The Frog Genus Atelopus in Ecuador

(Anura: Bufonidae)

JAMES A. PETERS

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 145

SERIAL PUBLICATIONS OF T H E S M I T H S O N I A N INSTITUTION
The emphasis upon publications as a means of diffusing knowledge was expressed
by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti-
tution, Joseph Henry articulated a program that included the following statement:
"It is proposed to publish a series of reports, giving an account of the new discoveries
in science, and of the changes made from year to year in all branches of knowledge."
This keynote of basic research has been adhered to over the years in the issuance
of thousands of titles in serial publications under the Smithsonian imprint, com-
mencing with Smithsonian Contributions to Knowledge in 1848 and continuing with
the following active series:
Smithsonian Annals of Flight
Smithsonian Contributions to Anthropology
Smithsonian Contributions to Astrophysics
Smithsonian Contributions to Botany
Smithsonian Contributions to the Earth Sciences
Smithsonian Contributions to Paleobiology
Smithsonian Contributions to Zoology
Smithsonian Studies in History and Technology
In these series, the Institution publishes original articles and monographs dealing
with the research and collections of its several museums and offices and of professional
colleagues at other institutions of learning. These papers report newly acquired facts,
synoptic interpretations of data, or original theory in specialized fields. These pub-
lications are distributed by mailing lists to libraries, laboratories, and other interested
institutions and specialists throughout the world. Individual copies may be obtained
from the Smithsonian Institution Press as long as stocks are available.

S. DILLON RIPLEY
Secretary
Smithsonian Institution
SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 145

The Frog Genus Atelopus in Ecuador

(Anura: Bufonidae)

James A. Peters

SMITHSONIAN INSTITUTION PRESS

City of Washington
1973
 ABSTRACT
Peters, James A. The Frog Genus Atelopus in Ecuador. Smithsonian Con-
tributions to Zoology, number 145, 49 pages, 31 figures, 1973.—The genus
Atelopus is represented in Ecuador by 15 species, of which 6 are here described
as new taxa. A key for identification, descriptions, and figures are presented for
all species, and a detailed analysis of variation in proportions attempts to
quantify certain subjective characteristics used by earlier frog systematists.
The genus includes two nonphylogenetic groups of species, the first, with
slender bodies and long limbs, usually found in the lowlands and lower moun-
tain slopes, and the second, with stocky bodies and short limbs, usually found
at higher altitudes in the Andes. The species appear to be highly responsive to
environmental changes, and successful invasions of the mountains by the low-
land species are accompanied by differentiation, evolution, and speciation.
Invasions are likely to be along major river drainages. The invaders are ef-
fectively isolated from adjacent drainages by factors such as the highly dissected
terrain, rain shadow, barren areas, and so on. High altitude species are restricted
in lateral distribution, but often have considerable altitudinal range, while low
altitude species have wide latitudinal range, but are restricted altitudinally.
The mountains in Ecuador are "continental islands," showing diversity and
adaptive radiation similar to that of the Hawaiian Islands, with the major dif-
ference that the Andes are not dependent upon "waifs" or "drifters" for invaders,
but are constantly supplied by the abundant life of the surrounding tropical
forests.

OFFICIAL PUBLICATION DATF is handstampcd in a limited number of initial copies and is recorded
in the Institution's annual report, Smithsonian Year. SI PRESS NUMBER 4775. SERIES COVER
DESIGN: The coral Montastrea cavernosa (Linnaeus).

Library of Congress Cataloging in Publication Data

Peters,' James Arthur, 1922-1972
The frog genus Atelopus in Ecuador.
(Smithsonian contributions to zoology, no. 145)
Bibliography: p.
1. Atelopus. 2. Frogs—Ecuador. I. Title. II. Series: Smithsonian Institution. Smithsonian
contributions to zoology, no. H.">.
QL1.S54 no. 145 [QI.668.E2] 591'.08s [597'.8] 72-13540

For sale by the Superintendent of Documents,

U.S. Government Printing Office, Washington, D.C. 20402
Price 85 cents, domestic postpaid; 60 cents, GPO Bookstore
 Contents

Page
Introduction 1
Acknowledgments 1
Species Groups within the Genus 2
Comparative Measurements 2
Ratio 3
Knee to Knee/Standard Distance 3
Tibiofibula/Standard Distance 4
Head Length/Standard Distance 5
Head Width/Head Length 6
Tibiofibula/Head Length 7
Tibiofibula/Knee to Knee 7
Sexual Dimorphism 8
Key to the Atelopus of Ecuador 9
Atelopus arthuri, new species 10
Atelopus balios, new species . 12
Atelopus bomolochos, new species 14
Atelopus boulengeri Peracca 16
Atelopus bufoniformis Peracca 18
Atelopus elegans (Boulenger) 20
Atelopus halihelos, new species 22
Atelopus ignescens (Cornalia) 24
Atelopus longirostris Cope 27
Atelopus mindoensis, new species 30
Atelopus nepiozomus, new species 32
Atelopus pachydermus (Schmidt) 34
Atelopus palmatus Andersson 38
Atelopus planispinus Espada 39
Atelopus pulcher pulcher (Boulenger) 41
Present Position of Taxa Formerly Assigned to Atelopus in Ecuador 43
Incertae Sedis . 43
Atelopus longirostris marmorata Werner 43
Zoogeography and Evolution 44
Literature Cited 48

in
 The Frog Genus Atelopus in Ecuador
(Anura: Bufonidae)

James A. Peters

Introduction have helped me to amass the material on which

this study is primarily based, and I wish particu-
It has been my good fortune to be able to spend larly to thank Manuel Olalla, Peter Spoecker, Rob-
considerable time in the field in Ecuador, where I ert Mullen, and Gonzalo Herrera, who have helped
collected representatives of most of the species of in the field. Gustavo OrceW. continues as an
Atelopus now known from that country. There invaluable friend and critic, and his collections,
are three new species in the material I obtained, now at the National Museum of Natural History,
and three more new species have been identified in Smithsonian Institution, were extremely useful in
the collections of other museums. In addition, the this study. The curators at the following institu-
taxonomic position of several of the previously tions have loaned specimens to me and have pro-
described forms has been revised as a consequence vided me with space to work during visits to their
both of the new material available and of reexam- institutions, and I appreciate both kindnesses (ab-
ination of the type specimens or series. The prep- breviations used in the text follow the name):
aration of this paper has been stimulated by the American Museum of Natural History (AMNH);
recent work on the genus Atelopus by Juan Rivero. British Museum (Natural History) (BMNH);
My collections strongly supplement his material, California Academy of Sciences (CAS, CAS-SU
and I have found his discussions of the Ecuadorian refers to numbers assigned to specimens from the
species cogent and stimulating, even with regard collections of Stanford University, now housed at
to those points on which we are not in agreement. the Academy); Field Museum of Natural History
The synonymies presented for each species are (FMNH); Museo ed Instituto di Zoologia Systema-
limited to only those references pertaining to Ecua- tica, Universita di Torino, Italy (UT); Museo Na-
dorian records and specimens, except where a cional de Ciencias Naturales, Madrid, Spain (MM);
change in status of the taxon is involved. I have Museum of Comparative Zoology, Harvard Uni-
made no attempt to document use of a taxonomic versity (MCZ); Museum National d'Histoire Na-
name in an inclusive manner. turelle, Paris, France (PM); University of Kansas
ACKNOWLEDGMENTS.—My field work in Ecuador Museum of Natural History (UKMNH); Univer-
was supported in 1954 by the American Philosophi- sity of Michigan Museum of Zoology (UMMZ);
cal Society and Brown University, in 1962 by Grant and University of Puerto Rico, Mayaguez
No. G21010 of the National Science Foundation, (UPR-M). The abbreviation used for the collec-
and in 1965-1966 by the Smithsonian Institution. tions in the National Museum of Natural History
I am indebted to each for this aid. Many people is USNM (United States National Museum). Un-
cataloged specimens from the James A. Peters
(JAP) and the Gustavo OrceW. (GOV) collections
James A. Peters, Department of Vertebrate Zoology, National
Museum of Natural History, Smithsonian Institution, Wash- are deposited at the National Museum of Natural
ington, D.C. 20560. {Deceased 18 December 1972.) History.
1
 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

SPECIES GROUPS WITHIN THE GENUS.—Although the long-legged and the short-legged species, and
one should not attempt the definition of species pointed out several anatomical characteristics
groups unless his study encompasses the entire which differentiate the two groups. Unfortunately,
range of variation in the genus, it has been most his samples were not large enough to permit full
convenient for me to recognize that there are two review of all the Ecuadorian species. Where he
different groups of species in the Ecuadorian Ate- had material it was usually only a single specimen,
lopus. I am not here according these groups any and he had no material at all for several of the
kind of taxonomic status, and I regard them as taxa.
phenetic rather than phylogenetic, so I would not COMPARATIVE MEASUREMENTS.—All measure-
like to have any taxonomic or phylogenetic inter- ments used in this study were made with a Helios
pretation placed on my treatment of them. But vernier caliper, measuring to the nearest tenth of
the groups do provide an evident basis for discus- a millimeter. The "standard distance" is the
sion and organization of data, as well as a primary straight-line distance between the tip of the snout
step in definition of the species and in the prepara- and the posterior end of the coccyx. Although
tion of a key for identification. many previous authors have defined the body
The longirostris group includes the slim-bodied, length as from the snout to the anus, preservation
long-legged taxa, while the ignescens group in- techniques have an effect on the position of the
cludes the taxa with short, heavy-set bodies and anus, and the tissues in that region can be easily
short limbs. Brighter colors and flash patterns tend compressed, so the measurement is more variable
to occur more commonly in the longirostris group, and not as accurate or consistent as the one defined
while the members of the ignescens group tend here. All other measurements are taken from a
to have darker, more sombre colors. Coloration is bony surface to another bony surface, insofar as
not invariable, however, and is not a good basis possible. The measurement from knee to knee is
for group recognition. The longirostris group is made with the femora at right angles to the body,
found at lower altitudes on both sides of the forming a straight line. The calipers are set solidly
Andes in Ecuador, while most of the taxa in the against the distal ends of the femora to record the
ignescens group are high altitude forms. measurement. The tibiofibula measurement is
The species in the ignescens group include from end to end of the bone. The head width is
nrthuri, bomolochos, bufoniformis, halihelos, ig- taken at the widest point of the head, with the
nescens, nepiozomus, and pachydermus. The mem- points of the caliper set to pass snugly over the
bers of the longirostris group are balios, elegans, head. The head length is from the snout tip to
longirostris, palmatus, planispinus, and pulcher. the posteriormost margin of the skull at the dorsal
The largest species of all, bonlengeri, and one of midline.
the smallest, mindoensis, are intermediate between As mentioned elsewhere, the two species groups
the two groups in many respects, and are difficult are clearly distinctive in their general appearance.
to assign in any unequivocal way. This intermedi- The differences are well reflected by calculations of
acy is referred to repeatedly in the subsequent proportional values, using the various measure-
sections, and reinforces my statement above that ments discussed above. In some proportions there
these are not taxonomic groupings, but simply is a considerable sexual dimorphism, and the sexes
represent convenient groupings. The species are are separated in all calculations, even where the
arranged on the graphs in the following sections differences were not significant.
by groups, with the two "intermediate" species be- The figures show the mean as a vertical line, the
tween them. total range of variation as the largest rectangle, the
While groups were not emphasized as such by mean it twice the standard error of the mean as
McDiarmid (1971) in his recent review of the the smallest rectangle, and the mean ± one stand-
genus, and he repeatedly pointed out the uni- ard deviation as the third rectangle. Only the
formity of characters within Atelopus in contrast mean and range are given for very small samples
to its differences when compared with other genera, (fewer than 6). The colors of the rectangles are
lie did recognize the existence of the two groups reversed from one sex to the other to permit rapid
discussed here. He referred in several places to recognition. The value found as a proportion is
NUMBER 145

always plotted in these figures as a percentage, and Ratio

it should be clear that all references to ratios, pro-
portions, and percentages are to these figures. KNEE TO KNEE/STANDARD DISTANCE.—In this ra-
All references in this paper to significant differ- tio, seen in Figures 1 and 2, higher values indicate
ences between the means of any two samples is a longer leg in proportion to body length, or stand-
based on non-overlap between the rectangles rep- ard distance. Although the two species groups are
resenting the mean ± twice its standard error. quite distinct from each other both in body shape
This spread of values includes slightly more than and leg length, as pointed out elsewhere, this is
95 percent of the possible values of the mean not strongly reflected by proportional differences.
itself, and nonoverlap between the two populations This is because both body and legs are elongated
indicates a very high likelihood that the two sam- in the longirostris group, while in the ignescens
ples as drawn are different. group both body and legs are short and stubby.

IGNESCENS
BOMOLOCHOS
ARTHURI
HALIHELOS
PACHYDERMUS
BOULENGERI
MINDOENSIS
PULCHER
BALIOS
PALMATUS
LONGIROSTRIS
ELEGANS

75 80 85 90 95 100 105
FIGURE 1 Species comparison, males, for the ratio knee to knee/standard distance x 100.
Numbers in this and Figure 2 refer to sample size, which are same for other figures.

