Mamm. biol. 67 (2002) 372±377 Mammalian Biology ã Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift fuÈr SaÈ ugetierkunde

Short communication Rediscovery of Enders's small-eared shrew, Cryptotis endersi (Insectivora: Soricidae), with a redescription of the species

By R. H. PINE,N.WOODMAN and R. M. TIMM Illinois Mathematics and Science Academy, Aurora, Illinois, and Field Museum, Chicago, Illinois; National Museum of Natural History, Washington, D.C.; Natural History Museum and Department of Ecology and Evolutionary Biol- ogy, University of Kansas, Lawrence, Kansas, USA

Receipt of Ms. 09. 05. 2001 Acceptance of Ms. 06. 05. 2002

Key words: Cryptotis, Soricidae, Insectivora, shrews, Panama

Of the 270 or so currently recognized spe- east, and adds to understanding of variation cies of of mainland Central in the species. America, the least known is Enders's During 15 May±10 July 1980, Pine,R.J.Izor, small-eared shrew, Cryptotis endersi. Setzer and J.A. Wagner conducted field work (1950) described this species from a single along and near the Continental Divide of subadult of unknown sex, obtained by the Cordillera de Chiriquõ (or SerranõÂa del R.K. Enders in western Panama in 1941. TabasaraÂ) in Bocas del Toro and ChiriquõÂ, For a half century, the holotype has re- PanamaÂ.On 18 June, Izor trapped the mained the only specimen known, and all C. endersi (field no.RHP 8354±now USNM literature citations (e.g., Miller and Kel- 541048) in a ªMuseum Specialº snap trap logg 1955; Handley 1966; MeÂndez 1970; baited with peanut butter and rolled oats, 1 Choate and Fleharty 1974; Corbet and on Cerro Bollo (3 /2 km E Escopeta), Hill 1980, 1986, 1991; Honacki et al.1982; 1800±1856 m, ChiriquõÂ/Bocas del Toro Nowak and Paradiso 1983; Burton 1987; boundary, PanamaÂ, about 70 km east of the Eisenberg 1989; Nowak 1991, 1999; Hut- type locality.The specimen was prepared terer 1993; Carleton and Musser 1995) as a skin + skull + body skeleton. and redescriptions (e.g., Hall and Kelson The holotype is of unknown sex.Although 1959; Choate 1970; Hall 1981) are based Setzer (1950) stated that it was an adult, upon it alone.Also, starting with the origi- the specimen appears young (subadult as nal description, there have been uncertain- noted by Choate 1970).It has but slightly ties regarding characteristics of the holo- worn dentition and is relatively small.The type and details of the type locality. new specimen (Fig.1) has more worn denti- We herein report a second specimen of tion, is larger, and has different proportions C. endersi, redescribe the species, describe than the holotype. the habitat, and review ambiguous and con- Cryptotis endersi is a large Cryptotis (Tab.1), flicting accounts of the holotype and type and it has a long tail (49 to 60% of head- locality.The new specimen is an adult male and-body-length).In the field, the new spe- that extends the known distribution to the cimen was obviously larger and longer

