HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

UNIVERSITY OF KANSAS

MUSEUM OF NATURAL HISTORY

PUBLICATIONS

The University of Kansas Publications, Museum of Natural History,
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 Museum of Natural History
Occasional Papers. The University of Kansas Museum of Natural His-
tory Miscellaneous Publications began with number 1 in 1946. Longer
research papers are published in that series. Monographs of the Museum
of Natural History 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 Microsoft" Word and Aldus
PageMaker" on a Macintosh computer. Museum of Natural History,
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, Museum of Natural History, The
University of Kansas. Lawrence, Kansas 66045-2454, USA.

Editor: Linda Trueb
Managing Editor: Joseph T Collins

Printed by

University of Kansas Printing Service

Lawrence, Kansas

OCCASIONAL PAPERS

of the

MUSEUM OF NATURAL HISTORY
The University of Kansas
Lawrence, Kansas

number 162. pages 1-76 24 november 1993

Systematics of the Leptodactylid Frog Genus
Telmatobius in the Andes of Northern Peru

John J. Wiens1

Museum of Natural History and Department of Systematics and Ecology,
The University of Kansas, Lawrence. Kansas 66045-2454

ABSTRACT The taxonomic status of populations of Telmatobius (Anura:
Leptodactylidae) in the Peruvian departments of Amazonas, Cajamarca, La
Libertad, and Piura is evaluated using an evolutionary species concept and data
from external morphology, osteology, morphometries, larval morphology, and
allozyme electrophoresis. Nine species are recognized, and a key for identification
is presented. The three named species are redescribed using the type specimens and
newly collected material. These species are: T brevipes Vellard, from the Cordillera
Occidental of Cajamarca and La Libertad; T. ignavus Barbour and Noble, from the
Cordillera Huancabamba in Piura; and T. latirostris Vellard, from the city of
Cutervo in Cajamarca. Lectotypes are designated for T. brevipes and T latirostris.
Six species are described as new. Telmatobius atahualpai inhabits high elevations
in the Cordillera Central in Amazonas. Telmatobius colanensis is known from a
single specimen from the Cordillera Colan in Amazonas. Telmatobius degener and
T. thompsoni are sympatric species from a locality in the Cordillera Occidental of
La Libertad. Telmatobius degener is characterized by several unusual osteological
features, and is the only known anuran that lacks premaxillary teeth but retains
maxillary teeth. Telmatobius necopinus is known from the lower slopes of the
Cordillera Central. Telmatobius truebae. found at several localities in the Cordillera
Central, is externally similar to T. necopinus, but is distinct osteologically and
biochemically, and is not the sister taxon of that species. Striking variability in the
morphology of the columella (middle ear bone) is described within and among
species. Generic assignment of the nine species is argued phylogenetically, and the
genus Telmalsodes is placed in the synonymy of Alsodes.

Key words: Leptodactylidae; Telmatobius ; T. atahualpai, T colanensis, T. degener,
T. necopinus, T. thompsoni, T. truebae; New species; Peru; Systematics.

'Present Address: Department of Zoology, The University of Texas, Austin, Texas
78712-1064.

© Museum of Natural History. The University of Kansas. Lawrence. ISSN:0091-7958

2 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

RESUMEN Se estudid la sistematica del genero leptodactflido de ranas
Telmatobius en los Andes del norte del Peru. Se analizo el status taxonomico de
poblaciones de Telmatobius en los departamentos de Amazonas, Cajamarca, La
Libertad, y Piura aplicando un concepto evolutivo de especie y usando datos de
morfologfa externa, osteologia, morfometrfa, morfologfa de renacuajos, y
aloenzimas. Los analisis revelaron la existencia de nueve especies en esta zona. Se
justifica filogeneticamente la asignacion de estas especies al genero Telmatobius.
Las tres especies descritas fueron redescrita usando el material tipo y material
nuevo. Estas especies son: T. brevipes Vellard, de la Cordillera Occidental de
Cajamarca y La Libertad; T. ignavus Barbour and Noble, de la Cordillera de
Huancabamba en Piura; y T. latirostris Vellard, se conoce solamente de la ciudad de
Cutervo en Cajamarca. Se designan lectotipos de T. brevipes y T. latirostris. Se
describen seis especies como nuevas. Telmatobius atahualpai es un habitante de las
partes altas de la Cordillera Central en el Departamento Amazonas. Telmatobius
colanensis se conoce solo un especimen de la Cordillera de Colan en Amazonas.
Telmatobius degener y T. thompsoni son especies simpatricas de una localidad en la
Cordillera Occidental en La Libertad. Telmatobius degener se caracteriza por
muchos rasgos osteologicos. Telmatobius necopinus se conoce de una ladera baja
de la Cordillera Central en Amazonas. Telmatobius truebae, se encuentra en varias
localidades en la Cordillera Central en Amazonas, es externamente similar a T.
necopinus, pero es distinta en caracteres osteologicos y bioqui'micos, y no es el
taxon hermano de esta especie. Se describe la notable variacion en la morfologfa de
la columela (hueso medio del oido) dentro y entre las nueve especies. Se presenta
tambien una clave de identificacion y datos sobre la ecologfa.

Palabras claves: Leptodactilidos; Telmatobius; T. atahualpai, T. colanensis, T.
degener, T. necopinus, T. thompsoni, T. truebae; Especies nuevas; Peru; Sistematica.

The leptodactylid frogs of the genus Telmatobius are a diverse and
problematic group from the central and southern Andes mountains of South
America. Thirty species and 30 subspecies of Telmatobius were listed by
Frost (1985). Since then, one species has been transferred to another genus
{Phrynopus juninensis, Cannatella. 1985) and five new species have been
described (T. scrocchii, Laurent and Lavilla, 1986; T. hypselocephalus and
T. platycephalus, Lavilla and Laurent, 1988a; T. pinguiculus Lavilla and
Laurent, 1988b; T. carrillae, Morales, 1989). The genus ranges from south-
ern Ecuador, through Peru and Bolivia, to the Andes of northern Argentina
and Chile. Although Telmatobius are found between 1300 and 5000 m
elevation (T. ceiorum and T. marmoratus, respectively), most species occur
near 3000-3500 m (Cei, 1986). In general, Telmatobius are nocturnal,
secretive, aquatic to semiaquatic frogs of moderate to large size.

As discussed by Trueb (1979) and Cei (1986), the systematics of
Telmatobius is complex and difficult. Descriptions and diagnoses of most
species are inadequate and not comparable among accounts. The existence
of numerous subspecies suggests that the number of (evolutionary) species
in the genus has been greatly underestimated. Cei (1986) and Trueb (1979)

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 3

discussed confusing and discordant patterns of inter- and intrapopulation
variation in Tel mat obi us, and Trueb (1979) suggested that morphological
data alone might be insufficient to resolve the status of some populations.
Furthermore, supraspecific groupings based on explicit hypotheses of phy-
logenetic relationships have not been proposed, numerous species have
been shown to have been incorrectly assigned to the genus (e.g.. Lynch,
1978; Cannatella, 1985), and the monophyly of the genus remains highly
suspect (e.g., Cei, 1986; Morales, 1989). Trueb (1979) began the onerous
task of revising the genus with her revision of the Telmatobius of Ecuador.
In this paper I review the systematics of the genus in the adjacent
geographic region, the Andes of northern Peru. More specifically, the
purposes of this paper are to: ( 1 ) evaluate the status of Telmatobius popu-
lations from the departments of Amazonas, Cajamarca, La Libertad, and
Piura using an evolutionary species concept and data from a diversity of
character systems (external morphology, osteology, larval morphology,
morphometries, allozyme electrophoresis); (2) redescribe the three previ-
ously named species from the region (T brevipes Vellard, T ignavus
Barbour and Noble, T. latirostris Vellard); (3) describe six new species; and
(4) justify inclusion of these species in Telmatobius.

STUDY AREA

The physiography of northern Peru is dominated by the Cordillera
Occidental, the Cordillera Central, and the Huancabamba Depression. The
Huancabamba Depression is a system of low arid valleys (ca. 1 000-2000 m
above sea level) and isolated mountain ranges (including the Cordillera
Colan and Cordillera Huancabamba) that separate the northern and central
Andes (Duellman, 1979). South of the Huancabamba Depression, the
central Andes are divided into the Cordillera Occidental and Cordillera
Central by the low, arid valley of the Rio Maranon. The lower Pacific
slopes of the Cordillera Occidental are relatively arid, with desert at low
elevations successively giving way to spiny forest, dry forest, cloud forest,
and paramo (montane grassland above treeline), at the highest elevations
(Tosi, 1960). The Cordillera Central is more lush, with paramo at the
highest elevations, cloud forest at lower elevations, and subtropical
rainforest on the lowest Amazonian slopes (Tosi, 1960). The geographic
distribution of each species is shown in Figure 1, and their altitudinal
distributions are shown in Figure 2.

MATERIALS AND METHODS

Decisions about the taxonomic status of populations should appeal to an
explicit definition of what species are believed to be (e.g., Cracraft, 1992).
In this study, I use an evolutionary species concept, as proposed by Simpson

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 1 . Distribution of species of Telmatobius in northern Peru. T. brevipes
(circles), T. colanensis (plus sign), T. ignavus (inverted triangles), T. latirostris
(square), T. necopinus (hexagon), T. truebae (triangles), sympatric T. degener and T.
thompsoni (diamond), and sympatric T. atahualpai and T. truebae (star). Localities
are approximated from distances by road; multiple localities in close proximity are
represented by a single symbol. Dotted lines show areas over 3000 m, solid lines
indicate limits of Peruvian departments. The inset shows the location of the study
area within Peru.

( 1961 ) and modified by Wiley (1978) and Frost and Hillis (1990). Thus, I
consider a biparental species to be the largest single lineage within which
there is reproductive cohesion and which maintains its identity apart from
other lineages through time. I favor this concept because it requires that the
recognized taxa be logically consistent with evolutionary history (in con-

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

CD

>

03
<D
c/>

<X>
>
O
-Q
CC

</>
Q)

CD
"O

3

4000—,

3000

^ 1000 —

2000 —

Fig. 2. Altitudinal distributions of species of Telmatobius in northern Peru.

trast to the biological species concept) and avoids the philosophical pitfalls
and potential overreductionism of the phylogenetic species concept (Frost
and Hillis, 1990). The evolutionary species concept provides a general
conceptual framework for recognition of species, rather than a set of strict
operational rules (Frost and Hillis, 1990). In this study, I considered the
following to be evolutionary species: ( 1 ) populations that apparently were
isolated geographically (especially if they were morphologically and/or
biochemically distinct); (2) diagnosable sympatric forms; and (3) groups of
homogeneous, geographically contiguous populations that were presumed
to be interbreeding and/or whose status as a single monophyletic lineage
was not rejected by phylogenetic analysis of the allozyme data. Because
delineation of real evolutionary entities can be a difficult and even subjec-
tive undertaking (e.g., Frost and Hillis, 1990), I consider it necessary to
give a more explicit justification and explanation for species-level deci-
sions than a mere listing of the distinguishing characters of the putative
species. Taxonomic decisions made in this study are explained and dis-
cussed near the end of the Results section.

Over the course of the study, 22 1 preserved adult and subadult frogs
(including the type material and newly collected specimens of the 3 de-

6 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

scribed species), 14 lots of tadpoles, 19 cleared-and-stained and dry skel-
etal preparations, and tissues from 21 individuals from northern Peru were
examined (Appendix I). Institutional abbreviations are as follow: Museum
of Natural History, The University of Kansas (KU), Louisiana State Uni-
versity Museum of Natural Science (LSUMZ), Museum of Comparative
Zoology, Harvard University (MCZ), Museo de Historia Natural,
Universidad Nacional Mayor de San Marcos, Lima (MHNSM, formerly
MHNJP), Florida State Museum, University of Florida (UF). Terminology
for external features generally follows Duellman (1970), and formulae for
toe webbing follow Savage and Heyer (1967) as modified by Myers and
Duellman (1982). Cleared-and-stained skeletons were prepared using a
modified version of the technique of Dingerkus and Uhler (1977). Termi-
nology for osteological features follows Trueb (1973) and Duellman and
Trueb(1986).

For descriptive purposes and morphometric analysis, 12 measurements
were recorded from adult specimens to the nearest 0.1 mm with needle-
tipped calipers. These are abbreviated throughout as follows: SVL (snout-
vent length), TIBL (tibia length; knee to heel), FOOT (foot length; proxi-
mal edge inner metatarsal tubercle to tip of fourth toe), HLEN (head length;
posterior corner of jaw to tip of snout), HWID (head width; from posterior
corner of jaw), IOD (interorbital distance), IND (internarial distance),
ENOS (eye-nostril distance; from anterior corner of eye), EYE (eye diam-
eter; between anterior and posterior corners of eye), HNDL (hand length;
proximal edge of outer palmar tubercle to tip of third finger), THBL (thumb
length; proximal edge inner palmar tubercle to tip of thumb), RDL
(radioulnar length; elbow to distal edge of outer palmar tubercle). Sexual
maturity was determined by presence of eggs in females (or relative size in
large females without eggs) and development of the nuptial excresence in
males. Descriptive statistics for these measurements were computed using
the SYSTAT statistical package. Multivariate analyses of the morphomet-
ric data were performed using the SAS statistical package on the IBM
mainframe computer system at the University of Texas at Austin. Multi-
variate techniques are described in the section on morphometries. In the
species descriptions, ranges of proportions are followed in parentheses by
the mean, one standard deviation, and the sample size.

Tadpoles were staged according to the table of Gosner (1960). Termi-
nology for larval characters follows Altig and Johnston (1989), and the
format for tadpole descriptions follows McDiarmid and Altig (1990).

Allozyme electrophoresis provides invaluable data for inferring gene
flow, phylogenetic relationships among closely related taxa, and species
boundaries (Hillis and Moritz, 1990). These data are particularly useful in
groups with complex patterns of inter- and intrapopulation morphological
variation, such as Telmatobius. (See Cei, 1986, and Trueb, 1979, for

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 7

examples.) Tissues for electrophoretic analysis were obtained from nine
populations of Telmatobius (representing 4 putative species) during the
spring of 1989. Throughout the paper, the populations of T. brevipes
sampled electrophoretically are numbered as follow: 1 = Cajamarca: 21 km
NNW Cajamarca; 2 = Cajamarca: 8 km S Cajamarca; 3 = Cajamarca: 9 km
S Celendin; 4 = La Libertad: 19 km SSE Cajabamba. Populations of T.
truebae are labelled as follows: 1 = Amazonas: 11 km SE Chachapoyas; 2
= Amazonas: 28 km SSW Leimebamba; 3 = Amazonas: Pomacochas
(Florida). Liver and muscle tissues were removed from animals freshly
killed using a 10% solution of benzocaine. Tissues were frozen immedi-
ately in liquid nitrogen for transport to the laboratory, were stored at -70°C,
and were used within 1 yr of collection.

Electrophoretic techniques and data analysis follow Wiens and Titus
(1991). Liver and skeletal muscle were homogenized separately with a
teflon homogenizer in a 1:1 (v:v) mixture of tissue and 0.01 M Tris-0.001
M EDTA-0.001 M mercaptoethanol. pH 6.8. Homogenates were centri-
fuged at 13,446 g for 10 min at 5°C. Tissue samples were run at 5°C on
horizontal starch gels composed of 12% hydrolyzed potato starch. Products
of the presumptive gene loci were visualized by histochemical staining
methods (Harris and Hopkinson, 1976; Selander et al., 1971; Siciliano and
Shaw, 1976). Buffer recipes are from Hillis (1985) and enzyme nomencla-
ture follows the recommendations of the International Union of Biochem-
istry Nomenclature Committee (1984). Loci were numbered from anode to
cathode and alleles were labelled a, b, c, etc., in order of increasing anodal
mobility. Enzymes, loci, tissue sources, and buffer systems are listed in
Table 1.

To estimate the evolutionary relationships among the populations of
Telmatobius sampled, I subjected the allozyme data to parsimony analysis.
Methods for phylogenetic analysis of allozyme data currently are contro-
versial, largely because of the problem of dealing with the abundance of
intraspecific variation (see Swofford and Berlocher 1987, for a recent
review). I prefer direct use of frequencies over strict "presence/absence"
treatment of allelic data, as the latter method discards potentially useful
information (e.g., presence of an alllele at a frequency of 1% as different
from 99%; Swofford and Olsen, 1990) and is highly subject to sampling
error (Swofford and Berlocher, 1987). However, the computer algorithm
for quantitative parsimony analysis of frequency data (FREQPARS;
Swofford and Berlocher, 1987) has a weak tree-searching routine and,
therefore, is unlikely to find the shortest tree for larger data sets (Swofford
and Berlocher, 1987). As a compromise, I used quantitative ("presence/
absence") coding of the data to determine the shortest and near-shortest
trees (using the PAUP software package; Swofford, 1990), and then input
these topologies into FREQPARS for evaluation.

8 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Table 1 . Enzymes, loci, International Union of Biochemistry Nomenclature Com-
mittee numbers, tissue sources and buffer systems.

Enzyme

Locus

IUBNC

Tissue

Buffer

Adenosine deaminase

Ada

3.5.4.4

muscle

TC7.0

Aspartate aminotransferase

Aat

2.6.1.1

liver

TBE9.1-NAD

Alcohol dehydrogenase

Adh

1.1.1.1

liver

TBE9.1-NAD

Glycerol-3-phosphate

dehydrogenase

G3pdh

1.1.1.8

liver

TBE9.1-NAD

Glucose-6-phosphate

isomerase

Gpi

5.3.1.9

liver

TBE9.1-NAD

3-Hydroxyisobutyrate

dehydrogenase

Hbdh

1.1.1.31

liver

TBE9.1-NAD

Isocitrate dehydrogenase

Icdh-2

1.1.1.42

liver

LiOH-NADP

Lactate dehydrogenase

Ldh-1

1.1.1.27

liver

TBE9.1-NAD

Ldh-2

1.1.1.27

liver

TBE9.1-NAD

Malate dehydrogenase

Mdh-2

1.1.1.37

liver

LiOH-NADP

Malic enzyme (NADP-

dependent malate

Me-1

1.1.1 .40

liver

TC 7.0-NADP

dehydrogenase)

Me-2

1.1.1.40

liver

TC 7.0-NADP

Mannose-6-phosphate

isomerase

Mpi

5.3.1.8

muscle

TC7.0

Peptidase-B

Pep-B-1

3.4.13.11

liver

LiOH-NADP

Pep-B-2

3.4.13.11

liver

LiOH-NADP

Phosphoglucomutase

Pgm

5.4.2.2

liver

Poulik

Sorbitol dehydrogenase

Sdh

1.1.1.14

liver

TBE9.1-NAD

Superoxide dismutase

Sod

1.15.1.1

liver

TBE9.1-NAD

Triosephosphate isomerase

Tpi

5.3.1.1

liver

TBE9.1-NAD

For the qualitative analysis, electromorphs were coded by considering
the locus as the character and the allelic array (combination of alleles
present in a terminal taxon) as the character state. Polymorphisms within
populations were weighted using the step-matrix option (following Mabee
and Humphries, 1993). The appearance of a "new" allele as a polymor-
phism was given an a priori weight of 0.5, as was the fixation of that allele
(or the loss of the plesiomorphic allele). Thus, the transformation between
the allelic array aa to ab would be weighted by 0.5, between aa and cc by
1 .0, between aa and cd by 1 .5, and between ab and cd by 2.0. The branch-
and-bound algorithm (Hendy and Penny, 1982) was used to guarantee
finding the shortest tree(s). To avoid considering ambiguously placed char-
acter states as support, I checked state assignments to each stem using both

SYSTEM ATICS OF THE FROG GENUS TELMATOBIUS IN PERU 9

the accelerated transformation (ACCTRAN; Farris, 1970; Swofford and
Maddison, 1987) and delayed transformation (DELTRAN; Swofford and
Maddison, 1987) optimization routines.

Because there is no evidence that the species examined constitute a
monophyletic group within Telmatobius (and therefore no appropriate
outgroup taxa could be identified), all trees were unrooted. Each population
was treated as a separate terminal taxon. Although some of these popula-
tions almost certainly do not represent distinct, separately evolving lin-
eages (thus violating a basic assumption of phylogenetic analysis), using
them as terminal taxa allowed testing of their relationships without making
a priori assumptions about their taxonomic assignment.

To avoid creating or exacerbating nonmonophyletic taxa, assignment of
newly described species to genera should be justified based on phyloge-
netic hypotheses rather than overall similarity and convenience. The prob-
lem of generic allocation without reference to phylogeny is particularly
evident in Telmatobius; numerous species have recently been removed
from Telmatobius and reassigned to different genera (e.g.. Lynch, 1978)
and even different tribes (e.g., Cannatella, 1985). In this study, generic
assignment is argued using derived characters that unite the new species
with the type species of Telmatobius (T. peruvianus Wiegmann 1835), and
is discussed at the end of the Results section. Telmatobiine specimens
examined for the purpose of determining generic placement are listed in
Appendix II.

In the Species Accounts, descriptions generally follow the format of
Trueb (1979). However, diagnoses only distinguish among the species
from northern Peru. The three Ecuadorian species were treated by Trueb
(1979); Telmatobius cirrhacelis and T vellardi have distinctive color pat-
terns, whereas T. niger can be distinguished from the northern Peruvian
species by its unusual arrangement of nuptial spines. Telmatobius niger has
small conical spines that are restricted to the dorsal and medial surface of
the thumb, with a disjunct patch of spines usually (71% of 11 males)
present near the inner margin of the inner palmar tubercle. The Telmatobius
in the adjacent region of central Peru (the department of Ancash) are being
revised by Antonio Salas (MHNSM; in prep.) and all seem to be distinct
from the species in northern Peru (pers. obs. and pers. comm.). I assume it
unlikely that more geographically distant taxa will be conspecific or poten-
tially confused with the species discussed in this study. Because so little is
known about the phylogenetic relationships of the species of Telmatobius
(and polarities of character states therefore are difficult to determine),
species are "diagnosed" in the traditional sense (diagnosis = characteriza-
tion) and the diagnoses are not restricted to derived character states. Termi-
nology and distribution of ecological life zones are based on Tosi (1960).

10

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

RESULTS

External Morphology

There are few consistent differences in overall head shape among the
species of Telmatobius from northern Peru. The shape of the snout (and
associated characters) seems to vary continuously within and among the
species. In general, the snout is long and sloping in T. thompsoni (Fig. 3C),
moderately long in T. brevipes (Fig. 3A), moderately short in T. truebae
(Fig. 4A) and T. degener (Fig. 3D), short and bluntly rounded in T.
atahualpai (Fig. 5), T. colanensis (Fig. 4C), T. ignavus (Fig. 3B), and T.
necopinus (Fig. 4B). and very short and sloping in T. latirostris (Fig. 4D).