BUFONIFORMIS
IGNESCENS
BOMOLOCHOS
ARTHURI
PACHYDERMUS
NEPIOZOMUS
BOULENGERI
MINDOENSIS
PULCHER
BALIOS
PALMATUS
PLANISPINUS
LONGIROSTRIS
ELEGANS
65 70 75 80 85 90 95 100 105
FIGURE 2. Species comparison, females, for the ratio knee to knee/standard distance x 100.
 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Even so, the longirostris group is characterized by the range of variation of the sexes. The males have
values greater than 89 for the proportion, while longer legs in proportion to the body length than
individual proportions in the ignescens group are the females in every species.
almost always below 90. A. pachydermus males, TIBIOFIBULA/STANDARD DISTANCE.—This charac-
of the ignescens group, range up to values of 94, ter is similar to the previous one (KK/SD) in that
and A. balios, of the longirostris group, has values it is an expression of a primary difference between
the two groups of species, but it expresses the con-
for most of the females and many of the males be-
trast much more strongly (Figures 3 and 4). Al-
low a ratio of 90. A. mindoensis and A. boulengeri
though the body and femora have changed in such
are fairly typical of the ignescens group, with only a fashion that the proportion between them is simi-
a few males falling above a value of 90. lar in the two groups, the region of the tibiofibula
Sexual dimorphism is fairly strong in this char- is much more elongated, proportionately, than the
acteristic. There is a significant difference between femoral region in the longirostris group, and this
the means for the sexes in every species except A. has been reflected in this ratio.
elegans, when samples are large enough for com- Individuals in the ignescens group are charac-
parison, although there is considerable overlap in terized by values below 45 for this ratio, while

IGNESCENS
B0M0L0CH0S
ARTHURI
HALIHELOS
PACHYDERMUS
BOULENGERI
MINDOENSIS
PULCHER
BALIOS
PALMATUS
LONGIROSTRIS
ELEGANS
— • • ••• --•• • • » — • —i » > — i i • • • i i i i i \ •

35 40 45 50 55
FICURE 3.—Species comparison, males, for the ratio tibiofibula/standard distance x 100.

BUFONIFORMIS
IGNESCENS
B0M0L0CH0S
ARTHURI
PACHYDERMUS
NEPIOZOMUS
BOULENGERI
MINDOENSIS
PULCHER
BALIOS
PLANISPINUS
PALMATUS
LONGIROSTRIS
ELEGANS
30 35 40 45 50 55
FIGUKE 4.—Species comparison, females, for the ratio tibiofibula/standard distance x 100-
NUMBER 145

those in the longirostris group almost invariably group, for in no case are the means of the two
have values above 45. In the latter group only sexes significantly different. In this regard, A. min-
males and females of A. pulcher give lower figures, doensis and A. boulengeri show similarities to the
while the remaining species have practically no ignescens group, since the means for the sexes are
individuals with values below 46. In the ignescens significantly different in both species.
group males of A. pachydermus and A. halihelos HEAD LENGTH/STANDARD DISTANCE.—This ratio
include individuals with values above 45, none of (Figures 5 and 6) produces differences between
which reach 46, and females of A. nepiozomus may species within the two major groups, rather than
have values up to 47. No males of A. nepiozomus
have yet been taken, but it might be predicted
that they would give higher values, as well. IGNESCENS
The two species A. mindoensis and A. boulengeri BOMOLOCHOS
are completely intermediate in this character. Most ARTHURI
of the females of both species lie below 46, most HALIHELOS
of the males of both fall above 45. Within the PACHYDERMUS
sexes, the mean for each of these species is signifi- BOULENGERI
cantly higher than that of all others within the MINDOENSIS
ignescens group, except for females of A. nepiozo- PULCHER
mus, but the same is true of a comparison with the BALIOS

members of the longirostris group, where the values PALMATUS

for these two species are significantly lower. LONGIROSTRIS
ELEGANS
Sexual dimorphism is apparent in all species of
the ignescens group for this character, with the 26 28 30 32 34 36 38 40
males having means significantly higher than the FIGURE 5 Species comparison, males, for the ratio head
females. The opposite is true of the longirostris length/standard distance x 100.

BUFONIFORMIS
IGNESCENS
ARTHURI
BOMOLOCHOS
PACHYDERMUS
NEPIOZOMUS
BOULENGERI
MINDOENSIS
PULCHER
BALIOS
PALMATUS
PLANISPINUS
LONGIROSTRIS
ELEGANS

26 28 30 32 34 36 38 40

FIGURE 6 Species comparison, females, for the ratio head length/standard

distance X 100-
 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

following group lines, and thus provides a basis ferences within A. boulengeri or A. pachydermus,
for separation of species within groups. Thus, all and the means for the sexes in A. ignescens are
individuals of A. balios and A. elegans, both of the very narrowly different at the 95 percent level. In
longirostris group, fall below a proportional value the longirostris group, significant sexual differences
of 33, as do the following species from the ignes- in this character exist in A. longirostris at a high
cens group: all females of A. bomolochos and A. level of significance, and in A. pulcher with only
ignescens and all males of boulengeri. In the marginal significance, but not at all in A. elegans.
longirostris group this is also true of most individ- It appears likely that sexual dimorphism exists in
uals of both sexes of A. pulcher as well as females A. balios, and perhaps not in A. palmatus, but the
of A. planispinis and A. longirostris, and in the samples are insufficient to be certain of either.
ignescens group, most of the males of A. ignescens HEAD WIDTH/HEAD LENGTH.—As seen in Figures
as well as the females of A. arthuri, A. boulengeri, 7 and 8, the members of the ignescens group are
and A. mindoensis. On the other hand, values of characterized by a broad head in relation to its
greater than 33 are found, in the ignescens group, length (measured from the tip of the snout to the
in all A. bufoniformis as well as the males of A. posteriormost point on the dorsal midline). This
arthuri, A. mindoensis, A. halihelos, and A. bomo- is most exaggerated in A. ignescens itself. The
lochos, while, in the longirostris group, values distinction between the two groups is not sharp,
above 33 are characteristic of both sexes of A. pal- and several species show an intermediate condition.
matus, and the females of A. longirostris. A pachy- Within the ignescens group, A. bufoniformis, A.
dermus, of the ignescens group, is completely ignescens, A. arthuri, A. pachydermus, and A. bou-
intermediate in this character, with as many in- lengeri are quite distinct, with all the individuals
dividuals in both sexes with values above 33 as showing a ratio of 95 or more. Also, within the
those below. The mean for females in this species longirostris group, A. planispinus, A palmatus, and
is 31.7, for males it is 33.2. A. longirostris are equally distinct, with ratios gen-
It is apparent from the figures above that sexual erally below 92. Three species in the ignescens
dimorphism occurs in some of the species, but group also show such low values, however, includ-
again it is a mixture between groups. In the ignes- ing A. bomolochos, A. halihelos, and A. nepiozo-
cens group significant differences exist between the mus, with the means for the last two significantly
means for the sexes in A. bomolochos, A. arthuri, different from all other species in the group, and
and A. mindoensis, while there are no sexual dif- with the latter two having mean values below 90.

IGNESCENS
BOMOLOCHOS
ARTHURI
HALIHELOS
PACHYDERMUS
BOULENGERI
MINDOENSIS
PULCHER
BALIOS
PALMATUS
LONGIROSTRIS
ELEGANS

115

FIGURE 7.—Species comparison, males, for the ratio head width/head length x 100.
NUMBER 145

BUFONIFORMIS BUFONIFORMIS
IGNESCENS
IGNESCENS
BOMOLOCHOS
BOMOLOCHOS ARTHURI
ARTHURI PACHYDERMUS
PACHYDERMUS NEPIOZOMUS
NEPIOZOMUS BOULENGERI
BOULENGERI MINDOENSIS
PULCHER
MINDOENSIS
BALIOS
PULCHER PLANISPINUS
BALIOS PALMATUS
PLANISPINUS LONGIROSTRIS
PALMATUS ELEGANS
LONGIROSTRIS 80 100 120 140 160 180
FIGURE 10.—Species comparison, females, for the ratio tibio-
ELEGANS fibula/head length x 100.
70 80 90
FIGURE 8.—Species comparison, females, for the ratio head those of the longirostris group are generally above
width/head length x 100. 135. A. bufoniformis is very distinct, overlapping
only one other species in its total range of varia-
In the longirostris group, A. pulcher, A. balios, tion.
and A. elegans all have higher ratios, indicating In those species for which there are sufficient
proportionately broader heads. individuals of both sexes for comparison, there is
There are no species with differences between no sexual dimorphism in the ignescens group, with
the means for the sexes that are statistically sig- the exception of A. mindoensis. In this regard A.
nificant, i.e., no sexual dimorphism exists in this mindoensis is more like the species of the longiros-
character. tris group, where a significant difference between
TIBIOFIBULA/HEAD LENGTH.—This ratio, as the means for the two sexes occurs in all species
shown in Figures 9 and 10, demonstrates the dichot- except A. elegans.
omy of the two groups, and the intermediacy of TIBIOFIBULA/KNEE TO KNEE.—This ratio (Fig-
A. mindoensis. Members of the ignescens group ures 11 and 12) is an attempt to express quantita-
generally have proportional values below 135, and tively a relationship often used in the earlier litera-

IGNESCENS
BOMOLOCHOS ELEGANS
ARTHURI LONGIROSTRIS
HALIHELOS PALMATUS
BALIOS
PACHYDERMUS
PULCHER
BOULENGERI
MINDOENSIS
MINDOENSIS BOULENGERI
PULCHER PACHYOERMUS
BALIOS HALIHELOS
PALMATUS ARTHURI
LONGIROSTRIS BOMOLOCHOS

ELEGANS IGNESCENS

100 ifc
120 140 160 44 46 48 50 52 54 56
FIGURE 9 Species comparison, males, for the ratio tibio FIGURE 11 Species comparison, males, for the ratio tibio-
fibula/head length x 100. fibula/knee to knee x 100.
 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

ELEGANS
LONGIROSTRIS
PALMATUS
PLANISPINUS
BALIOS
PULCHER
MINDOENSIS
BOULENGERI
NEPIOZOMUS
PACHYDERMUS
BOMOLOCHOS
ARTHURI
IGNESCENS
BUFONIFORMIS

42 44 46 48 50 52 54 56 58
FIGURE 12. Species comparison, females, for the ratio tibiofibula/knee to knee
X 100.

ture on anuran taxonomy. The lower limb is muscular, while in the longirostris group the leg
placed parallel to the femur, which is extended at is long, lean, and slender.
right angles to the body, and the relationship of
the heels in this position is usually stated as over- Sexual Dimorphism
lapping, touching, or separated. Quantitatively, if
the heels overlap, the value of this ratio is usually In addition to the several characteristics dis-
greater than 51, while if the heels are separated, it cussed under the heading of comparative measure-
is smaller than about 50. Most of the members of ments, there are other cases of distinctive sexual
the ignescens group give values below 50, while the dimorphism in Ecuadorian Atelopus. The most
species in the longirostris group usually show ra- conspicuous sexual difference is associated with the
tios above 51. A. bufoniformis shows the lowest act of clasping, which in this genus is a matter
values of the genus in Ecuador. A. mindoensis di- of some magnitude, in light of the length of time
verges sharply from the ignescens group in this that the clasp is maintained. Males of most species
characteristic, with all the males having ratios have a cornified edge on the first digit of the fore-
higher than 51, and only 2 of the 16 females with foot, and in several species this cornified area is
ratios below 51. A. boulengeri is intermediate, with expanded well onto the second digit and adjacent
values ranging from 48 to 52 in both sexes, al- areas of the palmar surface. This horny area per-
though in every case in this species the heels do mits external sex determination in almost every
not overlap, as a consequence of the method of case, since it is never present in females. The
measurement. amount of cornified tissue is considerably greater
Although the two principal groups are distin- than in most anuran species, and is presumably
guishable, and the means for this character are very useful to the male in the maintenance of his
significantly different between species of the groups, tenacious grip on the female while awaiting the
the ranges of variation often overlap extensively, deposition of the eggs. Since amplexing pairs may
and this is not a good character for population and remain that way for as long as two weeks (see "Re-
species differentiation. It appears that the propor- marks" under A. ignescens), a good grip is rather
tional relationships between these two parts of the important.
hind leg remain rather constant, even though in Rivero (1963:109) has remarked, "Spiny or tu-
the ignescens group the leg is short, stubby, and bercular skin is apparently not a sexual character
NUMBER 145

in Atelopus, although in tubercular species, the each case it is the female that is more heavily warty
male is usually more distinctly so than the fe- or densely spinose, quite the reverse of the situation
male." This has not been true in the species I in related amphibians, including other genera in
have examined, for all species with tubercles show
considerable sexual dimorphism in the number, the Bufonidae. In some Ecuadorian species, the
distribution, and prominence of warts, tubercles, males are almost smooth on the sides of the body
and particularly in spiny points, single or several, and the temporal region, while the females have
on the tops of the warts or glandular areas. In prominent spines in the same area.