1616-5047/02/67/063-372 $ 15.00/0. Rediscovery and redescription of Cryptotis endersi 373 haired than sympatric C. nigrescens, crista, paracone, precentrocrista, mesostyle, although the two species are similar in col- postcentrocrista, and metacone; a well de- or.Dorsal pelage is rather long (guard hairs veloped protocone is present, but a hypo- from 7 to 8 mm in length) and dark (Mum- cone is absent or very poorly developed. my Brown to Fuscous±capitalized color The mandible is relatively long with low to terms from Ridgway, 1912); ventral pelage moderately high coronoid process.The is somewhat paler than the dorsal pelage anterior border of coronoid process joins (Drab Gray to Light Drab).In contrast, the horizontal ramus of mandible at a rela- Setzer (1950) noted the color of the holo- tively low angle.Articular process is tall type as ªChaetura Black, no appreciable and narrow.Lower sigmoid notch shallow lightening on sides or bellyº. (but not very shallow).The lower third pre- The forefeet are not notably broad or en- molar is relatively long and low-cusped. larged.The foreclaws are long, but narrow The posterior border of I1 extends to pos- rather than broadened as in the C. gold- terior edge of P4's cusps, but not to its pos- mani group.There are two large, dorsal terior cingulum.Entoconid in the talonid foramina on the skull.A vestigial foramen of M3 vestigial.Dentition bulbous.The leading to a ventral extension of the sinus new specimen agrees in all non-age-related canal is present posterior to the dorsal ar- particulars with Choate's (1970) redescrip- ticular facet on at least one side of the skull tion except that the ªanterior element of in the new specimen; there is no foramen ectoloph of M1 (= M1)º is ªreduced relative dorsal to the dorsal articular facet.The to posterior elementº. anterior border of zygomatic plate is Two other species of Cryptotis are known aligned with the mesostyle-metastyle valley from western PanamaÂ, C. gracilis and C. ni- of M1; the posterior border is even with me- grescens.Reports of C. parva orophila from tastyle of M2 or parastyle of M3.Upper there (Choate 1970; Hall 1981; Eisenberg toothrows are uncrowded.U 4 is in line with 1989) were based on misidentified speci- the rest of the unicuspid toothrow and visi- mens of C. nigrescens (Woodman and Timm ble in labial view.The posterior borders of 1993).In western Panama Â, C. endersi, P4,M1, and M2 are only very slightly re- C. gracilis, and C. nigrescens are relatively cessed.M 3 is relatively complex, with para- easy to distinguish from each other on ex-

Fig. 1. Skull of adult male Enders's small-eared shrew, Cryptotis endersi (USNM 541048). 374 R. H. PINE et al. ternal characteristics. Cryptotis endersi and development and provided a map of the C. gracilis have long tails, while C. nigres- area. cens has a much shorter tail (Tab.1).Fully Shrews were captured in several different adult Cryptotis endersi can also apparently habitats at three trapping localities.Eight be identified externally by their greater to- Cryptotis nigrescens were trapped in what tal length.Although dorsal pelage of all is here called ªPacific Slope Forestº ± wet three is similar in color (dark brown to forest distributed on steep slopes from be- blackish), that of C. nigrescens is shorter low 1275 m to the crest of Cerro Colorado and less ªluxuriousº.Cranially, C. endersi at ca.1525 m.Although often foggy, ªPaci- differs from C. gracilis in its larger brain- fic