Except for Telmatobius ignavus and a few (7% of 67 specimens) T
brevipes, all Telmatobius in northern Peru lack a tympanum, and in T.
ignavus the tympanum is partially concealed by the supratympanic fold
and may be difficult to detect (Fig. 3B). A tympanum occurs as a polymor-
phism in three populations of T. brevipes near the city of Cajamarca.

Fig. 3. Lateral view of heads of Telmatobius. A. T. brevipes, KU 2 1 2440. B. T.
ignavus, KU 181442. C. T. thompsoni, UF 39733. D. T degener, UF 39751. Scales
= 5 mm.

SYSTEMATICS OF THE FROG GENUS TELMATOBWS IN PERU

11

Palmar supernumerary tubercles are present in Telmatobius colanensis,
T. degener, T. ignavus, T. necopinus, and T. thompsoni, are present in some
individuals of T. brevipes, T. latirostris, and T. truebae, and are absent in T.
atahualpai(Figs. 5, 7, 8). All specimens of T. truebae from Pomacochas
(the type locality) are uniquely characterized by having an outer palmar
tubercle that is concave at its outer distal margin (in dorsal view); a small
supernumerary tubercle is either adjacent to or fused to this concavity (Fig.
8A). Plantar supernumerary tubercles occur in all T. degener and in some T.
thompsoni and T. truebae, and are absent in the remaining species. Outer
metatarsal tubercles are present in all species, but show considerable inter-
and intraspecific variation in size. The outer metatarsal tubercles are con-
sistently small and round in T. necopinus (about one-fourth size of inner
metatarsal tubercle), tend to be small in T. truebae (usually about one-third
size inner tubercle), and are relatively large in T. degener and T. thompsoni
(about two-thirds to four- fifths size of inner tubercle); in the other species,
the outer tubercle is usually about half the size of the inner tubercle.
Variation in toe webbing is generally uninformative (see Species Accounts

D

Fig. 4. Lateral view of heads of Telmatobius. A. T. truebae, KU 181528. B. T.
necopinus, KU 212483. C. T. colanensis, LSUMZ 37102. D. T. latirostris. KU
212450. Scales = 5 mm.

12

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 5. Lateral view of head and palmar view of right hand of T. atahualpai,
UF 39715. Scales = 5 mm.

for webbing formulae of each species), but the toe webbing is relatively
extensive in T. colanensis, especially between the fourth and fifth toes (Fig. 6).

The size and distribution of keratinized nuptial excrescences on the
thumbs of adult males provide invaluable characters for distinguishing
among species of Telmatobius (Figs. 7-8). Telmatobius colanensis, T.
latirostris, T. necopinus, and T. truebae have large, conical nuptial spines
that are restricted to the dorsal and medial surface of the thumb. The
slightly raised nuptial pad is distinct from the inner palmar tubercle in each
of these species. Among these species, the spines are slightly smaller in T.
colanensis, T. latirostris, and one of the two male T. necopinus. Telmatobius
degener also has relatively large, conical spines, but the spines extend onto
the ventral surface of the thumb and the nuptial pad is continuous with the
inner palmar tubercle. Telmatobius brevipes, T. ignavus, and T. thompsoni
possess minute, close-set nuptial spines. The spines are slightly larger and
more distinctly conical in T. thompsoni than in T. brevipes and T ignavus.
The spines are more extensive on the ventral surface of the thumb in T.
brevipes than in T. thompsoni, and are more extensive ventrally in T.
thompsoni than in T. ignavus. In T. brevipes and T. thompsoni, the swollen
nuptial pad obscures the limits of the inner palmar tubercle, whereas in T.
ignavus the inner palmar tubercle remains distinct.

The dorsum is smooth in Telmatobius atahualpai, T degener, T ignavus,
T. necopinus, and T. thompsoni, and usually is smooth in T. brevipes (90%;
n = 67); it usually is coarsely granular in T. latirostris and tuberculate in T.
truebae (82%; n = 67). The venter is smooth in all species (pustular near the
hindlimbs in male T. colanensis and T. truebae), but the skin ventromedial
to the cloaca usually is pustular in all species except T. atahualpai.

SYSTEMATICS OF THE FROG GENUS TELMATOB1US IN PERU

13

Fig. 6. Plantar view of left feet of Telmatobius. A. T. tmebae, KU 181515. B.
T. colanensis, LSUMZ 37102. Scales = 5 mm.

Photographs of living Telmatobius atahualpai, T brevipes, T. ignavus,
T. latirostris, T. necopinus, and T. truebae are shown in Figure 9. The
dorsum is dull brown to gray in T brevipes, T ignavus, T. necopinus, and T.
truebae (in life and preservative) and T. colanensis, T. degener, and T.
thompsoni (preservative only, color in life unknown). Darker spots are
variably present in these species, and at least one individual of T. necopinus
had yellowish-tan flecks on the dorsum in life. Telmatobius latirostris is
unique in having a yellowish-orange dorsum with black spots and reticula-
tions (in life). In living T. atahualpai, the dorsum is black with greenish-
gold (or metallic green) flecks. The venter and undersides of the limbs are
dull gray to tan (in life and preservative) in T atahualpai, T. colanensis
(color in life unknown), T. ignavus, and T. necopinus. Bright mottling
(yellowish to lavender in life) occurs on the ventral surfaces of the limbs
(and sometimes on adjacent parts of the venter) in T brevipes (53%; n =
67), T. latirostris, and T. truebae (90%; n - 67). In preserved T. degener and
T. thompsoni, the venter is yellowish cream with gray speckling.

14

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 7. Palmar view of left hands of male Telmatobius. A. T. brevipes, KU
2 12435. B. T. ignavus, UF 34089. C. T. thompsoni, UF 39734. D. T. latirostris, KU

212450. Scales = 5 mm.

Osteology

The skulls of the seven species for which osteological material is
available are illustrated in Figures 10-12. The skull appears relatively wide
in Telmatobius latirostris (Fig. 1 1 A). The sphenethmoids of T. degener, T.
thompsoni, and T. truebae are poorly ossified and lack the anteromedial
ossification between the nasal capsules that is present in adults of the other
species. The single adult skeleton of T. thompsoni has a small, independent
ossification (or mineralization) center within the optic foramen (Fig. 10F);
I am not aware of a similar ossification in other anurans. The prootic and

SYSTEMATICS OF THE FROG GENUS TELMATOB1US IN PERU

15

Fig. 8. Palmar view of right hands of male Telmatobius. A. T. truebae, KU
212476. B. T. necopinus, KU 212481. C. T. colanensis, LSUMZ 37102. D. T.
degener, UF 39746. Scales = 5 mm.

exoccipital are fused to each other in adults of all species. The crista
parotica is well developed in all species except T. degener, in which it is
short and unossified (Fig. 12); however, a bony articulation between the
squamosal and crista parotica is absent in all species. In T. degener, the otic
capsules are relatively large (Fig. 12).

The frontoparietals are fused posteriorly in adults of all species (unfused
in some individuals of Telmatobius truebae) and the frontoparietal fontane-
ls always is at least partially exposed. The fenestra is most extensive in T.

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 9. Photographs of living Telmatobius. Upper left. Holotype of T.
atahualpai, subadult female, 43.2 mm SVL. KU 212585. Upper right. T. brevipes,
adult female, 58.0 mm SVL. KU 212413. Middle left. T. ignavus, juvenile, 39.3
mm SVL, KU 181438. Middle right. T. latirostris, adult female, 83.6 mm SVL,
KU 212448. Lower left. Holotype of T. necopinus, adult male, 7 1 .4 mm SVL, KU
212482. Lower right. Paratopotype of T. truebae, adult male, 62.3 mm SVL. KU
212465.

degener, T. necopinus, and T. truebae. The nasals are small, slender, and
clearly separated medially; in T. necopinus the nasals are more slender and
curved anteriorly (at midlength) than in the other species (Fig. 1 1C). Long,
slim, palatines (neopalatines of Trueb, 1993) are present in all species
(absent on one side of one individual of T. degener); the palatines are small
and do not reach the maxillae in T. degener (Fig. 12B). The lateral end of
the palatine is distinctly wider than the medial end in T. brevipes (Fig.
10B), whereas the medial end is slightly expanded in T. thompsoni (Fig.

S YSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 1 7

10F). The cultriform process of the parasphenoid extends anteriorly to (or
almost to) the level of the palatines. The parasphenoid alae are nearly
perpendicular to the longitudinal axis of the skull but are slanted slightly
posterolaterally. The anterior ramus of the pterygoid extends to the level of
the palatines, and overlaps the palatine dorsally. This overlap is most well
developed in T. brevipes (Fig. 10A). The anterior ramus of the pterygoid
bears a distinct dorsal process (presumably for attachment of a ligament
from the zygomatic ramus of the squamosal) in T. thompsoni and one of the
two skeletons of T. ignavus. The median ramus of the pterygoid extends to
the prootic, but does not abut the parasphenoid ala. In T. necopinus, the
medial ramus of the pterygoid is especially robust and its articulation with
the prootic is expanded (Fig. 1 ID).

The quadratojugal is present, well developed, and always articulates
with the maxilla. The otic ramus of the squamosal is small, and an otic plate
is absent. The zygomatic ramus is well developed and deflected medially.
The zygomatic ramus is relatively long in Telmatobius brevipes, T. ignavus,
T. necopinus, and T. thompsoni, moderately short in T. latirostris and T.
truebae, and very short (almost equal in length to the otic ramus) in T.
degener. Telmatobius degener is unique among the species examined in
that the jaw articulation is at the level of the crista parotica (Fig. 12),
whereas in the other species, the jaw articulation is just posterior to the
level of the crista parotica. This difference seems to reflect the relatively
short maxillary arch and the more nearly vertical orientation of the ventral
ramus of the squamosal in T. degener. Telmatobius latirostris also has a jaw
articulation that is unusually anterior in position (but not as far as in T.
degener): this may be correlated with the wider skull of this species. The
maxillary arcade is dentate in all species except T. degener, which lacks
teeth on the premaxillae (Fig. 12B). Although loss of teeth on the maxillary
arcade has occured in numerous anuran lineages (see Duellman and Trueb,
1986; Trueb, 1973; Trueb, 1993), T. degener seems to be the only anuran
that lacks premaxillary teeth but retains teeth on the maxillae. The alary
processes of the premaxillae are posterodorsally oriented. Typically, the
vomers are dentate in all species, but vomerine teeth are absent on one or
both sides of some individuals of T. brevipes, T. degener, and T. truebae.
The maxillary, premaxillary, and vomerine teeth are elongate, recurved,
and fanglike. Little variation was found in mandibular elements, but in T.
necopinus the posterior margin of the coronoid process of the
angulosplenial is concave and recurved; it is convex in the other species.

The columella is present in all species, but its size and shape differ
strikingly between and within species and individuals. In most anurans. the
columella consists of a long, rodlike stylus laterally and an expanded
footplate medially (abutting the operculum). In Telmatobius brevipes, this
morphology was observed in one of the specimens (KU 180489; Fig. 10B),

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 10. Skulls of Telmatobius in dorsal ( A, C, E, G) and ventral (B, D, F, H)
view.^A-B. T. brevipes, KU 181489. C-D. T. ignavus, KU 181440. E-F. T.
thompsoni, KU 218468. Septomaxillae and cartilage not shown. Scales = 5 mm.

whereas in another (KU 212447), the stylus is reduced laterally and very
thin. In both skeletons of T. degener, the columella is reduced to a small,
nearly spherical nubbin consisting only of the footplate (Fig. 12B). The
columella has the typical shape in the two skeletons of T. ignavus, although
there is a marked bilateral variation in the thickness of the stylus in one
(KU 181441). In the single specimen of T. latirostris, the columella is
normal on the right side (although unusually thin) and is reduced to a
minute, nearly spherical remnant on the left (Fig. 1 IB). In the exemplar of
T. necopinus (Fig. 1 ID), the columella on the left side is slightly shortened

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

19

Fig. 1 1. Skulls of Telmatobius in dorsal view (A, C, E) and ventral view (B, D,
F). A-B. T. latirostris, KU 2 1 245 1 . C-D. T. truebae, KU 2 1 2464. E-F. T. necopinus,

KU 212184. Septomaxillae and cartilage not shown. Scales = 5 mm.

and dumbell shaped, expanded both laterally and medially. The stylus is
absent on the right side in this specimen. The columella is of the usual
anuran type on the left side of the adult T. thompsoni skeleton (KU 218468;
Fig. 10F), on the right side the columella apparently was lost in prepara-
tion. The columella morphology is normal on both sides of four of the six
skeletons of T. truebae observed. In KU 212464 (Fig. 1 IF), the columella
is a spherical nubbin (stylus lost) on the right side; on the left side the stylus
is shortened but expanded laterally, and in KU 212478, the columella is
shortened on one side. The operculum seems to be present but cartilaginous
(sometimes poorly stained) in all species. A tympanic annulus is absent in

20

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 12. Skull of Telmatobius degener, KU 218466, in dorsal (A) and ventral
(B) view. Septomaxillae and cartilage not shown. Scale = 5 mm.

T. degener. reduced to one or two sliverlike remnants in T. necopinus, and
present and complete in the other species. The pars externa plectri seems to
be absent in T. latirostris (although the tympanic annulus is present) and T.
degener, but it is present in the other species.

Despite the limited sample sizes, the intraspecific and intraindividual
variation in columella morphology observed in these seven species of
Telmatobius is remarkable. Similarly, Trueb (1979) reported considerable
intraindividual variation in presence of the tympannic annulus in 32 speci-
mens of T. niger (but she did not determine the condition of the columella).
I am unaware of reports of similar levels of intraspecific and bilateral
variability in plectra! morphology in other anurans. Furthermore, there are
few (if any) other anuran groups in which the columella is altered nega-
tively in shape (Trueb, 1985), although it is lost in numerous anuran
lineages (including within Telmatobius). The extent and source of this
apparently unique variation are clearly deserving of further study.

The hyoids of four species {Telmatobius brevipes, T. latirostris. T.
necopinus. T. truebae) were examined. The hyale bears short, anterior
processes in T. brevipes. T. latirostris. and T. necopinus (Fig. 13B). These
processes are absent ( 1 specimen) or are medially oriented (2 specimens) in
T. truebae (Fig. 13A). An anterolateral hyoid process is present in all four
species and usually is distally expanded. Posterolateral processes also are
present and are variably expanded distally. The posterolateral processes are
short in T. latirostris (Fig. 13B). The posteromedial hyoid processes are
long and ossified. In T. latirostris. ossification of these processes is more
advanced on the hyoid plate than in the other species (Fig. 13B) and is
accompanied by some calcification of the hyoid plate.

There are eight procoelous, mostly nonimbricate presacral vertebrae.
Presacrals I and II are almost imbricate in some individuals (Fig. 14). The

SYSTEM ATICS OF THE FROG GENUS TELMATOBIUS IN PERU 2 1

Fig. 13. Hyoids of Telmatobius in ventral view. A. T. truebae, KU 212480. B.
T. latirostris, KU 212484. Stippled overlay indicates cartilage. Scales = 5 mm.

sacrum and Presacral VIII are fused in the single skeletal preparation of
Telmatobius necopinus. The transverse processes of the presacral vertebrae
are relatively short, thick, and uniform in width in T. degener (Fig. 14A). In
T. latirostris (Fig. 13E), the transverse processes of Presacral II are dis-
tinctly curved anteriorly. The atlantal cotyles are relatively narrowly sepa-
rated (Type II of Lynch, 1971 ). The sacral diapophyses are slightly dilated,
and have a marked dorsolateral orientation in T. thompsoni. The urostyle is
bicondylar, has a dorsal crest, and usually lacks a lateral flange (present in
an apparently aberrant T. ignavus and one side of the only skeleton of T.
necopinus).

The pectoral girdles are fully arciferal. with a cartilaginous sternum that
is distinctly notched posteriorly. The omosternum is elongate and cartilagi-
nous. The scapula is bicapitate, and the scapula, coracoid and clavicle are
robust and well ossified. The cleithrum is cleft distally into two approxi-
mately equal parts. Typically the omosternum, the cartilage of the scapula
and suprascapula, and the procoracoid and epicoracoid cartilages are calci-
fied to some degree. The lateral end of the clavicle usually extends to or
close to the anterior tip of the pars acromialis of the scapula, but the extent
of the clavicle on the scapula is distinctly reduced in T. necopinus. Lynch
(1978:fig. 15) illustrated the pectoral girdle of T. brevipes.

The ilium is cylindrical and bears a moderately well developed crest and
dorsal protuberance in all species (Fig. 15). The pubis is ossified in adults
of all species (unossified in individuals of Telmatobius truebae and T.
degener of questionable maturity) and is fused indistinguishably with the
ilium and ischium; hence, the union of these three elements is indistin-
guishable in most adults. The acetabular fossa is fully ossified in the best
ossified specimen of T. degener (Fig. 15D).

22

A

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162
B ^ C D

Fig. 14. Vertebral columns of Telmatobius in dorsal view. A. T. degener, KU
218466. B. T. thompsoni, KU 218468. C. T. brevipes, KU 181489. D. T. ignavus,
KU 181440. E. T. latirostris, KU 212451. F. T. truebae, KU 181536. G. T.
necopinus, KU 212484. Scales = 5 mm.

The humerus bears distinct posterior and anteroproximal crests (crista
medialis and crista ventralis, respectively) in the species for which male
skeletons are available (Telmatobius degener, T. brevipes, T. latirostris, and
T. truebae). The crista ventralis is distinctly hypertrophied distally in T.
latirostris. A small crista ventralis is present on the humerus of females
(condition unknown for T. latirostris). The hand bears three to five small
prepollical elements (with considerable inter- and intraspecific variation);
the most distal one to four elements are cartilaginous. There are five other

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

23

Fig. 15. Lateral view of pelvic girdles of Telmatobius. A. T. brevipes, KU
181489. B. T. ignavus, KU 181440. C.T. thompsoni, KU 218468. D. T. degener, KU
218466. E. T. latirostris, KU 212451. F. T. necopinus, KU 212484. G. T. truebae,
KU 212464. Scales = 5 mm.

carpal elements: the ulnare, radiale, centrale, Carpal 4, and an element
consisting of the fused Carpals 2- 4. A large sesamoid bone is present
dorsal to the radiale in all species. The phalangeal formula of the hand is 2-
2-3-3. A bony nuptial tuberosity is present on the medial surface of Metac-
arpal I in known males; this tuberosity is poorly developed in T. brevipes.
The foot bears four ossified tarsal elements; an ossified prehallical element
(articulating with Metatarsal I), a centrale between the prehallux and
fibulare. a small distal tarsal (proximal to Metatarsal II), and a larger distal
tarsal (proximal to Metatarsals II— IV). A variable number of small, incon-
spicuous, usually cartilaginous prehallical elements articulate with the
ossified proximal element. The terminal phalanges are knobbed and the
phalangeal formula for the foot is 2-3-3-4-3.

24

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Fig. 16. Lateral views of tadpoles of Telmatobius. A. T. brevipes, KU 181857
(Stage 38). B. T. ignavus, KU 181846 (Stage 34). C. T. truebae, KU 181855 (Stage
35). Scales = 10 mm.

Larvae

The tadpoles of three species {Telmatobius brevipes. T. ignavus, and T.
truebae) were examined and illustrated (Fig. 16); the following general
description applies to all three. The body is ovoid in dorsal and lateral
aspects, and is generally more streamlined and less globular in T. brevipes
than in T. ignavus and T. truebae. The snout is bluntly rounded in lateral
and dorsal profiles. The eyes are dorsolateral and located closer to the
laterally oriented nostrils than to each other. The spiracle is lateral and
sinistral, and the vent is median. The skin of tadpoles of T. ignavus and T.
truebae often has a loose, "baggy" appearance. The caudal musculature is
robust and well developed. The caudal fins do not extend onto the body and
the terminus of the tail is blunt. The mouth is anteroventral and the oral disc
ovoid. The lateral margin of the oral disc is transangular (vs. constricted;
sensu Lavilla, 1988b). Scattered intramarginal papillae are present in the

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 25

lateral angular area. Tadpoles of T. ignavus have a more extensive prolif-
eration of intramarginal papillae laterally than do T. truebae and most T.
brevipes; intramarginal mental papillae are otherwise absent. The marginal
labial papillae are broadly interrupted anteriorly. The tooth-row formula is
2/3 and the innermost dorsal and ventral tooth rows are cleft medially. The
jaw sheaths are keratinized, wider than high, and the dorsal sheath is wider
than the ventral sheath. The inner margins of these sheaths are finely
serrated and gently arced. Lavilla (1988b;fig. 1A) illustrated an oral disc
typical of Telmatobius and the species described here.

In preservative, the dorsum is light brown. Small, round, dark spots are
present on the dorsum of most Telmatobius brevipes and some T. ignavus.
The caudal musculature is marbled with brown pigment and the tail fins are
transparent and flecked with brown. The caudal musculature of T. brevipes
is usually more evenly pigmented than in the other species and the caudal
fins usually bear more dark pigment. In life, T. brevipes is dull olive-tan or
green dorsally with or without dark brown spots or greenish flecks (field
notes, W. E. Duellman, for KU 181842, 181844, 181857, and 212503). The
belly is gray, the tail tan, and the iris dull bronze or brown. The body and
caudal musculature of T. ignavus are olive-gray; the caudal fins are translu-
cent tan and the iris dull bronze (field notes, W. E. Duellman, for KU
181845). In living T. truebae tadpoles, the body and caudal musculature are
mottled olive-brown with greenish olive and black flecks. The venter is
gray and is lavender anteriorly. The fins are translucent tan with dark brown
flecks and the iris is pale yellowish bronze (field notes, W. E. Duellman, for
KU 181855).

MORPHOMETRICS

Descriptive statistics for the 12 measurements recorded for adult speci-
mens of Telmatobius brevipes, T. degener, T. ignavus. T. latirostris. T.
necopinus, and T. truebae are given in Table 2; data for T. colanensis and T.
thompsoni are given in the species descriptions. Aside from differences in
size, there are few obvious morphometric differences between the putative
species. In order to explore further the morphometric data and to determine
if they confirm the groupings based on other kinds of data, I performed a
series of multivariate analyses. For these analyses, the species were split
into two groups (based on their similarity and geographic proximity) to
facilitate differentiation of taxa and interpretation of results. The first group
consisted of T. colanensis, T. latirostris. T necopinus. and T. truebae. and
the second of T. brevipes, T. degener, T. ignavus, and T. thompsoni.
Telmatobius atahualpai was excluded from the morphometric analyses,
because thus far, it is known only from sexually immature specimens.

Two sets analyses were performed on each group of species; one ex-
plored the morphometric variation independent of taxonomic assignment

26

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Ob

o
o

<L>

a

on
+1

P-H

E-^

be y.

t- -2

>
o

■a

o

O P

CN

<u gj

e-

-
>

6c

"a

l>5

e-;

00
00

On

o
■sj-

n

ri

■st-

NO

CN
CM

o

en
CN

o

U-N

oo

in

NO

o

oo

NO

NO

g

en
II

-3-
+1

OO

1

n

+1

en

1

n

+1

00

en

1

+i

in

ri

1

+1

o

en

1

O
+1

NO

1

O
+1

1

c
+1

NO

1

o
+1

1

u

c

vO

cn

oc

-r

—

ON.

