 1858) (Carnivora: Felidae), a systematic and biogeographic review. American Museum Novitates 096:1 6. GARCíA-PEREA, R. 96. Patter of postnatal development in skulls of lynxes, genus Lynx (Mammalia, Carnivora). Journal of Morphology 229: GARCíA-PEREA, R., R. BAQUERO, R. FERNÁNDEZ-SAL- VADOR, AND J. GISBERT. 96. Desarrollo juvenil del cráneo en las poblaciones ibéricas de gato montés, Felis silvestris Schreber, Doñana, Acta Vertebrata 2: GARCíA-PEREA, R., J. GISBERT, AND F. PALACIOS. 85. Review of the biometrical and morphological features of the skull of the Iberian Lynx, Lynx pardina (Temminck, 1824). Säugetierkundliche Mitteilungen 2: GAUNT, W. A. 59. The development of the deciduous cheek teeth of the cat. Acta Anatomica 8: GREER, J. K. 65. Another record of the Andean highland cat from Chile. Journal of Mammalogy 46:507. JOHNSON, W. E., M. CULVER, J. A. IRIARTE, E. EIZIRIK, K. L. SEYMOUR, AND S. J. O BRIEN. 98. Tracking the evolution of the elusive Andean Mountain cat (Oreailurus jacobita) from mitochondrial DNA. Journal of Heredity 89: KITCHENER, A. 91. The natural history of the wild cats. Cornell University Press, New York. KUHN, H. J. 7. Zur Kenntnis der Andenkatze, Felis (Oreailurus) jacobita Cornalia, Säugetierkundliche Mitteilungen 21: MATSCHIE, P. 12. Über Felis jacobita, colocola und zwei ihnen ähnliche Katzen. Sitzungsberichten der Gesellschaft Naturforschender Freunde (Berlin) 4: NOWELL, K., AND P. JACKSON (EDS.). 96. Wildcats, status survey and coervation action plan. Cat Specialist Group (SSC, IUCN), Gland, Switzerland. PEARSON, O. P. 57. Additio to the mammalian fauna of Peru and notes on some other Peruvian mammals. Breviora 7:1 7. PHILIPPI, R. A Ueber Felis colocolo Molina. Archiv für Naturgeschichte 6: PHILIPPI, R. A Ueber Felis guiña Molina und über die Schädelbildung bei Felis pajeros und Felis colocolo. Archiv für Naturgeschichte 9:8 15. PINE, R. H., S. D. MILLER, AND M. L. SCHAMBERGER. 79. Contributio to the mammalogy of Chile. Mammalia 4:9 76. POCOCK, R. I. 16. The structure of the auditory bulla in existing species of Felidae. Annals and Magazine of Natural History, Series 8 18:26 4. POCOCK, R. I. 41. The examples of the Colocolo and of the Pampas cat in the British Museum. Annals and Magazine of Natural History, Series 11 7: POCOCK, R. I. 51. Catalogue of the genus Felis. British Museum (Natural History), London, United Kingdom. REDFORD, K. H., AND J. F. EISENBERG. 92. Mammals of the Neotropics. The southern cone: Chile, Argentina, Uruguay, Paraguay. University of Chicago Press, Chicago, Illinois 2:1 40. SANDERSON, J. G. 99. Andean mountain cats (Oreailurus jacobita) in northern Chile. Cat News 0: SCROCCHI, G. J., AND S. P. HALLOY. 86. Notas sistemáticas, ecológicas, etológicas y biogeográficas sobre el gato andino, Felis jacobita Cornalia (Felidae, Carnivora). Acta Zoologica Lilloana 8: SEIDENSTICKER, J., AND S. LUMPKIN (EDS.). 91. Great cats. Majestic creatures of the wild. Rodale Press, Emmaus, Penylvania. SEVERTZOW, M. N Notice sur la classification multisériale des carnivores, spécialement des félidés, et les études de zoologie générale qui s y rattachent. Revue et Magazine de Zoologie, 2nd Series 10:85 9. SEYMOUR, K. L. 99. Taxonomy, morphology, paleontology and phylogeny of the South American small cats (Mammalia: Felidae). Ph.D. dissertation, University of Toronto, Ontario, Canada. SPSS INC. 9. SPSS for Windows. Version 6.0. SPSS Inc., Chicago, Illinois. WOZENCRAFT, C. W. 9. Order Carnivora. Pp in Mammal species of the world. (D. E. Wilson and D. M. Reeder, eds.). 2nd ed. Smithsonian Ititution Press, Washington, D.C. YENSEN, E., AND K. L. SEYMOUR Oreailurus jacobita. Mammalian Species 644:1 6. YENSEN, E., T. TARIFA, AND S. ANDERSON. 94. New distributional records of some Bolivian mammals. Mammalia 58: YEPES, J. 29. Notas sobre algunos de los mamíferos descriptos por Molina, con distribución geográfica en Chile y Argentina. Revista Chilena de Historia Natural : ZIESLER, G. 92. Souvenir d un chat des Andes. Animan, Nature et Civilisatio 50: Submitted 4 May Accepted 9 May Associate Editor was Meredith J. Hamilton. APPENDIX I Collectio coulted for this study. American Museum of Natural History, New York (AMNH); The Natural History Museum, London, United Kingdom (BM); Colección Boliviana de Fauna, Mastozoología, La Paz, Bolivia (CBF); Collection of the Grupo de Ecología Comportamental de Mamíferos, Universidad Nacional del Sur, Bahía Blanca, Argentina (CGECM); Field Museum of Natural History, Chicago, Illinois (FMNH); Museo Argentino de Ciencias Naturales, Buenos Aires, Argentine (MACN); Museum of Comparative Zoology,

15 124 JOURNAL OF MAMMALOGY Vol. 8, No. 1 Harvard University, Cambridge, Massachusetts (MCZ); Museo Nacional de Historia Natural, Santiago de Chile, Chile (MNHN); Museo de Historia Natural Javier Prado, Lima, Peru (MUSM); Museum of Vertebrate Zoology, University of California, Berkeley, California (MVZ); Servicio Agrícola y Ganadero, Ministry of Agriculture, Chile (SAG); University of Gottingen, Gottingen, Germany (UG); National Museum of Natural History, Washington D.C. (USNM); Zoologisches Staatssammlung München, Munich, Germany (ZSM). Catalog numbers. Oreailurus jacobita. BM: ; ; CBF: 00445; 02018; 02227; CGECM: 027. MACN: ; ; 7.1; 7.2; 7.; 7.4; 7.6; 7.7; 7.8; 7.106; 7.107; 7.108; 7.109; 7.111; 7.112; 7.1; 7.116; 7.2; 7.; 42.11; 9 uncataloged. MNHN: 1 uncataloged. MUSM: 6015; 1 uncataloged. MVZ: SAG: 1 uncataloged. UG: D27. ZSM: 84/12. Lynchailurus pajeros. BM: ; ; ; ; ; CBF: 02224; 02229; 02274; 02960; 062; 064; 065; 066. CGECM: 001; 022. FMNH: 21677; 2550; 4975; 52488; MACN: 086; 17816; 25.; 26186; 26187; 29765; 010; 422; 426; 4565; 620; 4116; MUSM: 01 J.O.; 415; 417; 418; 420; 421; 2150; 2151; 8467; MVZ: 1777; 1942; 194; ZSM: 16/8; 74/1. Felis catus. AMNH: 079; 41557; 1972; CBF: 4 uncataloged. CGECM: 028; 1 uncataloged. MCZ: MVZ: ; 16. USNM: 176; ;