Slope Forestº was not wet enough to case, broader rostrum and palate, bulbous be, nor had the aspect of, cloud forest (e.g., dentition, shallower sigmoid notch, and there were fewer epiphytes), nor had it the only a vestigial entoconid of the lower third height or stratification of rainforest.This molar.Both species have complex M 3s with habitat seemed relatively undisturbed aside metacones, but the protocone and meta- from a number of paths and some signs of cone typically are more distinct in C. graci- past and current logging. lis.From the skull of C. nigrescens, that of Cryptotis nigrescen (5 specimens) also was C. endersi differs in being larger, with a re- trapped in primary cloud forest at Quebra- latively and absolutely longer, narrower pa- da Alicia and environs [25 km NNE San late, complex M3, a low and long mandible Felix (ChiriquõÂ), 1425±1525 m, Bocas del with anterior border of the coronoid pro- Toro].This site is on the Bocas del Toro cess joining the horizontal ramus at a lower (Caribbean Slope) side of the road on a angle.Both externally and cranially, C. en- mostly north-facing slope along the crest of dersi appears to be closely related to C. gra- Cerro Colorado. cilis, as suggested by Choate (1970). The highest locality trapped was at 1800± 1 Reproductive data for C. endersi are essen- 1856 m on Cerro Bollo, 3 /2 km E Escopeta tially nonexistent.The holotype is of a (Escopeta is at ca.23 km NNE San Felix), young of unknown sex and taken ChiriquõÂ/Bocas del Toro boundary.Both on 24 July.The new specimen is an adult Cryptotis endersi (1) and C. nigrescens (14) male (testis = 8 mm) trapped on 18 July, were taken here.OC onnor et al.(1982) and with well-developed lateral glands. and Fain and Lukoschus (1984) gave coor- Pine and Izor's collecting was along a bro- dinates for this locality as 81°50' W, 8°34' N, ken, irregular elevational transect with a but these may actually apply to Escopeta. variety of habitats up Cerro Colorado, from The summit of Cerro Bollo, a rounded Escopeta, a mining camp [ca.23 km NNE peak, was originally all cloud forest and el- San Felix, ca.900 m on the Pacific (south) fin forest, the latter on slopes facing N or slope, ChiriquõÂ], to the drainage of Quebra- NW into the very strong prevailing wind. da Alicia and environs (25 km NNE San Fe- Much of Cerro Bollo still had its original lix (ChiriquõÂ), 1425±1525 m) on the Conti- vegetation.There had been a good deal of nental Divide, Bocas del Toro.Other clearing to make a line-of-sight to a sur- nearby sites also were sampled, including veyor's marker at the summit on the Conti- Cerro Bollo, also on the Continental Di- nental Divide, however.Trapping was along vide.Roads and lodging in this precipitous a pre-existing path and paths cut by the col- and extremely windy region were main- lectors.One of the C. nigrescens may have tained for a planned copper strip-mining been trapped in the elfin forest, but the rest operation.Publications dealing with the of the shrews caught here, including the Cerro Colorado area and its biotas include C. endersi, were in cloud forest.At least Myers (1969), Myers and Duellman one C. nigrescens was caught in disturbed (1982), Savage and Donelly (1992), and cloud forest.Other mammals taken on Cer- Woodman and Timm (1993). Gjording ro Bollo were vegetus, Oryz- (1981) discussed social impact of the mining omys devius, Reithrodontomys mexicanus, Rediscovery and redescription of Cryptotis endersi 375