NO

cn

nC

oo

ON

NO

NO

IT;

en

r-

^H

C<~1

a

PL

r--

ON

NO

cn

ON

ri

in

en

CN

en

en

Cn|

ri
ri

oo

r-

CN

in

l/~,

IT)

in

in

-vt

r~

NO

ZJl

o
o

ON

—

r-

r)

—

ri

oc

"sT

00

r-

ri

o

■*

■*

, .

■vt

r-~

■— I

13

ri
II

r--

+1

"St"

1

en

+ 1

CN

en

1

+1

CO

en

1

cn
+1

1

ri

+1

ON

CN

1

+1

NO
1

o
+1

m

1

o
+1

in

1

o
+l

NO

1

s

^

o

o

ON

r-

o

r--

o

oo

O

cn

o

00

o

r-

ON

OO

NO

in

o

in

ON

t-

CN

o

(N

ON

oc

NO

m

en

in

-t

-t

<tf

NO

NO

O vO

CS II

1)

C/3

<U en

3 ii

o
—

X!
U

r-~ no ri ri en en ri — ri n

q no

en on

+1 I

r- in

en — '

OC
nC

en
r- O m

en

oo

r i
I

en

en m, CN oo

NO

ri en — •<*•

+1 I +1 I +1 I

— oc ri m in O

C h ei C \tVi

nO
en

d

+1

oc —

CN
I

O
+1

en

CN

d

+1

I

en oo on r-~ «n — '
cn — ri CN ri ri

o
-r

d
+l

oc r-

— on oc r~-

r-- o

O ri x

en 00 — en r) vC O en — nC

+1 I +1

en 3\ exj

ri ri — —

I +1 I +1 I +1 I +1

— ' O — no oc en en r) —

CN

I

O en

+1

CN

o
+1

C On 00

in cn cn C -rf ri ri — in ^t CN (N en en cn cn
■st ■** ri r-i ri ri — — — — i

v~ cn r~

oc

— NO o

^t 00 en

in — ri d ri en — on cn in d -vt

r- en en — cn

+1 I +1 I +1 I +1 I +1 I +1 I

oc o —

CN r~ en ri ri

+1 I +1
r- so — r-~ en o

— iy-, ir, ri r-~ ^t r- in — r~ en en en en Tt-
nO >n CN CN CN CN — i — ■ ei —

NO

r- m,

-r' ^

NO

+1 I

nO oc

NO

ri oc oc

— — no r-

oo
en

— 00

cn en en en

ON

I

o

r- </~, on m r i

>t oi h — tj- cn m — — — i in

^t

+1

en

- ^t v~i ~r r- -|- q
d ■*' d -t d no

l +1 I
in c", en
en in "st

ri

I +1
oc in

en

I

+1

I

r- -t —

— en
e^i

+1 I

nO oo

r I
m

in n on -st — in m oo in
■st cn — ri ri — — — —

-r
O en oo

d ^' d Nt d Nt d iri

+i i +l i +i I +i i

— -t en oo ri no —■ -vt
en ri en ri en ri in -f

-J
>

s 8

Z
J

i 2 S

O W

z ^

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

27

—

_0J en

~ n

c

c
U

•_

03

u

CN

o

<u

c3
—

_C

cd

cfl

JS

H

U

r<", —

— CN CO

— ri O
ri

+1 I

C*"l i/~,

+1

0\ CN

I +1 I

oc cn oc

C") — O

— ri ri

X X r^, — C 3~

r-- cn

ri O O

oc

•^r- r- o

— oc r-

Q

-J
CQ
X
H

Q
a:

£

IT)

-1-

DC

~t

r I

vC

+1

1

+ 1

1

+1

1

oc

o

—

OC

t~»

oc

05

ju

^C

' —

<N

—

oc

—

~-

cd

E

II

+i

1

+1

1

+1

1

*—

CT)

—

3

o

—

—

£

U-

<3

o>

c~~

^c

—

^^

c

—

— •

—

W

^3

CO

c;

,__,

OC

■*r

ri

cr-

CO

b-H

C/3
<L>

co

O

<N

r-

<=

CO

r3

II

+1

1

+1

1

+1

1

s

^Z

—

-r

o

</~,

co

«n

—

3

r-

sO

ri

—

o

,_H

r-

(N

,

oc

~

—

—

—

l-H

E

ii

+ 1

1

+1

1

+1

1

gj

sC

ro

ro

</~.

O

—

tu

■*t

CO

3

OC

vC

t

Q.

—

—

—

—

—

4j

-C

oc

(N

—

hi

G^

CN

oc

-5-

O

B3

—

o

vC

o

o

P~

o

—

•^

—

CO

ii

+1

1

+1

1

+ 1

1

s

~

—

vC

o>

'/".

^C

o>

28

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

•—

to

>

c

03

c

•a

c
o
o
u

CO

Q

+l
c

03

*)
s

•a
c

o3

to
3

.J

c

kj -°

0~ 03

■-3 o3

£ °

e-;

£

>
l—

03

i

03

T3

03

.S ,o

c ~

c ..

S -

<3

jo E

n ;s

u ~

— . r~

£ .g

to
s
c

I

<4j

r~i

c«

0)

U.

O
ri

JIJ —
03 ||

C8

o _

73 II

E "

2 «

g II

3 M

15 ll

o

03
—
03

u

ON

o
+1

q
oo

oo

CO

o

On

r-

+1

en
en

I

+1

in

en

+1

O
en

sD

en

— en — ir, o ri — Oi C if; C it,

+1

o
I

ON

NO

fN fi en r-~ ri en

+1

— Tfrririr~OON\OfNrir-~

r-

ON

p

en

in — oc n o

nC en ci en en ri —

on no ^r

ri ri

-I- ~t

+1

00

in no
no it i in r~

o
•C

+1

— rj —

en
+1 I +1

NO

en

I

+1

IT)

o

r I

O —
— ' \q

+1 I

■5t oo

<n oo n
■<* o

o
+1

p
in

+1
p
in

c nC

i +l i

CI On —
■5t in un

en rj en in
d m' o vd

I +1 I +1 I +1

— 1- ^t h n x it, x "■, oc no oc no on r-- ri

no oc o in — h a> it, Tf —
no m en rN en ci — — n n

't Wi ^ f) IT, IT,

ON_

ri
r-

oc

'nO

en

nC —
en ri

oc

NO
IT,

m —
r-

+1 I

r- —
in r- r~

C; rr, Q

en

+1

oc
r- ri —

+1 I

r i — m qn

en

I

o
+1

I

On ri ri

d

— — ri

oc
ri

in

d in

n
i/"i

if.

en

r-

nO

r) r- on r^

+1
in

+1

On [— rn CnJ Tt- ee, — — r-

vC o

X

nC

I +1

00 nO

I +1

+1 I

en en rn n ri n (N

ir. -I-

<n in

on no ^- ri

•t -t \d vd

— r~

ci

nC — o

+1
t|-

en
x

I

x c n n '

en

I

o
in rs <n oc

-t

+i

ri

I

en rr, vC nC
00 00 rr, r*-,

O

I +1

r-- rj

on ^t' "st r-i r4

en cn) ri en en

O en

rn
+1 I
oc in

+1

-t

I

o

e^, On en
d in

I +1

o
+l

r^

-r — onocnO'+oono
-r -I- in in in in \d no

en
If,

en en
00

+1

ON

p oc in —

— iri —

I +1 I +1

t~~ 0C On CN

nC

On

ee,
I

"3-

I

cni cm rM
d ^t

en

I

d ^6

00 o

d no

o
in -v^

d r^

+1 I +1

+11+11

e-OXr-(N,tv, O^i-

oc vC rr-, — ■ \c in e*-, rn O On rf
r~ r-~ en en en en r i r ) en rN

m in in <n ^t -o no

>

CO

_

r-

B 8

Z

^ o z

on

O UJ

z >

UJ UJ

SYSTEMATICS OF THE FROG GENUS TELMATOBWS IN PERU

29

&3

s

00

~3
-

C

o
U

X
CN

x

69

1/3

E

0

—

CN

ON

— 1

SO

oc

Tt"

m

II

0
+1

O
CN

1

+ 1

ir,
1

+1

(-1

r 1
1

=:

^c

ir,

—

On

—

00

u.

oc

r^

r<-,

, —

O

CN

00

2 11

PL,

3 II

o
S3

X

U

vO —

— OC

+1

r 1

+1 I
vO —
ri —

O
CN

o

vO

cn

O ON O
+ 1 I +1

O
+1

VD l/~, Ol ON r<";

on on cn ri

r-

O

+1

©
CN
CM

I

■C

— © tT -*
CN ri — —

(N

+1

CN

+1 I

ri i/',

-f —

— 01

ri

I

O

ri n

X

CN

oc

CN

it

-r

r<",

O
+1

n
1

+1

in

+1

c 1
1

ON

r-

CN

©

CN

—

ri ri m in c", CO
CN CN — — CN CN

oc

cn cn

— ci

CM

+1 I

O m

CN O

CN CN

-J

_J

Q

pq

-J

Z

X

Q

I

H

oi

30 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

of individuals (thereby allowing for unbiased discovery of groups), whereas
the other maximized difference between groups (species) delimited a priori.
First, the data were subjected to Principal Components Analysis (PCA).
Principal components are linear composite variables, uncorrelated with
each other, that explain the maximum amount of variation observed with-
out reference to a priori assessments of group membership.

For morphometric data, the first principal component generally de-
scribes differences in size, but there may be size effects in subsequent
principal components (Humphries et al., 1981). Several techniques have
been developed to deal with the confounding of shape and size. Rohlf and
Bookstein (1987) advocated Bumaby's (1966) method as superior for size
correction alone, but others (e.g., Humphries et al., 1981; Bookstein et al.,
1985) argued that results obtained using Burnaby's technique are difficult
to interpret and that "shearing" (Humphries et al., 1981) is therefore
preferable. These techniques are described, discussed, and compared by
Rohlf and Bookstein (1987). Both techniques were used in this study, but
because they yielded nearly identical results in terms of separation of
species, I have presented and interpreted the results of the shearing analysis
only. Shearing was implemented using SAS code written and provided by
Les Marcus. Burnaby's method was performed using a slightly modified
version of the SAS code for shearing, as described by Rohlf and Bookstein
(1987). Burnaby's technique was used to generate an adjusted data set with
between-group size differences effectively held constant; the adjusted data
were then subjected to PCA and canonical discriminant analysis. For both
techniques, principal components were derived from variance-covariance
matrices (as opposed to correlation matrices) and log|()-transformed data,
following the recommendations of Bookstein et al. ( 1985).

Scores of individual specimens of the large-spined and eastern species
(Telmatobius colanensis, T. latirostris, T. necopinus, and T. truebae) on the
size component, and the sheared second and third principal components
were plotted successively against each other. The least overlap of principal
component scores among species is seen when the second and third princi-
pal components are compared (Fig. 17, top). This scatterplot is almost
identical to that for PC I versus PC II for the Burnaby adjusted data (not
shown). Telmatobius colanensis and T. latirostris are separated from T.
necopinus and T. truebae along Sheared Principal Component II (SPC II).
The correlation of individual variables with the principal components
reveals the relative importance of these variables in accounting for separa-
tion along the principal components. These correlations are given by the
loadings in Table 3. Interorbital distance, eye diameter, and internarial
distance have the highest loadings for SPC II. Telmatobius colanensis and
T. latirostris are well differentiated along SPC III, which has high loadings
for interorbital distance, eye diameter, and thumb length. Telmatobius
necopinus and T. truebae are only poorly differentiated.

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

31

<

o

5

4

3

2

1

0

-1

-2

-3

V

v v
• • • •

0 2

CAN

8

A colanensis
+ latirostris

v necopmus
• truebae

Fig. 17. Plot of sheared principal component scores (top) and canonical
discriminant scores (bottom) for Telmatobius colanensis. T. latirostris, T. necopinus,
and T. truebae. SPC = sheared principal component. CAN = canonical discriminant
axis. The single individual of T. colanensis could not be included as a group in the
canonical discriminant analysis.

Canonical discriminant analysis creates uncorrelated linear composite
variables (discriminant functions) that maximize differences between
groups, where group membership of individuals is specified a priori. Analy-
ses were performed with data transformed using Burnaby's technique

0.059

-0.070

0.329

-0.239

0.592

0.509

-0.088

0.447

0. 1 75

0.231

0.229

-0.019

-0.525

0.525

0.476

0.191

-0.362

0.240

0.536

0.151

-0.151

0.029

0.569

-0. 1 1 5

-0.118

0.017

0.430

-0.178

-0.092

-0.199

0.496

-0.074

0.068

-0.402

0.321

-0.089

0.023

-0.281

0.340

0.026

-0.140

-0. 1 78

0.394

0.008

-0.190

-0. 1 80

0.480

-0.021

32 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Table 3. Variable loadings for sheared principal components and discriminant
function-variable correlations (pooled within-class canonical structure) from prin-
cipal components analysis and canonical discriminant analysis of Telmatobius
latirostris, T. necopinus, and T. truebae.

Sheared Sheared Canonical Canonical

Variable PC II PC III Axis I Axis II

SVL

IOD

ENOS

EYE

INAD

HWID

HLEN

HNDL

THBL

RDL

TIBL

FOOT

(between-group size held constant) and with the data logH)-transformed
only. The degree of separation of taxa is nearly identical in both analyses,
although the relative contribution of different variables to the discrimina-
tion can differ. I have chosen to present and interpret the results using the
log-transformed data. Canonical discriminant analysis requires that each
group contain more than one individual. Therefore, Telmatobius colanensis
could not be included, but the morphometric distinctness of this species is
evident from the PCA. Unlike the PCA. canonical discriminant analysis
showed T. necopinus and T. truebae to be morphometrically distinct. This
is evident in the plot of scores for Canonical Axes I and II (Fig. 17, bottom)
for the log-transformed data (which is identical to the plot for the Burnaby
adjusted data). These two species show separation along both canonical
axes. Based on their large values for pooled within-group canonical struc-
ture (the correlation between the canonical discriminant function and the
original variables, equivalent to loadings in PCA; Table 3), differences in
head width, internarial distance, and hand length are most important to
separation along Canonical Axis I (CAN I). Telmatobius latirostris and T.
truebae are separated along CAN II, which is influenced most heavily by
interorbital distance. Note that some variables have different rankings in
separating the same species in PCA versus canonical discriminant analysis;
presumably this reflects the different contributions of the variables to
overall variance versus between-group variance.

PCA of the western and small-spined species (Telmatobius brevipes, T
dc»ener, T. ignavus, and T. thompsoni) is largely unsuccessful in differen-

SYSTEMATICS OF THE FROG GENUS TELMATOB1US IN PERU

33

tiating the species. The plot of sheared PC II versus sheared PC III illus-
trates the considerable overlap among the scores (Fig. 18, top); plots of size
versus the sheared components (not shown) display similar degrees of
overlap. The plot of PC I versus PC II for the Burnaby-adjusted data is
virtually identical to the plots of the sheared components.

O
D_
CO

go

4-

A A

3-

+

+

A -

2-

+

*

▲

A

1-

0-
1 ■

+

+

+

+

• \ m

• ••••

V

■

2-

V

■

3-

■ ■ i ■ i

-8

-4-2 0 2

CAN I

8

• brevipes v thompsoni
+ degener a ignavus

Fig. 18. Plot of sheared principal component scores (top) and canonical
discriminant scores (bottom) for Telmatobius brevipes, T. degener, T. ignavus, and
T. thompsoni. SPC = sheared principal component. CAN = canonical discriminant
axis.

34 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

In contrast, the plot of canonical discriminant scores for the log-trans-
formed data (Fig. 18, bottom) shows little overlap among the four species.
Using Burnaby's technique, the only difference is that there is slightly
greater overlap between Telmatobius brevipes and T. thompsoni. Most of
the discrimination occurs along CAN I. Based on the values for canonical
structure (pooled within-group, Table 4), the variables that contribute most
heavily to discrimination on this axis are hand length, radia length, and
head length. Discrimination along CAN II is influenced most strongly by
internarial distance, as well as eye diameter and hand length. The large
separation between the scores for the two individuals of T. thompsoni
apparent on CAN II most likely is the result of sexual dimorphism.

In summary, the morphometric data can distinguish among Telmatobius
species in northern Peru when the groups are determined a priori and
differences between them are maximized (using canonical discriminant
analysis). Without delimiting groups beforehand (using PCA), little mor-
phometric differentiation is apparent except among some of the large-
spined species (T. colanensis, T. latirostris, T. necopinus, T truebae).

Allozymes

Nineteen variable loci were scored (Table 5). These data first were
coded qualitatively for phylogenetic analysis (Table 6). Parsimony analysis
(using PAUP) yielded two shortest trees (Fig. 19A.B), each with a length of
34 steps. Differences between these trees involve rearrangements of the

Table 4. Variable loadings for sheared principal components and canonical dis-
criminant function-variable correlations (pooled within-class canonical structure)
from principal components analysis and canonical discriminant analysis of
Telmatobius brevipes. T. degener, T. ignavus, and T. thompsoni.

Sheared Sheared Canonical Canonical

Variable PC II PC III Axis I Axis II

SVL

IOD

ENOS

EYE

INAD

HWID

HLEN

HNDL

THBL

RDL

TIBL

FOOT

0.035

0.019

0.440

0.078

0.756

-0.261

0.559

0.192

0.529

- 0.353

0.414

0.162

0.229

0.367

0.523

0.264

0.038

0.661

0.466

0.560

0.088

0.117

0.576

0.157

0.022

-0.016

0.615

0.169

0.029

0.036

0.632

0.255

0.075

- 0.406

0.536

0.163

0.007

- 0.034

0.630

0.107

0.120

0.180

0.395

0.229

0.242

0.143

0.445

0.193

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 35

three populations of Telmatobius truebae. The optimality (lengths) of these
two shortest trees and several near-shortest trees were compared using
FREQPARS; there are two trees of 34 steps, four of 34.5 steps, and 1 7 of 35
steps. Differences among these trees involve rearrangements among the
populations of T. brevipes and T. truebae, and placement of T. latirostris
with one or more populations of T. truebae. Allelic frequencies were
calculated from the table of genotypic frequencies (Table 4) and are not
presented separately. FREQPARS ranked two of the 23 input trees as
shortest, each with a length of 58.08 (Fig. 19A,C). One topology has a
length of 34 steps using qualitative character coding (Fig. 19A), whereas
the other is 34.5 steps long (Fig. 19C). Differences between these trees
involve different placements of populations of T. brevipes. In summary,
one tree is shortest by both qualitative and quantitative analyses of the
electrophoretic data (Fig. 19A). The support for this tree is detailed in
Appendix III. However, this particular "phytogeny" of populations within
T. brevipes and T. truebae is chosen somewhat arbitrarily for description; if
these populations are interbreeding (reticulating) it may be inappropriate to
try to recover (or impose) hierarchical relationships among them.

Several of the morphological characters discussed in previous sections
also could support some stems of this tree. Telmatobius brevipes (Stem 3)
has minute nuptial spines and an extensive nuptial pad that conceals the
inner palmar tubercle in males. Both T. latirostris and T. truebae (Stem 4)
have a relatively short zygomatic ramus of the squamosal and usually have
tuberculate dorsums and bright mottling on the underside of the limbs,
although T. necopinus and T. truebae tend to have smaller outer metatarsal
tubercles. Telmatobius truebae (Stem 5) is characterized by a poorly ossi-
fied sphenethmoid and processes on the hyale that are medially rather than
anteriorly oriented (or are absent). These eight characters were not in-
cluded in the parsimony analyses because data are unavailable for popula-
tions that were sampled electrophoretically but not osteologically (e.g.,
brevipes- 1 . brevipes-2, truebae-I ) or from which male specimens were not
collected (e.g., truebae- J): FREQPARS does not work with missing data.

The results of the parsimony analyses of the electrophoretic data show
that the populations of Telmatobius brevipes and T. truebae cluster together
in the unrooted tree. Furthermore, all four species examined appear to have
at least one "fixed" and "unique" electromorph, although these loci may
actually be more variable (if more individuals were sampled) and/or the
alleles may diagnose more inclusive groups (if additional species were
included).

Population 3 of Telmatobius brevipes also seems to have two unique,
fixed alleles (Ldh-lee, Pep-B-ldd). These apparently fixed alleles could
suggest that this population is a reproductively isolated lineage that re-
quires recognition as a separate species. However, the sample sizes used in

36

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

fc,

-3

-o

<4j

>

"■*

"O

^

^3

on

<4H

o

O

C/J

S-H

-^

CU

X>

S

3

ii

(1;

p

*—

cj

_2

T3
3

^

^i

cu

C/J

^

<u

S)

X3

-c

3

b-H

•—

II

&,

3

^)

03

-c

cu

X)

<o

E

3

z

-♦—

a

c/i

3

£

^O

<^

03

a

3

c/j

Q-

n

O

o

D-

*—*

T)

a

CD

3

CD

u.

O

u

a,

-i

CU

c

3

<j—

c

O

1—

&

.Si

c>

o

03

CJ

03

O

X)

Ih

03

o

s-

"1

>

(1)

O

^

^«

CU

J.J

eu

03

oo

t/3

CU

<u

o

a

c

-c.

cu

!»)

3

cj

f-;

u.

■4— •

o

3

o

n

OX)

D.

>1

w

T3

cu

o

^

C

u

cu

(U

a

3

in

CJ

03

_aj

Co

CU
X

03

3

a

3-

o
a.

s

CN

-1-

5

-5

CN

i

«^j

3

cj
O

J

CN m

cn m in m

in -h ^t in

- 't it, in

Xj

S X T3

o

cj XI X)

03 X

CJ

03

Xj X Xj

cj

X

cS X "O

Kj

CJ X X>

03 X

CJ

03

03 XJ X)

cj

(N CN

CO CN

03 03
03 03

X ^
XJ 03

CN

X) o3

X 03

CN (N

Xj 03
Xj 03

CN

CN

CN CN

CN

CN — — '

X)

03

T3

o

Xj Xj

03

Cj

03

Xj

Xj

O "O

Xj

03

■a

cj

Xj Xj

03

CJ

03

Xj

Xj

cj cj

Xj

03

-tu-

o

Xj Xj

03

CJ

03

Xj

Xj

cj

Xj

03

ts

03

Xj Xj

03

o

03

Xj

Xj

CJ

en CN — ■ cn — — cn m en

-a

T3

o

CJ

03 Xj
03 Xj

03 cj "O 03 "O Xj -3
03 O "O 03 "O 03 Xj

cj O
O O

en

CN

ri

ri

ri ci

CN

m

cn

CN

cn ri

Xj

X

03

03

T3
T3

o

CJ

03 "O
03 X!

