o HARVARD UNIVERSITY Ernst Mayr Library of the Museum of Comparative Zoology OCCASIONAL PAPERS of the HFn 9 i _ NATURAL HISTORY MUSEUM ^^^^ The University of Kansas Lawrence, Kansas ^^^"^^^^/rv NUMBER 174, PAGES 1-20 20 SEPTEMBER 1995 A New Species of Hyla (Anura: Hylidae) From Cloud Forest in Oaxaca, Mexico, With Comments on the Status of the Hyla bistincta Group Kevin R. Toal III' and Joseph R. Mendelson III Natural Histoiy Museum and Department of Systematics and Ecology, The University of Kansas. Lawrence. Kansas. 66045-2454. USA ABSTRACT A new species of Hyla. allocated to the H. bistincta group, is de- scribed from the cloud forest of the Sierra de Juarez and the Sierra Mixe in northern Oaxaca. Mexico. These populations previously were referred to H. siopela Duellman, but they differ in features of external morphology, osteology, and larval morphology. The taxonomic status of H. hogertae Straughan and Wright is re- viewed, and this taxon is placed in the synonymy of//, crassa (Brocchi). A revised key and a discussion of the taxonomic status of the H. bistincta group are presented. Key words: Anura: Hylidae: Hyla bistincta group: Hyla siopela: Hxia celata: Hyla bogertae: Hyla crassa: New species: Sierra de Juarez: Sierra Mixe: Oaxaca: Mexico. RESUMEN Se describe una nueva especie del genero Hyla. del grupo H. bi.stincta, de los bosques nublados de la Sierra de Juarez y la Sierra Mixe, norte de Oaxaca, Mexico. Estas poblaciones se asignaron anteriormente a H. siopela Duellman, pero son distintas en ciertas caracteristicas de la osteologia y la morfologia de adultos y renacuajos. Se examina el estado taxonomico de H. bogertae Straughan and Wright y se ubica a esta especie en la sinonimi'a de H. crassa (Brocchi). Se presenta una clave dicotomica y una discusion sobre el estado taxonomico del grupo H. bistincta. Palabras claves: Anura; Hylidae; Grupo Hyla bistincta; Hyla siopela; Hyla celata; Hyla bogertae: Hxla crassa; nueva especies; Sierra de Juarez; Sierra Mixe: Oaxaca: Mexico. 'Present Address: Department of Biological Sciences, Southeastern Louisiana Uni- versity. Hammond, Louisiana 70402. USA. © Natural llistmy Museiiiii, The Uiiiversily of Kan^as. Lawrence. ISSNiOOyi-Vy.SS 2 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 Currently, there are 49 recognized species of HyUt in Mexico (Liner, 1994; Mendelson and Campbell, 1994; Toal. 1994). 40 of which are en- demic to the country. This high number of endemics is likely attributable to the extensive mountain ranges that have produced islands of microhabitats along elevational gradients, a pattern known to exist in birds (Vuillemier, 1970) and mammals (Brown. 1971 ). The Sierra de Juarez in central Oaxaca is such an example. A subsidiary mountain chain of the Sierra Madre Oriental, the Sierra de Juarez contains hve floral associations along its northern slopes (Bogert. 1968; Wake et al., 1992) that provide a wide variety of microhabitats in which a rich fauna of amphibians and reptiles exists. For example, the type localities for several anurans (e.g., H. calvicolliiui. H. cyanomma, Eleiitherodactyliis polymniae), salamanders (e.g.. Thorhis smithi, Pseiidoeurycea saltator), and squamates {e.g.,Ab)-onia mitchelli, Exiliboa placata, Ciyophis hallbergi) are found along this slope, and several taxa remain undescribed (Hanken and Wake, 1994). Caldwell (1973) conducted an elevational transect study of the anuran community structure associated with the various habitats along the slopes of the Sierra de Juarez. Part of the material presented herein originated from collections that she made during that study. Species in the Hyla bistincta group (sensu Duellman, 1970) are endemic to the Mexican highlands northwest of the Isthmus of Tehuantepec (Duellman and Campbell, 1992), and range from southwestern Durango east to Puebla and south to the Pacific versant of Oaxaca. These species are of moderate size (>40 mm), and breed in cold mountain streams at eleva- tions from 1400 m {H. bistincta) to 3050 m (//. wbertsorwn). Many of these species have restricted distributions; several are known only from the type locality. Duellman (1968) described Hyla siopela from Cofre de Perote in open, dry pine forest of the Sierra de Madre Oriental in central Veracruz from 2500-2550 m. (See Wake et al. [1992] for vegetation types along the north and south slopes of Cofre de Perote.) Caldwell (1974) discovered a popu- lation of frogs in cloud forest on Cerro Pelon, the highest peak of the Sierra de Juarez, between 2640 and 2670 m. She referred these specimens to H. siopela, thereby documenting an isolated population about 200 km south of the only other known population. Since Caldwell's work, few additional specimens have been collected from this population. In light of the apparently restricted distributions of many of the species that occur along the slopes of the Sierra de Juarez, we were suspect of the disjunct distribution of the two populations of Hyla siopela. Critical exami- nation of the specimens of H. siopela frotn Veracruz and Oaxaca revealed consistent interpopulational differences that led to the investigation of the taxonomic status of the Oaxacan population that we present herein; this investigation supports recognition of this population as a distinct species. A NEW CLOUD FOREST HYLA FROM OAXACA 3 MATERIAL AND METHODS Measurements (in millimeters) were made using a Wild M5 dissecting microscope and digital calipers (0.01 mm accuracy) and were rounded to the nearest 0.1 mm; measurements and ratios follow those of Duellman (1970). Terminology follows that of Duellman (1970) and Harris (1979). Characters and measurements used in preparing the taxonomic key were taken from Duellman (1970), and those discovered by direct examination of specimens. Osteological observations were made from radiographs and cleared-and-double-stained specimens (Appendix). Sex was determined by direct observation of the gonads. Museum abbreviations follow stan- dard usage (Leviton et al.. 1985). Webbing formulae follow those of Savage and Heyer (1967), as modified by Myers and Duellman (1982). Snout-vent length is abbreviated SVL throughout. A list of specimens examined is given in the Appendix. SYSTEMATICS Hyki celata new species Figures 1-3 Holotype.