Key to the Atelopus of Ecuador

1. Legs long, thin; body usually slender; heels overlap or touch when tibiofibulae are parallel
to femora at right angles to body; ratio tibiofibula/SD usually more than 45 2
Legs short, stout; body stout; heels do not touch when tibiofibulae are parallel to femora
at right angles to body; ratio tibiofibula/SD usually below 45 9
2. First finger almost entirely hidden in webbing; tip may be free 3
First finger distinct; webbing somewhat incised between first and second fingers 6
3. Belly unicolor, without spotting 4
Belly with clear-cut, marked, round black spots pulcher
4. Dorsum whitish with irregularly parallel, broad, black or brown stripes elegans
Dorsum not colored as above 5
5. Dorsum with broad band of light yellowish brown, filled with reddish brown marbling
and spotting ndndoensis
Dorsum chocolate brown, dorsolateral streak and dorsal spotting of light gray, bordered by
narrow brown or black edge palmatus
6. Limbs smooth 7
Limbs not smooth; may be shagreened, pustulate or warty 8
7. Ground color of dorsum of body light tan, with many small, rounded, regular darker
brown spots balios
Ground color of dorsum dark brown, with small, light yellowish spots often bilaterally
symmetrical longirostris
8. Entire body and limbs light brown, with small, scattered, dark spots above and below
halihelos
Dorsum light tan, vermiculated with light to dark brown, venter yellowish white, no
spotting ptanispimu
9. First finger distinct, free at least in part from webbing 10
First finger indistinct, entirely enclosed in webbing bufoniformis
10. Areas of spinules, warts, or pustules on some parts of head and/or body, usually on sides
and temporal region 11
Skin of head and body entirely smooth, no pustules, warts, or spinules boutengeri
11. Dorsum not unicolor, or, if so, not black 12
Dorsum unicolor, usually black, color extending well down sides and onto venter ..ignescens
12. Dorsum with strongly contrasting light (white or yellow) and dark (usually black) areas,
legs dark brown to black, usually with yellow spotting pachydermus
Dorsum not as described 13
13. Middorsum yellowish or greenish, with few scattered dark spots or uniform; no pustules
or warts on humerus, tibiofibula, or tarsometatarsus; dorsum of body usually smooth,
but few individuals have low warts posteriorly 14
Color not as described; all parts of limbs usually with warts, pustules, or spinules; dorsum
may or may not be warty 15
14. Humerus smooth, no pustules or glandular areas; ratio head length/SD x 100 greater
than 33 bomolochos
Humerus pustulate in females, with small rounded glands in males; ratio HL/SD X 1 0 °
smaller than 33 ignescens
15. Middorsum of body with many low, rounded warts, which may have one to many whitish
spicules on them, venter usually dark brown with scattered darker spots nepioxomus
Middorsum of body smooth, no warts or pustules; venter yellow, without spots (bright
red in life) arthuri
10 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

The males are smaller than the females in those spotting or vermiculation on chin and throat in
species where both sexes have been taken. The some individuals.
forelimb of the male is stouter and more muscular DESCRIPTION.—Snout from above with pointed
than in the female, presumably also associated with tip; distinct angle in canthus at level of nostril,
the mating behavior, at least as observed in A. canthus forms straight line angled outward from
ignescens (see elsewhere). The humeral area of nostril to eye; snout, canthus, and upper eyelid
the forearm is usually noticeably thinner than the heavily fleshy and raised, with continuous fleshy
radioulna area, which is quite thick. As a result fold from posterior corner of eye to back of head.
the males have arms reminiscent of the cartoon Muzzle between canthi distinctly concave, rest of
character Popeye the Sailor. dorsum of head flat. Snout from side distinctly
protruding beyond upper lip and lower jaw, nos-
tril directed laterally and slightly posteriorly; lo-
Atelopus arthuri, new species real region vertical anteriorly, distinctly concave
FIGURE 13 immediately before eye; temporal and tympanic
areas glandular and pustulate. Dorsum of body
HOLOTYPE.—USNM 193470, female, from 15 quite smooth or finely wrinkled, without pustules;
kilometers north of Pallatanga, Chimborazo Prov- sides of body and most of upper surfaces of limbs
ince, Ecuador, at an altitude between 2800 and heavily pustulate; females with whitish spinules
2860 meters, collected by Arthur J. Peters, Febru- singly or in clusters on top of pustules. Ventral
ary 14, 1959. surfaces of body, chin, and limbs smooth or with
DIAGNOSIS.—A member of the ignescens group very fine wrinkles, no or very few pustules.
of Ecuadorian Atelopus, with a vermiculate dorsal Forelimb very short; if pulled back along body,
pattern of light brown and dark brown on both tip of longest finger does not reach vent. Forearm
body and limbs; belly clear yellow (tomato red in slightly larger in diameter than humeral region.
life) with a brown patch on buttocks and brownish Forefoot fleshy, with fleshiness extending into web,

FICUKE 13.—Holotype, USNM 193470, of A. arthuri, new species.

NUMBER 145 11
which thus appears reduced, although actually in- rounded by a bright light green, which also covers
cluding first phalanx of each digit; three phalanges much of the ventral surfaces of arms and legs.
of third and fourth digits are free of web. No Dorsum of limbs usually as back, joints light
tubercles on ventral surface of forefoot in most orange. Large metatarsal tubercle on outer edge
individuals; some have series of fleshy swellings of foot bright yellow, lower surfaces of digits dull
both sub- and interarticular. Palmar tubercle large, orange, as is same surface of forefoot. Pupil black,
low, rounded; plantar tubercle very weakly defined. bright green ring around it, iris black with minute
Hind limb short, fleshy; when brought forward green punctulations.
along body heel may reach to points between pos- HOLOTYPE.—The holotype agrees with the de-
terior margin of head and axilla. Heels are well scription above in all characteristics for which no
separated when tibiofibulae are parallel to femora variation is described. Where the species is varia-
at right angles to body. Foot with fleshy web, ex- ble, the holotype has the following characteristics:
tending to tip of toe on all except fourth digit; skin on dorsum of body wrinkled in nuchal re-
deeply incised between digits. Sole of foot smooth, gion, smooth on rest of back. Pustules on sides of
with poorly defined subarticular tubercles; outer body and limbs with lighter areas, not clearly
metatarsal tubercle smaller than inner, but raised spinulate. Ventral surfaces of body wrinkled, a
higher, making it more prominent; inner meta- few pustules on throat. Scattered fleshy pustules
tarsal tubercle flattened. on ventral surface of forefoot. Plantar tubercle
COLOR IN ALCOHOL.—Dorsum light grayish not recognizable. Heel reaches to level of axilla.
brown with darker brown spotting, vermiculation, Subarticular tubercles on hind foot somewhat more
or streaking; canthus, margin of eye, and parotid prominent than in other individuals. Head with
swelling usually unicolor grayish brown (in some brown vermiculations and spotting. No distinc-
individuals almost entire head is unicolor). Dor- tive colorless areas on feet. Palm yellow, sole
sum of limbs and feet as body, with same light brown, with lighter tubercles. Measurements (in
and dark vermiculations, except in some individ- millimeters): snout to vent, 46.9; snout to end of
uals, which show clear, almost colorless, slightly coccyx, 40.4; knee to knee, 32.3; tibiofibula, 15.5;
yellow areas on first two digits and webbing of both head length, 12.5; head width, 12.8; eye diameter,
fore and hind feet. Sides of body as dorsum, with 3.0; snout to eye, 5.9; interorbital distance, 3.5;
tops of pustules appearing as white spots in some width of eyelid, 3.5.
individuals. Chest, belly, and ventral surfaces of VARIATION.—Since the entire type series was col-
thighs and humerus light yellow, with viscera lected by my son and me, we were able to gather
and/or eggs showing through the skin in many in- notes on the variation of the species in life. In-
dividuals. A more or less enlarged brown patch dividuals are extremely variable in the distribution
on the thighs below the anus, extending onto ven- of the colors of the ventral surfaces, which are to-
tral surface for short distance. Chin and throat mato red and bright green. One or two have the
yellow, almost always vermiculate or spotted with dorsum almost totally green, with l^|le or no ver-
dark brown, extending onto girdle region. Ven- miculation. Others have more yellowish orange on
tral surfaces of forearm, crus, and ankle areas of the back than dark brown. In addition, individuals
limbs dark brown, occasionally punctate with yel- may have different colors than the majority, or dif-
low. Ventral surfaces of digits entirely yellow, with ferent distribution of colors and pattern. One or
palm and sole also yellow in some individuals. more individuals were taken with the following
When palm and sole are partly or wholly brown, variants: orange belly and flash areas; green spots
tubercles are yellow topped and contrast strongly on chest, belly, and chin; entire venter green; dark
with brown areas. One individual (USNM 193466) brown stripe or chevron across sternum; greenish
has a reddish brown venter rather than yellow. yellow reticulations on dorsum; bright green dor-
COLOR IN LIFE.—Ground color of dorsum dark solateral line from eye to groin or from tip of snout
brown, heavily reticulate with orange-yellow. to groin; and dorsum of feet green.
Ground color on sides dull reddish orange, with ECOLOGY.—The entire type series was collected
dark brown reticulations. Large tomato-red spot on February 14 and 15, 1959, at altitudes between
ventrally, on thighs, belly, and chest, entirely sur- 2800 and 2860 meters. The rainy season was late
12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

during that year in that part of Ecuador, and the SPECIMENS EXAMINED.—The following are para-
frogs were actively breeding. Several mated pairs types: ECUADOR: Chimborazo Province, 15 km
were taken. The species is entirely diurnal, and N of Pallatanga, between 2800 and 2860 m altitude
individuals were found wandering about over quite (AMNH 79461, UKMNH 108938-39, UMMZ
dry terrain, using a slow walk and hopping only 131683, CAS 134199, MCZ 84199, USNM 193463-69,
when disturbed. The establishment of pairs ap- 193471-75).
pears to be entirely on a visual basis, and there
is no calling or chorus on the part of the exploring Atelopus balios, new species
males. The latter will raise themselves up high on
FIGURE 14
the forelimbs to look about for a female. Mated
pairs were found in pools formed in a fairly rap- A.[telopus] longibrachius.—Rivero, 1968:22.
idly flowing mountain stream, but no egg deposi- A.[telopus] spurrelli.—Rivero, 1968:22.
tion was observed. The voice in the species is char-
HOLOTYPE.—AMNH 17638, female, from the Rfo
acteristic of Ecuadorian Atelopus where known,
Pescado, Guayas Province, Ecuador, collected by
which is a low, chicklike peep, often not distin- G. H. H. Tate, May 28, 1922.
guishable at any distance. Occasionally males of DIAGNOSIS.—A member of the longirostris group,
this species will trill the peep slightly. with long, thin legs and with heels overlapping
ETYMOLOGY.—The species name is derived from when tibiofibulae are parallel to femora at right
the first name of the collector of the holotype, who angles to body; first digit on forefoot distinct, not
provided me with pleasant company on the long hidden in webbing; limbs smooth. Similar to
trip to Papallacta and back, and who collected longirostris, but differs in having a lighter ground
the first specimens of the new taxon while I was color with darker spotting, while longirostris has
busy fixing our supper. a dark ground color with yellowish spots.

•".

FIGURE 14.—Holotype. AMNH 17638, of A. balios, new species.

NUMBER 145 13

DESCRIPTION.—Snout from above rounded or irregularly broken into spots along lower sides.
with blunt point; outward curve of muzzle becom- Upper lip light, rest of body below stripe usually
ing inward curve from level of nostril to eye; snout, unicolor light yellowish white, a few individuals
canthus, upper eyelid, and temporal region some- with scattered spots on belly. Ventral surfaces of
what fleshy. Dorsum of snout weakly concave, rest limbs yellowish white, of feet dull yellow.
of dorsum of head flat, lightly shagreened. Eye HOLOTYPE.—The holotype agrees with the de-
slightly bulging dorsally. Snout from side protru- scription above in all characteristics for which no
sive over lower jaw and upper lip, forming blunt variation is described. Where the species shows
point; nostril lateral, loreal region vertical or very variability, the holotype has the following charac-
slightly concave. Eye slightly overhung by fleshy teristics: snout from above with blunt point, loreal
margin of upper eyelid; temporal region more region vertical. Tibiotarsal articulation reaches
heavily shagreened than dorsum, and also heavily posterior corner of eye; heels overlap when tibio-
wrinkled. Dorsum almost smooth, very finely fibulae are parallel to femora at right angles to
wrinkled; sides with wrinkles forming deep creases; body. Outer metatarsal tubercle present, distin-
chin and belly so heavily wrinkled that areas be- guishable. Color pattern as shown in Figure 14,
tween wrinkles look like pustules. lateral brown streak broken into series of spots, no
Forelimb very slender, with forearm more fleshy spots on belly. Measurements: (in millimeters)
than humeral region, dorsum thickly shagreened, snout to vent: 35.2; snout to end of coccyx: 33.5;
ventral area wrinkled. When forearm is pulled knee to knee: 28.4; tibiofibula: 15.7; head length:
back along body, second and fourth digits reach 9.2; head width: 8.6; eye diameter: 3.0; snout to
vent. Sole of forefoot fleshy, fleshy webbing en- eye: 4.6; interorbital distance: 3.2; width of eyelid:
closes all but tip of first digit, extends as fringe to 2.2.
ends of other digits, giving them a flattened ap- REMARKS.—The AMNH catalog shows the type
pearance, webbing between toes incised to a point locality as in Guayas Province, and notes made
about one quarter of distance to digit tips. All by Tate would indicate it is near the Guayas-Azuay
subarticular tubercles almost totally indistinguish- border. Tate did not indicate what province he
able, as are both plantar and palmar tubercles, was in when he collected the type series, but he
with latter occasionally forming low rounded was on the mule trail from Naranjal to the high-
hump. Males with horny excrescence on first digit lands, and he estimated the altitude at the river
of forefoot. to be about 1400 feet above Naranjal, or approxi-
Hind limb slender, finely but densely shagreened mately 500 meters. He noted that it was three hours
dorsally, less dense laterally, giving way to wrin- by trail from Naranjal to the river, which was at
kling on ventral surface. Tibiotarsal articulation the foot of the "real hills." He found the area
may reach to points between anterior edge of eye through which the river flows to be very wet, but
and anterior margin of tympanic area when carried from there on the area became much drier. Since
forward along body. Heels touch or overlap when the trail passes the 500-meter contour well before
tibiofibulae are parallel to femora at right angles the Guayas-Azuay border is reached, I agree the
to body. Entire foot fleshy with thick webbing to specimens were probably collected in Guayas Prov-
tips of toes, first three toes almost totally obscured ince.
by webbing, which forms fringe on fourth toe from Rivero discussed the type series, and figured the
level of tip of third and fifth toes. No subarticular specimen here designated as the holotype (1968:22;
tubercles, very reduced outer metatarsal tubercle, pi. 3: fig. 5). On the legend of the plate the AMNH
almost indistinguishable in some individuals, no catalog number is given as 17738 rather than 17638,
inner metatarsal tubercle. which can be read on the photograph. Rivero con-
COLOR IN ALCOHOL.—Dorsal ground color very sidered the specimens possibly to be conspecific
light tan, with numerous rounded spots, completely with Atelopus longibrachius Rivero, a Colombian
separated from each other. Dorsum of legs, sides species. When A. balios is compared with descrip-
of body as dorsum of body, side of head with dark tions of the type material (Rivero, 1963:112; Coch-
brown line from tip of snout through eye along ran and Goin, 1970:135), the differences in color
side of head to shoulder, line either continuous or pattern are distinctive, as pointed out by Rivero,
14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