03

03

u

CJ

CJ
CJ

Xj
Xj

X o
Xj cj

CN CN

CN ^ — CN

oj

CN i— i i— i

03 -3 Xj
03 Xj Xj

03 cj
03 cj

Xj 03 Cj Xj
Xj 03 cj Xj

Xj
Xj

03 Xj Xj
03 03 Xj

CJ
CD

Xj
X)

03 cj
03 cj

Xj
Xj

Xj
Xj

Xj
Xj

03 Xj

03 03

— — CN

CN — — CN CN

CN

CU CD XJ

O CU X)

"1

CJ

cj Xj

CJ

Xj

Xj

03 -3

03

03

cj XJ

CJ

Xj

Xj

03 Xj

CN

CN

—

— CN

ri

CN

CN

—

-

CN

CN

^ ~

Xj

X

03

X

X

cj X
cj X

03
7i

o
u

X
X)

o

o

X
X

X
X

03
03

03 X
03 XI

X

"3

XJ

ri H

ri

■
X
XJ

— ri

03 *_,
T3 03

< <

X

-o

<

D.

m

o

& X)

Icdh
Ldh

X!

"3
-J

CU CU

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

37

c
o

_73
3

a.
o
a.

i
3

Sj

=:

-5

i

:-

i — i T)- i/-, \r, rr. r) m, i/~, ro

u T3 s 73 X> X> cd 73 73

O O 73 73 73 X^ 73 73 73

— ' — ri tN (N (N ri (N

T3 "O 73 73 X 73 73 73

O "O 73 73 X 73 73 73

U 73 73 X 73 73 X

O 73 73 X> 73 73 C3

m

o
o

ri — <•«-,

73 X) X)
73 73 X

73
73

rn

ci

<N

o

X

x

X)

73

73

— CI

X

X

— — ri

73 X X
73 X X

X T3

X) T3

rn

73

73

(N

ri —

73
73

73
73

— — ri

— — «N

X X 73

73 X) 73

~ 73 X X
"O 73 73 X

73
73

CO

X

X

— — C)

73 X 73

73 X 73

73

X o
X u

73

—

ri

ir.

3
U

o

CQ

CQ

'5*

ex

PL,

a.

PL,

X
CO

T3

o

CO

38 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

the analysis are too small to assess whether or not these alleles are really
"fixed" or not and, therefore, whether or not there is gene flow between this
population and others. Without additional allozymic data and in the ab-
sence of morphological differences between this population and others, I
consider it premature to consider this population as a separate species.

The results of the allozyme analysis also demonstrate that, despite their
external similarity, Telmatobius necopinus and T. truebae are not sister
taxa. These two species occur in adjacent areas of the Cordillera Oriental
and there are no external morphological features that can distinguish them
consistently. (All external characters are variable in one or the other spe-
cies.) This illustrates the potential danger of assigning allopatric popula-
tions to a given species on the basis of geographic proximity and/or overall
external similarity.

Taxonomic Decisions

In this section. I justify and discuss the recognition of each of the nine
evolutionary species of Telmatobius in the Andes of northern Peru.
Telmatobius atahualpai, T. degener, and T. latirostris are known from
single populations and have unique character states that easily and consis-
tently separate them from their congeners. These are clearly distinct evolu-
tionary lineages, even though T. atahualpai is known to date only from
subadults.

Telmatobius ignavus is restricted to an isolated mountain range and
seems to represent a single interbreeding population. Although similar to T.
brevipes and T. thompsoni (but nevertheless morphologically distinct), the
geographical isolation of the Cordillera Huancabamba makes interaction
between T. ignavus and these species seem extremely unlikely. Telmatobius
thompsoni is known from a single locality and is similar to, but allopatric
with respect to and morphologically distinct from, T. brevipes (although
there are only two known adults of T. thompsoni). Therefore, T. thompsoni
is also hypothesized to be a distinct lineage, but at this point it is not certain
whether the geographical separation (and morphological distinctness) of T.
thompsoni and T. brevipes is real or the result of inadequate sampling.

Because of the allozyme data discussed previously and the absence of
obviously derived character states uniting the numerous populations, the
status of T. brevipes as a single lineage (evolutionary species) is somewhat
problematic. Note, however, that the geographical separation of T. brevipes
into northern and southern groups of populations (apparent in Fig. 1 ) is not
supported by allozymic or morphological data. Assignment of populations
to T. brevipes was based largely on their morphological and biochemical
similarity and geographical proximity to each other — certainly not ideal
criteria. However, if there are multiple lineages within T. brevipes, their
existence and limits are not obvious from data currently available.

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU

39

o

x:

.s

U",

-o
-J

£

d

x

rl

T3

,

P-

T3

r^

,-H

<u

G
n

■a

>>

0/J

■53

- — *

o

■s

n

oc

a.

>
id

ri

<u

!—

■

o.

£

X

x

H

Xi

73

T3

re

o

■ — -

C

c

^«

UN

re

c

r-

^

CJ

X

■ — -

£3

5

X

CO

p.
o

<n

X

oc

i)

no

- "

o

^_ '

O-

'a.
n

x
-a
c/3

re

<u

^^

ON

ir,

r-

' — '

• — '

■ '

T3

c

O

x

73

r-

C

re

P,
en

a

<

a,

ri

CQ

O
rf

CJ

-T-

, — ,

i

re

C

m

u.

X

q

■ — ■

aj

U

^

re

, — ,

« ,

<r

*0

K

;

o

<u

<N|

C

G

—

re

r*

re

-a

CQ

o

<

D.

aj

^M

Cu

y:

y.

* — "

, ,

CO

«^>

C/3

"+

c

re

C

<D

3

re

O

r.

re

x
cj

On

T3

xl
5

o

(xi

„

•,"■ '

C

rs)

c:

£

tU

5

oj

o

->

OJj

!

oo

O

n

- — -

— ^

(N

1

x

~

(N

—

Ofl

—

T3

,u

<U

re

X

3

S

0\

H

re

r-

C/2

4J

~

1

oc

c

iJ

NO

re
re

"re
o

ri

"*

Q

o

X
T)

m

^1

2

CJ

re

cu

X

O

x

c

- — -

re

,re

=

„

c

r-

e

ri

CJ

ON

r--

nC

ON

oc

nO

m

ri

re
re

re

re

re
re

re
re

re
re

re
re

X

re

re

re
re

>— i

—

— i

— i

—

—

cs

—

— '

X5

re

o

XJ

re

re
re

re

re

c3

re
re

r3

re

re

re

re
re

(N

rr-,

O

X

X

X

CJ
CJ

X

X

re
re

x>

X

re

re

C3

r3

re
re

r*",

CnJ

-t

n

—

n

—

*— ■

—

re

re

Xo
re

re

Xj

X
X

X

X
X

JZ

X
X)

1—1

r^i

—

ri

r<~,

ri

n~,

c<~,

<-«-,

~

"O

re

re

re

X

X

CJ

o

re
re

c^

~

— ri

— rr, -f — — —

x re -o re

3 CJ
CJ

re re re re

ri — ^t- — r«-, — — — —

XXXX cj CJ CJ-OT3
X re x X o o cj cj o

ri — ri ri en ft rr, -if -if

XXXX Cj Cj Cj "O Cj

rexrerecjooocj

i — c 1 •— ' ~- rr, rr, rr, ^J- f^l

re re re rex oxxx
re re re rex oxxx

— — — — r] rr, ri ri ri

XXXXXXXXX

xxxxx rexx re
ri ri ri ri ri — ri ri —

xxxx-j^3rerere
xxxxorererere

en en en en ^f cs — — i —

cj cj aj cj cj "O cj cj cj

U O II S JJ O J u H

CN (N u~i — ri en ri ri -^i

xxxxrerererere
xxxxrerererere

ri ri ri ri — — — — —

Cj cj O cj"OXXXX
cj re CJ CJ "p X X X X
ri — ri ri ~-i rr, rr, rr, rr,

rererererereXXX
rerererererexxx

— — — — — — ri (N tN

X X X X CJ CJ CJ CJ CJ

xxxxooreore
en en en en ri ri — ri —

U QJ ajX"0"0"0"0"0
X CJ d) S 13 "3 "3 T! T3

(N rr, -if — <fi <*~i >/~i >0 m

rerererererererex
.•* oQ re re re w ro ot re

_-j_:_:_;_;_;^^-;ri

XiXXXX ^xxx
xxxxx rexxx

cn ri oi rsi r-i — ri r^i r-i

— - r,

I I

T-

rN

^ ^j >^>

40

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

brevipes-2

brevipes-4
brevipes-3 \ jr j

brevipes-1

B

brevipes-2
brevipes-4 v \-
brevipes-3 v jr

brevipes-1

c

brevipes-4
brevipes-2
brevipes-3

brevipes-1

truebae- 1
truebae-2

truebae-3

< truebae-2
y truebae- 1

truebae-3

truebae- 1
truebae-2

truebae-3

Fig. 1 9. Unrooted trees of shortest length from parsimony analysis of allozyme
data using PAUP (A and B; length = 34 steps), and FREQPARS (A and C; length =
58.08). The support for the numbered stems of Tree A are detailed in Appendix III.
The localities of the numbered populations of Telmatobius brevipes and T. truebae
are given in the Materials and Methods; T. latirostris and T. necopinus are known
only from single localities.

SYSTEMATICS OF THE FROG GENUS TELMATOB/US IN PERU 41

The remaining three species (T. colanensis, T. necopinus, and T. truebae)
are morphologically distinct from all their congeners in northern Peru and
are geographically separated from them (except T. atahualpai) by the
valley of the Rio Maranon. Telmatobius truebae presently is known from at
least three allopatric populations, whereas T. colanensis and T. necopinus
are each known from single allopatric localities. The distinctness of T.
necopinus from T. truebae is clear from osteology and the allozyme data;
they are not sister taxa (see above). There are no apomorphies that unite the
populations of T. truebae to the exclusion of T. colanensis, but there is little
variation between the populations of T. truebae to suggest that they are
separate lineages. Judging by the distribution of cloud forest over 2000 m
in the Cordillera Central, it seems reasonable to expect future collecting to
show T. truebae to be distributed more continuously among the three
populations than currently is indicated.

Telmatobius colanensis is considered a distinct evolutionary species
because it is allopatric and morphologically diagnosable from the other
species. However, discrimination of T. colanensis from T. truebae and T.
necopinus is difficult, because only a single specimen of T. colanensis
exists, the external differences are subtle, and there are no data on its color
in life, osteology, larval morphology, or allozymes. Also, the geographical
isolation of T. colanensis from the other species may be more artifactual
than real, as the region is poorly known herpetologically. However, if T.
colanensis is not a distinct species, it is not clear with which species (T.
necopinus or T. truebae) it is conspecific. Recognition of this specimen as
representing a separate species is the best hypothesis given the data at hand.

Generic Assignment

In this section I discuss evidence for placement of the nine species
studied in Telmatobius. Unfortunately, no synapomorphies have been pro-
posed to support the monophyly of Telmatobius, Telmatobiini.
Telmatobinae, or Leptodactylidae. Lynch (1978) performed a cladistic
analysis of the so-called "lower telmatobines," a group of doubtful mono-
phyly that includes Telmatobius. Certain aspects of Lynch's analysis were
clearly problematic (e.g., no outgroups specified). However, for the pur-
poses of this discussion, I accept his hypothesis that Telmatobius and
Alsodes are sister taxa, and note that this arrangement is likely supported by
the presence of an enlarged crista medialis on the humerus in males
(Lynch's Character 23, not included in his analysis). The recently described
genus Telmalsodes Diaz 1989 (for A. montanus and A. pehuenche) is
indistinguishable from Alsodes and herein is considered a synonym. The
only character that purportedly differentiates the two genera is the presence
of plantar webbing in Telmalsodes (probably plesiomorphic at this level),
but plantar webbing also occurs in Alsodes gargola (Cei, 1980), A.

42 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

monticola (pers. obs.), and A. vanzolinii (pers. obs.). Alsodes (sensu lato) is
likely monophyletic based on the presence of paired patches of spines on
the chest in males (also in Insuetophrynus) and an enlarged crista lateralis
on the humerus in males (Lynch, 1978; pers. obs.).

Assuming that Alsodes and Telmatobius are sister taxa, character states
present in both Alsodes and other telmatobiine genera are likely to be
plesiomorphic in Telmatobius. I suggest that assignment of each of the
northern Peruvian species (except T. atahualpai) to Telmatobius is sup-
ported by two presumptive synapomorphies: ( 1 ) frontoparietals fused pos-
teriorly, and (2) nuptial excresence on Finger I only. These derived charac-
ter states are not universal within Telmatobius, but because they are shared
by the species of interest and the type species of the genus (T. peruvianus),
they are the ones most relevant to generic allocation. The frontoparietals
are fused posterior to the frontoparietal fontanelle in all species of
Telmatobius for which I have seen skeletal material (Appendix II), except
in T. cirrhacelis and T. vellardi. Frontoparietal fusion is intraspecifically
variable in T. truebae, Hylorina sylvatica, and Telmatobufo venustus, but
this variation does not endanger the polarity or applicability of this charac-
ter.

Except for Telmatobius arequipensis , all male Telmatobius that I have
examined (Appendix II) lack a patch of keratinized nuptial excresences on
Finger II. Note that I make a distinction between the patches of spines on
Fingers I and II and the small cornified denticles that are widespread on the
skin in many telmatobiines. In Alsodes, there are distinct nuptial
excresences on both Fingers I and II, but this character is variable among
other "lower telmatobiines," with nuptial excresences present on Finger II
in some Atelognathus (A. patagonicus, A. reverberii, A. salai),
Insuetophrynus, Batrachyla, Caudiverbera, and Eupsophus, but absent in
some Atelognathus (A. praebasalticus), Hylorina, Lynchophrys, and
Somuncuria (the last based on Cei. 1980). Batrachophrynus lack nuptial
excresences on both fingers. A rigorous, species-level phylogenetic analy-
sis is needed to address the monophyly of Telmatobius, especially the
possible relationship of certain Telmatobius to Lynchophrys and
Batrachophrynus (see Cei, 1986; Morales, 1989). Regardless, there is at
least some character evidence to support the generic assignment of eight of
the species from northern Peru.

Allocation of Telmatobius atahualpai is more difficult, as there are no
adults, males, or skeletal material available. However, the two larger speci-
mens have relatively long, recurved, fanglike teeth on the maxilla, premax-
illa, and vomer. This condition occurs in all Telmatobius that I have
examined (Appendix II; excepting the edentate T carrillae), but occurs
also (to a greater or lesser extent) in Alsodes and Insuetophrynus. Based on
biogeography, it seems more likely that T. atahualpai is closely related to

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 43

northern Andean Telmatobius rather than the Patagonian genera, but this
admittedly is not a strong argument. More confident assignment of this
species must await the acquisition of additional material.

KEY TO ADULT TELMATOBIUS OF NORTHERN PERU

1. Premaxillary teeth present; adult size > 50 mm 2

Premaxillary teeth absent; adult size < 50 mm T. degener

2. Dorsum black in life, with yellow-orange reticulations or greenish-gold
flecks 3

Dorsum dull brown to gray in life (with or without darker spots); no
yellow-orange reticulations or greenish-gold flecks 4

3. Dorsum black with yellow-orange reticulations; dorsum coarsely granu-
lar; skin pustular ventromedial to cloaca; northern Cordillera Occidental

T. latirostris

Dorsum black (gray to brown in preservative) with greenish-gold flecks;
skin smooth; Cordillera Central T. atahualpai

4. Tympanum present, nuptial spines minute but not extensive on ventral
surface of thumb; Cordillera de Huancabamba T. ignavus

Tympanum absent (except some T. brevipes) 5

5. Nuptial spines minute, extensive on dorsal, lateral, and ventral surface
of thumb; nuptial pad continuous with inner palmar tubercle; Cordillera
Occidental 6

Nuptial spines large, conical, restricted to dorsal and lateral surface of
thumb; separated from inner palmar tubercle: Cordillera Central 7

6. Nuptial spines larger, less extensive on ventral surface of thumb, larger
size (males to 68.9 mm SVL, females 77.3) T. thompsoni

Nuptial spines smaller, more extensive on ventral surface of thumb;
smaller size (males to 63.5 mm SVL, females 71.3) T. brevipes

7. Webbing extending to level of ultimate phalange of fifth toe; eyes
distinctly protuberant on top of head; inner palmar tubercle with distinct
inner margin; inner metatarsal tubercle raised, subcircular; outer

metarsal tubercle roughly one-half size of inner; Cordillera Colan

T. colanensis

Webbing usually not extending to level of ultimate phalange on fifth toe;
eye not distinctly protuberant on top of head: inner palmar tubercle with
indistinct inner margin; inner metarsal tubercle ovoid; outer metatarsal

44 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

tubercle roughly one-third to one-fourth size of inner 8

8. Dorsum usually tuberculate; underside of limbs mottled with yellow to
lavender blotches; yellowish-tan flecks absent on dorsum ....T. truebae

Dorsum smooth, underside of limbs uniform gray; yellowish-tan flecks
present on dorsum in life (in some) T. necopinus

SPECIES ACCOUNTS

Telmatobius atahualpai new species
Figure 9

Holotype. — KU 212485 a subadult female from the north slope of the
Abra Barro Negro, 28 km SSW Leimebamba (7°48' S, 77°51' W), 3470 m.
Departamento Amazonas, Peru, collected 23 January 1989 by William E.
Duellman.

Paratypes — KU 182084 from 28.3 km SSW Leimebamba, 3450 m,
Departamento Amazonas, Peru, collected 7 March 1979 by William E.
Duellman, and UF 39715 from 41 km SW Leimebamba, 3600 m,
Amazonas, Peru, collected 27 April 1972 by Fred G. Thompson.

Diagnosis. — ( 1 ) Premaxillary teeth present; (2) tympanum absent; (3)
condition of nuptial spines unknown; (4) dorsum (in life) black with
greenish flecks; (5) venter gray, unpatterned; (6) skin of dorsum smooth;
(7) maximum size of adults unknown.

This species is easily distinguished from the other northern Peruvian
Telmatobius by its unique dorsal coloration. Telmatobius atahualpai also
can be distinguished from the partially sympatric T. truebae by its skin
texture and ventral color. The skin of the former species is always smooth,
whereas T. truebae usually has warts on the dorsum and ventral to the
cloaca. In T. truebae there is usually (90%) bright yellow to lavender
coloration on the underside of the limbs, but in T. atahualpai the ventral
surfaces are uniform gray.

Description. — (Based on the two largest individuals) head slightly
narrower than body; head wider than long (HLEN 77.2-83.9% HWID; x =
80.5 ± 4.7, n = 2); head length 28.2-28.7% of SVL (x = 28.5 ± 0.3, n = 2);
head width 34.2-36.6%. of SVL (3c = 35.4 ± 1.7, n = 2); nostril not
protuberant, located at anterior terminus of snout; canthus rostralis indis-
tinct, short, slightly concave in dorsal profile, elevated in lateral profile;
loreal region concave; snout short, bluntly rounded; eye anterolateral,
protuberant or not on top of head; eye diameter 34.8-37.7% of head length
(x = 36.3 ± 2.0, // = 2); tympanum absent; condition of tympanic annulus
unknown; supratympanic fold well developed. Maxillary and premaxillary
teeth fanglike, embedded in labial mucosa; dentigerous processes of vomers
closer to large, subcircular choanae than to each other; processes medial to

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 45

choanae; oriented perpendicular to anteroposterior axis of skull (with slight
posteromedial inclination); each process bearing 4-5 (x = 4.2, n = 4)
fanglike teeth embedded in buccal lining for most of their lengths. Tongue
large, subcircular. shallowly notched posteriorly; attached through two-
thirds of its length anteriorly, free posteriorly. Condition of vocal slits in
males unknown.

Forelimbs relatively slender; dermal wrist fold weakly developed; rela-
tive lengths of fingers: III > I > IV > II; webbing and lateral fringes absent;
tips of fingers spherical; inner palmar tubercle oval, depressed; outer
palmar tubercle slightly larger than outer, subcircular; one round, indistinct
subarticular tubercle present proximally on each finger (larger tubercle on
Finger I); distal subarticular tubercles barely discernible on Fingers III and
IV; supernumerary palmar tubercles absent; condition of nuptial spines
unknown. Hind limb length (foot plus tibia) 91.2-94.57f of SVL (x = 92.8
±2.3. /; = 2): relative lengths of toes: V > IV = III > II > I; webbing formula
(n = 2): I (2--2)— 21/: II Va— (3-3+) III (2+-2'/3)— (32A-4~) IV (3^-3Vi)—
(l2/3-2)V; webbing diminishing distally to form fringes along lateral mar-
gins of toes; tips of toes spherical, approximately equal in size to fingertips;
inner metatarsal tubercle distinct, ovoid; outer metatarsal tubercle distinct,
subcircular. approximately half size of inner metatarsal tubercle;
subarticular tubercles round, mostly distinct, distributed on toes as follows:
1(1), 11(1), 111(2), IV(3). V(2); plantar supernumerary tubercles absent;
tarsal fold indistinct extending approximately half length of tarsus,
confluent distally with fringe along inner margin of Toe I.

Skin smooth; cloacal opening round, unornamented, at upper level of
thigh; transverse fold of skin dorsal to cloacal opening.

Color in preservative: Gray or brown with fine yellow to off-white
speckling on face, dorsum, and dorsal surfaces of limbs; slightly paler gray
or brown ventrally (with paler speckling on throat, flanks and underside of
limbs in KU 2 1 2485 ); tips of fingers and toes yellowish cream; palmar and
plantar tubercles pale gray.

Color in life: Dorsum nearly black with metallic green streaks and
flecks; venter uniform dark gray (W. E. Duellman field notes for KU
212485). Dorsum black with greenish-gold markings; venter gray; iris dull
bronze (W. E. Duellman field notes for KU 212485). Dorsum black with
gold flecks (F G. Thompson field notes for UF 39715).

Measurements of holotype (in mm): SVL: 43.2, TIBL: 18.1, FOOT:
21.3, HLEN: 12.2, HWID: 16.2, IOD: 3.2, IND: 3.3. ENOS: 3.4. EYE: 4.6,
HNDL: 11.9, THBL: 7.6, RDL: 12.1.

Distribution. — This species is restricted to the westernmost ridge of the
Cordillera Central (Fig. 1) at elevations of 3450-3600 m (Fig. 2).

Ecology. — All specimens were found under rocks by day; KU 212485
was under a rock at the edge of a small stream. Sympatric anurans include

46 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Telmatobius truebae and an undescribed Eleutherodactylus of the
unistrigatus group. Although sympatric with T. truebae at the type locality,
T. atahualpai apparently ranges to slightly higher elevations (to 3600 m)
than T. truebae (up to 3470 m), whereas T. truebae ranges to lower
elevations (down to 2180 m) than T. atahualpai (to 3450 m). Telmatobius
atahualpai is found in Very Humid Subalpine Paramo.

Etymology. — The specific name is a patronym of Atahualpa, the last
ruler of the Incas, and is a noun in the genitive case.