— KU 137103, an adult male, from 0.9 km N Cerro Pelon, Sierra de Juarez, Oaxaca, Mexico, 2670 m ( 17°30' N, 96°30' W), obtained on 10 June 1970 by Janalee P. Caldwell and Paul B. Robertson. Field number JPC 2474. Paratypes.— KU 137090-92, 137095-97, 137099-100, 137105-10, 137112-14, 137116-18, all males, and KU 137093-4, 137098, 137101-2, 137104, 137115, all females, from 0.9-1.3 km N Cerro Pelon, Sierra de Juarez, Oaxaca, Mexico, 2640-2670 m (17°30' N, 96°30' W), collected between 18 December 1969 and 12 July 1970 by Janalee P. Caldwell and Paul B. Robertson: UTA A-5742 (male), UTA A-5743-44 (females) from 0.7 mi N crest of Cerro Pelon, N slope Sierra de Juarez, Oaxaca, Mexico, collected on 2 April 1978, by J. A. Campbell: UTA A- 13380 (female) from 25.4 km S Metates, Sierra de Juarez, Oaxaca, collected on 30 June 1 98 1 , by J. A. Campbell. Referred specimens. — See Appendix. Diagnosis. — A moderate-sized, stream-breeding frog distinguished from other Middle American Hyla by the following combination of charac- ters: (1) quadratojugal absent; (2) prepollex ossified, enlarged, blunt, bear- ing small nuptial excrescences in adult males; (3) vertical rostral keel pronounced; (4) skin moderately thick; (5) flanks mottled; (6) vocal slits absent; (7) tarsal fold absent; (8) tympanum and tympanic annulus evident; (9) fingers long with vestigial webbing; and (10) transverse processes on UNIV. KANSAS NAT. HIST. MUS, OCC. PAP. No. 174 Fig. 1. Hyla celata in life. A female (UTA A- 13380. paratype). from Ceno Pelon. Oaxaca. Mexico: SVL 36.4 mm. Figure reproduced from UTA color trans- parency 18153. photographed by W. W. Lamar; specimen collected by J. A. Campbell. the most posterior presacral vertebra slightly curved, angled 90° with respect to the vertebral column. Hyla celata is most similar to frogs placed in the H. hlstincta group (sensu Duellman, 1970) in lacking a quadratojugal and having a broad, ossified prepollex. Among these species. H. celata is most similar to H. siopela (sensu stricto) in coloration and body proportions, but differs by usually lacking a tarsal fold, having a well-defined vertical rostral keel, and slightly curved transverse processes on the most posterior presacral verte- bra; the processes are perpendicular to the long axis of the vertebral column. These processes are straight and acutely angled anteriorly in H. siopela (Fig. 4). Hyla celata can be distinguished from all other members of the H. bistincta and H. arborescanclens groups (sensu Duellman. 1970) by having the following combination of characters: tarsal and thoracic folds absent; tympanum and tympanic annulus evident, distinct; vocal slits absent in adult males; nuptial excrescences present only on prepollex; and rostral keel distinct. Description of holotype. — Adult male; body slender; head as wide as body; head width 31.47f SVL; head length 29.5% SVL. Snout truncate in A NEW CLOUD FOREST HYLA FROM OAXACA 5 protile (Fig. 2). slightly inclined posteroventrally. bluntly rounded in dorsal view with a well-pronounced vertical rostral keel; snout length (anterior margin of eye-anterior margin of head distance) 36.8% head length; ros- trum flat; canthus rostralis distinct, rounded; loreal region concave; lips rounded; nostrils ovoid, barely protuberant, directed dorsolaterally. situ- ated 2/3 distance from anterior corner of eye to tip of snout; internarial region slightly sloped ventrally. Top of head flat; interorbital distance 31.5% head width; eye diameter greater than eyelid width. 27.8% head width, eyelid width 33.8% head width. Supratympanic fold extending from middle of posterior edge of orbit, adjacent to and obscuring dorsal margin of tympanic annulus. terminating above forelimb at level of ventral margin of tympanic annulus, thick anteriorly, becoming thinner posteriorly; tym- panic annulus thin, distinct, protruding above slightly differentiated skin of tympanum; tympanum diameter 54.0% eye diameter; distance between tympanum and eye 1 .5 times tympanum diameter. Axillary membrane absent; thoracic fold absent; fold on wrist slight. Fingers long, bearing slight lateral fringes, terminal discs round. Forearm robust; length of longest finger equal to forearm length; relative length of fingers 1 < 2 < 4 < 3; disc on first finger as large as tympanum; nuptial excrescences fine, brown, located only on ossified, enlarged, blunt prepollex; subarticular tubercles small, one-third size of thenar tubercle, round, none bifid; supernumerary tubercles numerous, diffuse, present on palmar surface and proximal phalanges of digits; thenar tubercle flat, elliptical, situated upon prepollex midway between proximal phalange of first finger and wrist; palmar tubercle flat, bifid, large lobe three times size 5 mm Fig. 2. Profile aspect of head of Hyla celata (KU 137103, holotype). 