and there are additional differences in the number I do not regard them as typical, and they are not
and distribution of pustules, the length of the fore- designated as para types.
limb, the presence and size of tubercles on the ETYMOLOGY.—The specific name is derived from
palm and sole, and in other characters. Geograph- the Greek word balios, meaning "spotted" or
ically the localities for the two taxa are widely "dappled," and refers to the dorsal pattern of the
separated. Much collecting has been done by many members of this taxon.
workers in the area between the two type localities, SPECIMENS EXAMINED.—ECUADOR: (Guayas
but nothing resembling either taxon has turned up Province; Rio Pescado (AMNH 17619-30, 17636-
in collections as yet. 38, USNM 193957-58). All are paratopotypes. As-
In the same paper Rivero (1968:22) indicated signed to this species, but not designated paratypes:
that some of the AMNH specimens from the Rio ECUADOR: Guayas Province; Rio Pescado
Pescado belonged to Atelopus spurrelli Boulenger, (AMNH 17548-49).
a Choc6an species. He has reference to AMNH
17548-49, which are marked differently, and the Atelopus bomolochos, new species
pattern is certainly reminiscent of that of spurrelli.
FIGURE 15
I have noted, as well, that these two specimens have
a somewhat more prominent snout tip, which is HOLOTYPE.—CAS 93910, female, from Ecuador,
slightly hooked downward, as is seen in A. longi- Azuay Province, Sevilla de Oro, approximately
rostris, and the third digit on the forefoot seems 2800 meters. Collected January 9, 1958, by G. Fry-
to be markedly longer than in the type series. The mire, A. Embree, and S. Strobel.
color pattern could be formed from fusion of the DIAGNOSIS.—A member of the ignescens group,
discrete spots seen in the type series, however, and most similar to ignescens, from which it is distin-
this is what I think has happened, so I am regard- guished by the light dorsal coloration often marked
ing the specimens as members of the new taxon. with irregular dark spots; ratio head length/stand-

'

FIGURE 15.—Holotype, CAS 93910, of A. bomolochos, new species.

NUMBER 145 15
ard distance X 100 usually greater than 33; warts Hind limb short, stout; femoral region spinose
and pustules absent on upper arm, crus, and ankle, on upper thigh only, close to body, rest of limb
but present on sides from tympanic area to groin, smooth or slightly wrinkled, occasionally with
usually absent on dorsal and ventral surfaces of glandular areas on thigh and crus. Tibiotarsal
body; first digit on forefoot distinct, almost totally articulation reaches to points at arm insertion when
free of webbing. carried forward along body; heels are separated
DESCRIPTION.—Snout from above gently rounded when tibiofibulae are parallel to femora at right
or with blunt point; distinct angle above nostril; angles to body. Foot fleshy, with very indistinct
snout, canthus, and upper eyelid fleshy and raised, scattered pustules on sole; webbing fleshy, extends
with flattened, glandular, fleshy ridge from pos- to digit tips as fringe except on first toe, which is
terior corner of eye to end of head. Dorsum of almost completely hidden in webbing; webbing
head between fleshy areas slightly concave or flat, between all digits except first and second deeply
smooth. Eyes protrude laterally, dorsal area of eye incised. Outer metatarsal tubercle raised and prom-
in same plane as rest of head. Snout from side inent, inner metatarsal tubercle indistinct.
extends short distance beyond level of lower jaw COLOR IN ALCOHOL.—Dorsum of head, body, and
and upper lip, nostril lateral, occasionally directed limbs light, may be yellowish, yellow-green, or light
slightly backward, on raised, rounded area. Loreal brown; occasionally uniform, but usually with few
region distinctly concave; fleshy area over eye fused to many irregular, scattered, dark brown to black
with upper eyelid; upper lip slightly fleshy. Tem- spots. Dark dorsolateral line begins on side of
poral area studded with glandular prominences, snout just below fleshy fold, runs through eye and
rounded in males, covered with spiny points in across temporal region onto sides, where it may be
females. Middorsum of body completely free of continuous and broad, narrow and jagged, or al-
warts, glands, or pustules; usually quite smooth. most totally absent. Ventral surfaces unicolor
Dorsolateral area of neck with heavy, glandular white, yellowish white or yellow; occasional indi-
fold, shelflike in appearance, extending back from vidual with one or two small brown spots on
dorsolateral temporal fold, sides wrinkled and throat or below anus. Spots often present at limb
folded, covered by small white or yellow pustules articulations, may extend around limb, but in no
from tympanic area to groin, and on upper arms instance continuous ventrally. Sole of forefoot
and thighs; pustules may be single and evenly scat- always unicolor, sole of hind foot usually unicolor,
tered or bunched together. Chin, chest, and belly
may be spotted or streaked.
with wrinkles and folds, most prominent on chest.
HOLOTYPE.—This specimen conforms with the
No pustules or spicules on ventral surfaces, al-
description given above for the species, where no
though occasional individuals show very tiny spic-
alternatives are given. In the case of alternatives,
ulations on chest.
it has the following characteristics: snout rounded;
Forelimb short and heavy; upper humeral region dorsum of head slightly concave, temporal region
with spicules, but lower humeral region and fore- lightly pustulate. Skin on chest and throat lightly
arm free of spicules, smooth or coarsely wrinkled. spiculate. Dorsum yellowish, with many black
When forelimb is stretched along body, fingertips spots, dorsum of limbs with black reticulation; two
do not extend beyond posterior margin of thigh. black spots on throat at chest. Measurements of
Flesh of forefoot swollen; all digits free, with fleshy type (in millimeters): snout to vent: 50.4; snout
webbing only at base. Sole of forefoot covered with to end of coccyx: 45.0; knee to knee: 34.6; tibio-
raised, rounded, glandular areas, subarticular tu- fibula: 17.0; head length: 12.7; head width: 12.6;
bercles not differentiated from these pustular areas; eye diameter: 4.0; snout to eye: 5.7; interorbital
palmar tubercle larger and well defined. Males distance: 3.8; width of eyelid: 3.6.
with horny excrescence on first and second digits,
first usually strongly flexed, with prominent, thick, REMARKS.—The tendency seen in these individu-
rounded fleshy hump dorsally, keratinized area als to have spots of color at the articulations of
wide and extending from tip of digit to its base; the limbs is reminiscent of the stronger develop-
second digit with cornified area narrow, extending ment of rings around the joints in A. botdengeri.
from base of digit to terminal articulation. This species is comparatively large, and could cer-
16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

tainly be considered as a close relative of bou- in Sevilla de Oro in July 1962, but we took no
lengeri. Atelopus at all. The ignescens material at the
Of the seven species described as new in this University of Kansas was collected in the months
paper, bomolochos is perhaps of most questionable of May, June, and July. The California Academy
validity. It is clearly closely related to ignescens, collectors of the type material were in the area in
and, as is pointed out in the discussion of that the months of January and February. This seems
species, the variability of ignescens is considerable. to indicate a seasonal occurrence and disappear-
It could easily encompass the differences on which ance for bomolochos, but may simply reflect the
bomolochos is based, although I have as yet seen breaks of the game.
specimens with the particular constellation of char- ETYMOLOGY.—The specific name is directly from
acters used to define bomolochos only from a the Greek word bomolochos, meaning "toady," or
comparatively limited area. There is a distinct pos- "lickspittle." While perhaps a stretch of the imagi-
sibility that the two taxa are sympatric, although nation for those who believe in strict interpreta-
there seem to be no existent collections where a tion of the literal meanings of words, it delights
single series contains both taxa. Both the Cali- me to think of the species as a "toady type frog,"
fornia Academy and the University of Kansas have or, for that matter, since it is a bufonid, a "toady
had active collectors working along the Pan- type toad."
American Highway in Canar, Azuay, and Loja SPECIMENS EXAMINED.—The following are desig-
Provinces in the last few years, and a fairly con- nated as paratypes: Azuay: Sevilla de Oro, 2800 m
tinuous series of ignescens can be put together from (CAS 85341-42, 93910; holotype: 93912, USNM
the available material. While these southern speci- 193956). The following are tentatively identified
mens are not identical with northern populations, as bomolochos, but are not designated as para-
they are not so distinct that I think any purpose types: Azuay: Cuenca, 2535 m (CAS 85139-41,
would be served by describing them as different. 85161). Canar: between Juncal and General Mo-
A single individual tentatively called bomolochos, rales, approximately 2500 m (CAS 93906).
CAS 93906, is from a locality well to the north of
the remaining sample, and was also collected along
Atelopus boulengeri Peracca
the Pan-American Highway. There is a small series
from Cuenca which I feel is legitimately associated FIGURE 16
with the new taxon. This locality, too, is on the Atelopus Boulengeri Peracca, 1904:20 [type locality: "di
highway. Specimens called ignescens in this paper Gualaquiza e parecchi di S. Jose, Ecuador orientale"].
come from both north and south of these two lo- Atelopus bicolor Noble, 1921:3, fig. 4 [type locality: Cordillera
calities on the highway, and one gains a distinct Kutuku, east of Macas, Ecuador, 1800-2000 m].—Rivero,
impression of either interdigitation or sympatry. 1963:107.
A.[telopus] boulengeri.—Nieden, 1926:82.—Rivero, 1963:107.
The two taxa, bomolochos and ignescens, are not
distinguishable on the basis of any of the measure- SYNTYPES.—Six specimens, UT An. 560, from
ments shown in the figures, with the possible ex- San Jose, and eleven specimens, UT An. 559, from
ception of head width/head length. There are in- Gualaquiza, in the Zoological Museum of the Uni-
sufficient males of bomolochos available for statis- versity of Torino, Italy.
tical analysis, but the females of bomolochos are DESCRIPTION.—Snout from above with rounded
distinct from those of ignescens with regard to this tip, outward curve of muzzle becoming inward
character. Otherwise, the variation in bomolochos curve from nostril to eye; snout, canthus, and upper
is almost completely within that expected or ob- eyelid fleshy and raised, with fleshy ridge from pos-
served in ignescens. terior corner of eye to edge of head. Dorsum of
It is worth noting that both workers from the head between these raised, fleshy areas flat, broad
University of Kasnas and several of my field par- and smooth, lying in same plane as broad, flat back.
ties have collected in the area from which bomo- Snout from side slightly protrusive beyond upper
lochos is recorded without obtaining specimens. lip and lower jaw, nostril lateral, below fleshy area
My party, including two students, an experienced of canthus, and directed slightly posteriorly. Loreal
Ecuadorian collector, and Dr. C. F. Walker, was region strongly concave behind nostril; eye over-
NUMBER 145 17

FIGURE 16.—A representative individual, USNM 193491, of A. boulengeri Peracca.