Telmatobius brevipes Vellard
Figure 9

Telmatobius brevipes Vellard, 1951:71.— Lectotype: MHNSM 3743,
from Huamachuco, 3350 m, Departamento La Libertad, Peru.

Diagnosis. — (1) Premaxillary teeth present; (2) tympanum usually
(93%) absent; (3) nuptial spines minute, extensive, nuptial pad continuous
with inner palmar tubercle; (4) dorsum (in life and preservative) dull brown
or gray, usually with darker spots; (5) venter pale brown or gray, underside
of limbs frequently (53%) with yellowish color; (6) skin of dorsum usually
(90%) smooth; (7) snout-vent length in males to 63.5 mm, in females to
71.3 mm.

The presence of minute nuptial spines on the thumbs of males distin-
guishes Telmatobius brevipes from the other northern Peruvian species,
except for T. ignavus and T. thompsoni. Telmatobius brevipes can be
distinguished from these two species in the following characters: ( 1 ) nup-
tial spines more extensive on ventral surface of thumb in T. brevipes, (2)
larger adult size in T. ignavus and T. thompsoni, (3) anterior ramus of
pterygoid substantially overlaps palatine in T. brevipes, whereas this pro-
cess only contacts the palatine in the other two species, and (4) sacral
diapophyses with a more dorsal orientation in T. ignavus and T. thompsoni.
Telmatobius brevipes also can be distinguished from T. ignavus in that in T.
ignavus the tympanum is always present (usually absent in T. brevipes), the
quadratojugal is thickened vertically, and the anterior margin of the dorsal
lamina of the atlas forms approximately a straight line (indented posteri-
orly in T. brevipes). The body of the tadpole of T. ignavus is more globular
and less streamlined and the tail is less extensively pigmented. Telmatobius
thompsoni also can be distinguished from T. brevipes in that the nuptial
spines are slightly larger, the palatine is more slender laterally than medi-
ally (more slender medially in T. brevipes), the anterior ramus of the
pterygoid bears a distinct raised process, and the sphenethmoid is less
extensive.

Description. — Head narrower than body; wider than long (HLEN 75.4-
88.3% HWID; x = 82.5 ± 3.4, n = 41); head length 26.6-32.1% of SVL (x

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 47

= 29.0 + 1.4, n = 41); head width 30.7-39.9% of SVL ( x = 35. 1 ± 2.1, n =
41); nostril not protuberant, located at anterior terminus of snout; canthus
rostralis indistinct, concave dorsally; in lateral profile posteriorly elevated;
loreal region concave, slightly depressed; snout long, bluntly rounded in
lateral profile; eye oriented anterolaterally, protuberant on top of head; eye
diameter 27.7-38.2% of head length (x = 32.5 ± 2.7; n = 41 ); tympanum
usually absent (absent in 93% of 67 individuals), tympanic annulus present;
supratympanic fold well developed to rudimentary, extending from poste-
rior corner of eyelid to insertion of forelimb; vertical fold posterior to
corner of jaw weakly developed, if present. Maxillary and premaxillary
teeth fanglike, embedded in labial mucosa with only tips protruding;
dentigerous processes of vomer closer to small, subcircular choanae than to
each other; medial or slightly posterior to choanae; oriented perpendicular
to longitudinal axis of skull or with slight posteromedial inclination; each
bearing 0-6 (x = 2.6, // = 82) recurved, fanglike teeth embedded in buccal
lining with only tips portruding. Tongue large, subcircular posterior border
free, unnotched or only shallowly notched; tongue attached anteriorly
through about two-thirds of its length. Vocal slits absent.

Forelimbs relatively slender, slightly more robust in males: dermal wrist
fold weakly developed in males, absent or poorly developed in females;
relative lengths of fingers: III > I = IV > II; palmar webbing absent; lateral
fringes rudimentary or absent, but well developed lateral fringe present
along outer margin of Finger I in males; tips of fingers spherical: in
females, inner palmar tubercle approximately elliptical, extending onto
proximal portion of Finger I, medial margin usually indistinct; in males,
inner palmar tubercle continuous with nuptial pad; outer palmar tubercle
distinct, flat, ovoid; one large, round subarticular tubercle present proxi-
mally on each finger, tubercle on Finger I very large; smaller, less distinct
subarticular tubercles present distally on Fingers III and IV; supernumerary
palmar tubercles present or absent proximally on each finger, in some
specimens, additional tubercle present along outer margin of inner palmar
tubercle; minute, densely packed nuptial spines on dorsal, medial, and
ventral surfaces of thumb. Hind limb length (foot plus tibia) 75.0-100.2%
of SVL (X = 88.6 + 5.8, n = 41); relative lengths of toes: IV > V = III > II
> I; webbing formula (range followed by mode; n = 36): I (l'/2-2; \2A) —
(2'/3-22/3; 2V2) II ( 1-2 ; V/2)—(22A-3W, 3+) III ( l'/2-2+; 2) — (3+-32/?;
3[/:) IV (3-3'/:; 3'/2) — (l+-2: l2A) V; webbing diminishing distally to form
lateral fringes along edges of toes; tips of toes spherical, about equal in size
to fingertips; inner metatarsal tubercle small, oval, usually flat, distal edge
free; outer metatarsal tubercle subcircular to elliptical, approximately half
size of inner metatarsal tubercle; small, round subarticular tubercles dis-
tributed on toes as follows: K 1 ). IK 1 ). 111(2), IV(3), V(2); supernumerary
plantar tubercles absent; tarsal fold well developed usually extending two-

48 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

thirds length of tarsus, confluent distally with fringe along inner margin of
Toe I.

Skin usually smooth dorsally and ventrally; dorsum with small warts in
10% of 67 individuals; skin usually pustular ventral to cloaca; cloacal
opening round, unornamented, directed posteriorly at upper level of thighs;
thick, transverse fold of skin dorsal to the cloaca in many specimens.

Color in preservative: Face and dorsal surfaces of head, body, and limbs
dull brown or gray, with darker markings on dorsum (small flecks to a few
large blotches) in most individuals and pale spots in two individuals (KU
212412 and 212429); venter pale brown or gray; ventral surfaces of limbs
frequently (53% of 67) yellowish tan, overlain by brown or gray spots
(especially medially); yellowish color extending onto entire venter in some
individuals; tips of fingers and toes yellowish white; hand and foot tu-
bercles light gray.

Color in life: Dorsum pale to medium olive-green with small to me-
dium-sized dark olive blotches; dorsal surface of thighs with dark flecks
(usually) or dark blotches (rarely); posterior surface of thighs light olive;
venter greyish olive; chins of smaller individuals with slight yellow tint;
iris yellowish green (T. H. Fritts field notes for KU 131684-94 from
Huamachuco). Dorsum dull olive-green or tan with darker green or brown
spots; lighter tan flecks in some individuals; venter dull cream, olive-
cream, or creamy yellow; underside of hind limbs dull yellowish tan or dull
orange and gray; iris tan, copper, or bronze with reddish-brown or black
reticulations or flecks ( W. E. Duellman field notes for KU 2 1 24 1 2-24 from
various localities in Departamento Cajamarca).

Measurements oflectotype (in mini: SVL47.6, TIBL 19.8. FOOT 22.5,
HLEN 14.0, HWID 16.0. IOD 2.7. IND 3.0, ENOS 2.8. EYE 5.0. HNDL

11.6, THBL 7.7. RDL 12.9.

Larvae. — The following description is based on an individual of Stage
38 from the series KU 181857. The individual is illustrated in Fig. 16Aand
has the following measurements (in mm): total length 73.4. body length

28.7, basal tail muscle height 8.4, basal tail muscle width 4.9, maximum
dorsal fin height 4.4 located 35.7 from body terminus, maximum ventral fin
height 3.7 located 29.1 from body terminus, body height 12.1, body width

14.8, eye diameter 2.0, pupil diameter 0.7, interorbital distance 2.8, narial
diameter 0.7. internarial distance 3.8, snout-naris 5.6, naris-eye 2.5, trans-
verse oral disc diameter 5.4.

Body ovoid in dorsal and lateral view; snout bluntly rounded; naris
small, round, oriented anterolaterally; eyes small, oriented dorsolaterally;
spiracular tube lateral, sinistral, located slightly anterior to midlength of
body, short, oriented posterodorsally, pigmented only slightly; vent me-
dian, unpigmented. Dorsal fin slightly narrower and ventral fin narrower
than depth of caudal musculature at midlength of tail; caudal fins not

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 49

extending onto body; tail relatively elongate, gradually tapering to blunt
terminus; caudal musculature robust. Mouth anteroventral; oral disc ovoid;
outer margin of disc pigmented; marginal papillae arranged biserially
posteriorly, in single series laterally and anteriorly, unpigmented; anterior
gap in marginal papillae wide; oral disc with transangular margin;
intramarginal papillae large, present laterally; intramarginal mental papil-
lae absent; jaw sheaths keratinized, gently arced, finely serrated, dorsal
sheath wider than ventral; labial tooth row formula 2/3, tooth rows adjacent
to jaw sheaths interrupted medially.

Color in preservative': Dorsum of body uniform brown with scattered
round, dark brown spots; venter largely unpigmented, intestine visible
through skin; caudal musculature evenly covered with brown pigment;
dorsal fin with numerous dark blotches scattered throughout its length:
ventral fin lightly and irregularly pigmented along ventral margin and
terminus.

Color in life: (W. E. Duellman field notes for KU 181857) Body and tail
dull olive-tan with dark brown spots; belly grayish white; iris bronze.

Variation: Other individuals included in KU 181857 are generally simi-
lar to the specimen described. KU 1 8 1 843 and 1 8 1 844 consist of individu-
als in earlier developmental stages. In life, individuals of KU 181843 were
olive-green dorsally and greenish gray below (W. E. Duellman field notes).
Tadpoles from Abra Quilsh (26 km NNW Cajamarca. KU 212503) are less
streamlined, have a shorter and broader tail, and smaller dark spots on the
dorsum. Tadpoles from Abra Comulica (KU 212525) are also less stream-
lined than those of KU 181857, appear to be slightly larger, are more lightly
pigmented, and have only indistinct dark spots on the dorsum. In life, these
tadpoles have a silvery-bronze tint on the venter and pale gray fins with
black flecks (W. E. Duellman field notes). KU 212526 consists of two
tadpoles tentatively assigned to this species (no adults are known from the
locality); these individuals are relatively lightly pigmented (especially on
the ventral fin) and the dorsum lacks dark brown spots.

Distribution. — Telmatobius brevipes is relatively widely distributed in
the Cordillera Occidental (Fig. 1 ) at elevations from 2000-3520 m (Fig. 2).

Ecology. — Telmatobius brevipes occurs in or along streams or irrigation
ditches in paramo and in cultivated areas. Almost all specimens have been
collected by day; most specimens were found under rocks or in streamside
vegetation. The distribution is within the Humid and Very Humid Montane
Forest (cloud forest) and the Very Humid Subalpine Paramo life zones.
Sympatric anurans include At el opus peruensis, Bufo cophotis, B. limensis,
Gastrotheca peruana , and Phrynopus simonsii. The stomachs of four indi-
viduals (KU 131690, 181432, 181435, 212446) contained a planthopper
(Homoptera, Membracidae), ants (Hymenoptera, Formicidae), several
beetles (Coleoptera: Chrysomelidae, Curculionidae). a springtail

50 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

(Collembola, Poduridae), coleopteran larvae (Cantharidae, Elateridae), lepi-
dopteran larvae, a centipede, a nematode, a spider, an earthworm, a slug, a
small, unidentifiable anuran, and an unidentifiable tadpole. Eggs (KU
181842) attributed to this species were found adherent to vegetation in a
narrow rivulet in March. Tadpoles were collected in pools in slow-moving
streams, in grassy ponds, and in irrigation and roadside ditches in January
and March.

Remarks. — Vellard (1951) designated the holotype of Telmatobius
brevipes as specimen #3 of the series MHNJP (= MHNSM) 246 (given as
249 by Morales et al., 1990). In the original description, this series con-
tained 13 individuals, of which 246(3) was the largest adult male. When I
examined the type series of this species in 1989, I found that these speci-
mens lacked individual tags and numbers. Two specimens of the original
type series seem to be missing. Comparison of the sizes of the specimens
listed by Vellard (1951) and those of the remaining specimens indicates
that the missing specimens could be Vellard's holotype and allotype (the
largest male and female). Morales et al. (1990) also suggested that the
holotype of this species is missing and listed new numbers for the 1 1
paratypes (MHNSM 3739-3749). I designate one of these (MHNSM 3743.
the largest male in the series of 1 1 specimens) as the lectotype of T.
brevipes.

Telmatobius colanensis new species

Holotype— LSUMZ 37102. an adult male, from the Cordillera Colan.
SE La Peca (ca. 5°34' S, 78° 19' W), 2410 m, Departamento Amazonas,
Peru, collected 16 October 1978 by Thomas S. Schulenberg.

Diagnosis. — ( 1 ) Premaxillary teeth present; (2) tympanum absent; (3)
nuptial spines moderately large, distinctly conical, restricted to dorsal and
lateral surface of thumb and separate from inner palmar tubercle; (4)
dorsum (in preservative) dull gray; (5) venter tan; (6) skin of dorsum
smooth; (7) snout-vent length in male 62.5 mm.

The large size, large nuptial spines, and dull gray dorsum of T. colanensis
readily distinguish it from the other species in the region, except for T.
necopinus and T. truebae. Telmatobius colanensis is distinct from these two
species in having: ( 1 ) toes more extensively webbed (interdigital webbing
extending to level of ultimate phalange of fifth toe in T. colanensis, not
extending as far in T. necopinus and T truebae), (2) inner metatarsal
tubercle nearly spherical (vs. elliptical), larger, and more distinctly raised,
(3) outer metatarsal tubercle larger (about one-half size inner vs. about one-
third to one-fourth size inner in T. necopinus and T. truebae), (4) inner
palmar tubercle with distinct inner margin, and (5) eyes more protuberant
on top of head. Telmatobius colanensis can be distinguished further from T.

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 5 1

truebae in having: (1) skin smooth on dorsum (tuberculate in 82% of T.
truebae), (2) nuptial spines smaller, and (3) snout usually shorter.

Description. — Head slightly narrower than body; head wider than long
(HLEN 78.5% HWID); head length 31.5% SVL; head width 40.2% SVL;
nostril not protuberant, located just posterior to anterior terminus of snout;
canthus rostralis short, indistinct, and slightly concave in dorsal profile;
slightly elevated posteriorly in lateral profile; loreal region concave; snout
short, bluntly rounded; eye large, anterolaterally oriented, protuberant on
top of head; eye diameter 36.5% of head length; tympanum absent, condi-
tion of tympanic annulus unknown; supratympanic fold well developed,
smooth. Maxillary and premaxillary teeth fanglike, embedded in labial
mucosa; dentigerous processes of vomers closer to small, circular choanae
than to each other; located medial to choanae; processes perpendicular to
longitudinal axis of skull with slight posteromedial inclination; each bear-
ing five or six recurved, fanglike teeth embedded in buccal lining for most
of their lengths. Tongue moderately large, free, and unnotched posteriorly,
attached through two-thirds of its length anteriorly. Vocal slits absent.

Forelimbs moderately robust; dermal wrist fold distinct; relative lengths
of fingers: III > I > IV > II; palmar webbing absent; weak lateral fringes
present on fingers; tips of fingers spherical; inner palmar tubercle oval,
distinct; outer palmar tubercle roughly same size as inner, subcircular,
slightly raised; one round, distinct subarticular tubercle present proximally
on each finger, slightly larger tubercle on Finger I; round subarticular
tubercles distally on Fingers III and IV; indistinct supernumerary palmar
tubercle proximal to Finger I, adjacent to outer edge of inner palmar
tubercle; two smaller, less distinct supernumerary tubercles present proxi-
mally on Finger II; tips of fingers roughly spherical; nuptial spines, moder-
ately large, conical, keratinized on dorsal and lateral surface of thumb,
spines on raised pad proximally. Hind limb length (foot plus tibia) 102.1%
of SVL; relative lengths of toes: IV > III > V > II > I; webbing formula: I
V-A— 2+II 1'/^— 3 III 2—3 '/UV 3—1 V; webbing diminishing distally to
form wide fringes along lateral margins of toes; tips of toes spherical,
relatively large, roughly equal in size to fingertips; inner metatarsal tu-
bercle ovoid, raised, relatively large; outer metatarsal tubercle subcircular,
raised, approximately one half size of inner; subarticular tubercles round,
distinct, distributed on toes as follows: 1(1), 11(1), 111(2), IV(3). V(2);
supernumerary plantar tubercles absent; tarsal fold extending approxi-
mately three-fourths length of tarsus, confluent distally with fringe along
inner margin of Toe I.

Skin of dorsum smooth; skin pustular and creased ventromedial to
cloaca and on underside of forelimbs; creased on underside of hindlimbs
and abdomen; cloacal opening round, unornamented, at upper level of
thigh; short transverse fold of skin present dorsal to cloaca.

52 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Color in preservative: Dorsum, dorsal surfaces of limbs, and palmar and
plantar surfaces dull gray; venter throat and underside of limbs limbs tan;
tips of fingers and toes yellowish tan; palmar and plantar tubercles pale
gray. Color in life unknown.

Measurements ofholotype: SVL 62.5, TIBL 32.1, FOOT 31.7, HLEN
19.7, HWID 25.1, IOD 3.8, IND 5.1. ENOS 4.3, EYE 7.2. HNDL 16.9,
THBL 10.4, RDL 18.6.

Distribution. — Known only from the type locality in the Cordillera
Colan (Fig. 1 ).

Ecology. — Collected in a rocky, high-gradient stream in undisturbed
cloud forest (T. Schulenberg, pers. comm.).

Etymology. — The specific name is an adjective formed from the name
of the mountain range to which the species is endemic.

Telmatobius degener new species

Holotype. — UF 39736, adult male, collected 13 km ESE Yamobamba,
18 km SE Agallpampa on road from Otuzco to Huamachuco (7°59' S.
78°36' W). Departamento La Libertad, Peru, 3290 m, on 9 May 1972 by
Fred G. Thompson.

Paratypes— KU 218463-66, UF 39738^11, 39743-45, 39748-49,
39752-53, and 39755, same locality and collector as holotype. KU 218466
is a cleared-and-stained skeleton, and KU 218465 is a dry skeletal prepara-
tion.

Diagnosis. — (1) Premaxillary teeth absent; (2) tympanum absent; (3)
nuptial spines moderately large, distinctly conical, on dorsal and ventral
surface of thumb; nuptial pad continuous with inner palmar tubercle; (4)
dorsum (in preservative) gray to brown; (5) venter yellowish cream with
gray speckling; (6) skin of dorsum smooth; (7) size small; snout-vent
length in males to 48.9 mm, in females to 49.6 mm.

This species can be distinguished from the other northern Peruvian
Telmatobius on the basis of its small size, large, conical, nuptial spines, and
the following osteological characters: (1) premaxillary teeth absent, (2)
palatine reduced, (3) crista parotica short, (4) otic capsule large, (5)
sphenethmoid poorly ossified, (6) jaw articulation at (not posterior to) level
of crista parotica, (7) zygomatic ramus of squamosal very short, (8) col-
umella reduced to small spherical nubbin. (9) tympanic annulus and pars
externa plectri absent, and ( 10) transverse processes of posterior presacral
vertebrae very short and thick.

Description. — Head equal to or slightly narrower than body; head
wider than long (HLEN 79.7-88.0% HWID; x = 83.6 ± 3.1, n = 9); head
length 25.1-29.2% of SVL (x = 27.8 ± 1.4, // = 9); head width 31.1-34.3%
of SVL (x = 33.3 ± 1.0, n = 9); nostril not protuberant, located at anterior

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 53

terminus of snout; canthus rostralis indistinct in dorsal view, slightly el-
evated in lateral view; loreal region barely concave, slightly depressed;
snout short, and rounded in lateral profile, slightly acuminate in dorsal
view; margin of upper lip flared slightly; eye anterolaterally oriented;
protuberant on top of head; eye diameter 33.1-39.2% of head length (x =
35.4 ±2.1,// = 9); tympanum and tympanic annulus absent; supratympanic
fold poorly developed. Premaxillary teeth absent; maxillary teeth fanglike,
embedded in labial mucosa with tips barely protruding; dentigerous pro-
cesses of vomers closer to choanae than to each other; anterior to level of
choanae: roughly perpendicular to longitudinal axis of skull; each bearing
(Ml (x - 1.9, // = 26) fanglike, recurved teeth; teeth absent on one side of
two individuals; teeth embedded in buccal lining for most of their lengths;
choana varying from large subcircular opening to small, posterolaterally
oriented slit. Tongue large, subcircular, posterior border free, unnotched;
attached anteriorly through approximately three-fourths of its length. Vocal
slits absent.

Forelimbs moderately robust in males (dermal wrist fold usually
present), relatively slender in females (dermal wrist fold absent or rudi-
mentary); relative lengths of fingers: III > IV > I > II; palmar webbing
absent; lateral fringes on fingers absent or rudimentary; tips of fingers
spherical; in females, inner palmar tubercle large, raised slightly, extending
onto base of Finger I; in males, inner palmar tubercle continuous with
nuptial pad; outer palmar tubercle large, ovoid, raised slightly; one large,
round subarticular tubercle proximally on each finger; distal subarticular
tubercles on Fingers III and IV; one to three supernumerary tubercles
proximal to each proximal subarticular tubercle: nuptial spines keratinous,
moderately large, distinctly conical, on dorsal and ventral surface of thumb.
Hind limb length (foot plus tibia) 85.4-108.4% of SVL (x = 94.2 ± 7.2. //
= 9): relative lengths of toes: IV > III > V > II > I; webbing formula (range
followed by mode; // = 10): I (1-VA; T)— (2+-2'/2; 2lA) II (\+-Va\ 1+)— (3~
-3; 3) III (l'/2-2+; l2A)—(3-3lA; 3+) IV (3-3 Vs; 3) — (1— lVft; D V: webbing
diminishing distally to form lateral fringe along edges of toes; tips of toes
spherical, roughly equal in size to fingertips; inner metatarsal tubercle
large, elliptical, distinctly raised; outer metatarsal tubercle large (two-
thirds to four-fifths size of inner metatarsal tubercle), elliptical to ovoid,
elevated; subarticular tubercles well developed, round, distributed on toes
as follows: I( 1 ), II( 1 ). 111(2), IV(3), V(2); row of one to four small supernu-
merary tubercles present proximally on Toes II-V; tarsal fold distinct,
extending approximately two-thirds length of tarsus, confluent distally
with fringe along inner margin of Toe I.

Skin smooth dorsally and ventrally: skin pustular ventral to cloaca;
cloacal opening round and unornamented. directed posteriorly at upper
level of thish.

54 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Color in preservative: Dorsal surfaces of head and limbs and ventral
surfaces of tarsus, forearm, hands, and feet dull gray to brown; venter,
throat, underside of limbs, and tips of fingers and toes yellowish cream
with gray speckling; hand and foot tubercles pale gray. Color in life
unknown.