6 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 of subarticular tubercles, small lobe equal in size to thenar tubercle; ulnar tubercles and fold absent; webbing on fingers vestigial (Fig. 3). Heels overlapping about one-fourth length of shank when hind limbs adpressed; transverse dermal fold on heel distinct; tarsal fold absent; tibia length 47.6% SVL; foot length 45.9% SVL; inner metatarsal tubercle elongate anteroposteriorly. flat, two times size of subarticular tubercle of first toe; outer metatarsal tubercle absent; length of longest toe equal to length of shank; length of toes 1 < 2 < 5 < 3 = 4; toe discs slightly smaller than those on fingers; subarticular tubercles of toes moderately small, round; supernu- merary tubercles numerous, small, flattened, absent from two most distal phalanges of digits; webbing on toes: I V/i — 2'/: II 2 — 3 III 2 — 3'/: IV 3 — 2 V(Fig. 3). Cloacal opening directed posteriorly at midlevel of thighs; cloacal sheath short, directed posteriorly; transverse cloacal fold prominent, situated dorsomedially to top of thighs; cloacal tubercles present on ventral surfaces of inner thighs. Skin on dorsum smooth, finely scabrous at midbody (possibly an artifact of preservation); skin on venter finely verrucose; skin smooth on remaining surfaces. Tongue broadly cordiform. free posteriorly, shallowly notched; dentigerous processes situated slightly posterior to level of choanae, transverse, ovoid, bearing five and four teeth, separated medially by a distance equal to distance between ovoid choanae. Vocal slits absent. ' ■" \i 5 mm Fig. 3. Ventral aspect of the hand and foot of holotype of Hyla celaki (KU 137103, holotype). Note the vestigial webbing on the hand. Nuptial excrescences on the ventral surface of the prepollex have been abraded, and are thus not evident from this position. A NEW CLOUD FOREST HYLA FROM OAXACA 7 Measurements: Snout-vent length 42.4, forearm length 16.0. tibia length 20.2, foot length (measured from proximal edge of inner metatarsal tu- bercle to tip of longest toe) 19.5. head length 12.5. head width 13.3. head depth 5.4. snout length 4.6, interorbital distance 4.2, eye diameter 3.7, eyelid width 4.5. tympanum diameter 2.0. internarial distance 4.0. Coloration in preser\'ative: Dorsum and flanks olive-brown with nu- merous black punctations. reduced on thighs; flanks mottled with white blotches. Venter dull yellow, densely mottled throughout with small brown flecks under 25x magniflcation; posterodorsal surface of cloacal region bearing small white spots; ventral surface of forearm possessing single row of three white spots on left arm, four on right; white spots present around cloacal region; palpebral membrane clear, superior edge densely mottled. Coloration in life: ( Fig. 1 ; UTA color transparency 18153, female paratype UTA A- 13380): All dorsal surfaces bronze-brown with scattered distinct black flecks that are most concentrated on the lateral surfaces of the body; canthus rostralis dark brown; dorsal surfaces of fingers dull yellow distally; eye dark brown with thin black reticulations and few, small, red spots near periphery of upper eyelid. The Held notes of J. P. Caldwell ( 10 June 1970) provide the following description: "Dorsum tan with black mottling, covered evenly with bronze chromatophores; abdomen and chin iridescent white; ventral surfaces of limbs pinkish gray; toe pads dull yellow; canthus bronze above, brown below; eyes bronze with black reticulations." Variation: Variation in body measurements between males and females of Hyla celata and H. siopela are listed in Table 1 . Ranges of ratios of those measurements of//, celata (males, followed by females in parentheses) are as follows: tibia length/SVL 4.59-5.37 (4.59-5.20); foot length/SVL 4.02- 4.67 (4.32-4.70); head length/SVL 2.60-3.27 (2.88-3.12); head width/ SVL 2.98-3.23 (2.94-3.27); snout length/head length 3.01-4.33 (2.97- 4.70); internarial distance/head width 1.78-2.99 (2.00-2.56); interorbital distance/head width 3. 12-3.57 (2.97-3.58); eye diameter/head length 2.58- 4.35 (2.76-3.67); tympanum diameter/eye diameter 3.14-6.45 (3.96-6.68). Information on variation in color-pattern variation among six specimens was taken from the field notes of J. P. Caldwell (11 May 1970; 10 June 1970). No statement of sex was given, but we were able to discern colora- tion of the holotype by her field notes on 10 June as the "smaller [male] specimen" collected along with a "larger [female] specimen." This pro- vides unequivocal color-in-life data for one male and one female. Color-in- life data for five paratypes are as follow: KU 137102. female, dorsal coloration was leaf-green with brown mottling with bronze chromato- phores localized at edges of green on the sides of body and the sides of limbs; KU 137098-101 had dorsal mottling varying from dark green and brown in two individuals to light and dark brown in one individual, and light and dark green in another; sides of the body were delineated from the 8 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 Table 1. Comparison of measurements of the type series adults of Hyki ceUtta (Pelon) and H. siopela (Perote). Mean and standard deviation are given over range in parentheses. TL = tibia length, FT = foot length, HL - head length. HW = head width, ED = eye diameter, EN = eye-naris distance. TD = tympanum diameter. IND = internarial distance, lOD = interorbital distance. Males H. celata H. siopehi Females H. celata H. siopela Character /; = 24 u = l n= 15 SVL 42.0 ± 1.8 42.5 ± ."^.1 45.7 ± 4.0 47.9+ 2.9 (37.8-56.2) (39.9-44.4) (37.8-51.1) (44.0-51.4) TL 20.7 ±0.9 21.1 ± 0.7 22.8 ± 1.6 24.0 ± 0.7 (18.5-21.7) (20.1-22.3) (19.1-24.9) (22.9-24.9) FT 18.6 ± 0.7 20.0 ± 0.8 21.2 ± 1.6 22.4 10.9 (17.3-19.9) (18.7-21.2) (17.5-24.3) (21.1-23.5) HL 12.5±0.9 12.5±0.6 13.9± 1.1 14.210.6 (10.6-13.9) (11.6-13.5) (11.1-15.4) (13.3-15.0) HW 13.0 1 0.6 (11.4-14.3) 12.71 0.3 (12.0-13.1) 14.3 1 0.9 14.2 11.1 (12.4-15.6) (12.5-15.5) ED 4.1 1 0.4 (3.4-4.9) 3.7 1 0.3 (3.3-4.1) 4.3 1 0.6 (3.3-5.7) 4.01 0.3 (3.6-4.5) EN 3.3 1 0.2 (3.0-4.0) 3.4 1 0.2 (2.9-3.6) 3.7 1 0.3 (3.1-4.1) 3.5 1 0.2 (3.2-3.7) TD 1 .9 1 0.2 (1.4-2.3) 2.0 1 0.2 (1.7-2.3) 2.2 1 0.3 (1.7-2.6) 2.2 1 0.2 (2.0-2.4) IND 2.9 1 0.4 (2.2-4.0) 2.9 1 0.2 (2.7-3.1) 3.2 1 0.2 (3.0-3.4) 3.2 1 0.3 (3.0-3.7) lOD 4.3 1 0.2 (3.7-4.6) 4.5 1 0. 