Table 1. Measurements (in mm) of Cryptotis from the ChiriquõÂ Cordillera, PanamaÂ. (C. gracilis includes ones from interconnected Talamancan Cordillera, Costa Rica). Measurements as in WOODMAN and TIMM (1993) ± terminology not the same in all cases, but abbreviations (CBL, CB, etc.) are the same.

C. endersi C. gracilis C. nigrescens (subadult ANSP holotype; (n = 31) (PanamaÂ) USNM 541048) (n = 23) head-and-body length 73; 81 67 Ô 7 73 Ô 6 57±77 60±83 tail length 36; 49 36 Ô 3 31 Ô 4 29±42 19±37 tail as % of head-and-body length 49; 60 55 Ô 7 43 Ô 5 45±68 28±51 condylobasal length (CBL) 20.4; 21.3 19.4 Ô 0.6 19.1 Ô 0.5 18.0±20.4 18.2±20.2 breadth of braincase (CB) 9.8; 10.5 9.7 Ô 0.4 9.6 Ô 0.3 9.0±10.3 (n = 30) 8.7±10.3 breadth of zygomatic plate (ZP) 2.0; 1.9 1.7 Ô 0.1 2.0 Ô 0.2 1.5±2.0 1.5±2.5 postorbital breadth (IO) 4.6; 4.9 4.6 Ô 0.2 4.6 Ô 0.2 4.1±4.6 4.3±4.9 breadth across upper first unicuspids (UIB) 2.7; 2.6 2.3 Ô 0.2 2.6 Ô 0.1 2.0±2.6 2.4±2.8 breadth across upper second molars (M2B) 5.8; 5.8 5.2 Ô 0.2 5.7 Ô 0.2 4.9±5.6 5.3±6.0 palatal length (PL) 8.6; 9.3 8.7 Ô 0.4 8.1 Ô 0.3 7.8±9.3 7.5±8.8 length of maxillary toothrow (TR) 7.9; 8.3 7.5 Ô 0.3 7.3 Ô 0.3 6.9±8.0 6.9±7.8 length of upper unicuspid toothrow (UTR) 2.7; 2.9 2.5 Ô 0.2 2.4 Ô 0.1 2.3±2.9 2.2±2.6 length of molariform toothrow (MTR) 5.6; 5.6 5.2 Ô 0.2 5.3 Ô 0.2 4.8±5.4 4.8±5.7 posterior width of upper first molar (WM1) 1.8; 1.8 1.6 Ô 0.1 1.8 Ô 0.1 1.4±1.8 1.7±1.9 length of mandible from lower sigmoid 6.3; 7.2 6.3 Ô 0.3 6.4 Ô 0.3 notch to mental foramen (ML) 5.7±6.8 5.6±6.9 height of coronoid process (HCP) 4.8; 4.6 4.0 Ô 0.2 4.7 Ô 0.2 3.3±4.4 4.2±5.0 height of coronoid valley (HCV) 2.9; 2.8 2.7 Ô 0.1 2.8 Ô 0.1 2.3±2.8 2.5±3.0 height of articular condyle (HAC) 3.8; 4.0 3.7 Ô 0.2 3.8 Ô 0.2 3.3±4.0 3.5±4.1 breadth of articular condyle (BAC) 3.0; 3.4 2.9 Ô 0.1 3.0 Ô 0.1 2.7±3.2 2.9±3.2 articular condyle to posterior edge of lower 4.8; 5.5 5.1 Ô 0.2 4.6 Ô 0.3 third molar (AC3) 4.4±5.5 3.8±5.1 length of mandibular toothrow (TRD) 6.0; 6.5 5.9 Ô 0.2 5.8 Ô 0.2 5.5±6.3 5.3±6.2 length of lower first molar (m1L) 1.9; 1.8 1.6 Ô 0.1 1.7 Ô 0.1 1.5±1.8 1.6±1.9 376 R. H. PINE et al.