Measurements of holotype (in nun): SVL 48.9, TIBL 23.7. FOOT 26.2,
HLEN 13.4, HWID 16.7, IOD 2.4, IND 3.5, ENOS 3.1, EYE 4.5, HNDL
12.0, THBL 7.2. RDL 13.3.

Distribution. — Known only from the type locality in the Cordillera
Occidental (Fig. 1 ).

Ecology. — According to field notes of F. G. Thompson, the type series
was collected in a densely vegetated roadside marsh in an "open pasture."
The marsh was fed by a brook draining into the Rio Moche, and was in the
Very Humid Subalpine Paramo life zone. The frogs were active and abun-
dant in pools and running water by day (13:30 h). and were found simulta-
neously with Telmatobius thompsoni. The stomach of one individual (KU
218466) contained two caddisfly larvae (Trichoptera; Hydroptilidae and an
undetermined family).

Etymology. — The specific name degener is a Latin adjective meaning
degenerate or departing from its kind, and is used in reference both to the
small size and distinctive osteological features of this species.

Telmatobius ignavus Barbour and Noble

Figure 9

Telmatobius ignavus Barbour and Noble. 1920:414. — Holotype: MCZ
4093 (subadult female) from the city of Huancabamba, 2000 m,
Departamento Piura, Peru.

Diagnosis. — ( 1 ) Premaxillary teeth present; (2) tympanum present; (3)
nuptial spines minute, mostly on dorsal and medial surface of thumb;
nuptial pad distinct from inner palmar tubercle; (4) dorsum (in life and
preservative) dull tan to brown with darker spots; (5) venter tan, gray, or
grayish brown; (6) skin of dorsum smooth; (7) snout-vent length in males
to 74.9 mm, in females to 78.9 mm.

This species is uniquely characterized among northern Peruvian
Telmatobius in having a tympanum. A tympanum also is found in some
(7%) T. brevipes, but T. brevipes is distinct in having: (1) nuptial spines
more extensive on the ventral surface of the thumb. (2) adult size smaller,
(3) anterior ramus of pterygoid with substantial overlap on palatine, (4)
quadratojugal more slender, (5) sacral diapophyses more laterally oriented
(rather than dorsolaterally), (6) anterior margin of dorsal lamina of atlas
posteriorly indented, (7) tadpole body shape more streamlined, and (8)
tadpole with caudal fins and musculature usually more extensively pig-

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 55

merited. Telmatobius ignavus also is similar to T. thompsoni in having small
nuptial spines, but that species is characterized by ( 1 ) snout usually longer,
(2) nuptial spines slightly larger, (3) venter yellowish cream with gray
speckling (in preservative), (4) palatine more slender laterally, (5)
sphenethmoid less extensively ossified anteriorly, and (6) orbitonasal fora-
men in sphenethmoid large, oblong (vs. small, round foramen in T. ignavus).

Description. — Head equal to or slightly narrower than body; head
wider than long (HLEN 75.3-85.3% HWID; x = 80.8 ± 3.6, n = 5); head
length 30.4-32.4% of SVL (x = 31.6 ± 1.0. n = 5); head width 37.0-40.5%
of SVL (x = 39.1 ± 1.5, n = 5); nostril not protuberant, located at anterior
terminus of snout; canthus rostralis distinct dorsally, elevated in lateral
profile; loreal region concave; snout short, bluntly rounded in lateral pro-
file; eye oriented anterolaterally, somewhat protuberant on top of head; eye
diameter 27.7-33.0% of head length (x = 30.1 ± 2.6. n = 5); tympanum
present; tympanic annulus barely visible under skin; tympanum largely
concealed posteriorly and dorsally by supratympanic fold; supratympanic
fold robust, thick, glandular, extending from posterior corner of eyelid to
insertion of forelimb; vertical fold extending ventrally from supratympanic
fold to throat, posterior to tympanum and corner of jaw. Maxillary and
premaxillary teeth fanglike, embedded in labial mucosa with tips protrud-
ing or not; dentigerous processes of vomer approximately equidistant
between small, subcircular choanae and each other; processes medial to
choanae, perpendicular to longitudinal axis of skull with slight posterome-
dial orientation; each bearing 2-7 (x = 4.3, n = 6) fanglike, recurved teeth;
teeth embedded in buccal lining for most of their lengths. Tongue large,
subcircular, posterior border free and shallowly notched, attached anteri-
orly for approximately two-thirds of its length. Vocal slits absent.

Forelimbs moderately slender; dermal wrist fold absent; relative lengths
of fingers: III > I > IV > II; webbing absent; lateral fringes rudimentary or
absent; tips of fingers spherical; inner palmar tubercle elliptical, elongate,
extending onto Finger I. slightly larger than outer palmar tubercle, distinct
from nuptial pad in males; outer palmar tubercle large, ovoid; one large,
round subarticular tubercle present proximally on each finger; distal
subarticular tubercles on Fingers III and IV; small, indistinct supernumer-
ary palmar tubercle proximal to each proximal subarticular tubercle; tu-
bercle of Finger I situated along outer margin of inner palmar tubercle;
nuptial spines minute, keratinous, mostly on dorsal and medial surface of
thumb. Hind limb length (foot plus tibia) 82.4-95.2% of SVL (x = 89.3 ±
4.8, // = 5); relative lengths of toes: IV > III > V > II > I; webbing formula
(range followed by mode: n = 3): I 2— Va II ( Va-12a; Va)— (3-3 '/j;
3) III (2-21/:; 2+)— (3-3'/2; Va) IV {Va-V/i\ 31/:)— 2 V; webbing diminish-
ing distally to form lateral fringes along edges of toes; tips of toes spheri-
cal, about equal in size to tips of fingers: inner metatarsal tubercle ovoid.

56 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

flat (except in UF 34089); outer metatarsal tubercle oval, slightly raised,
one-half to two-thirds size of inner; subarticular tubercles small, round,
distributed on toes as follows: I( 1 ), II( 1 ), 111(2), IV(3), IV(2); supernumer-
ary plantar tubercles absent; tarsal fold well developed, extending at least
two-thirds length of tarsus, confluent distally with fringe along inner mar-
gin of Toe I.

Skin of dorsum and venter smooth; skin pustular ventral to cloaca;
cloacal opening directed posteriorly at upper level of thighs; opening round
and unomamented; transverse fold of skin dorsal to cloaca, usually extend-
ing anterolaterally to insertion of hindlimbs.

Color in preservative: Dorsum dull brown or gray; venter lighter brown
or gray; tips of fingers and toes yellowish cream; palmar and plantar
tubercles pale gray.

Color in life: (for KU 181438-181442, W. E. Duellman field notes)
Dorsum dull tan. olive-tan or brown with dark brown, olive-brown, or
olive-green spots; venter dull tan. dull gray or grayish brown; iris bronze
with black reticulations.

Measurements ofholotype (in mm): SVL 54.5, TIBL 23.9, FOOT 26.0.
HLEN 17.2, HWID 23.0. IOD 4.5. IND 4.3. ENOS 3.9, EYE 4.7, HNDL
14.5. THBL 10.3, RDL 14.0.

Larvae. — The following description is based on an individual (from the
series KU 181846) in Stage 34 (Fig. 16B) with the following measurements
(in mm); total length 66.6, body length 28.1, basal tail muscle height 7.6,
basal tail muscle width 5.4, maximum dorsal fin height 6.0 located 30.4
from body terminus, maximum ventral fin height 5.1 located 24.3 from
body terminus, body height 10.5. body width 16.8, eye diameter 2.0, pupil
diameter 0.5, interorbital distance 3.7. narial diameter 0.7, internarial dis-
tance 4.4. snout-naris 4.5, naris-eye 2.6. transverse oral disc diameter 6.8.

Body ovoid in dorsal and lateral view; snout bluntly rounded; naris
small, round, oriented anterolaterally; eyes small, oriented dorsolaterally;
spiracular tube lateral, sinistral, located at about midlength of body, short,
opening posterodorsally, lightly pigmented; vent median, short,
unpigmented. Dorsal and ventral fins narrower than depth of caudal muscu-
lature at midlength; caudal fins not extending onto body; tail tapering to
broad, blunt terminus; caudal musculature robust, extending onto body;
skin on body loose, baggy. Mouth large, opening ventrally, subterminal;
oral disc ovoid; marginal papillae irregularly arranged, at least two papillae
thick posteriorly, single row anteriorly and laterally; papillae unpigmented;
anterior gap in marginal papillae wide; oral disc with transangular margin:
intramarginal papillae present laterally, extending medially for short dis-
tance, intramarginal mental papillae absent; jaw sheaths keratinized, gently
arced, margins finely serrated; dorsal sheath wider than ventral; labial tooth
row formula 2/3; tooth rows adjacent to jaw sheaths interrupted medially.

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 57

Color in preservative: Dorsum uniform brown; venter unpigmented,
intestine readily visible through skin; caudal musculature with scattered
brown pigment; dorsal fin with numerous brown blotches, fewer blotches
on ventral fin.

Color in life: (W. E. Duellman field notes) Body and caudal musculature
olive-gray: caudal fins translucent tan with gray blotches; iris dull bronze.

Variation: Individuals from El Tambo (KU 181845) are slightly darker
in preservative than those closer to Huancabamba (KU 181846-847). In
life, tadpoles of KU 181847 (W. E. Duellman field notes) were brown with
greenish flecks ventrolaterally; the caudal musculature and fins were tan
with gray brown flecks (metallic green flecks proximally on ventral fin) and
the iris was greenish bronze.

Distribution. — Known only from the isolated Cordillera Huancabamba
(Fig. 1 ) at elevations from 1840-3080 m (Fig. 2).

Ecology. — Individuals were under rocks in streams by day or in pools
by day or night. Telmatobius ignavus occurs in Very Humid Montane
Forest (most localities). Humid Lower Montane Forest, and possibly Dry
Lower Montane Forest (at Huancabamba), and has been found with
Colostethus sylvaticus, Gastrotheca lateonota, Eleutherodactylus
cajamarcensis, E. colodactylus, E. cryptomelas, E. rhodoplichus, at least
one undescribed Eleutherodactylus (Duellman and Wild), Phrynopus
nebulanastes, and P. parkeri. The stomach of one individual (KU 181441)
contained remains of an adult and a larval lepidopteran. two beetles (Co-
leoptera, Curculionidae), a fly (Diptera, Nematocera), and a larval midge
(Diptera, Chironomidae). Tadpoles of this species have been found in both
rocky and muddy pools and in rocky streams in January and February.

Remarks. — Duellman and Wild (1993) gave the altitudinal distribution
of this species as 2320-3080 m. However, the type locality is at 2000 m (in
the city of Huancabamba), and a specimen (UF 52148) has been collected
at 1840 m.

Telmatobius latirostris Vellard
Figure 9

Telmatobius latirostris Vellard, 1951:68.— Lectotype: MHNSM 3733.
from the city of Cutervo, 2620 m, Departamento Cajamarca, Peru.

Diagnosis. — (1) Premaxillary teeth present; (2) tympanum absent; (3)
nuptial spines moderately large, conical; nuptial pad on dorsal surface of
thumb and separate from inner palmar tubercle; (4) dorsum (in life) yellow-
orange with black spots and reticulations; (5) venter gray (in preservative)
with yellowish-cream blotches on underside of limbs; (6) skin of dorsum
usually coarsely granular; (7) size large, snout-vent length in males to 83.0
mm, in females to 83.6.

58 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

The unique color pattern of Telmatobius latirostris readily distinguishes
it from congeners. This species also is distinct osteologically in having a
wide skull, short posterolateral processes of the hyoid, extensive ossifica-
tion of the posteromedial hyoid processes onto the hyoid plate, transverse
processes of the second presacral vertebrae with a marked anterior curve,
and hypertrophy of the distal portion of the anteroproximal (crista ventralis)
humeral crest (in males). Telmatobius latirostris also is distinct morpho-
metrically from the other large-sized, large-spined species in the region, T.
colanensis, T. necopinus, and T. truebae, and is distinct electrophoretically
from T. necopinus and T. truebae.

Description. — Head narrower than body; head wider than long (HLEN
73.5-80.0% of head width; x = 76.7 ± 2.8, n = 5): head length 28.8-31.1%
of SVL (x = 29.8 ± 0.9, // = 5); head width 37.6-39.9% of SVL (x = 38.8
± 0.8, // = 5); nostril not protuberant, located at anterior terminus of snout;
canthus rostralis short, indistinct, and concave in dorsal view, elevated in
lateral profile; loreal region concave; snout very short, sloping, and bluntly
rounded in lateral profile; in dorsal profile bluntly rounded; eye
anterolaterally oriented: not protuberant on top of head; eye diameter 26.7-
32.2% of head length (x = 28.8 ± 2.2, n = 5): tympanum absent, tympanic
annulus present; supratympanic fold present, robust; extending from poste-
rior corner of eyelid to just anterior to insertion of forelimb; thick vertical
fold of skin posterior to corner of jaw, extending from beneath supratym-
panic fold to throat. Maxillary and premaxillary teeth fanglike; embedded
in labial mucosa with only tips protruding; dentigerous processes of vomer
slightly closer to each other than to small, subcircular choanae; processes at
level with or slightly posterior to choanae; oriented perpendicular to longi-
tudinal axis of skull with posteromedial inclination; each process bearing
3-6 (x = 4.3, /; = 6) fanglike, recurved teeth; teeth embedded in buccal
lining for most of their lengths. Tongue large, circular, attached anteriorly
approximately two-thirds to three-fourths of length; posterior border free
and unnotched. Vocal slits absent.

Forelimbs robust in males (dermal wrist fold present), moderately ro-
bust in females (dermal wrist fold absent); relative lengths of fingers: III >
IV = I > II; webbing absent; lateral fringes usually present along distal
margins of fingers, indistinct; tips of fingers spherical; inner palmar tu-
bercle oval, usually equal in size to outer; outer palmar tubercle large,
ovoid, slightly depressed; one large, round subarticular tubercle proximally
on each finger (tubercle of Finger I very large); distal subarticular tubercles
on Fingers III and IV; one or more small supernumerary tubercles present
at base of each finger in some individuals; supernumerary palmar tubercle
also present along outer, anterior edge of inner palmar tubercle; nuptial pad
present on dorsomedial surface of Finger I in males, separate from inner
palmar tubercle; nuptial spines moderately large, keratinous. conical,

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 59

densely packed, in contact at their bases. Hind limb length (tibia plus foot)
87.2-93.6% of SVL (x = 90.2 ± 2.7. n = 5); relative lengths of toes: IV >
III > V > II > I; webbing formula (range followed by mode; n = 3): I ( \-\vr,
1+)_(2+-21a; 2^)II(1-1+; 1) — (3-3; 3) III (IV: -2; l:/i)— (3-3'/>;
3+) IV (3+-3'/^; 3+)— (Va-2: l2A) V; webbing diminishing distally to form
lateral fringes along toes; tips of toes spherical, roughly equal in size to tips
of fingers; inner metatarsal tubercle ovoid, elevated, large: outer metatarsal
tubercle oval, slightly elevated, approximately half size of inner;
subarticular tubercles well developed, distributed on toes as follows: I( 1 ).
11(1). 111(2). IV(3), V(2): plantar supernumerary tubercles absent; tarsal
fold distinct, extending approximately one-half to two-thirds length of
tarsus, confluent distally with fringe along inner margin of Toe I.

Skin of dorsum usually coarsely granular; skin ventral to cloaca pustular;
skin otherwise smooth; cloacal opening round and unornamented, directed
posteriorly at upper level of thigh.

Color in preservative: Dorsum, flanks, face, and dorsal surfaces of
limbs cream reticulated with black; venter gray; ventral surfaces of limbs
and posterior of belly creamy white with dark gray mottling; tips of fingers
and toes yellowish white; palmar and plantar tubercles pale gray.

Color in life: Dorsum yellow-orange with black spots and reticulations;
iris bronze with fine black reticulation.

Measurements of holotype (in mm): SVL: 76.7, TIBL: 31.9, FOOT:
35.0, HLEN: 23.0, HWID: 29.7. IOD: 3.8. IND: 6.0. ENOS: 4.5. EYE: 7.4.
HNDL: 21.3, THBL: 14.5, RDL: 20.5.

Distribution. — Known only from the type locality in the Cordillera
Occidental (Fig. 1 ).

Ecology. — All specimens have been taken in drainage ditches at night
inside the city of Cutervo. The vegetation in the immediate area consists of
remnant cloud forest along streams, in the Very Humid Montane Forest life
zone. Gastrotheca monticola and at least one undescribed Colostethus also
were found in this area.

Remarks. — A similar situation exists for the type material of
Telmatohius latirostris as was described for T. brevipes. Vellard (1951)
designated the holotype of T. latirostris as specimen #1 of the series
MHNJP (= MHNSM) 325. the largest male of a series of 13 specimens.
When I examined the type series of this species in 1989. I found 10
specimens in the series, all of which lacked individual tags. One of Vellard's
paratypes is now in the Museum National d'Histoire Naturelle (Lavilla,
1988a). thereby leaving two adult specimens unaccounted. Morales et al.
(1990) suggested that the holotype and allotype were the missing speci-
mens and listed the remaining 10 specimens as paratypes (new numbers
MHNSM 3729-38). I designate the largest male of this series (MHNSM
3733) as the lectotype.

60 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Although I have seen only specimens from Cutervo, John Cadle (pers.
comm.) has recently collected this species at other localities in
Departamento Cajamarca, and is currently describing the tadpole.

Telmatobius necopinus new species
Figure 9

•■£>*

Holotype. — KU 212482, an adult male from the east slope of the Abra
Pardo de Miguel, 39 km ENE Pomacochas (= Florida) on road to
Moyobamba (5°46' S, 77°42' W), 2050 m, Departamento Amazonas. Peru,
collected 31 January 1989, by John J. Wiens.

Paratypes. — KU 212483 and 212484 (cleared-and-stained skeleton)
collected with the holotype, and 212481 collected 1 February 1989, at the
same locality.

Diagnosis. — ( 1 ) Premaxillary teeth present; (2) tympanum absent; (3)
nuptial spines large, conical; nuptial pad on dorsal surface of thumb and
separate from inner palmar tubercle; (4) dorsum (in life) dull brown, with
yellowish-tan flecks in some; (5) venter olive-gray; (6) skin of dorsum
smooth; (7) snout-vent length in males to 71.4 mm, in females to 72.9 mm.

The large size, large nuptial spines, and dull brown dorsum distinguish
this species from all other Telmatobius in the region except for T. colanensis
and T. truebae. Telmatobius necopinus differs from T. colanensis in having
the eye not protuberant dorsally, less extensively webbed toes, smaller
tarsal tubercles, and an indistinct medial margin of the inner palmar tu-
bercle (in males). Telmatobius necopinus is difficult to distinguish from T.
truebae on the basis of external morphology, but differences include: ( 1 )
dorsum smooth in T. necopinus versus tuberculate in most (82%) T. truebae,

(2) dorsum with yellowish flecks in some (25%) T. necopinus, absent in T.
truebae, (3) yellow to lavender mottling on underside of limbs in most
(90%) T. truebae, absent or indistinct in T. necopinus, (4) nuptial spines
smaller in one of the two males of T necopinus, and (5) metatarsal tubercle
always small (about one-fourth size of inner metatarsal tubercle) in T.
necopinus, usually larger in T. truebae (greater than one-fourth inner
tubercle in about 70% of adult T. truebae). Osteologically, T necopinus is
distinct from T. truebae in having: (1) nasals more slender and distinctly
curved anteriorly at midlength, (2) zygomatic ramus of squamosal longer,

(3) median process of sphenethmoid extending between nasals, absent in T.
truebae, (4) median ramus of pterygoid bent posteriorly and expanded at
articulation with prootic, (5) posterior margin of coronoid process of
angulosplenial concave, convex in T. truebae, (6) tympanic annulus re-
duced to paired remnants, versus normal and unpaired in T. truebae, (7)
anteriorly oriented processes on hyale, absent or medially oriented in T.
truebae, (8) Presacral VIII and sacral vertebrae fused, unfused in T truebae.

SYSTEM ATICS OF THE FROG GENUS TELMATOBIUS IN PERU 61

and (9) clavicle less extensive on pars acromialis of scapula, lateral end of
clavicle extending to tip of pars acromialis in T. truebae. Telmatobius
necopinus and T. truebae also are distinct electrophoretically; they share no
alleles at five loci (Ada, Gpi. Me-1, Pep-B-2, Sdh) and parsimony analysis
of the allozyme data shows they are not sister taxa.

Description. — Head narrower than or as wide as body; head wider than
long (HLEN 73.4-78.7% HWID; x = 76.4 ± 2.7, n = 3); head length 29.0-
31.4% of SVL (x = 30.0 ± 1.3, n = 3): head width 38.4-39.9% of SVL (x
= 39.3 ± 0.8, n - 3); nostril not protuberant; located at anterior terminus of
snout; canthus rostralis distinct and slightly concave in dorsal view; el-
evated in lateral profile; loreal region concave; snout short, obtuse, bluntly
rounded in dorsal and lateral view; eye anterolaterally oriented, not protu-
berant on top of head; eye diameter 30.0-33.3%- of head length (x = 31.7 ±
1.7, n = 3); tympanum absent; tympanic annulus reduced but present;
supratympanic fold well developed. Maxillary and premaxillary teeth
fanglike, embedded in labial mucosa; dentigerous processes of vomers
slightly closer to small, semicircular choanae than to each other, located
medial or slightly posterior to choanae: processes perpendicular to longitu-
dinal axis of body with slight posteromedial orientation; each process
bearing 3-6 (x = 4.2, // = 6) fanglike teeth; teeth embedded in buccal lining
with only tips protruding. Tongue large, subcircular; attached anteriorly for
approximately four-fifths of its length; tongue free and unnotched (or
shallowly notched) posteriorly. Vocal slits absent.

Forelimbs moderately robust in males, relatively slender in females;
weak dermal wrist fold present; relative lengths of fingers: III > I = IV > II;
palmar webbing absent; lateral fringes present, extending to tips of fingers;
fringe on inner margin of thumb well developed in males; tips of fingers
spherical; inner palmar tubercle oval, flat, round, approximately equal in
size to outer palmar tubercle; inner margin of tubercle indistinct, continu-
ous with skin of palm; inner palmar tubercle separate from nuptial pad in
males; outer palmar tubercle oval, flat; subarticular tubercles small, round,
distinct, present proximally on Fingers I-IV; larger tubercle on Finger I;
distal subarticular tubercles on Fingers III and IV; one supernumerary
palmar tubercle at base of each finger; supernumerary palmar tubercle
adjacent to outer anterior margin of inner palmar tubercle small, flat;
nuptial spines moderately large (KU 212482) to large (KU 212481), kera-
tinous, conical, on pad on dorsomedial surface of thumb of males. Hind
limb length (foot plus tibia) 94.7-99.3% of SVL (x = 96.6 ± 2.4, // = 3);
relative lengths of toes: IV > V > III > II > I; webbing formula (range
followed by mode; n = 3): I (2-2; 2~)— (2]a-2]a\ 2[A) II {\l/i-\2A; \2a)—
(3-3+; 3+) III (2+-2^; 2lA)— (3lA-32A; Va) IV (3lA-32A: 3^)— (1-/3-2: 2) V;
webbing diminishing distally to form lateral fringes along edges of toes;
tips of toes spherical, approximately equal in size to fingertips; inner meta-

62 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

tarsal tubercle ovoid, flat; outer metatarsal tubercle small, round, approxi-
mately one-fourth size of inner metatarsal tubercle; subarticular tubercles
large, round, distributed on toes as follows: 1(1), 11(1), 111(2), IV(3), V(2);
plantar supernumerary tubercles absent; tarsal fold well developed, extending
approximately two-thirds length of tarsus, confluent distally with fringe of
webbing along inner margin of Toe I.