1 (4.3-4.6) 4.8 1 0.4 (4.2-5.4) 4.9 1 0.3 (4.5-5.4) donsum by a variable mottling pattern consisting of dark brown, iridescent copper, light green, and yellow coloration; chin mottling varied from solid brown in some individuals to absent in others; mottling varied from sparse to absent on the abdomen, never covering its entire surface; black dots on the dorsum and flanks of certain individuals arc replaced by brown blotches. A NEW CLOUD FOREST HYLA FROM OAXACA Fig. 4. Dorsal aspect of the presacral vertebrae of Hyla siopekt (A. KU 137103. holotype) and H\la celata (B. KU 100976. paratype) to show differences in the condition o\' the most posterior transverse processes. Shaded areas represent cartilage. Note that KU 105629 has an anomalous lateral process on the urostyle. Ulnar tubercles vary from absent (condition of the holotype) to a series of three to five small bumps on each arm. Males have two to six teeth on each dentigerous process, whereas females have three to six teeth. One adult male (KU 1 37 1 1 8) and an adult female (KU 1 37 1 1 5) lack dentigerous processes altogether. Nuptial excrescences vary (in preservative) from dull yellow (color of prepollex) to tan or dark brown. In one adult male (KU 137109). the nuptial excrescences have been removed, presumably after preservation. Tadpoles. — Caldwell ( 1974) provided a complete description and illus- tration of the tadpole of Hyla celata (KU 139841, as H. siopela). The tadpoles of H. celata are moiphologically similar to those of H. siopela (described and illustrated by Duellman. 1970), except that those of H. celata are larger than those of H. siopela at each Gosner stage (Caldwell, 1974). Caldwell (1974) suggested that interspecific competition with sym- patric species of stream-breeding frogs at Cerro Pelon, or differences in water temperature between the two localities, might account for the ob- served differences in overall size of the tadpoles; no other hylids are known to be sympatric with Hyla siopela. This hypothesis never has been tested explicitly in these species. Distribution and ecology. — Hyla celata is known only from two locali- ties in cloud forest: Cerro Pelon in the Sierra de Juarez and near Santa Maria Totontepec in the Sierra Mixe. This species is known to breed in mountain streams at elevations of 2640 m and 2670 m on Cerro Pelon. On two separate occasions, specimens were collected from a mossy rock wall 10 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 at night sitting in the spray of a small waterfall (field notes of J. P. Caldwell, 1970). J. A. Campbell (pers. comm.) found this species active at 1400 h in direct sunlight on rocks in the middle of a stream, along with individuals of H. cyanomma. Bogert (1968) provided a detailed description of the vegeta- tion along the slopes of the Sierra de Juarez. The cloud forest habitat along the slopes of the Sierra de Juarez has been severely disturbed by logging. J. A. Campbell (pers. comm.) has visited the type locality of Hyla celata a number of times since 1980, and has not located individuals of H. celata, H. calvicollina, or H. cyanomma. If these frogs are sensitive to environmental change, as many stream-breeding frogs are, these species may have been extirpated from Cerro Pelon. Additional field work is needed to determine the status of these frogs. There is little, if any, undisturbed cloud forest within ca. 15 km of Santa Maria Totontepec (field notes of J. R. Mendelson, 17 July 1992). Active searches for hylid frogs were not made during that visit. Etymology. — The specific epithet is derived from the Latin celo mean- ing hidden or concealed, in reference to the way in which specimens of this species were "concealed" under the name Hyla siopela. Remarks. — Among some of the referred specimens of Hyla celata. we noticed some variation in the diagnostic characters. For example, six individuals (KU 148664-69) from the type locality have more robust bodies, unpatterned venters, and weak tarsal folds. In addition, MVZ 161028 has a tarsal fold and lacks a distinct rostral keel: MVZ 180384 lacks a rostral keel, and CAS 14 10 1 3 has a distinct tarsal fold. The latter three specimens are from Cerro Felon. Among the specimens from the Sierra Mixe, UTA A-6843, 6846, and 6936 lack rostral keels. In our experience, some soft anatomical characters, such as tarsal folds, vary among specimens of many species in the H. bistincta group and may be affected by different preservation techniques. Nonetheless, we tentatively refer the aforementioned specimens to H. celata until more knowledge of variation is obtained from additional material. In her review of anuran osteological characters, Trueb (1977) suggested that the angles formed by the transverse processes of the vertebrae might be too variable to be informative for elucidating phylogenetic relationships. Although this is the case in her study on Hyla lanciformis, we found no intrapopulational variation in the shape, or angle of the leading edge, of the processes on the most posterior presacral vertebra. Thus, we considered the condition exhibited by H. celata to be a unique diagnostic character. Field work by J. A. Campbell and associates from UTA in the Sierra Mixe has resulted in the discovery of populations of Hyla sabrina, H. crassa, H. celata, and H. arborescandens. Hyla sabrina and H. celata otherwise are known only from the Sierra de Juarez, and H. arborescandens is known from many localities throughout the Sierra Madre Oriental of A NEW CLOUD FOREST HYLA FROM OAXACA 1 1 Puebla, Veracruz, and Oaxaca. and in the Sierra de Juarez (Duellman, 1970). Historically. Hylo crassa is known from only a few localities in central Oaxaca. (See discussion below.) These findings come as little surprise because a continuous comdor of cloud forest habitat exists through both the Sierra de Juarez and the Sierra Mixe (Campbell and Frost. 