hung by fleshy margin of eyelid, upper lip some- last three toes and reaching tip of fourth toe as
what fleshy. No tubercles or pustules anywhere on fringe only; first and second toes almost totally
head or body. Dorsum and sides of body smooth or hidden in webbing, tips of all toes free and some-
finely wrinkled, belly with more prominent and what prominent, with slight swelling. Subarticular
distinct wrinkles. tubercles weak, outer metatarsal tubercle raised
Forelimb slender, forearm usually somewhat and prominent but smaller than the flattened,
more fleshy than humeral region, dorsum weakly rounded inner metatarsal tubercle.
pustulate, rest of arm smooth. When forelimb is COLOR IN ALCOHOL.—Dorsum usually unicolor,
pulled back along body, palm of hand reaches vent. chocolate brown, dark brown, or blackish; extend-
Entire hand fleshy, quite heavy around first digit, ing slightly onto sides of body, where it is sharply
which is clearly distinct and prominent, second set off by yellow of sides and belly, both of which
digit shorter than fourth, subarticular tubercles are also unicolor. Dorsum of head with dorsal color
at base of digits only. Tips of digits prominent, extending to varying degrees anteriorly; in some
slightly swollen ventrally. Palmar tubercle distinct, it.ends at eye level, in others it covers snout. Side
plantar tubercle present. Edge of prepollex with of head, lips, chin, and throat all monochrome
horny excrescence in males. yellowish. Limbs usually slightly darker yellow
Hind limb slender, smooth except for few low, than venter, with rings of brown around all joints.
very indistinct pustules on dorsum. Tibiotarsal Dorsum of first finger and first two toes often
articulation reaches to points between posterior lighter than rest of foot. Ventral surfaces of feet
margin of eye and arm insertion when carried usually dark, with lighter tubercles. Brown of
forward along body; heels do not overlap or touch dorsum usually extends over posterior margin of
when tibiofibulae are parallel to femora at right legs and forms brown "seat" or "diaper" on rear
angles to body. Entire foot with fleshy webbing to of venter, with anus usually bright yellow in center
tips of toes; webbing somewhat excised between of this brown area.
18 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

COLOR IN LIFE.—Rich chocolate brown dorsally, dard head length measurements that might give
with bright yellow laterally and ventrally. a larger value, and thus a smaller proportion, for
ECOLOGY.—The two specimens collected in 1962 Noble's specimens.
were found along a trail in fairly heavy subtropical Although there may have been some geographical
forest, in full daylight, about 11:00 A. M. Both validity for retaining the two species prior to our
were walking, not hopping, along the trail, fully field work in 1962, the two specimens we obtained
exposed, and made no effort to avoid being cap- weaken this validity considerably. The type series
tured. for boulengeri came from Gualaquiza, a town in
SYNTYPES.—The 17 syntypes agree completely the valley of the Rio Zamora, at 950 meters. Noble's
with the description above, with the series exhibit- specimens came from quite a bit higher in an
ing all the variations mentioned in the description. isolated Cordillera to the east of the valley of the
Some of the specimens from Gualaquiza are some- Rio Upano, a tributary of the Namangoza, which
what dried, and the digits do not have either as joins the Zamora to form the Rio Santiago. The
thick or as full a webbing as in fresh specimens, two new specimens come from the eastern slopes
but the San Jose series is in good condition, and of the main Andean chain at an altitude of 1560
the digits are as described. Only one specimen in meters, and some distance to the west of the valley
the series shows a yellow area at the anus; the of the Namangoza. This indicates an altitudinal
rest have the anal region the same color as the range of about 600 meters, with the species occupy-
dorsum. ing the foothills and mountain slopes of the head-
COMPARISONS.—The differences between boul- water streams of the Santiago up to an altitude
engeri and bicolor appear to be the consequence of 1560 meters.
of differences between investigators, rather than Rivero (1963:110) regarded A. bicolor Noble
between the types. The primary distinguishing as probably not referable to any other Ecuadorian
characters pointed out by Noble in the description species, but later (1968:23) indicated that it is
of bicolor were the development of the webbing "certainly a synonym of A. boulengeri Peracca."
and several proportions. Noble said the digits of the He records four specimens from Turula, which he
forefoot were webless, while Peracca described the indicates as being east of Macus, but there is only
webbing between the second, third, and fourth one specimen in the AMNH from Turula, which
fingers as rudimentary, and as relatively well de- is a "tambo" on one of the mule trails from Rio-
veloped between the first and second fingers. The bamba to the Amazonian region, and lies northwest
area between these digits is actually quite fleshy, of Macas. The bicolor paratypes, AMNH 13133-35,
and it appears that Peracca considered this fleshi- are not from Turula (see below).
ness equivalent to webbing, while Noble, who SPECIMENS EXAMINED.—ECUADOR: no addi-
illustrated this area clearly in his figure, thought it tional data (AMNH 17427-28, 17430-31). Morona-
should not be called webbed at all. The description Santiago Province; Cordillera Cutucii (AMNH
of the webbing of the hind foot by Peracca is 13132—holotype of bicolor, AMNH 13133-35—
quite consistent with Noble's figure, although Per- paratypes of bicolor, UKMNH 108940, USNM
acca described the foot as strongly webbed, while 193493), Gualaquiza (UT An. 559—syntypes of
Noble said, "toes webbed to the tips . . . but the boulengeri), Rio Piuntza, 1830 m (UKMNH
web deeply indented, making the toes appear half 147071-81), San Jose (UT An. 560—syntypes of
webbed." The differences in proportions are quite boulengeri), Turula (AMNH 33893), between
minor, and could easily be the result of different Mirador and Copal, on mule trail from Sevilla de
techniques in measuring. Thus, in boulengeri the Oro to Mendez, 1600 m (USNM 193491-92). Loja
head length is contained in the body length 2i/2 Province; San Ramon (JAP 8934).
to 3 times, according to Peracca, while in bicolor,
according to Noble, the body is 3i/2 times as long
as the head. But Noble clearly indicated that the Atelopus bufoniformis Peracca
head length meant from snout to angle of jaw, and FIGURE 17
no one knows for sure how Peracca made his head
Atelopus bufoniformis Peracca, 1904:20 [type locality: El Pun,
length measurement. There are several other stan- Ecuador].
NUMBER 145 19

FIGURE 17.—A representative individual, USXM 193537, of A. bufoniformis Peracca.

SYNTYPES.—Two specimens, UT An. 561, in the warts, increasing in prominence over tympanic
Zoological Museum of the University of Torino, area. Entire body, limbs, chin, throat, and belly
Italy. heavily pustulate or warty, with pustules low and
DESCRIPTION.—Snout from above angulate and flattened.
pointed, with a sharp angle at level of nostril; Forelimb very short, slender. When limb is car-
canthi from nostril to eye almost parallel, without ried along body, digits barely reach insertion of
a curve; eye strongly protruding and directed hind leg. Hand with very fleshy webbing enclosing
slightly forward; canthus fleshy and raised above first digit completely, second and fourth almost
median area of snout; dorsum of eye fleshy with as completely, and on third digit to tip as fringe,
many wrinkles and grooves. Dorsolateral area of with whole foot giving appearance of toad wear-
head behind eye with a prominent ridge from eye ing a mitten. Entire lower surface of hand as
level extending posteriorly along back, converging heavily pustulate as rest of body, obscuring sub-
on ridge from opposite side to a point above shoul- articular tubercles completely; plantar tubercle
der, at which point they diverge until they disap- very vaguely defined. Hind limb short, slender;
pear. Second, more glandular ridge arises at corner tibiotarsal articulation reaches to points on side
of eye and runs over tympanic region, either ending well behind axilla when leg is carried forward
in small boss at eiul of head, or continuous with a along body; heels widely separated when tibio-
glandular dorsolateral fold. Rest of head dorsally fibulae are parallel to femora at right angles to
covered with wrinkles, grooves, and pustules. body. Entire foot with very thick, fleshy webbing,
Snout from side extends very slightly beyond end enclosing fully all digits except fourth, which is
of lower jaw, and is either bluntly rounded or fringed to tip by web. No distinguishable tubercles
weakly pointed. Loreal region vertical, not con- on sole, which is fleshy and wrinkled; outer meta-
cave; entire side of head with rather thick skin, tarsal tubercle vaguely defined, inner metatarsal tu-
wrinkled, grooved, and with small, fleshy, flattened bercle even more vaguely defined.
20 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

COLOR IN ALCOHOL.—Dorsum of body, head, and 132126), N slope of Nudo de Mojanda, 3400 m
limbs entirely very dark brown to black; only occa- (UKMNH 131797-98).
sional individuals with somewhat lighter brown
vermiculations on back. Venter of body brown, Atelopus elegans (Boulenger)
with more or less vague dull yellowish mottling.
Sole of feet lighter than rest of ventral surfaces. FIGURE 18
COLOR IN LIFE.—USNM 193537 was very reddish
Phryniscus elegans Boulenger, 1882a: 155 [type locality: Tanti,
in life, and the warty pustules on the venter were Ecuador, 2000 feet]; 1882b:464, fig. 1.
a very bright red. USNM 193538 had yellow pus- Atelopus elegans.—Boulenger, 1894:375; 1898:119.—Peracca,
tules on the venter. 1904:22.
SYNTYPES.—The two syntypes are badly dried, A.[telopus] elegans.—Nieden, 1926:81.
but it is still possible to identify them as bufoni- Atelopus [elegans] elegans.—Rivero, 1963:109.
formis in the sense of this paper. The larger of the HOLOTYPE.—British Museum (Natural History),
two syntypes is the individual for which measure- one female.
ments are presented in the original description. DESCRIPTION.—Snout from above rounded, not
The second specimen is about 20 millimeters long. ending in point; distinct angle in canthus at nos-
The larger specimen fits the description above tril level; canthus and upper eyelid not fleshy,
except as follows: the glandular ridges arising at dorsum of head smooth and flat. Snout from side
the corners of the eye are continuous with the with distinct projection, ending in blunt point,
glandular dorsolateral folds. The snout from the extending beyond symphysis of lower jaw; loreal
side is bluntly rounded. The specimen is too badly region vertical, not concave. Head dorsally and
dried to carry foot forward, but Peracca said that laterally smooth, without spines, warts, or pustules.
the tibiotarsal articulation "raggiunge scarsamente Body slender, skin completely smooth.
la spalla." Forelimb long, slender; when laid back along
REMARKS.—This is a very small species. The body, palm reaches anus. Webbing fleshy, enclosing
largest individual seen measured 33 millimeters and almost completely obscuring first digit; full
from snout to end of coccyx, and most specimens webbing between first and second digits; web ex-
were less than 30 millimeters long. The shortened tends to tips of other digits but is deeply incised
limbs and increased padding or fleshiness of the between them. Fourth digit extending beyond level
feet seem to be highly adaptive, providing reduced of penultimate joint of third; considerably longer
surfaces for heat loss. The species seem well adapted than second digit. Palm of hand fleshy, with deep
for high altitude life, and has been taken as high folds; palmar tubercle low, rounded; plantar tuber-
as 3700 meters. cle weak and indistinct.
Cochran and Goin (1970:120) described and Hind limb long, slender; when brought forward
figured a single specimen of this species from the along body, tibiotarsal articulation reaches points
Paramo de las Papas, Cauca, Colombia. It is cor- between loreal region and posterior margin of
rectly identified, and thus A. bufoniformis shows eye; heels overlap when tibiofibulae are parallel
a wider distribution than any other highland
to femora at right angles to body. Hind foot with
species except A. ignescens. As I have pointed out
thick, fleshy webbing; first and second digits al-
elsewhere in this paper, these are the only two
most totally enclosed, with only their tips dis-
species that have been able to break out of the
tinguishable; other digits with full webbing that
isolating mechanisms faced by all high altitude
is shallowly incised, and extends to tip of fourth
species of Atelopus, and extend their distribution
digit as fringe only. Fifth toe longer than third.
into higher areas of both Ecuador and Colombia.
Sole of foot quite fleshy with deep grooves or folds,
SPECIMENS EXAMINED.—ECUADOR: Napo Pro-
no clearly defined subarticular tubercles; outer
vince; El Pun (UT An. 561-syntypes), Environs of
metatarsal tubercle poorly defined, inner metatarsal
Santa Barbara (USNM 193537-40). Carchi Pro-
vince; Paramo del Angel, 23 km SW of Tulcan, tubercle almost indistinguishable.
3700 m (UKMNH 117880). Imbabura Province: COLOR IN ALCOHOL.—Dorsal ground color white,
Cordillera de Intag, La Delicia, 2700 m (UKMNH with more or less regular pattern of brownish or
NUMBER 145 21

FIGURE 18 A representative individual, UMMZ 51270, of A. elegans Boulenger.