Skin smooth dorsally and ventrally; skin ventral to cloaca pustular;
cloaca! opening round and unornamented. directed posteriorly at upper
level of thigh; transverse fold of skin dorsal to cloacal opening (except KU
212482).

Color in preservative: Dorsum dull brown with numerous minute black
flecks and larger dark brown spots; venter pale olive-gray with minute dark
flecks (absent in KU 212481). ventral surfaces of limbs, hands, and feet
darker gray; tips of fingers and toes off-white; palmar and plantar tubercles
pale gray.

Color in life: (W. E. Duellman held notes for KU 212482) Dorsum dull
brown with yellowish-tan flecks; venter olive-gray; iris bronze with black
reticulations.

Measurements ofholotype (in mm): SVL 7 1 .4. TIBL 33.7. FOOT 34.7.
HLEN 20.7, HWID 28.2, IOD 5.1. IND 5.4, ENOS 4.6, EYE 6.2, HNDL
19.6, THBL 13.5, RDL21.6.

Distribution. — Telmatobius necopinus is known only from the eastern-
most ridge of the Cordillera Central (Fig. 1 ) at an altitude of 2050 m.

Ecology. — Three specimens were in a muddy roadside ditch at night;
the fourth specimen ( KU 2 1 248 1 ) was under a rock adjacent to a stream by
day. The type locality is situated in cloud forest, probably within the Very
Humid Lower Montane Forest zone. Other anurans collected at this site
include Scinax oreites, Eleutherodactylus bromeliaceus, three Eleuthero-
dactylus sp.. and two undescribed Colostethus. The stomach of KU 212484
contained an adult Colostethus sp., a large slug, a nematode, an opalione
arachnid, two beetles (Coleoptera: Chrysomelidae. Carabidae) and a beetle
larva (Hydrophilidae).

Etymology. — A Latin adjective meaning unexpected, referring both to
the surprise of finding a Telmatobius on the lower Andean slopes and to the
unanticipated distinctness of this species, biochemically and osteologi-
cally, from the geographically adjacent and externally similar T. truebae.

Telmatobius thompsoni new species

Holotype. — UF 39734, adult male, collected 13 km ESE Yamobamba,
18 km SE Agallpampa on road from Otuzco to Huamachuco (7°59' S,
78°36' W), Departamento La Libertad, Peru. 3290 m. on 9 May 1972 by
Fred G. Thompson.

SYSTEMATICS OF THE FROG GENUS TELMATOBWS IN PERU 63

Paratypes.— KU 218467. adult female, KU 218468, adult female (dry
skeleton), and KU 218469 (cleared-and-stained skeleton; juvenile), UF
39747, 39754, and 39756 (juveniles) same data as holotype.

Diagnosis. — ( 1 ) Premaxillary teeth present; (2) tympanum absent; (3)
nuptial spines small, distinctly conical, on dorsal and ventral surface of
thumb: nuptial pad continuous with inner palmar tubercle; (4) dorsum (in
preservative ) dull gray or brown with fine speckling of small dark spots; ( 5 )
venter dull yellow with small, dull gray or brown spots; (6) skin of dorsum
smooth; (7) snout-vent length of males to 68.9 mm. in females to 77.3 mm.

This species is similar to two other species in the region with small
nuptial spines. T. brevipes and T. ignavus. Telmatobius thompsoni is unique
among these three species in having slightly larger nuptial spines, more
slender palatines, and a poorly ossified sphenethmoid. Telmatobius
thompsoni can be further distinguished from T. brevipes in having less
extensive proliferation of nuptial spines on the ventral surface of the
thumb, a raised process on the anterior ramus of the pterygoid, more dorsal
orientation of the sacral diapophyses, and larger size. Telmatobius ignavus
has a tympanum, a small, round, orbitonasal foramen in the sphenethmoid
(vs. large and oblong in T. thompsoni), a plain tan to gray venter, and
usually has a short, bluntly rounded snout (long and sloping in T.
thompsoni). This species is also similar to the sympatric T. degener, but that
species is smaller and lacks premaxillary teeth; the latter character can be
assessed without dissection and is consistent in juveniles of T. thompsoni as
well as adults. Telmatobius degener also has larger nuptial spines and
differs in several other osteological characters.

Description. — Head equal to or slightly narrower than body; head
wider than long (HLEN 73.3-84.4% HWID; x = 78.9 ± 7.9, n = 2); head
length 28.7-33.1% of SVL (x = 30.9 ± 3.1, n = 2); head width 39.2% of
SVL (« = 2); nostril not protuberant, located at anterior terminus of snout;
canthus rostralis indistinct in dorsal view, not distinctly elevated in lateral
profile; loreal region barely concave, depressed; snout long and sloping in
dorsal and lateral profile; eye anterolaterally oriented; eye slightly protu-
berant on top of head; eye diameter 26.8-28.4% of head length (x = 27.6 ±
1.1. n = 2); tympanum absent; tympanic annulus present: supratympanic
fold moderately well developed, extending from posterior corner of eyelid
to anterior to insertion of forelimb; vertical fold of skin posterior to corner
of jaw, extending from supratympanic fold to throat. Maxillary and pre-
maxillary teeth fanglike, embedded in labial mucosa with only tips protrud-
ing; dentigerous processes of vomers closer to subcircular choanae than to
each other, located medial or slightly anterior to choanae; processes ori-
ented laterally, each process bearing 2-6 (x = 3.7. // = 6) fanglike teeth
embedded in buccal lining for most of their lengths. Tongue large,
subcircular, posterior border free and unnotched; attached anteriorly

64 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

through about three-fourths of its length. Vocal slits absent.

Forelimb moderately robust in males, less so in females; dermal wrist fold
present; relative lengths of fingers: III > IV > I > II; palmar webbing absent;
lateral fringes weak or absent; tips of fingers spherical; inner palmar tubercle
large, oval, extending onto Finger I; in male, tubercle continuous with nuptial
pad; outer palmar tubercle large, distinct, subcircular; subarticular tubercle,
large, round, present proximally on each finger, tubercle of Finger I very
large; distal subarticular tubercles present on Fingers III and IV; each finger
with at least one small supernumerary palmar tubercle proximally; larger
supernumerary tubercle between inner and outer palmar tubercles; nuptial
spines small, keratinous, distinctly conical, present on dorsal and ventral
surface of thumb. Hind limb length (foot plus tibia) 88.4-93.3% of SVL (x =
90.8 ± 3.5; n =2); length of toes in decreasing order: IV > III > V > II > I;
webbing formula (range followed by mode; // = 3): I 1 x/i — (2+-2'/i; 2+) II ( 1+-
IW, Va)—3 III (2-2+; 2) — (3+-3Vi; 3+) IV 3+— 1 V; webbing diminishing dis-
tally to form lateral fringes along edges of toes; tips of toes spherical, roughly
equal in size to fingertips; inner metatarsal tubercle large, elliptical, raised;
outer metatarsal tubercle ovoid, large, two-thirds size of inner; subarticular
tubercles well developed, round, distributed on toes as follows: 1(1), 11(1),
111(2), IV(3), V(2); supernumerary plantar tubercles proximally on Toes II— V
in some individuals, small, indistinct; tarsal fold weak, extending approxi-
mately two-thirds length of tarsus, confluent distally with fringe along inner
margin of Toe I.

Skin of dorsum and venter smooth; skin ventral to cloaca pustular,
forelimbs slightly pustular, skin on forelimbs and throat slightly creased;
cloacal opening round, unornamented, and directed posteriorly at upper
level of thigh.

Color in preservative: Dorsum, dorsal surfaces of limbs, ventral surface
of forearm and tarsus, palms, and soles dull gray or brown; dorsum with
fine speckling of small dark flecks; venter, throat and underside of limbs
dull yellow with small, dull, gray or brown spots (these spots obscure most
of the yellow ventral color in one individual, KU 218467); tips of fingers
and toes yellowish cream; tubercles on hands and feet pale gray. Color in
life unknown.

Measurements of holotype (in nun; measurements of adult female
paratype in parentheses): SVL 68.9 (77.3), TIBL 30.1 (33.7), FOOT 34.2
(34.6), HLEN 22.8 (22.2), HWID 27.0 (30.3), IOD4.5 (4.8), IND 4.7 (4.3),
ENOS 4.4 (4.8), EYE 6.1 (6.3), HNDL 16.8 (17.3), THBL 11.0 (11.7),
RDL 19.3 (20.0).

Distribution. — Known only from the type locality in the Cordillera
Occidental (Fig. 1 ).

Ecology. — According to field notes of F. G. Thompson, the type series
was collected in a densely vegetated roadside marsh in an "open pasture."

SYSTEMATICS OF THE FROG GENUS TELMATOBWS IN PERU 65

The marsh was fed by a brook draining into the Rio Moche, and was in the
Very Humid Subalpine Paramo life zone. The frogs were active and abun-
dant in pools and running water by day (13:30 h) and were found simulta-
neously with Telmatobius degener.

Etymology. — The specific name (a noun in the genitive case) is a
patronym for Fred G. Thompson of the University of Florida, who col-
lected the type series of this species and of T degener, and made several
other important collections of Telmatobius in northern Peru.

Telmatobius truebae new species
Figure 9

Holotype. — KU 181528, an adult male from Pomacochas (= Florida;
5°49' S, 77°55' W), 2180 m, Departamento Amazonas, Peru, one of a series
collected on 4 March 1979 by Thomas J. Berger, David C. Cannatella, and
William E. Duellman.

Paratypes — KU 181492-527, 181529-36 (181536 is a dry skeleton)
collected with holotype; 181855 (series of tadpoles) collected by W. E.
Duellman, same data as holotype; 212465-72, 212465-76, 212477-80
(cleared-and-stained skeletons) from the type locality collected 29-30
January 1989 and 14 February 1989 by Fernando M. Cuadros, William E.
Duellman, and John J. Wiens.

Referred Specimens. — See Appendix I.

Diagnosis. — (1) Premaxillary teeth present; (2) tympanum absent; (3)
nuptial spines large, conical, on dorsal surface of thumb; nuptial pad
separate from inner palmar tubercle; (4) dorsum (in life) dull brown, gray
or olive-green with or without darker spots; (5) venter gray, usually (90%)
with yellow to lavender mottling on underside of limbs; (6) skin of dorsum
usually (82%) with warts; (7) snout-vent length of males to 68.9 mm,
females to 82.0 mm.

The large nuptial spines, large size, and dull gray-brown dorsal color
easily distinguish T. truebae from all other species in the region except for
T. colanensis and T. necopinus. Telmatobius truebae can be distinguished
from T colanensis in having: (1) nuptial spines larger, (2) eye less protu-
berant on top of head, (3) snout usually longer, (3) dorsum usually (82%)
tuberculate, (4) toes less extensively webbed, (5) inner palmar tubercle
usually smaller, less distinct (in males), and (6) inner metatarsal tubercle
smaller, lower, more elliptical. From T. necopinus, T. truebae differs in the
following characters: (1) dorsum usually (82%) tuberculate (smooth in T.
necopinus), (2) dorsum lacking yellowish flecks (variable in T. necopinus),
(3) underside of limbs usually (90%) with bright mottling, (4) nuptial
spines always large (smaller in one of two males of T. necopinus), (5) outer
metatarsal tubercle usually larger (greater than one-fourth size inner meta-

66 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

tarsal tubercle), (6) nasal not slender and not distinctly curved anteriorly (at
midlength), (7) zygomatic ramus of squamosal shorter, (8) median process
of sphenethmoid absent, not extending between nasals, (9) median ramus
of pterygoid not bent posteriorly or expanded at articulation with prootic,
( 10) tympanic annulus not reduced to paired vestiges, (11) posterior margin
of coronoid process of angulosplenial convex, (12) process on hyale medi-
ally oriented or absent, (13) Presacral VIII and sacral vertebrae not fused,
(14) clavicle extending to dorsal tip of pars acromialis of scapula.
Telmatobius truebae also is distinct electrophoretically from T. necopinus;
they share no alleles at five loci (Ada, Gpi, Me-1, Pep-B-2. Sdh) and
parsimony analysis of the allozyme data shows they are not sister taxa.

Description. — Head as wide or narrower than body; head wider than
long (HLEN 72.5-85.8% of HWID; x = 79.3 ± 3.3. n = 36); head length
26.6-31.5% of SVL (x = 29.0 ± 1.1, n = 36); head width 34.4^0.0% of
SVL (x = 36.6 ± 1.3, n = 36); nostril on anterior terminus of snout, not
protuberant; canthus rostralis indistinct and concave in dorsal profile;
loreal region concave; snout moderately short and bluntly rounded in
dorsal view and lateral profile; eye anterolaterally oriented, not or only
slightly protruding on top of head; eye diameter 25.9-38.0% of head length
(x = 29.0 ± 2.3, n = 36); tympanum absent; tympanic annulus present;
supratympanic fold usually well developed, often strongly warty; vertical
fold posterior to jaw thick. Maxillary and premaxillary teeth fanglike,
embedded in labial mucosa; dentigerous processes of vomers closer to
small, subcircular choanae than to each other; processes medial or slightly
posterior to level of choanae; oriented perpendicular to longitudinal axis of
head (or with slight posteromedial inclination); each process bearing 0-5
(x = 2.6, // = 108) fanglike teeth; vomerine teeth absent on one process in
eight individuals, two individuals lacking vomerine teeth entirely (all 10
individuals lacking teeth are from type locality); teeth embedded in buccal
lining for most of their lengths. Tongue large, subcircular, attached anteri-
orly for three-fourths of its length; free and unnotched (or only shallowly
notched) posteriorly. Vocal slits absent.

Forelimbs robust in males, relatively slender in females; dermal wrist
fold usually present; relative lengths of fingers: III > I > IV > II; palmar
webbing absent; lateral fringes absent or very poorly developed; fringe on
outer margin of thumb in males well developed; tips of fingers spherical;
inner palmar tubercle oval, approximately equal in size to outer; outer
margin distinct; inner margin indistinct, continuous with skin of palm;
inner palmar tubercle separate from nuptial pad in males; outer palmar
tubercle large, flat, circular to ovoid; inner distal margin concave with
small supernumerary tubercle either adjacent to margin or fused to it in all
specimens from Pomacochas; subarticular tubercles small, round, present
proximally on Fingers II— IV; subarticular tubercle on Finger I approxi-

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 67

mately twice size of tubercles on other fingers; distal subarticular tubercles
small, round, present on Fingers III and IV; at least one supernumerary
tubercle present at base of each finger; supernumerary tubercle along outer
margin of inner palmar tubercle large; males with large, conical nuptial
spines on raised pad on dorsomedial surface of thumb. Hind limb length
(foot plus tibia) 80.1-99.9% of SVL (x = 90.6 ± 5.0. // = 36); relative
lengths of toes: IV > V > III > II > I; webbing formula (range followed by
mode; n = 42): I ( Va-2~; \2a)—(2+-22a\ 2'a) II ( l-l'/v Va)—(22a-3+:
3) III (l2A-2+; 2)— (3+-3'/i; 31*) IV (3+-3fc; 3Vs) — (l+-2; Va) V; webbing
diminishing distally to form lateral fringes along edges of toes; tips of toes
spherical, roughly equal in size to fingertips; inner metatarsal tubercle
small, oval, raised slightly; outer metatarsal round to oval, one-fourth to
two-thirds size of inner (usually one-third); subarticular tubercles round,
distinct, distributed on toes as follows: 1(1), 11(1), 111(2), IV(3). V(2);
plantar supernumerary tubercles usually absent; tubercles present in some
individuals between metatarsal tubercles and at bases of toes; tarsal fold
extending approximately one-half length of tarsus, confluent distally with
fringe along inner margin of Toe I.

Dorsum usually (82%) with warts; skin of venter, flanks, and limbs
smooth; skin ventral to cloaca pustular; forelimbs pustular in some males;
cloacal opening round and unornamented at mid-upper level of thigh;
transverse fold dorsal to cloaca in some individuals.

Color in preservative: Dorsum gray or brown (with darker spots in some
individuals) with similar color on limbs, palms, and soles; venter lighter
brown or gray with fine gray and white speckling in some individuals;
ventral surfaces of limbs usually (90%) with irregular yellow-white
blotches; tips of fingers and toes yellowish white; palmar and plantar
tubercles pale gray.

Color in life: (KU 181492-536 from Pomacochas, W. E. Duellman field
notes) Dorsum dull olive-green to dark gray with fainter dark spots by day;
at night, olive-green to tan with dark olive-green, brown, or black spots;
venter and posterior surfaces of thighs dark gray; throat with faint lavender
or olive suffusion in some; ventral surfaces of thighs (some) and/or shanks
(few) with pale pinkish-orange marks; iris dull bronze with black reticula-
tions. Dorsum dull olive-green to brown with or without darker spots; belly
gray; ventral surfaces of limbs mottled dull orange-yellow and black; iris
gold with black flecks (from W. E. Duellman field notes for KU 212458-61
from Abra Barro Negro).

Measurements ofholotype (in mm): SVL 68.1, TIBL 30.0, FOOT 30.6,
HLEN 20.0, HWID 25.6, IOD 4.8, IND 4.5, ENOS 5.1, EYE 5.7, HNDL
17.6. THBL 12.6, RDL 20.2.

Larvae. — The following description is based on an individual (one of
the series KU 181855; Fig. 16C) at Stage 35 from the type locality with the

68 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

following measurements (in mm): total length 75.2, body length 29.5, basal
tail muscle height 8.6, basal tail muscle width 6.6, maximum dorsal fin
height 5.2 located 27.5 from body terminus, maximum ventral fin height
5.0 located 28.6 from body terminus, body height 13.6, body width 17.4,
eye diameter 2.4, pupil diameter 0.8, interorbital distance 3.9, narial diam-
eter 0.6, internarial distance 4.9, snout-naris 5.2, naris-eye 3.2, transverse
oral disc diameter 6.6.

Body ovoid in dorsal and lateral view; snout bluntly rounded; naris
small, round anterolaterally oriented; eyes small, dorsolateral^ oriented;
spiracular tube lateral, relatively ventral in position, short; vent short,
median, unpigmented. Dorsal and ventral fins slightly narrower than depth
of caudal musculature at midlength; caudal fins not extending onto body;
tail gradually tapering to blunt terminus; caudal musculature very robust,
extending onto body dorsally. Mouth large, anteroventral; oral disc ovoid;
marginal papillae arranged in double row posteriorly, single row laterally
and anteriorly; some papillae darkly pigmented; anterior gap in marginal
papillae wide; oral disc with transangular margin; few, scattered
intramarginal lateral papillae in angular area; intramarginal mental papillae
absent; jaw sheaths keratinized, gently arced, with finely serrated margins;
dorsal sheath wider than ventral; labial tooth-row formula 2/3; tooth rows
adjacent to jaw sheaths widely interrupted medially.

Color in preservative: Dorsum of body brown; venter mostly
unpigmented; intestine visible through skin; caudal musculature with pig-
ment restricted to scattered brown blotches; dorsal fin with few blotches
posteriorly, ventral fin with many small blotches throughout its length.

Color in life: (W. E. Duellman field notes for KU 181855) Body and
caudal musculature mottled olive-brown with greenish-olive and olive-
black flecks; venter gray; venter lavender anteriorly; fins translucent tan
with dark brown flecks; iris pale yellowish bronze.

Variation: Individuals from near Chachapoyas (KU 212504) appear to
have longer tails but are otherwise similar to individuals in the series KU
181855. Two large tadpoles (KU 181856, Stage 38, total length 82.2 and
83.2) from the Abra Barro Negro may be assignable to this species (but T.
atahualpai also occurs at this locality). These specimens differ from KU
181856 and KU 212504 in having more heavily pigmented caudal muscu-
lature and a dark peritoneum covering the intestine.

Distribution. — Telmatobius truebae is known from several localities in
the Cordillera Oriental (Fig. 1) at elevations from 2150-3470 m (Fig. 2).

Ecology. — Specimens were under rocks in or along streams by day and
in drainage ditches by night (in Pomacochas). This species occurs in
Humid Lower Montane Forest, Very Humid Montane Forest, and Very
Humid Subalpine Paramo. Near the Abra Barro Negro, Telmatobius truebae
was collected with Gastrotheca monticola, Eleutherodactylus sp.

SYSTEMATICS OF THE FROG GENUS TELMATOB/US IN PERU 69

{unistrigatus group), Phyllonastes lynchi, Phrynopus sp., and T. atahualpai.
In Pomacochas, T. truebae was collected with Colostethus sp., Gastrotheca
monticola, and Scinax oreites. The stomachs of three individuals (KU
212464, 212477, and 212480) contained several beetles (Coleoptera:
Carabidae, Curculionidae, Staphylinidae), a leafhopper (Homoptera,
Cicadellidae), various flies (Diptera: Nematocera and Syrphidae), an ant
(Hymenoptera. Formicidae), lepidopterans (adult parts and larvae), two
roaches (Blatoidea), a spider, and several millipedes (Diploda:
Polydesmidae). Tadpoles were found in pools in streams in late January
and early March.

Etymology. — The specific name (a noun in the genitive case) is a
patronym for Linda Trueb, in recognition for her many contributions to
anuran morphology and systematics, especially of Telmatobius.