1993); several other species of amphibians and reptiles (e.g.. H. chaneqite, Abronia fuscolabialis, Exiliboa placata) exhibit similar distribution patterns (Campbell and Frost, 1993). Although the area between the Sierra de Juarez and Cofre de Perote remains poorly known herpetologically. recent field work in the interven- ing Sierra de Huautla by several parties has not revealed any frogs referable to the Hyla bistincta group (Mendelson. pers. observ.; A. Nieto, pers. comm.). This corroborates our hypothesis that H. siopela and H. celata are indeed allopatric. TAXONOMIC NOTES ON THE HYLA BISTINCTA GROUP The status of Hyla bogertae. — While preparing a key to the species of the Hyla bistincta group, we noticed many similarities between the descrip- tions of H. crassa (Brocchi) and H. bogertae Straughan and Wright. Duellman (1964; 1970) redescribed H. crassa and referred H. robustofemora Taylor to the synonymy of this species; at this time only two specimens were available for study. Straughan and Wright (1969) de- scribed H. bogertae from a series of adults, juveniles, and tadpoles; this species was diagnosed from H. robustofemora (= H. crassa) based on putative differences in the tarsal fold, foot webbing, and the condition of the prepollex. Caldwell (1974) reported on variation among additional specimens (KU 125354, 139850, 139854, 148696-700) that she referred to H. crassa and described the tadpole of this species. All specimens referred to H. crassa by Duellman ( 1 970) and Caldwell ( 1 974) are from north of the city of Oaxaca in the Sierra Aloapaneca and the southern slope of the Sierra de Juarez, whereas the type series of H. bogertae is from a tributary of the RioAtoyac in the Sierra de Cuatro Venados, about 50 km (by air) SWof the city of Oaxaca. We have examined an additional series of specimens referable to H. crassa (UTA A-5786. 5842-43, 5894) that were collected in 1978 in the Sierra Mixe. Examination of this more recently collected material indicates that the diagnostic characters of Hyla bogertae (viz., absence of tarsal fold, less webbing on fourth toe, presence of a single bony element in the prepollex, and absence of webbing on the hand) are present, albeit variable, among specimens referred to H. crassa. Among specimens of H. crassa, tarsal folds vary from weak (UTA A-5842, male; KU 148696, female) to strong (KU 148697, male; KU 148698. female) and webbing on the fourth toe 12 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 may reach the base of the disc (KU 148697, male; KU 125354, female), or not (UTA A-5843. 5894, females). Similarly, the prepollex of H. crassa may contain one (UTA A-5842, male) or two (KU 148697. male) elements. Ossification of the prepollex may progress through ontogeny; the larger (presumably older) male specimen that we examined (UTA A-5842, SVL 57.9 mm) has a single prepollical element, whereas the smaller male (KU 148697, SVL 52.9 mm) has two such elements. However, the male allotype of H. bogertae is shown by Straughan and Wright ( 1969) to have a single prepollical element and it is smaller than the specimen of H. crassa with which they compared it. Number of elements in the prepollex also may vary without respect to body size (such as in Pseudacris, D. Meinhardt, pers. comm.). or may be easily obscured through different radiograph techniques and positioning of specimens (Mendelson. pers. observ.). The amount of webbing on the hand in all specimens examined by us closely matches that of three of the paratypes (KU 152338-39, and LACM 44401 as illustrated by Adler and Dennis 1 1972:14]). The variation in the degree of concealment of the tympanum by the supratympanic fold noted by Caldwell ( 1974) is also evident among the UTA specimens from the Sierra Mixe. Based on this evidence, we hereby place Hyla bogertae Straughan and Wright as a junior synonym of Caiiphias crassiis Brocchi 1877. The distribution of this species is now known to include the Sierra Aloapaneca, Sierra de Cuatro Venados, and the Sierra Mixe (sensu Binford, 1 989). in the state of Oaxaca, Mexico. The status of the Hyla histincta group. — A number of definitions of the Hyla bistincta group have been proposed either explicitly (e.g., Duellman. 1964, 1970; Caldwell, 1974) or implicitly (Duellman and Campbell, 1992). However, none of these is satisfactory under the tenets of the phylogenetic system (Farris, 1979). Each definition promotes the rec- ognition of paraphyletic groups based on combinations of plesiomorphic characters, as is the case with most species groups in Hyla (Duellman and Hoogmoed, 1992). In their outgroup analysis of the phylogenetic relation- ships of Plectroliyla, Duellman and Campbell (1992) created a new defini- tion of the H. bistincta group; they noted four synapomorphies that united the H. bistincta group and Plectrohyla as sister taxa. They also identified six synapomorphies that corroborated the monophyly of Plectrohyla, but were unable to provide evidence for the monophyly of the H. bistincta group (Duellman and Campbell. 1992). In their reanalysis of the phylogenetic relationships of Plectrohyla. Wilson et al. ( 1994) placed a new species {P. chrysopleiira) in the genus. Although they recognized the problems of Duellman and Campbell's ( 1992) outgroup analysis (i.e.. the nonmonophyly of the Hyla bistincta group), Wilson et al. (1994) did not address these problems. Their cladogram exhibited a greater degree of resolution than that of Duellman and A NEW CLOUD FOREST HYLA FROM OAXACA 1 3 Campbell, but their result is flawed by the same problems pertaining to outgroup structure and does not clarify the status of the H. bistincta group. Given this, the phylogenetic hypothesis proposed by Wilson et al. (1994) for relationships of Plectwhyla species should be viewed as preliminary. As the current taxonomy stands, the Hyla bistincta group is defined as those species possessing ( I ) a long medial ramus of the pterygoid articulat- ing with the otic capsule. (2) thick dorsal skin. (3) continuous fringing papillae on the upper lips of tadpoles, (4) expanded labia with one or more rows of large submarginal papillae medial to the fringing papillae of tadpoles (Duellman. 1970: Duellman and Campbell. 