R. creper, and Scotinomys xerampelinus. however, are isolated barrens with stunted The last two species and the toad Bufo peri- trees and sparse ground cover.According patetes were taken only at this locality. Per- to the collector, R.K. Enders, the holo- omyscus nudipes, trapped in cloud forest in type of endersi was caught in such a bar- and around Quebrada Alicia, was not taken renº (Choate 1970). Choate (1970) also here. stated that C. endersi ªprobably occurs in Setzer (1950) stated that the holotype was isolated barrens on upper Caribbean taken at ªCylindroº in the ªProvince of slopes of the Cordillera de Talamanca in Chiriquõº. Miller and Kellogg (1955) and western Panama and southeastern Costa Hall and Kelson (1959) accepted place- Ricaº and ªis known only from a soggy ment of Cylindro in ChiriquõÂ, but Fairchild barren with stunted trees and sparse and Handley (1966), Handley (1966), and ground cover on the upper northern slope Choate (1970) unequivocally placed Cylin- of the Cordillera de Talamancaº. Handley dro in Bocas del Toro Province. Fairchild (1966) gave the type locality as in ªfog for- and Handley (1966) wrote ªCylindro (NE estº (= cloud forest).This, along with the of Boquete on upper Caribbean slope, near new specimen's habitat, suggests that Boquete ... above 4000 feetº, Bocas del C. endersi is, indeed, a cloud forest species. Toro.They gave Boquete as in ChiriquõÂat Carleton and Musser (1995) listed C. en- ª2000±7500 feetº and at 8°47' N, 82°25' W. dersi as one of 17 terrestrial mammals en- Choate (1970) placed Cylindro at ªAp- demic to the highlands of western Panama proximatelyº 8°49' N, 82°24' W.There and/or of Costa Rica, and supposed that it seems to be no reason not to accept Cylin- might occur in the Cordillera de Talaman- dro's placement as given by Fairchild and ca of the latter country. Handley (1966), Handley (1966), and Cryptotis endersi remains one of the most Choate (1970).However, confusion has poorly known species of Recent mammals, persisted. Burton (1987) wrote ªOnly and as deforestation continues in high-ele- known from Bocas del Toro, Chiriqui, Pa- vation forests of Costa Rica and PanamaÂ, it namaº. Burton presumably followed Hall poses a serious threat to C. endersi and (1981), who stated ªtype from Bocas del other cloud forest specialists.It is listed as Toro, ChiriquõÂ, Panamaº±omitting the ac- Endangered by the IUCN (IUCN 1999) un- tual type locality, Cylindro, and treating Bo- der the spelling Cryptotis ender. cas del Toro as a specific locality rather than a province. Honacki et al.(1982) gave ªPanama, Chiriqui Bocas del Toro (sic), Cy- lindroº. Eisenberg (1989) wrote C. endersi Acknowledgements ªis known only from the area of Bocas del R.J.I zor and J.A.W agner did field work. Torro (sic) in northeastern (sic) Panamaº. J.S.B urt gave invaluable help, as did A.F ornero Hutterer (1993) placed the type locality in and others of Texasgulf Panama Inc., RTZ Devel- quotes, as follows, ªCylindro, above opment Enterprises, Ltd., and the Empresa de 4000 ft., Bocas del Toro, Panamaº, presum- Cobre Cerro Colorado, S.A. These firms, the ably quoting Choate (1970). Smithsonian Institution (USNM), and Field Mu- Setzer (1950) wrote that the holotype was seum provided monetary and/or logistic support. ªtaken in heavy forest near a rotting logº. The late C.O.H andley,Jr., gave useful advice, ªThe northern [Caribbean] slope supports and, with L.K.G ordon, H.L.K afka, and M.D.C arleton, loaned USNM specimens. much more luxuriant vegetation than the T.D aeschler made the holotype available to us. southern [Pacific] slope because it receives D.E.W ilson,E.MeÂndez,P.W.Freeman, substantially more rainfall ... Much of the J.K ethly,B.Morey, and the late F.S.L ukoschus area thus is covered with relatively undis- were helpful in various ways.J.R.C hoate dis- turbed rainforest, and is characterized by cussed the habitat of the holotype with us. almost inaccessible terrain and a paucity of human habitation.Within the rainforest, Rediscovery and redescription of Cryptotis endersi 377