black consisting of parallel stripes beginning on REMARKS.—Rivero (1963:109) regarded Atelopus

eyelid, where they may be joined by bar between gracilis Barbour, from Gorgona Island, Colombia,
eyes, and extending to just short of anus, which is as a subspecies of Atelopus elegans. Cochran and
usually in isolated spot of brown narrowly separ- Goin (1970:133) made elegans a subspecies of A.
ated from stripes. Dorsal stripes irregularly con- varius, with gracilis a strict synonym. They did not
nected by vermiculations and lateral bars, or sep- refer to Rivero's action in their synonym. Neither
arated by additional stripes running lengthwise, Rivero nor Cochran and Goin presented any docu-
by light spots surrounded by brown borders, or mentation for the changes they made. Rivero
by brown spots surrounded by white. Solid white stated that specimens from Ecuador "seem suffi-
line from eye to groin along dorsolateral line in ciently distinct from Gorgonan material to make
most individuals, bordered below by dark line the island form deserve the racial name A. elegans
beginning at tip of snout, passing along lateral gracilis." He does not describe the differences or
area of canthus, and continuous from eye to groin. define the race. Cochran and Goin reduced several
No pigment on sides or belly below this lateral species to the subspecific level within A. varius,
dark streak. Dorsal areas of limbs and feet mottled, but again the reasons for doing so were not clearly
streaked, and striped with dark brown to black, as stated. Even though it is entirely possible that
body; ventral areas of appendages unicolor yel- elegans represents the southern end of a genetic
lowish white. continuum which embraces both high and low
COLOR IN LIFE.—Unknown, but several speci- altitude populations, I have seen no evidence to
mens are very yellowish in ground color, indicat- support the concept. A. elegans is a distinct, easily
ing that this might be the living color. recognized, well marked taxon, and I prefer to
22 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

recognize it as a full species until intergradation Bogota from the east about halfway between Sali-
with the subspecies of varius can be demonstrated. dero and Pulun. I am, therefore, using the name
The relationship between gracilis and elegans may "Pulun" for the place where the AMNH specimens
be subspecifk, but I have not examined the gracilis were collected.
material, and leave that question in abeyance for SPECIMENS EXAMINED.—ECUADOR: Esmeraldas
other investigators. Province; Pulun (AMNH 10581-84, UMMZ
In 1955, I referred to the locality called "Buliin" 51270), Quininde (USNM 193534), Rio Cupa
in the American Museum catalogs as unlocated. (UKMNH 108911, USNM 193535), Rio Durango
Levi-Castillo, in an editorial footnote to my re- (AMNH 10585-88), Salidero (AMNH 10577,
marks, suggested that it was in reference to "Bulu- 10579-80, MCZ 2615, 7605). Imbabura Province;
Bulu," a river in Guayas and Caiiar Provinces. The Lita, Rio Mira (USNM 193536).
United States Board on Geographical Names (1957,
p. 138) lists a place called "Pulun" at the coordi-
nates l°05'N, 78°40'W, and it is shown at those Atelopus halihelos, new species
coordinates on the 1:1000000 map of the Instituto FIGURE 19
Geografico Militar del Ecuador (1950). The Ameri-
Atelopus [species].—Rivero, 1968:21.
can Museum specimens were collected at about the
same time as material from Salidero and Rio HOLOTYPE.—AMNH 16716, male, from Ecuador,
Durango, both of which are shown on a map of Morona-Santiago Province, Cordillera Cutucii,
Esmeraldas Province in Acosta-Solis (1944, opp. p. collected by E. Feyer, March 1921.
574). Both localities are very close to Pulun, which DIAGNOSIS.—A member of the ignescens group,
also appears on the map in the work by Acosta- although limbs are more slender than other species;
Solis. The coordinates for Salidero would be ap- most similar to nepiozomus, from which it differs
proximately l°04'N, 78°38'W, and the town lies in having an almost indistinct first digit on the
on the Rio Bogota. The Rio Durango enters the forefoot and in coloration, particularly in the

5 I t
<

FIGURE 19_Holotype, AMNH 16716, of A. halihelos, new species.

NUMBER 145 23

presence of a light whitish patch below the anus, thighs and belly immediately below anus very light
in contrast to the dark "diaper" of nepiozomus. colored, often whitish, without spots.
DESCRIPTION.—Snout from above with rounded HOLOTYPE.—This specimen conforms with the
point, distinct angle in canthus rostralis at nostril; description given above for the species in those
snout, canthus, and eyelid fleshy and raised, with characteristics for which no variation is mentioned.
either fleshy ridge or row of raised pustules from In the case of variable characters, it has them as
corner of eye along dorsolateral margin of head. follows: a row of raised pustules along dorsolateral
Dorsum of head flat, with scattered, low, glandular margin of head behind eye; loreal region slightly
pustules. Snout from side protrusive, pointed, ex- concave. Warts in temporal region do not bear
tending well beyond upper lip and lower jaw; spicules. Chest and throat with low, indistinct
nostril lateral, below fleshy area of canthus, at top pustules. Heel carried forward along body reaches
of rounded prominence. Loreal region flat or temporal region; heels do not touch when tibio-
slightly concave; eye overhung by flap formed of fibulae are parallel to femora. Soles of both fore
fleshy area, flap has free edge and can be lifted away and hind foot with few indistinct tubercles, sub-
from the eyelid; upper lip swollen. Temporal articular tubercles on hind foot slightly more
region with low, glandular pustules, which may prominent. Measurements of type (in millimeters):
bear pointed spicules in either sex. Dorsum and snout to vent: 25.3; snout to end of coccyx: 23.4;
sides of body smooth, with many low, rounded, knee to knee: 20.3; tibiofibula: 9.8; head length:
scattered warts. Ventral surfaces may be either 7.8; head width: 7.2; eye diameter: 2.6; snout to
comparatively smooth or with flattened, indistinct eye: 4.0; interorbital distance: 2.6; width of eyelid:
glandular patches separated by wrinkles and folds. 2.2.
Forelimb slender in females, shorter and more REMARKS.—This series of specimens was discussed
muscular in males, particularly in the forearm. by Rivero (1968:21), and he felt their affinity was
When limb is carried along body, digit tips reach closest with A. bufoniformis. Since there are neither
beyond posterior margin of thigh. Dorsal and lat- enough females of A. halihelos nor males of A.
eral surfaces of limb covered with warts, more bufonifmmis available for statistical analysis, the
prominent in males than in females. Hand with two species cannot be directly compared on the
fleshy webbing, first digit almost entirely enclosed, graphs provided earlier in this paper. Relative po-
other digits free. Males with keratinized pad on sition with regard to other species, however, shows
dorsum of first digit, no horny excrescences on sec- that halihelos is more slender, with longer limbs,
ond digit. Sole of forefoot smooth or with few low and it appears more closely related to A. nepiozo-
pustules; plantar tubercle weak, defined by absence mus and other species on the lower Amazonian
of pigment in area. slope than to the highly specialized, high altitude
Hind limb long, slender, covered dorsally and A. bufoniformis. As in the latter species, it appears
laterally with low warts; tibiotarsal articulation that halihelos is a diminutive species, with the
reaches to area between eye and end of head when largest specimen in the type series less than 32 mil-
leg is carried forward along body; heels touch or are limeters long.
barely separated when tibiofibulae are parallel to There is no way to be sure about the actual dis-
femora at right angles to body. Webbing thick but tribution or altitudinal relationships of this new
not heavily fleshy, extending to tips of all but species. The Cordillera de Cutucii is a fairly isolated
fourth digit, incised between all digits except first mountain mass to the east of the valley of the Rfos
and second; only tip of first digit free of web. Sole Upano and Namangoza, with the foot of the Cor-
almost entirely smooth, subarticular tubercles us- dilleras at approximately 1000 meters, and the high-
ually present, but low and often indistinct; outer est point somewhat more than 2000 meters. I would
metatarsal tubercle fairly prominent, distinct; in- predict that the species will be found at altitudes
ner metatarsal tubercle larger than outer, but low above the lower edge of the Cordillera, and pos-
and indistinct. sibly to the highest points. It seems to be a situation
COLOR IN ALCOHOL.—Entire body and limbs light similar to that of A. planispinus, which inhabits
brown, with small, scattered, irregular dark brown the slopes of a similar isolated mountain mass to
or blackish spots on all surfaces. Ventral area of the north.
24 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

ETYMOLOGY.—The specific name is derived from Atelopus carinatus Andersson, 1945:15, fig. 5 [type locality:
"Eastern Ecuador"].—Rivero, 1969:142, fig. 2.
the Greek hales or halis, meaning "in abundance, Atelopus ignescens ignescens.—Rivero, 1965:137.—Duellman
in heaps, crowds or swarms," and helos, a "wart" and Lynch, 1969:236.
or "callus." The reference is to the large number
of glandular swellings on the body and limbs of HOLOTYPE.—In the "Museo Mediolanense,"
individuals of this taxon. present state unknown.
SPECIMENS EXAMINED (in addition to holotype).— DESCRIPTION.—Snout from above not prominent
The following are paratopotypes: ECUADOR: Mo- or projecting, gently rounded, with little or no
rona-Santiago Province; Cordillera Cutucii (AMNH difference in angle between tip of snout and eye;
16717-24, USNM 193959). snout, canthus, and upper eyelid heavily fleshy and
raised, with a glandular, fleshy ridge from posterior
corner of eye to corner of head. Dorsum of head
Atelopus ignescens (Cornalia) between fleshy areas flat, smooth, or very finely
wrinkled. Snout from side only slightly protrusive
FIGURE 20
beyond lip and lower jaw, nostril lateral and below
Phryniscus ignescens Cornalia, 1849:316 [type locality: "in fleshy area of canthus. Loreal region vertical or
locis humidus circa Latacunga prope Quito," Ecuador]. slightly concave; eye overhung by heavily fleshy
Phryniscus laevis Giinther, 1858:43, pi. 3: fig. A [type locality: upper eyelid, upper lip not fleshy. Temporal area
"Panama. Quito. Chili"]; 1859:417.—Boulenger, 1882:151.
Phrynidium laeve.—Cope, 1867:196.
studded with pointed tubercles in females, or with
Atelopus laevis.—Cope, 1868:117.—Orton, 1871:693. fleshy, glandular bumps in males. Middorsum of
Atelopus ignescens Espada, 1875:139, pi. 1: figs. 7-9.—Boul- body completely free of warts, glands, pustules;
enger, 1880:46; 1898:119 Werner, 1901:600.—Peracca, either smooth or finely wrinkled. Dorsolateral area
1904:19.—Parker, 1934:268; 1938:440.—Rivero, 1963:109; and sides of body in females heavily studded with
1969:142. rounded glands that have one to several raised,
Phryniscus laevis exigua Boettger, 1892:22 [type locality:
"Zurucuchu, W-Andes of Cuenca, 3250 m, Ecuador"].
pointed spicules on them, giving spiny appearance;
A.[telopus] ignescens.—Nieden, 1926:83. same area in males covered with low, rounded