1992). and lacking the synapomorphies of Plectwhyla. However, these characters prove to be problematic with regard to their appli-^ability to many species in the H. bistincta group. Many species are known from few specimens (e.g.. H. calvicollina, celata, cembra, mykter, pachyderma), thereby rendering os- teological comparisons unfeasible. Moreover, larvae of H. calvicollina, cembra, charadricola, chryses, mykter. pachyderma, and sabrina are un- known. In addition, the condition of the dorsal skin seems to be defined subjectively among species of the H. bistincta group. Compared to other members of this group. H. charadricola and H. chryses have "thinner skin" (Duellman, 1970:457) and, on this basis, were excluded from the H. bistincta group by Duellman and Campbell (1992). Adier and Dennis (1972:15) recognized two "series" of species within the H. bistincta group — those with "more glandular skin" {H. bogertae, crassa, pachyderma) and those with "thinner skin" (//. mykter robertsorum. siopela). Within Plectrohyla, dorsal skin is thick and glandular (Duellman and Campbell, 1992). From our examinations, as well as those by the aforementioned workers, dorsal skin thickness is a poorly defined character for these species and varies continuously among them: therefore, we ques- tion the utility of this character as providing synapomorphic information for this group of frogs (but see Thiele, 1993). We agree with Duellman and Campbell ( 1992) that, in the absence of a phylogenetic analysis, the historical value of the name '^Hyla bistincta group" should be preserved. However, we think it important to emphasize its typology as being paraphyletic with respect to Plectrohyla. As such, we recommend that the definition of Duellman ( 1970) be followed as a tenta- tive, but more suitable, definition of the H. bistincta group. Thus, we place H. charadricola, chryses, calvicollina, celata, cembra, cyanomma, mykter, and sabrina into the H. bistincta group, although perhaps as a "divergence from the main evolutionary line within the [H. bistincta] group" (Duellman, 1970:457). The addition of these frogs to the Hyla bistincta group brings the number of species in the group to 14, and others remain undescribed (Toal and Mendelson, pers. obs!: J. A. Campbell, pers. comm). A rigorous cladis- tic analysis conducted with additional specimens (adults, larvae) of the H. 14 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 bistincta group and Plectwhyla, along with additional data sets, hopefully will yield a better understanding of the evolution of this group of Middle American treefrogs. When such an analysis is performed, it should also include H. arborescandens and H. miohinpamim. These two species share many characters with species of the H. bistincta group (Duellman. 1970:371-372) and. in fact, may be related closely to these frogs. Only at that time will any hope of delineation of the H. bistincta group come to fruition. KEY TO THE SPECIES OF THE HYLA BISTINCTA GROUP This key is designed to aid in the identification of live and preserved adults of the Hyla bistincta group. We have included H. arborescandens in this key because it may be mistaken easily for members of the H. bistincta group. (See comments in text.) Many of the characters used in this key are variable; as such, we encourage all users of this key to verify their identi- fications with the original descriptions or Duellman (1970). 1 . Toes webbed to base of disc on all toes (only to middle of penultimate phalange of fourth toe in some specimens); snout distinctly round in dorsal view; heavy supratympanic fold, often concealing tympanum; large size (males to 57.9 mm SVL. females 63.1) H. crassa Only fifth toe webbed to base of disc, or not; snout shape and size variable; supratympanic fold weak to moderate, never concealing entire tympanum 2 2. Tympanum and tympanic annulus not evident 3 Tympanum and tympanic annulus evident, distinct or indistinct.... 5 3. Nuptial excrescences as enlarged spines; tarsal and thoracic folds present, strong; known only from type locality (Pan de 011a, Sierra Madre Oriental. Veracruz) H. pachyderma Nuptial excrescences as fine granules, or absent 4 4. Tarsal and thoracic folds absent; dorsum mottled green, or uniform green; venter tan with dark brown mottling; flanks dark brown, posterior surfaces of thighs dull yellow; eyes red-brown; size small (males to 29.9 mm SVL, females 41.3); known only from the Sierra de Juarez and Sierra Mixe, Oaxaca H. sabrina Tarsal and thoracic folds present, weak; dorsum uniform green, sometimes with tiny yellow spots; venter gray, unmarked; flanks and posterior surfaces of thigh pale blue; eyes pale blue; size large (males A NEW CLOUD FOREST HYLA FROM OAXACA 15 to 56.0 mm SVL. females 64.5); known only from Cerro Pelon, Sierra de Juarez. Oaxaca H. cyanomma 5. Vocal slits present; tarsal fold prominent 6 Vocal slits absent; tarsal fold weak or absent 7 6. Rostral keel absent; posterior surfaces of thighs with brown reticulate markings; nuptial excrescences covering prepollex and inner surface of second finger in males; size large (males to 53.8 mm SVL. females 67.8) H. bistincta Rostral keel present; posterior surface of thighs without brown re- ticulate markings; nuptial excrescences covering prepollex only in males; size small (males to 40.0 mm SVL, females 51.6) H. arborescandens 7. Webbing extending to base of disc on fifth toe 8 Webbing extending to middle of penultimate phalange of fifth toe ^ 11 8. Tympanic annulus distinct, at least interiorly, diameter at least 40% eye diameter 9 Tympanic annulus indistinct, diameter less than 40% eye diameter 10 9. Cloacal sheath long; thoracic fold very weak; known from three localities in the Sierra Madre del Sur of Guerrero (Asoleadero, Puerto del Gallo, and Omilteme) H. mykter Cloacal sheath short; thoracic fold weak; known from several locali- ties in the pine and pine-fir forests of the Sierra Madre Oriental of Hidalgo and Puebla H. robertsorwn 10. Nuptial excrescences present; single male 37.0 mm SVL; known only from type locality (33.8 rd km N Candelaria Loxicha. 