References MeÂndez, E.(1970): Los Principales Mamõ Âferos Silvestres de PanamaÂ.ªPanamaº: Published Burton, J.A.(1987): The Collins Guide to the privately by the author. Rare Mammals of the World.Lexington, Miller, G. S., Jr.; Kellogg, R.(1955): List of Massachusetts: Stephen Greene Press. North American Recent mammals.U.S.Nat. Carleton, M. D.; Musser, G.G.(1995): Systema- Mus.Bull. 205, 1±954. tic studies of oryzomyine (Muridae: Myers, C.W.(1969): The ecological geography of ): definition and distribution of cloud forest in Panama.Am.Mus.Novit. Oligoryzomys vegetus (Bangs, 1902).Proc. 2396, 1±52. Biol.Soc.Wash. 108, 338±369. Myers, C. W.; Duellman, W.E.(1982): A new spe- Choate, J.R.(1970): Systematics and zoogeogra- cies of Hyla from Cerro Colorado, and other phy of Middle American shrews of the genus tree frog records and geographical notes from Cryptotis.Univ.Kans.Publ.,Mus Nat.Hist. western Panama.Am.Mus.Novit. 2752, 1±32. 19, 195±317. Nowak, R.M.(1991, 1999): Walker's Mammals of Choate, J. R.; Fleharty, E.D.(1974): Cryptotis the World.Vol.I.Baltimore: Johns Hopkins goodwini..Species 44, 1±3. Univ.Press. Corbet, G. B.; Hill, J.E.(1980, 1986, 1991): A Nowak, R. M.; Paradiso, J.L.(1983): Walker's World List of Mammalian Species.London: Mammals of the World.Vol.I.Baltimore: British Mus.(Nat.Hist.). Johns Hopkins Univ.Press. Eisenberg, J.F.(1989): Mammals of the Neotro- OConner, B. M.; Lukoschus, F. S.; Giesen, K.T. pics.The Northern Neotropics: Vol.I: Pana- M.(1982): Two new species of Marmosopus ma, Colombia, Venezuela, Guyana, Suriname, (Acari: Astigmata) from rodents of the genus French Guiana.Chicago: Univ.Chicago Press. Scotinomys () in Central America. Fain, A.; Lukoschus, F.S.(1984): New observa- Occas.Pap.Mus.Zool.Univ.Mich. 703, 1±22. tions on the genus Prolistrophorus Fain, 1970 Ridgway, R.(1912): Color Standards and Color (Acari: Astigmata: Listrophoridae).Syst. Nomenclature. Washington, D.C.: Published Parasitol. 6, 161±185. privately by the author. Fairchild, G. B.; Handley, C. O., Jr. (1966): Ga- Savage, J. M.; Donnelly, M.A.(1992): The sec- zetteer of collecting localities in Panama.In: ond collection of, and variation in, the rare Ectoparasites of Panama.Ed.by R.L. Wen- Neotropical toad Bufo peripatetes.J.Herpe- zel and V.J. Tipton.Chicago: Field Mus.Nat. tol. 26, 72±74. Hist.Pp.9±20. Setzer, H.W.(1950): Two new shrews of the Gjording, C.N.(1981): The Cerro Colorado Pro- genus Cryptotis from Panama.J.Wash.Acad. ject and the Guaymõ Indians of Panama.Cul- Sci. 40, 299±300. tural Survival Occas.Pap. 3, 1±45. Woodman, N.; Timm, R.M.(1993): Intraspecific Hall, E.R.(1981): The Mammals of North and interspecific variation in the Cryptotis ni- America.Vol.I.New York: John Wiley and grescens species complex of small-eared Sons. shrews (Insectivora: Soricidae), with the de- Hall, E. R.; Kelson, K.R.(1959): The Mammals scription of a new species from Colombia. of North America.Vol.I.New York: Ronald Fieldiana: Zool.(New Series) 74, 1±30. Press. Woodman, N.; Timm, R.M. (1999): Geographic Handley, C. O., Jr. (1966): Checklist of the mam- variation and evolutionary relationships mals of Panama.In: Ectoparasites of Panama. among broad-clawed shrews of the Cryptotis Ed.by R.L. Wenzel and V.J. Tipton.Chica- goldmani-group (Insectivora: Soricidae). go: Field Mus.Nat.Hist.Pp.753±795. Fieldiana: Zool.(New Series) 91, 1±35. Honacki, J. H.; Kinman, K. E.; Koeppl,J.W. (Eds.) (1982): Mammal Species of the World. Authors' addresses: Lawrence, Kansas: Allen Press and Associa- Dr.R onald H.P ine, IMSA, 1500 West Sullivan tion of Systematics Collections. Road, Aurora, Illinois 60506-1039, USA.(e-mail: Hutterer, R.(1993): Order Insectivora.In: Mam- [email protected]); Dr. Neal Woodman, USGS mal Species of the World: a Taxonomic and Patuxent Wildlife Research Center, National Geographic Reference.Ed.by D.E. Wilson Museum of Natural History, MRC-111, 10th St. and D.M. Reeder. Washington, D.C.: Smith- and Constitution Ave., NW, Washington, D. C. sonian Institution Press.Pp.69±130. 20560-0111, USA; and Dr.R obert M.T imm, Nat- Iucn (1999): Red Data Book.Mammalia.Gland, ural History Museum, University of Kansas, Switzerland: IUCN publ. Lawrence, Kansas 66045-2454, USA.