Acknowledgments: I am grateful to my field companions in the spring of
1989 and winter of 1991 — Fernando M. Cuadros V., William E. Duellman,
Michael E. Morrison, and Erik R. Wild — for braving the cold and murky
waters of the Peruvian Andes to collect Telmatobius and for stoically
enduring my company in the field. B. Anthony Luscombe and the
Asociacion de Ecologia y Conservacion (ECCO) provided invaluable lo-
gistic support in Peru. Collecting and export permits were issued by Ing.
Gonzalo Bravo Mejfa Munoz and Jose Purisaca of the Direccion General
Forestal y de Fauna, Ministerio de Agricultura, Lima, Peru. Field and
laboratory work were supported by a grant from the National Science
Foundation (BSR-8805920) to William E. Duellman. Additional support
was provided by a REU (Research Experience for Undergraduates) fellow-
ship from the National Science Foundation (under William E. Duellman),
two Undergraduate Research Awards from The University of Kansas Hon-
ors Program, and a NSF Graduate Fellowship. Electrophoresis was carried
out in the Biochemical Systematics Laboratory of the Museum of Natural
History, The University of Kansas, which is funded in part by NSF grant
BSR-8722562 to Douglas Siegel-Causey and Edward O. Wiley. I thank the
following individuals for loan of specimens, provision of working space,
and/or permission to skeletonize material: William E. Duellman and John
E. Simmons (KU). Douglas Rossman (LSUMZ), Jose Rosado (MCZ),
Nellie Carillo de Espinoza, Victor Morales, and Antonio Salas (MHNSM),
and David Auth (UF). I am grateful to Tom Titus for teaching me starch gel
electrophoresis, and to Paula Mabee, Andrew Simons, and Tom Titus for
invaluable discussions on phylogenetic analysis of allozyme data. William
E. Duellman and Fred G. Thompson allowed me to quote from their field
notes, Thomas S. Schulenberg provided valuable information on
Telmatobius colanensis, and Linda Trueb graciously granted me use of
unpublished notes from her preliminary study of Peruvian Telmatobius.

70 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Finally, I thank Les Marcus for providing the SAS code for the shearing
technique, Adrian Nieto Montes de Oca for correcting my translation of the
Spanish summary, Richard Leschen for identifying invertebrates removed
from stomachs, Luis A. Coloma and John E. Simmons for assistance with
the photographic prints, Antonio W. Salas and Linda Trueb for discussions
of Telmatobius systematics, and David C. Cannatella, Paul Chippindale,
William E. Duellman, David A. Kizirian, Linda Trueb, and an anonymous
reviewer for helpful criticism of the manuscript.

LITERATURE CITED

Altig, R., and G. F. Johnston. 1989. Guilds of anuran larvae: relationships among

developmental modes, morphologies, and habitats. Herpetol. Monogr. 3:81 —

109.
Barbour, T., and G. K. Noble. 1920. Some amphibians from northwestern Peru.

with a revision of the genera Phyllobates and Telmatobius. Bull. Mus. Comp.

Zool. Harvard 63:395-427.
Bookstein, F L., B. C. Chernoff, R. L. Elder, J. M. Humphries, G. R. Smith, and

R. E. Strauss. 1985. Morphometries in Evolutionary Biology. Academy of

Natural Sciences of Philadelphia Special Publ. 15:1-277.
Burnaby, T. P. 1966. Growth-invariant discriminant functions and generalized

distances. Biometrics 22:96-110.
Cannatella, D. C. 1985. The systematic status of Syrrophus juninensis (Anura:

Leptodactylidae). Proc. Biol. Soc. Washington 98:774-777.
Cei, J. M. 1980. Amphibians of Argentina. Monitore Zool. Ital. (NS) Monogr. 2.
Cei, J. M. 1986. Speciation and adaptive radiation in Andean Telmatobius frogs. Pp.

374-386 in Vuilleumier, F, and Monasterio, M. (eds.). High Altitude Tropical

Biogeography. New York: Oxford Univ. Press.
Cracraft, J. 1992. The species of the birds-of-paradise (Paradisaeidae): applying

the phylogenetic species concept to a complex pattern of diversification. Cladis-

tics 8:1-43.
Diaz, N. F. 1989. Phenetic and phylogenetic relationships of the Chilean Alsodes

and Telmatobius (Amphibia, Leptodactylidae) and proposal of a new genus.

Studies on Neotropical Fauna and Environment 24:25-33.
Dingerkus, G., and L. D. Uhler. 1977. Enzyme clearing of alcian blue stained

whole small vertebrates for demonstration of cartilage. Stain Technol. 52:229-

232.
Duellman, W. E. 1970. The Hylid Frogs of Middle America. Univ. Kansas Mus.

Nat. Hist. Monogr. 1:1-753.
Duellman, W. E. 1979. The herpetofauna of the Andes: patterns of distribution,

differentiation, and present communities. Pp. 37 1^459 /'// Duellman, W. E. (ed).

The South American Herpetofauna: Its Origin, Evolution, and Dispersal. Univ.

Kansas Mus. Nat. Hist. Monogr. 7:1—485.
Duellman, W. E., and L. Trueb. 1 986. Biology of Amphibians. New York: McGraw-
Hill Book Company.

SYSTEM ATICS OF THE FROG GENUS TELMATOBWS IN PERU 7 1

Duellman, W. E., and E. R. Wild. 1993. Anuran amphibians from the Cordillera de

Huancabamba, northern Peru: systematics, ecology, and biogeography. Univ.

Kansas Mus. Nat. Hist. Occ. Pap. 157:1-53.
Farris, J. S. 1970. Methods for computing Wagner trees. Syst. Zool. 19:83-92.
Frost, D. R. (Editor). 1985. Amphibian species of the World. A taxonomic and

geographic reference. Lawrence, Kansas: Allen Press and Association of Sys-
tematics Collections.
Frost, D. R., and D. M. Hillis. 1990. Species in concept and practice: herpetologi-

cal applications. Herpetologica 48:87-104.
Gosner, K. L. 1960. A simplified table for staging anuran embryos and larvae with

notes on identification. Herpetologica 16:183-190.
Harris, H., and D. H. Hopkinson. 1976. Handbook of Enzyme Electrophoresis in

Human Genetics. Amsterdam: North-Holland Publishing Co.
Hendy, M. D., and D. Penny. 1982. Branch and bound algorithms to determine

minimal evolutionary trees. Math. Biosc. 59:277-290.
Hillis, D. M. 1985. Evolutionary genetics of the Andean lizard genus Pholidobolus

(Sauria: Gymnophthalmidae): phylogeny, biogeography, and a comparison of

tree reconstruction techniques. Syst. Zool. 34:109-126.
Hillis, D. M., and C. Moritz. 1990. An overview of applications of molecular

systematics. Pp. 502-515 in Hillis, D. M. and Moritz, C. (eds), Molecular

Systematics. Sunderland, Massachusetts: Sinauer Associates.
Humphries, J. M., F. L. Bookstein, B. Chernoff, G. R. Smith, R. L. Elder, and S.

G. Poss. 1981. Multivariate discrimination by shape in relation to size. Syst.

Zool. 30:291-308.
International Union of Biochemistry Nomenclature Committee. 1984. Enzyme

Nomenclature, 1984. New York: Academic Press.
Laurent, R. E, and E. O. Lavilla. 1986. Redescripcion de Telmatobius hauthali

Koslowsky (Anura: Leptodactylidae) y descripcion de una nueva especies del

mismo genero. Cuad. Herpetol. 2:1-24.
Lavilla. E. O. 1988a. Telmatobius (Anura: Leptodactylidae): the name bearing

types of five Vellard's taxa. Alytes 7:6-18.
Lavilla, E. O. 1988b. Lower telmatobiinae (Anura: Leptodactylidae): generic

diagnoses based on larval characters. Univ. Kansas Mus. Nat. Hist. Occ. Pap.

124:1-19.
Lavilla, E. O., and R. F. Laurent. 1988a. Deux nouvelles especes du genre

Telmatobius (Anura: Leptodactylidae) en provenance de El Moreno (Provincia

de Jujuy, Argentine). Alytes 7:77-89.
Lavilla, E. O., and R. F. Laurent. 1988b. A new species of Telmatobius (Anura:

Leptodactylidae) from Catamarca (Argentina). Alytes 7:90-96.
Lynch, J. D. 1971. Evolutionary relationships, osteology, and zoogeography of

leptodactyloid frogs. Univ. Kansas Mus. Nat. Hist. Misc. Publ. 53:1-238.
Lynch, J. D. 1978. A re-assessment of the telmatobiine leptodactylid frogs of

Patagonia. Univ. Kansas Mus. Nat. Hist. Occ. Pap. 72:1-57.
Mabee, P. M., and J. Humphries. Coding polymorphic data: examples from

allozymes and ontogeny. Syst. Biol. 42:166-188.

72 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

McDiarmid, R. W., and R. Altig. 1990 ("1989"). Description of a bufonid and two

hylid tadpoles from western Ecuador. Alytes 8:51-60.
Morales, V. R. 1989. Una nueva especie de Telmatobius (Anura: Leptodactylidae)

de Ancash. Peru. Rev. Bras. Zool. 5:603-608.
Morales, V. R., N. Carrillo, and H. Ortega. 1990. El material tipo de peces,

anfibios y reptiles en el Museo de Historia Natural de la Universidad Nacional

Mayor San Marcos. Publ. Mus. Hist. Nat. Univ. Nac. Mayor de San Marcos (A)

33:1-7.
Myers, C. W., and W. E. Duellman. 1982. A new species of Hyla from Cerro

Colorado, and other tree frog records and geographical notes from western

Panama. Am. Mus. Novit. 2752:1-25.
Rohlf, F. J., and F. L. Bookstein. 1987. A comment on shearing as a method for

"size correction." Syst. Zool. 36:356-367.
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.
Selander, R. K., M. H. Smith, S. Y. Yang, W. E. Johnson, and J. B. Gentry. 1971.

Biochemical polymorphism and systematics in the genus Peromyscus. I. Varia-
tion in the old-field mouse (Peromyscus polionotus). Stud. Genet. VI. Univ.

Texas Publ. 7103:49-90.
Siciliano, M. J., and C. R. Shaw. 1976. Separation and localization of enzymes on

gels. Pp. 184-209 in Smith, I. (ed). Chromatographic and Electrophoretic

Techniques. 4th Ed. Vol. 2. London: William Heimermann Medical Books.
Simpson, G. G. 1961. Principles of Animal Taxonomy. New York: Columbia Univ.

Press.
Swofford, D. L. 1990. PAUP: Phylogenetic Analysis Using Parsimony, Version

3.0. Champaign: Illinois Natural History Survey.
Swofford, D. L.. and S. H. Berlocher. 1987. Inferring evolutionary trees from

gene frequency data under the principle of maximum parsimony. Syst. Zool.

36:293-325.
Swofford, D. L., and W. P. Maddison. 1987. Reconstructing ancestral character

states under Wagner parsimony. Math. Biosc. 87:199-229.
Swofford, D. L., andG. J. Olsen. 1990. Phylogeny reconstruction. Pp. 411-501 in

Hillis, D. M., and Moritz, C. (eds). Molecular Systematics. Sunderland, Massa-
chusetts: Sinauer Associates.
Tosi, J. A. 1960. Zonas de vida natural en el Peru. Bol. Tec. Inst. Interamericano

Cien. Agric. 5:1-271.
Trueb, L. 1973. Bones, frogs, and evolution. Pp. 65-132 in Vial, J. L. (ed),

Evolutionaiy Biology of the Anurans. Columbia: Univ. Missouri Press.
Trueb, L. 1979. Leptodactylid frogs of the genus Telmatobius in Ecuador with the

description of a new species. Copeia 1979:714-733.
Trueb, L. 1985. A summary of osteocranial development in anurans with notes on

the sequence of cranial ossification in Rhinophrynus dorsalis (Anura: Pipoidea:

Rhinophrynidae). S. African J. Sci. 81:181-185.
Trueb, L. 1993. Patterns of cranial diversity among the Lissamphibia. Pp. 255-343

in Hanken, J., and Hall, B.K. (eds.), The Vertebrate Shull. Vol. 2. Structural and

Systematic Diversity. Chicago: Univ. Chicago Press.

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 73

Vellard, J. 1951. Estudios sobre batracios andinos. I. — El grupo Telmatobius y
formas afines. Mem. Mus. Hist. Nat. "Javier Prado" 1:1-89.

Wens, J. J., and T. A. Titus. 1991. A phylogenetic analysis of Spea (Anura:
Pelobatidae). Herpetologica 47:21-28.

Wiley, E. O. 1978. The evolutionary species concept reconsidered. Syst. Zool.
27:17-26.

APPENDIX I
SPECIMENS EXAMINED

All localities are in Peru. Kilometer distances are by road. Abbreviations
are as follows: (AA) alizarin-alcian-stained skeletal specimen, (D) dry
skeletal specimen, (E) specimen examined electrophoretically, (T) series of
larvae.

Telmatobius atahualpai. — Amazonas: 41 km SW Leimebamba, 3600
m, UF 39715; 28.3 km SSW Leimebamba, 3450 m, KU 182084; N slope
Abra Barro Negro, 28 km SSW Leimebamba, 3470 m, KU 212485.

Telmatobius brevipes. — Cajamarca: 23 km SW Celendin, 3050 m, KU
181425-34, 181435 (AA), 181836-37, 181842-44 (T); Sendamal, 30 km
S Celendin on Cajamarca road, ca 2530 m, KU 196611, LSUMZ 32261-
62; 24 km N Cajamarca, UF 39709-14; S slope Abra Quilsh, 21 km NNW
Cajamarca, 3400 m, KU 212412-13 (E); S slope Abra Quilsh, 26 km
NNW Cajamarca, 3500 m, KU 212414; 4 km E Banos del Inca, 2900 m,
KU 212415-19, 212503 (T); 7 km S Cajamarca, 3010 m, KU 212420-22;
8 km S Cajamarca, 3050 m, KU 212423-24 (E), 212425-27; 3 km SW
Encanada, 2980 m, KU 2 1 2428-3 1 ; Abra Comulica, 20 km NE Encanada,
3420 m, KU 181486-87; S slope Abra Comulica, 20 km NE Encanada,
3420-3520 m, KU 181857 (T), 212432-34; 2 km E Celendin, 2650 m,
KU 212441-43; 9 km S Celendin, 2750 m, KU 212444 (E), 212445,

212446 (AA), 212525 (T); S slope Abra El Gavilan, 18.5 km N San Juan,
2920-3100 m,KU 181489(D), 181491; 12 km SSENamora, 2000 m, KU
2 1 2528 (T); 1 0 km SSE Cajabamba, 2900 m, KU 2 1 2439^10. La Libertad:
Huamachuco, 3350 m, KU 131684-89, 131690 (AA), 131691-94,
MHNSM 3739^19; 21 km SW Huamachuco, 3500 m, UF 39757-60,
39762; Laguna Sacsacocha, 12 km E Huamachuco, 3050 m, KU 131695;
19 km SSE Cajabamba, 2690 m, KU 212435, 212436-37 (E), 212438,

212447 (AA, E).

Telmatobius colanensis. — Amazonas: SE La Peca, (ca. 5°34'S,
78°19'W), Cordillera Colan, 2410 m, LSUMZ 37102.

Telmatobius degener. — La Libertad: 13 km ESE Yamobamba, 18 km
SE Agallpampa on road from Otuzco to Huamachuco, 3290 m, KU 2 1 8463-
64, 218465 (D), 218466 (AA), UF 39736, 39738^11, 39743-45, 39748-
49, 39752-53, 39755.

Telmatobius ignavus. — Piura: 29 km E Canchaque, 2530 m, UF 34089;

74 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

29 km E Canchaque. 1840 m, UF 52148; 16 km E Canchaque, 2042 m, UF
52 147; El Tambo, 3 1 .5 km E Canchaque, 2770 m, KU 1 8 1440 (D), 1 8 1441
(AA), 181442-43, 181845 (T), 219821; 12.7 km E El Tambo, 2820 m, KU
219822, 219823 (T), MHNSM 15466-67, 15468 (T); 9.5 km SW
Huancabamba, 2320 m, KU 181846 (T); 25.5 km SW Huancabamba, 3010
m, KU 181438-39; 31 km SW Huancabamba, 3080 m,KU 181847 (T);
Huancabamba, 2000 m, MCZ 4093.

Telmatobius latirostris. — Cajamarca: Cutervo, 2620 m, KU 212448,
212449-50 (E), 212451 (AA, E), MHNSM 3729-338.

Telmatobius necopinus. — Amazonas: E slope Abra Pardo de Miguel,
39 km ENE Pomacochas (Florida) on road to Moyobamba, 2050 m, KU
212481 (E), 212482, 212483 (E), 212484 (AA, E).

Telmatobius thompsoni. — La Libertad: 1 3 km ESE Yamobamba, 1 8 km
SE Agallpampa on road from Otuzco to Huamachuco, 3290 m, KU 2 1 8467,
218468 (D), 218469 (AA), UF 39734, 39747, 39754, 39756.

Telmatobius truebae. — Amazonas: E slope Abra Chanchillo, 42 km
ENE Balsas, 2870 m, KU 212454; 44 km ENE Balsas, 2940 m, KU 212464
(AA); 11 km SE Chachapoyas, 2840 m, KU 212463 (E), 212504 (T); N
slope Abra Barro Negro, 28 km SSW Leimebamba, 3470 m, KU 212455
(E), 212456 (E), 212457-61; 28.3 km SSW Leimebamba, 3450 m, KU
181485; 27 km SSW Leimebamba, 3440 m, KU 212462; Pomacochas (=
Florida), 2150 m, KU 181492-535, 181436 (D), 181855 (T), 212465-72,
212474-76 (E), 212477-80 (AA), MHNSM 12363-70.

APPENDIX II
OTHER TELMATOBIINES EXAMINED

All specimens are in The University of Kansas Museum of Natural
History (KU). Skeletal preparations stained for alizarin-red only (bone
only, cartilage not stained) are noted by AR; dry skeletal preparations are
noted by DS. For alcoholic specimens, only the males examined are listed.

Osteological. — Alsodes gargola 203317 (AR). Alsodes montanus
203350 (AR). Alsodes vanzolinii 162247 (DS). Atelognathus patagonicus
203332 (AR); 160467-68 (DS). Atelognathus reverbii 203342 (AR).
Batrachophrynus macrostomus KU 98127-28 (DS). Batrachyla leptopus
161448 (AR). Batrachyla taeniata 161489, 161496 (AR). Eupsophus
roseus 162137-38 (AR); 162204, 187646 (DS). Eupsophus vertebralis
162238 (DS). Hylorina sylvatica 161407-08 (DS). Insuetophrynus
acarpicus 161429-30. 203352 (AR). Lynchophrys brachydactylus 217152
(AR). Somuncuria somuncurensis 203348^19 (AR). Telmatobufo venustus
159811, 161439 (DS). Thoropa miliaris 92856, 186760 (DS).

Telmatobius arequipensis 164078 (DS). Telmatobius carrillae 214825
(DS). Telmatobius ceiorum 160848 (DS). Telmatobius cirrhacelis 165989

SYSTEMATICS OF THE FROG GENUS TELMATOBIUS IN PERU 75

(DS). Telmatobius culeus 135864 (DS). Telmatobius hintoni 160190-91
(DS). Telmatobius jelskii 164081, 181835 (DS). Telmatobius marmoratus

135903-04, 164079-80 (DS). Telmatobius niger 131796 (DS). Telmatobius
oxycephalus 128880 (DS). Telmatobius peruvianas 162113-14 (DS).
Telmatobius schreiteri 160885-86 (DS). Telmatobius simonsi 160139 (DS).
Telmatobius vellardi 120292 (DS).

External. — Alsodes monticola 160568. 160573. Alsodes vanzolinii
162207. 162222. Atelognathus patagonicus 160427-28, 160430.
Atelognathus praebasalticus 180992. Atelognathus salai 192117.
Atelognathus reverbii 180997. Batrachyla antarandica 161497-98,
161501. Batrachyla leptopus 161440, 161446. Caudiverbera caudiverbera
161411-12. Eupsophus roseus 160562, 160564-65. Eupsophus vertebralis
162241. Hylorina sylvatica 161399, 161405, 161410. Insuetophrynus
acarpicus 161413. Lynchophrys brachydactylus 217151, 217153.
Telmatobufo venustus 161438. Thoropa lutzi 92846—48. Thoropa miliaris
74203, 92854. Thoropa petropolitana 92858.

Telmatobius arequipensis 194900, 194903. Telmatobius atacamensis
179739-40. Telmatobius carrillae 214823. Telmatobius ceiorum 160846-
47. Telmatobius culeus 135869, 183138, 183143. Telmatobius halli 159840.
Telmatobius hauthali 182848—49. Telmatobius hintoni 160160. Telmatobius
jelskii 181460, 181463. Telmatobius laticeps 182837. Telmatobius
marmoratus 135893, 135895. 135898. Telmatobius niger 131782, 131785,
165976. Telmatobius oxycephalus 128875-76. Telmatobius peruvianas
162058, 162060. Telmatobius schreiteri 160856, 160861. 160883.
Telmatobius stephani 206702.

APPENDIX III
APOMORPHY LISTS

Character state optimizations and FREQPARS branch lengths are given
for each stem of the unrooted tree preferred by PAUP and FREQPARS
analyses (Fig. 17A). Allelic changes are superscripted (but left-to-right
sequence does not necessarily indicate primitive-to-derived transition);
ambiguously placed transformations are followed by the optimization rou-
tine that places them at that stem.

Stem 1.— PAUP: Pgmabua, Sdhbbbc. FREQPARS: 3.00.

Stem 2.— PAUP: Adheebe(DELTRAN),Me-2bb ab. FREQPARS: 2.50.

Stem 3.— PAUP: Adhdd"ee, G3pdhcc"bb, Hbdhbbcc, Icdh-2aabb. Ldh-2aabb,
Me-lbbaa(ACCTRAN), Me-2ccbb. Mpiccbb. FREQPARS: 16.00.

Stem 4.— PAUP: Me- laabb (DELTRAN), Pgmabbb, Sdhbbaa.
FREQPARS: 3.500.

Stem 5.— PAUP: G3pdhctac(ACCTRAN), Gpiaabb. FREQPARS: 2.66.

Stem 6.— PAUP: Mpicccd. FREQPARS: 0.800.

76 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP. No. 162

Telmatobius brevipes-1 — PAUP: Adhbeee (ACCTRAN), Pep-B-labaa,
Sodaa-ab FREQPARS: 0.50.

Telmatobius brevipes-2 — PAUP: Adheece, Hbdhccac, Mpibbab, Pep-B-
2aaad, Sodaaabc. FREQPARS: 2.00.

Telmatobius brevipes-3 —PAUP: Adhbeee (DELTRAN), Ldh-lccee,
Pep-B-laadd, Sdhcbcc. FREQPARS: 6.00.

Telmatobius brevipes-4— PAUP: Adhbeab (ACCTRAN), Adhee ab
(DELTRAN), Ldh-lcc ac. FREQPARS: 2.50.

Telmatobius latirostris— PAUP: Hbdhbb dd, Ldh-2aa "cc, Pep-B-laacc,
Pep-B-2aabb. FREQPARS: 8.00.

Telmatobius necopinus— PAUP: Adabbaa, Ldh-lcccd, Ldh-2aaad, Mdh-
2bb-ab Me-lbbcc, Pep-B-2aacc. FREQPARS: 5.16.

Telmatobius truebae-1— PAUP: G3pdhccac (DELTRAN), Tpiaaab.
FREQPARS: 2.00.

Telmatobius truebae-2— PAUP: G3pdhaccc (ACCTRAN), Me-2cccd.
FREQPARS: 1.20.

Telmatobius truebae-3— PAUP: Aataa ab, G3pdhcc ac (DELTRAN),
Ldh-lcccb, Mdh-2bbab, Pgmbbab. FREQPARS: 2.26.

2044 093 361 715