2160 m. Sierra Madre del Sur, Oaxaca) H. cembra Nuptial excrescences absent; size small (males to 44.4 mm SVL, females 50.9); known from several localities in the pine-oak forests of the Sierra Madre Oriental of Hidalgo and Puebla H. charadricola IL Lateral surfaces of head, body, and limbs darker than dorsum, with distinct dark brown markings; size large (males to 52.1 mm SVL, females 56.4) H. pentheter 16 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 Lateral surfaces of head, body, and limbs paler than dorsum, without distinct dark markings 12 12. Distinct thoracic fold and tarsal fold present; rostral keel absent; size small (males to 41.1 mm SVL. females 44.0) H. chiyses Thoracic fold absent; tarsal fold and rostral keel present or absent; size variable 13 13. Tarsal fold absent; dorsum of rostrum concave in profile; rostral keel present; tympanum indistinct; fingers short, with broad lateral fringe; known only from Cerro Pelon. Sierra de Juarez. Oaxaca H. calvicolliiui Tarsal fold and rostral keel present or absent; rostrum flat in profile; tympanum distinct; fingers long, without broad lateral fringe 14 14. Rostral keel weak or absent; tarsal fold weak; known only from Cofre de Perote. Veracruz H. siopela Rostral keel strong; tarsal fold absent; known only from the Sierra de Juarez and Sierra Mixe, Oaxaca H. celata Acknowledgments: We thank J. A. Campbell (UTA). R. W. McDiarmid (USNM), J. Vindum (CAS). A. G. Kluge and G. Schneider (UMMZ). S. Sroka (UIMNH), and D. B. Wake and H. W. Greene (MVZ) for loans of critical specimens. We thank K. Adler, J. A. Campbell, and L. Trueb for critical reviews of the manuscript, and B. I. Crother. W. E. Duellman. and D. A. Kizirian for helpful comments. J. A. Campbell provided much assistance, including valuable information on the ecology and current status of Hyla celata, as well as providing the photograph of H. celata in life. J. P. Caldwell and R. Altig provided important information. W. E. Duellman. H. da Silva. I. de la Riva. A. Maglia. and E. R. Wild offered assistance with the key. D. P. Lawson provided observations from several UTA specimens. J. E. Simmons provided helpful assistance. JRM received financial support from the Theodore Roosevelt Fund of the American Museum of Natural History. The Panorama Society of the KU Natural History Museum, and the Stearns Travel Grant Program of the California Academy of Sciences. LITERATURE CITED Adler, K.. and D. M. Dinnis. 1972. New tree frogs of the genus Hyla from the cloud forests of western Guerrero, Mexico. Univ. Kansas Mus. Nat. Hist. Occas. Pap. 7:1-19. BiNFORD, L. C. 1989. A distributional survey of the birds of the Mexican state of Oaxaca. Ornithol. Monoszr. 43:14-18. A NEW CLOUD FOREST HYLA FROM OAXACA 17 BoGERT, C. M. 1968. Anew genus and species of dwarf boa from southern Mexico. Am. Mus. Novit. 2354:1-38. Brown. J. H. 1971. Mammals on mountaintops: nonequilibrium insular biogeogra- phy. Am. Nat. 105:467-478. Caldwell. J. P. 1973. Tropical tree fw^ comiuiiiiities: patferns of reproditction. size, aiul utilization of structural habitat. Doctoral dissertation. Lawrence: Univ. Kansas. Caldwell, J. P. 1974. A re-evaluation of the Hyla bistincta species group, with descriptions of three new species (Anura: Hylidae). Univ. Kansas Mus. Nat. Hist. Occas. Pap. 28:1-37. Campbell. J. A., and D. R. Frost. 1993. Anguid lizards of the genus Abronia: revisionary notes, descriptions of four new species, a phylogenetic analysis, and key. Bull. Am. Mus. Nat. Hist. 216:1-121. DuELLMAN. W. E. 1964. A review of the frogs of the Hyla bistincta group. Univ. Kansas Publ. Mus. Nat. Hist. 15:469-491. DuELLMAN. W. E. 1968. Descriptions of new hylid frogs from Mexico and Central America. Univ. Kansas Publ. Mus. Nat. Hist. 17:559-578. DuELLMAN, W. E. 1970. Hylid frogs of Middle America. Univ. Kansas Mus. Nat. Hist. Monogr. 1:1-752. DuELLMAN, W. E., AND J. A. CAMPBELL. 1992. Hylid frogs of the genus Plectrohyla: systematics and phylogenetic relationships. Univ. Michigan Mus. Zool. Misc. Publ. 181:1-32. DuELLMAN. W. E., AND M. S. HoociMoED. 1992. Somc hylid frogs of the Guiana highlands, northeastern South America: new species, distributional records, and a generic reallocation. Univ. Kansas Mus. Nat. Hist. Occas. Pap. 147:1-21. Farris, J. S. 1979. The information content of the phylogenetic system. Syst. Zool. 28:483-519. Hanken. J., AND D. B. Wake. 1994. Five new species of minute salamanders, genus Thorius (Caudata: Plethodontidae), from Oaxaca, Mexico. Copeia 1994:573- 590. Harris, R. A. 1979. A glossary of surface sculpturing. Occas. Pap. Entomol. (Sacramento). 28:1-31. Leviton, a. E., R. H. Gibbs. Jr., E. Heal, and C. E. Dawson. 1985. Standards in herpetology and ichthyology: Part L Standard symbolic codes for institutional resource collections in herpetology and ichthyology. Copeia 1985:802-832. Liner. E. A. 1994. Scientific and common names for the amphibians and reptiles of Mexico in English and Spanish. Soc. Study Amphibians Reptiles Herpetol. Circ. 23. Mendelson, J. R.. III. and J. A. Campbell. 1994. Two new species of the Hyla sumiclirasti group (Amphibia: Anura: Hylidae) from Mexico. Proc. Biol. Soc. Wash. 107:398-409. Myers, C. W.. and W. E. Duellman. 1982. A new species of Hvla from Ceno Colorado, and other tree frog records and geographical notes from western Panama. Am. Mus. Novit. 2752:1-25. Savage, J. M., and W. R. Heyer. 1967. Variation and distribution in the treefrog genus Phyllomedusa in Costa Rica, Central America. Beitr. Neotrop. Fauna 5:111-131. 18 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 Straughan, I. R.. AND J. W. Wright. 1969. A new stream breeding frog from Oaxaca, Mexico (Anura: Hylidae). Contrib. Sci. Los Angeles County Mus. 169:1-12. Theile. K. 1993. The holy grail of the perfect character: the cladistic treatment of morphometric data. Cladistics 9:275-304. ToAL, K. R., III. 1994. A new species of Hyla (Anura: Hylidae) from the Sierra de Juarez, Oaxaca, Mexico. Herpetologica 50:187-193. Trueb, L. 1977. Osteology and anuran systematics: intrapopulational variation in Hyla kmciformis. Syst. Zool. 26:165-184. VuiLLEMiER, F. 1970. Insular biogeography in continental regions. I. The northern Andes of South America. Am. Nat. 104:373-388. Wake, D. B., T. J. Papenfuss, and J. F. Lynch. 