FIGURE 20.—A representative individual, JAP 4176, of A. ignescens Cornalia

NUMBER 145 25
glandular areas without spicules. Chin and belly color along a line running from corner of mouth
finely wrinkled and occasionally lightly shagreened along the lateroventral line. Ventral surfaces highly
on chest in males, same in females, but with ten- variable, from a light cream unicolor to light cream
dency in some individuals to have darkened spiny with yellowish or orange border to unicolor yellow-
area across chest. ish to unicolor orange. The eye is a deep black,
Forelimb short, stubby, and fleshy; humeral re- no pupil color can be distinguished. Often with
gion heavily studded with spicules in females, cov- a bright orange or yellow spot on ventral side of
ered by small rounded glands in males; forearm in crus, and some have same on dorsum of foot. In
both sexes heavily fleshy but quite smooth, con- this same series of 25, the following variation in
trasting strongly with upper arm. When forelimb color of the ventral surfaces was noted (individuals
is pulled back along body, tips of digits do not can occur in more than one tabulation): venter
reach past hind leg in males or past groin in fe- mostly dull orange-5; cream with orange in axilla
males. Entire forefoot fleshy, all digits free and and inguinal region-4; yellowish, with brightest
prominent, with fleshy webbing only at base. Sole yellow in limb insertions-2; yellow spot on dorsum
of foot covered with pustules, subarticular tuber- of foot-6; yellow spot on ventral surface of crus-8;
cles not differentiated from these pustules, palmar yellow venter with reddish orange in limb inser-
tubercle larger and clearly defined. Males have tions-2; unicolor yellow-2; venter of limbs macu-
horny excrescences on both first and second digits; lated with bright orange-1. A series from 5
first digit usually strongly bent under and flexed, kilometers west of Papallacta all had bright, to-
with prominent thick, rounded, fleshy hump dor- mato-red bellies in life but of course, now are
sally, keratinized area extends from tip of digit to identical with other alcoholics in belly color.
base; second digit with cornified area only at base. REMARKS.—Tadpoles presumably of this species
Hind limb short, stout, fleshy; femoral region have been described by Duellman and Lynch
studded with spines in females, heavily pustulate (1969:236). The material was taken at 3620 meters
in males, crural area smooth and fleshy. Tibiotarsal on the slopes of Cotopaxi Volcano. The tadpoles
articulation reaches to level of arm insertion when have a large suctorial disc behind the mouth which
carried forward along body; heels are well separated is used to anchor the individual to the underside
when tibiofibulae are parallel to femora at right of stones in swiftly flowing water.
angles to body. Foot fleshy with many pustules on It will be noted that I have continued the syno-
sole, webbing fleshy, extends to tips of digits as nym of A. carinatus Andersson with A. ignescens.
fringe except on first toe, which is almost com- Rivero (1969:142) has recently reexamined the
pletely hidden within webbing; webbing between holotype of Andersson's species, the only known
all digits, including first and second, deeply incised. specimen, and thought that the species should be
Subarticular tubercles not clearly distinguishable considered valid on the basis of several characters,
from pustules, outer metatarsal tubercle small but including the "nuchal tank" formed by the pro-
raised and prominent, inner metatarsal tubercle truding posterior margin of the frontoparietal and
larger but not as raised. the raised edges of the suprascapulae, the shape of
COLOR IN ALCOHOL.—Dorsum of head, body, and the snout, and the size of the eye. The first of these,
limbs uniform black, extending down sides and I am convinced, is the result of the way the speci-
ending abruptly at edge of belly, which is yellow men was preserved. The others are extremely
to yellowish white. Some individuals have spotting difficult to evaluate or quantify, seem of minimal
on chest. Ventral surfaces below anus and on limbs value taxonomically, and are, to my eye, at least,
often light brown to black; pigment distribution occasionally duplicated in large series of ignescens
ventrally differs considerably from one individual from a single locality. It should also be pointed out
to the next. Lower lip usually dark, same color as that the type locality, vague though it is, is an area
upper lip, and sharply distinct from rest of chin. of intensive activity by herpetological collectors,
COLOR IN LIFE (from a series of 25 individuals and no additional specimens resembling A. carina-
collected near Quito June 6, 1959).—Dorsum dark tus has ever been obtained in eastern Ecuador,
brown to black, lightening slightly on dorsum of while hundreds of animals looking very much like
limbs. Dorsal color sharply set off from ventral the type continue to be collected in the inter-
26 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Andean valleys. Until added material from the the bottom of the puddle and out the other side.
supposed type locality is available, I shall continue Experiments performed later with mated pairs in
to regard this weakly denned form as a synonym of the compound of Robin Copping's house in Quito
A. ignescens. demonstrated that once the male has mounted the
ECOLOGY.—A. ignescens is adapted to the most female he will remain in place no matter what the
rigorous climates found in Ecuador where am- female does until the eggs are deposited. This
phibians occur. It inhabits the high inter-Andean may be a matter of several weeks, during which
valleys, and has sufficient eurytopic abilities that it time the female feeds and moves about normally,
has crossed the inter-Andean nudos to invade but the male does not feed while he waits. The
neighboring valleys. It has also moved up the period of waiting for the male may not be so long
slopes of the highest mountains above tree line, under normal conditions, of course, but I suspect
where it is found in the cold, windy, and dry it is usually much longer than in other frogs.
paramos. It is tolerant of temperatures well below The activity period is diurnal, and I have never
levels at which other species of Atelopus become seen members of this species active at night. The
inactive, except perhaps A. bufoniformis. males seek out the females entirely visually. While
The species engages in mass migrations, appar- they do make a slight sound, it is a very low, chick-
ently at the beginning of the breeding season, and like peep, difficult to hear even at short distances,
can be seen in the hundreds at such a time. I have and made while walking about looking for a fe-
see two such migrations, the first of which was male, rather than made in a stationary position to
moving across the Pan-American Highway north which a female could be attracted by the "call."
of Latacunga on December 31, 1958. Large num- As they walk about they will occasionally pause,
bers of individuals were smashed on the road over raise themselves high on the forelimbs, and look
a distance of four or five miles. All living indi- about, scanning the horizon for other members of
viduals were crossing the highway in the same direc- their species. There is no sexual discrimination in
tion. Some mated pairs were seen. this scanning, because the individual moves off in
The second such migration was seen in the the direction of another frog as soon as the second
paramos on the road from Quevedo to Latacunga, is seen, and attempts to mount. If the second is a
on January 22, 1959, at an altitude of 3600 meters. male, the first is repelled by slow kicking and soft
In this case, the movement was again in a single peeping. If the second is a female, it can usually
direction, and definitely downhill. A light drizzly be successfully mounted. If the second is a female
rain was falling, and sufficient rain had already with another male already in mounted position,
fallen to make the road somewhat treacherous, the approaching animal is not put off for a second
with standing puddles. In both cases, the migra- by the fact of prior possession. He will, if per-
tion seems to be toward standing water, presumably mitted to do so, mount the first male, and clasp
the place of egg deposition, but this was not veri- him. Almost invariably, however, he is repelled by
fied by direct observation. the first male. Since all the movements by this
COURTSHIP AND MATING BEHAVIOR.—The migra- species are slow and deliberate, the repulsion al-
tion seen in the paramos included many clasped most appears to be in slow motion, but it still is
pairs as well as males ready to clasp. The males, effective. The mounted male raises his hind foot
always smaller than the females, clasp with an on the side of the approaching male, puts it very
axillary grip. I followed one such pair for almost deliberately in the face of the intruder, and softly
two hours, hoping to see egg deposition, but the but forcefully kicks him away. The intruder usu-
female simply wandered about in what appeared ally describes a gentle arc, landing on his back,
to be a random way, never pausing anywhere long and the mated pair moves away as he slowly strug-
enough for any action to take place. The rate of gles to turn over. If they have not gone far enough
progress is a slow, deliberate walk, never a hop or by the time he recovers his equilibrium, a second
a jump. At one time the female entered a small attempt will be made and the same sequence en-
puddle about five feet across and at most a foot acted.
deep, but she never broke either her stride or her SPECIMENS EXAMINED.—I have seen over 500 in-
pace, continuing to walk very deliberately across dividuals of this species, from all of the inter-
NUMBER 145 27

Andean valleys of Ecuador, from Tulcan south to almost parallel from nostril to point just before
the region of southern Cuenca Province. It seems corner of eye where each canthus flares outward;
of little value to list them in detail until someone area from snout tip along canthus, outer eyelid
has examined them more carefully for infraspecific border, and to corner of head raised and fleshy.
variation. Dorsum of head flat, shagreened; eyes prominent,
protrusive. Snout from side strongly protrusive
beyond lower jaw, with upper lip distinct from
Atelopus longirostris Cope
protrusion as well; nostril lateral, posteriorly di-
FIGURE 21 rected, slightly below fleshy canthus, and almost
directly above end of lower lip. Loreal region
Atelopus longirostris Cope, 1868:116 [type locality: "Valley of
Quito," Ecuador].—Orton, 1871:693.—Espada, 1875:155.—
sunken, eye overhung by fleshy eyelid margin; ver-
Nieden, 1926:81.—Guibe, 1948:32. tical ridge behind eye and horizontal ridge from
Atelopus varius [not Stannius] Boulenger, 1880:46. eye to corner of head; side of head and tympanic
Phryniscus longirostris.—Boulenger, 1882:153. region shagreened, with few larger, white-topped
Phryniscus boussingaulti Thominot, 1889:28 [type locality: pustules. Dorsum and sides of body heavily sha-
"Entre Latacunga et Guayaquil, au sud de Quito," Ecua- greened, with scattered white-topped pustules on
dor].
sides only, belly almost smooth or with finely wrin-
HOLOTVPE.—Apparently lost. Not located in kled surface. Vertebral column forms distinct,
either the Academy of Natural Sciences of Phila- raised middorsal ridge from shoulder level to coc-
delphia or the National Museum of Natural His- cyx.
tory. Forelimb very slender, forearm more fleshy than
DESCRIPTION.—Snout from above strongly pro- humeral region, pustulate on shoulder and dor-
jecting, with comparatively sharp point, distinct sum of humeral region, shagreened elsewhere. En-
inward curve from snout tip to nostril, canthi tire hand somewhat fleshy, heaviest around first

FIGURE 21. A representative individual, USNM 193834, of A. longirostris Cope.

28 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

digit, which is distinct although almost totally en- eggs show brightly through the belly flesh, giving
meshed in fleshy webbing. Tips of digits slightly a strong orange tint to ventral surfaces.
swollen; second digit shorter than fourth; second, ECOLOGY.—I have collected this species in several
third, and fourth digits almost free of webbing. places on the western slope, and always found the
Subarticular tubercles present only at digit base, specimens in rocky areas on the edges of streams.
weakly developed; large palmar, small plantar tu- I have found individuals under rocks and also out
bercle. walking about in and over the rocks in full sun-
Hind limb long, slender, shagreened but without light. The habitat is not usually a shaded one, but
pustules; tibiotarsal articulation reaches points be- is very exposed. A pair in amplexus was taken at
tween nostril and posterior corner of eye when Apuela on April 17, 1959, which was at the end
carried forward along body; heels overlap or touch of the rainy season. The female was heavy with
when tibiofibulae are parallel to femora held at eggs.
right angles to body. Fleshy webbing of foot ex- COMPARISONS.—An obviously close relative of
tends to swollen tips of digits, but is incised be- longirostris is A. varius glyphus, described by Dunn
tween digits; first digit completely hidden in (1931). The type and paratype series are in the
webbing, tip of second distinct. Subarticular tu- National Museum of Natural History, and I have
bercles present but weak; outer metatarsal tubercle compared them directly with the series of longi-
raised, prominent; inner metatarsal tubercle rostris available from Ecuador. In addition to the
elongate, flattened, kidney shaped. general similarity in color pattern, both species
COLOR IN ALCOHOL.—Dorsum brown, from red- have rather slender bodies, elongate, thin limbs,
dish brown to very dark brown, snout tip light and a comparatively prominent, overhanging
brown, sharply distinct from top of head, which snout. It is possible that the relationships of
usually has few scattered white spots, characteris- glyphus are not with varius at all, and it should be
tically one on back of each eyelid. Dorsum with regarded as a distinct species. The two paratypes
scattered, often bilaterally paired light spots; dor- from Porto Obaldia, noted as quite different by
solateral area over shoulder usually with elongate Dunn (1931:396), are very reminiscent of A. varius
white spot, almost forming stripe in some indi- loomisi Taylor (1952).
viduals, usually followed dorsolaterally by one or I have examined the holotype of Phryniscus
two additional elongated spots or streaks. Sides boussingaulti Thominot, and see no reason not to
usually unicolor, occasionally lightly vermiculated, continue its synonymy with longirostris. The speci-
abruptly changing into very light, immaculate yel- men is badly faded dorsally, retaining no trace of
low of belly, from chin to anus, with occasional pattern. There is a series of clear, rounded, whit-
individual showing brown vermiculations ven- ish spots laterally, which, although not typical of
trally. Legs as body, usually unicolor brown above, longirostris, is almost duplicated in some of my
spots laterally and dorsolaterally, light yellow be- specimens (e.g., USNM 193833). Photographs of
low, posterior surface of thigh usually vermicu- the holotype were taken and are on deposit at the
lated. Tips of digits yellow, soles of feet usually National Museum of Natural History.
dark, but some with yellow areas. REMARKS.—The type locality of this species,
given as the "Valley of Quito" by Cope, is open
COLOR IN LIFE.—Dorsum blackish brown, shad-
to the same question as for the other four species
ing into a light chocolate brown low on sides.
described from the same place and collected by
Spots on head, back, sides, and dorsum of legs dull
the Orton Expedition. Although one of these spe-
yellow. Those on back of femur white. Tip of
cies (Leimadophis alticolus) has been found since
snout yellowish brown, brighter than rest of spots.
in the immediate environs of Quito, the other three
Series of tiny white stipples along sides and be-
(Bufo andianus, Dryadophis pulchriceps, and
tween eye and arm, behind temple. Eye black all
Dipsas oreas) are clearly western slope species, usu-
over. Belly white, not flesh colored. Tips of digits ally at altitudes well below that of Quito. The
very light orange. Stripe along upper lip yellow road to Nanegal, on the western slope, is an easy
as dorsal spots, but grades into white of belly near one to travel, and a three-hour drive today puts
corner of mouth. In a female heavy with eggs, the one into the range of all of the species. I believe
NUMBER 145 29
that a local collector brought into Quito the ma- that Rivero (1963:112) was correct when he indi-
terial that eventually served as the types of these cated that the Colombian specimens he tentatively
species for Cope, just as today several local collec- assigned to longirostris (CNHM 43850-51) proba-
tors still find valuable and interesting material bly would not prove to be conspecific with that
around Mindo, Tandayapa, and Nanegal. species. Part of his doubt came from the difference
This problem is complicated by the fact that Or- in known altitudes for Quito and Cisneros, in
ton (1871) indicated that the type came from Ha- Colombia, but I have shown above that Quito is
cienda Antisana, which he said was at an altitude not a valid locality for longirostris. The series of
of 13,300 feet. This locality, according to Cope specimens from Santander called longirostris by
(1868:96) is on the western slope of Volcan An- Cochran and Goin (1970:143) unequivocally do
tisana. Cope indicated that the total collection at not belong to that taxon. Therefore, longirostris
this locality was "a small black frog," a descrip- is not known from Colombia at the present time,
tion more appropriate to A. ignescens than longi- but is confined to the lower western slopes of the
rostris. I think we can safely assume that Orton Andes in Ecuador.
was depending upon his memory when he indi- A series of specimens in the American Museum
cated the type locality, and in this case it was (AMNH 49943-46, 49948-50) are recorded as hav-
faulty. It is of additional interest to note that the ing been collected at "Mera, Rio Pastaza," which
USNM catalogs do not include any records of is a town on the principal road descending from
specimens from Antisana collected by Orton, so it the high Andes to lowland Amazonia. This is
appears that the "small black frog" from there was such a zoogeographic unlikelihood that I challenge
deemed too insignificant to catalog. the validity of the original data rather than record
A series of specimens collected by P. Hershko- the occurrence of the species on the Amazonian
vitz in 1935 at Carolina, Imbabura Province, at slope. The records show that the specimens were
900 meters, is tentatively assigned to this taxon. purchased from a professional collector, and the
In general appearance and morphology they closely history of Ecuadorian herpetology is replete with
resemble longirostris, but they are all quite small. erroneous records based on such purchased ma-
The difficulty in calling them longirostris is caused terial. The specimens are in my opinion clearly
by their color pattern. They are quite similar to members of the species longirostris, and I am sure
each other, in the occurrence of many more that they came from the lower western slopes of
rounded and more closely approximated yellow the Andes in Ecuador. In addition to the fact that
spots dorsally than is typical of longirostris, and the locality does not make zoogeographic sense,
the large, elongate spot immediately behind the it should be pointed out that collectors have been
head dorsolaterally, so characteristic of longirostris, delivering specimens to European, United States,
is absent in all four specimens, as is the lighter and Latin American collections from the Mera
coloration of the projecting snout. There is a region for over 200 years, since it lies on one of
more or less distinct, continuous yellow dorsolat- the main trails to the east, but there are no records
eral line on the specimens, while in longirostris of other specimens of this species from the area.
the same area is occupied by a discontinuous series I do not include this locality as part of the known
of spots. The specimens are the northernmost rec- range of the species until verification of its occur-
ord for the species as I define it, and may show rence there by further collections has been made.
incipient divergence from the parent taxon. The SPECIMENS EXAMINED.—ECUADOR (PM 207,
possibility that the specimens represented the taxon holotype of boussingaulti): Esmeraldas Province;
called Atelopus longirostris marmorata by Werner Alto Tambo (USNM 193853), Carondolet (USNM
occurred to me, but the dorsal pattern cannot be 193847-48). Imbabura Province; Apuela (USNM
called "marmorate," in my opinion, although it is 193819-21, 193823); near Penaherrera (UMMZ
closer to that than are typical longirostris, and I 83662-66). Pichincha Province; Chontapamba
have considered Werner's taxon unidentifiable (see (MCZ 84203-04, USNM 193834, 193854-64),
elsewhere, under incertae sedis). Llambo, on road to Gualea (BMNH 1972.737,
The large series of specimens available from USNM 193824-31), Mindo (UMMZ 55516, USNM
Ecuador provides a sufficient basis to demonstrate 193832, 193841, 193866), 10 km northwest of Nane-
30 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