1992. Distribution of salamanders along elevational transects in Mexico and Guatemala. Tulane Stud. Zool. Bot. Supp. Pub. 1:303-319. Wilson, L. D., J. R. McCranie, andG. A. Cruz. 1994. Anew species of P/^crro/7y/rt (Anura: Hylidae) from a premontane rainforest in northern Honduras. Proc. Biol. Soc. Wash. 107:67-78. APPENDIX Specimens examined and their localities. All specimens are post-meta- morphic and preserved in fluid, unless otherwise noted. Holotype and paratypes of Hyla celata are not included herein. Hyla arborescaudens MEXICO: Oaxaca: Sierra Mixes, 13.6 mi [21.9 km] (by road) NE Tamazulapan, UTA A-5761-62. 5891-92; Sierra Mixes, 3.6 mi [5.8 km] (by road) W Totontepec, UTAA-5763-67. 5774-75, 5777-81; Sierra Mixes, 11.4 mi [18.4 km] W (by road) W Totontepec. UTA A-5801-06: Sierra Mixes, Totontepec, UTA A-5838-39; Si- erra Mixe, 3. 1 mi [5.0 km] W Totontepec, UTA A-6846, 6933: Sierra Mixe, 5.6 km W Totontepec. 2121 m, UTA A- 1 3 1 96. Hyla calvicollina- MEXICO: Oaxaca: Sierra Juarez, 4 km N Ceno Pelon, 2518 m. KU 186356 (holotype); Sierra de Juarez. 0.7 mi [1.1 km] N Cerro Pelon, 27 1 2 m, UTA A-5059 (paratype). Hyla celata MEXICO: Oaxaca: Sierra Mixes, 3.1 mi [5.0 km] W Totontepec. UTA A-6936. 6843-46; bus stop shed, Cerro Pelon, 15.0 km SW Vista Hermosa on MX Hwy. 175, MVZ 1 80384; 52-54 km (by road) N of Guelatao, N slope Cerro Pelon on MX Hwy. 175, MVZ 161028; Distrito Sola de Vega, 2 km below Cerro Pelon, CAS 141013; 0.9 km N Cerro Pelon, 2670 m, KU 137089 (cleared-and-stained). ^Toal (1994) described Hxia calvicallina from the Sierra de Juarez, Oaxaca. A NEW CLOUD FOREST HYLA FROM OAXACA 19 Hyki ceinhra MEXICO: Oaxaca: Campamento Rfo Molino, 2160 m. KU 137035 (holotype). Hyki charadricola MEXICO: Pitc'hla: 14.4 km W Huachinango, 2280 m, KU 58414 (holotype), 53813-15. 152367-68. HyUt chryses MEXICO: Guerrero: between Puerto Chico and Asoleadero. 2540-2600 m, UMMZ 125374 (holotype). KU 106306 (paratype); Sierra Madre del Sur. 0.6 km SW Carrizal de Bravos, 2200 m. AS 142452-53; Sieira Madre del Sur, 5.5 km SW Carrizal de Bravos. 2523 m. CAS 142938-50; Sierra Madre del Sur. Carrizal de Bravos. 2505 m, CAS 143926-29; 4 km WNW Carrizal de Bravos. 2579 m. MVZ 112330-32. Hyla crassa MEXICO: Oaxaca: 9 mi [ 1 4.5 km] NE Oaxaca. KU 1 25354; 1 .9 km S El Estudiante, ca. 1 850 m. KU 148696-700; tributary of Rio Atoyac below Vivero El Tapanal. 1 .6 km S La Cofradia, 2652 m. KU 152338-40 (paratypes of H. bogertae: 152340 = tadpoles); 6.3 km SE Ixtlan. 1910 m. KU 139850 (tadpole); 2.3 km E and 11.6 km NE Oaxaca. 1720 m. KU 139854 (tadpoles); Sierra Mixes, 3.6 mi [5.8 km] W (by road) Totontepec. UTA A-5786. 5842-43. 5894; CeiTO San Felipe. UIMNH 25050 (holotype of//, robustofemora). Hyla cyanomma MEXICO: Oaxaca: 31.2 mi [50.2 km] N Guelatao, 9600 ft, KU 100507; 1.2 km N Cerro Pelon. 2650 m, KU 148656-63. 152370-72; Ceno Humo Chico (= adjacent to Cerro Felon). UIMNH 73892-98. Hyla luykter MEXICO: Guerrero: 11.4 km (by road) SW of Puerto del Gallo, 1985 m. KU 137553 (holotype); Asoleadero. 2520 m. KU 137552 (paratype); 1.6 km N Puerto del Gallo. 2438 m, UTA A-4108, 41 10-1 1; 0.8 km SW Omiiteme. UTA A-4904; 12.4 km SW Filo de Caballo, near intersection with Chichihualco road. UTA A- 13364. Hyla pachyderma MEXICO: Veracruz: Pan de Olla. S of Tezuitlan (Puebla). USNM 115026-27 (paratypes). Hyla pentheter MEXICO: Oaxaca: 32.9 km N San Gabriel Mixtepec, 1530 m, KU 136886-88. 20 UNIV. KANSAS NAT. HIST. MUS. OCC. PAP. No. 174 Hvla robertsorum MEXICO: Hiclali^o: El Chico Parqiie Nacional. 3050 m, KU 71265-95. Hyla sahhua MEXICO: Oaxaca: 1 1 . 1 km S Vi.sta Hermosa. 1 840 m. KU 1 37044 (paratype); I 1 .6 km S Vi.sta Hermosa, 1910 m. KU 137053 (paratype); 1 1 .9 km S Vista Hermosa, 1920 m, KU 137059-60, 137064. 137066, 137069-72 (paratypes); 12.3 km S Vista Hermosa. 1920 m, KU 137076 (paratype): 15.8 km S Vista Hermosa. 1990 m. KU 137085 (paratype); 16.6 km S Vista Hermosa, 2020 m KU 137087 (paratype); 25 mi [40.3 km] S of Valle Nacional, UTAA-4357; Sierra Mixes, 3.6 mi [5.8 km] (by road) Totontepec, UTA A-5771, 5776, 5782-83; Sierra Mixes, Totontepec, UTAA- 5838. Hyla siopela MEXICO: Veracni:: W slope Cofre de Perote, 2500-2550 m, KU 100976-80 (paratypes). 100981 (holotype), 100982-85 (paratypes), 100986-94. 105628 (paratype) 105629 (cleared-and-stained, paratype), 105630-34; Rancho El Capulio, 7 km SE Perote, 2920 m, KU 129163. UNIVERSITY OF KANSAS NATURAL HISTORY MUSEUM PUBLICATIONS The University of Kansas Publications. Natural History Museum, beginning with volume 1 in 1946. was discontinued with volume 20 in 197 1 . Shorter research papers formerly published in the above series are now published as The University of Kansas Natural History Museum Occasional Papers. The University of Kansas Natural History Museum Miscellaneous Publications began with number 1 in 1946. Longer research papers are published in that series. Monographs of the Natural History Museum were initiated in 1970. Authors should contact the editor regarding style and submission procedures before manuscript submission. All manuscripts are subjected to critical review by intra- and extramural specialists; final acceptance is at the discretion of the Direc- tor. This publication is printed on acid-free paper. Occasional Papers and Miscellaneous Publications are typeset using Aldus PageMaker"" on a Macintosh computer. ® Natural History Museum. The University of Kansas, Lawrence. Institutional libraries interested in exchanging publications may obtain the Occasional Papers and Miscellaneous Publications by addressing the Exchange Librarian, The University of Kansas Library, Lawrence, Kansas 66045-2800, USA. Individuals may purchase separate numbers from the Office of Publications, Natural History Museum, The University of Kansas. Lawrence, Kansas 66045-2454, USA. Editor: Linda Trueb Managing Editor: Joseph T. Collins Printed by LlNIVIiKSITY OF KaN.SAS PRINTING SERVICE Lawrence, Kansas 3 2044 093 361 723