gal Chico (USNM 193817), Nanegal Chico adult females usually shorter than 30 millimeters
(UKMNH 108942-13, UPR-M 5090-91, USNM in standard length. Males usually shorter than 20
193833, 193837-10), Rio Blanco (USNM 193846, millimeters. First digit on forefoot almost entirely
193852), Rio Mindo (USNM 193842-43), Rio hidden in webbing; ventral surfaces unicolor, no
Toachi (USNM 193849), San Tadeo, near Mindo spotting, dorsum with a broad yellowish brown
(USNM 193865), Tandayapa (BMNH 1972.736, band and reddish brown spotting and marbling.
CAS-SU 11388-89, USNM 193822, 193835-36, DESCRIPTION.—Snout from above bluntly
193851-52). Province unknown: Miligali (USNM pointed, with canthus forming straight line from
193844-45). tip of snout to nasal eminence, from that point
Tentatively allocated to taxon: Ecuador: Imba- to corner of eye it forms a gentle inward curve.
bura Province, Carolina, .900 m (AMNH 50969, Snout and edge of canthus fleshy anterior to nasal
UMMZ 83656, 83658-60). eminence, but no fleshiness on canthus from nostril
to eye, on eyelid, or on temporal region. Slight
depression in center of snout, as consequence of
Atelopus mindoensis, new species fleshy margin; rest of dorsum of head flat. Snout
FIGURE 22 from side prominent, projecting over lower jaw;
nostril closer to tip of snout than eye, on raised
HOLOTYPE.—USNM 193554, from Mindo, Pichin- area and directed laterally; loreal region vertical,
cha Province, Ecuador, 1200 meters, collected by gently rounded, no loreal depression anterior to
Manuel Olalla, May 1959. eye. Temporal region with scattered pustules and
DIAGNOSIS.—A diminutive member of the longi- spines, differing in number and prominence be-
rostris group, most similar to A. palmatus, with tween individuals. Body liberally studded dorsally

FIGURE 22— Holotype, USNM 193554, of A. mindoensis, new species.

NUMBER 145 31

and laterally with tiny pustules widely scattered low color of foot. No spotting of any sort on throat,
but at least in part organized into tracts; scattered belly, or soles, except in one female with few light
larger, pointed, usually pigmentless, rounded pus- brown spots on chin. Side of head dark brown, in-
tules along dorsolateral line and laterally; limbs cluding upper lip; area below eye often with light
also studded with tiny spicules, concentrated in yellow crescent. Lower jaw entirely yellow.
raised, fleshy areas along length of limbs. HOLOTYPE.—The holotype conforms with the
Forelimb slender in females, shorter and heavier description given above for the species, where no
in males, forefoot has somewhat fleshy webbing, variation is mentioned. Where alternatives appear
with first digit completely covered by it, second in the species description, the holotype has the
digit included by a border of web; webbing deeply following characteristics: temporal region with
incised between second and third and between about six large, white-tipped pustules and many
third and fourth digits, with fleshy nature of web- scattered spicules. Heel reaches to the posterior
bing making it appear absent between these three margin of the head when hind limb is brought for-
digits. No subarticular tubercles on slightly fleshy ward; heels overlap slightly when tibiofibulae are
ventral surfaces of digits; plantar tubercle low, parallel to femora at right angles to body. Dorsum
rounded, almost indistinct. almost unicolor, with irregular, somewhat lighter
area dorsolaterally that is vermiculated on its inner
Hind limb slender, heel reaches to points be-
margins. Clear yellowish spots present on lower
tween posterior margin of eye and axilla when
sides, some fused with light yellow ventral color.
hind limb is brought forward along body; heels
Eggs visible in body cavity. Ventral surfaces of
touch or slightly overlap when tibiofibulae are limbs dark brown, with few scattered lighter spots.
parallel to femora at right angles to body. Web- Light crescentic area present below eye, lower lip
bing on foot extensive and somewhat fleshy, with with dark brown margin, rest of chin and throat
first toe almost completely indistinguishable within clear. Measurements (in millimeters): snout to
it, second and third digits completely and continu- vent: 29.0; snout to end of coccyx: 27.2; knee to
ously webbed; fourth toe webbed to tip, but web- knee: 22.0; tibiofibula: 11.3; head length: 8.6; head
bing is hardly more than a border on ultimate width: 7.7; eye diameter: 4.0; snout to eye: 5.2;
phalanx; full webbing on fifth digit. Ventral sur- interorbital distance: 3.8; width of eyelid: 3.9.
face of foot is devoid of tubercles; fairly large, ETYMOLOGY.—The species name is derived from
raised, prominent, pigmentless outer metatarsal the name of the type locality, a most pleasant little
tubercle present. town where Beatriz Moisset and I spent several
COLOR IN ALCOHOL.—Dorsal ground color rather delightful days in the field in 1965.
light yellowish brown, forming broad band length SPECIMENS EXAMINED.—The following are para-
of back, filled with reddish brown marbling, spot- types: ECUADOR: Pichincha Province; Milpe, 900
ting, and vermiculation in younger individuals, m (USNM 193562-63), Mindo (BMNH 1968.115,
practically totally obscured in older individuals, UKMNH 108944-45, UMMZ 131684, USNM
which tend to have back almost unicolor reddish 193556-59), Rio Blanco, near mouth of Rio Yambi,
brown. Some individuals have irregular stripe from about 700 m (CAS 134204, USNM 193561), Rio
eye to groin, bordered below by lighter stripe be- Lelia (AMNH 79462), region below Sigchos (MCZ
ginning at posterior margin of orbit and disappear- 84200, USNM 193564), Rio Toachi(USNM 193555,
ing on sides. Sides are usually reddish brown, 193560). The following are not designated as para-
sharply distinct ventrolaterally from belly color; types: ECUADOR: Esmeraldas Province; Carondo-
some specimens have series of clear, yellowish spots let (USNM 193570), Alto Tambo (BMNH
on lower sides. Venter is clear light yellow, usually 1968.116, USNM 193571-72, UPR-M 5092). Pichin-
quite transparent, with eggs visible in body cavity cha Province: Mindo (USNM 193565), region be-
of ripe females. Ventral surfaces of limbs often low Sigchos (USNM 193569), region of Gualea
completely invaded by dorsal color, but in some (USNM 193566-68), Tandayapa (GOV 8696,
individuals yellow of belly is continuous with yel- USNM 193573).
32 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Atelopus nepiozomus, new species and raised, with nostril below it on a lateral promi-
nence; snout area between canthi slightly de-
FIGURE 23
pressed. Upper eyelid fleshy, followed by fleshy
HOLOTYPE.—USNM 193543, female, from ridge from corner of eye to back of head. Frontal
slightly above Suro Rancho (also called Guale- and occipital areas flat, markedly lower than fleshy
cenita), a single house on the trail between Limon margins of head. Snout viewed from side blunt,
and Gualeceo, approximately 16 kilometers airline slightly projecting beyond lower jaw; nostril closer
west of Limon (which is now known as General to tip of snout than eye, located on lateral emi-
Plaza), Morona-Santiago Province, Ecuador, ap- nence, directed laterally and slightly posteriorly,
proximately 3000 meters, collected by James A. loreal region distinctly concave. Area behind eye
Peters, August 23, 1962. See Figure 24 for map to shoulder with raised, rounded warts, each of
showing locality. which may have several white-topped spinules on
DIAGNOSIS.—A member of the ignescens group, it. Dorsum and sides of body and legs with similar
very similar to ignescens, bomolochos, and arthuri, low, rounded warts (more prominent in females
differs from all three in presence of scattered dark than in males), which may bear one to many whit-
spots ventrally; dorsum with many low, rounded ish spinules or lack them entirely; generally more
warts and pustules which may be spiculate; dorsum and larger warts laterally than dorsally. Chin,
usually spotted and variegated with darker brown throat, belly, and lower surfaces of limbs usually
on lighter ground color; also very similar to hali- finely wrinkled, but devoid of warts and spinules.
helos (for contrasting characteristics, see diagnosis Forelimb slender in females, heavier in males;
of halihelos). fingertips do not reach beyond body when leg is
DESCRIPTION.—Snout viewed from above roundly carried back along body. Webbing very reduced
pointed, with canthus forming slight inward curve on forefoot, wTith first digit clearly distinguishable;
from tip of snout to nostril and then much deeper webbing appears completely absent, but is actually
inward curve from nostril to eye; canthus fleshy very fleshy between digits at base; greatest extent of

FIGURE 23—Holotype, USNM 193543, of A. nepiozomus, new species.

NUMBER 145

webbing between first and second digits, where it HOLOTYPE (Figure 23).—The holotype agrees
includes about half of first digit. First digit quite with the species description above in all characters
short, fourth longer than second, third much longer for which no variation is mentioned. For those
than fourth. Many low, rounded, fleshy tubercles characters which vary, the holotype is as follows:
on sole of forefoot and at digit articulations; outer temporal and tympanic areas and sides of body
metatarsal tubercle larger than others, but not pro- with white-tipped spinules on warts, those on body
nounced or raised. not multiple on a single wart. Warts on legs lack
Hind limbs short; tibiotarsal articulation reaches white tips, except very near arm and leg inser-
to arm insertion or slightly farther when carried tions. Tibiotarsal articulation reaches point of
forward along body; heels do not overlap when insertion of forelimb; heels separated when tibio-
tibiofibulae are parallel to femora at right angles fibulae are parallel to femora at right angles to
to body. Webbing extends to tips of all toes, but body. No definable dark stripe from eye to groin,
is deeply incised between all except first and sec- whole side unicolor dark brown. Ventral surfaces
ond, where it is complete and includes all of first of feet with dark areas. Measurements (in milli-
toe; webbing forms distinct border on all toes, nar- meters): snout to vent: 32.4; snout to end of coccyx:
rowing toward tips, giving tapered appearance to 29.3; knee to knee: 24.9; tibiofibula: 11.5; head
individual toe. Low, rounded tubercles at articu- length: 11.0; head width: 9.5; eye diameter: 4.8;
lations and on sole; inner metatarsal tubercle low snout to eye: 6.9; interorbital distance: 5.2; width
and rounded as others, outer metatarsal tubercle of eyelid: 4.0.
smaller but distinctly raised, pointed, and hard. REMARKS.—All of the specimens in die type series
COLOR IN ALCOHOL.—Ground color of dorsum of were collected during our 1962 field work, which
legs and body pinkish gray, with marbling, spot- was made up of a series of transects on the eastern
ting, and variegations of reddish brown. Ground slopes of the Andes. Since the routes followed were
color usually predominates, and in one individual often only mule trails, it is not easy to identify
darker area is confined to thin pencil stripe of stopping places and collecting localities where we
color on vertebral column. Darker areas associated worked during the week or so that a descent or
with warty swellings, so that where wartiness is ascent might take. Figure 24 shows all of our col-
heavier there is greater amount of darker marbling. lecting spots during the final transect, which took
Sides of head dark brown, with distinct and sharp us down one mule trail and back up a second a
color change at canthus and along dorsolateral litle farther south. Most of the names on the map
ridge of head. In few individuals a poorly defined are "tambos," or single houses which provide over-
stripe of ground color from eye along dorsolateral night facilities for mule drivers, and cannot be
margin to groin, usually not present or indistinct. found on any other maps or in other gazetteers of
Ventral surfaces of body yellowish to yellow-green, Ecuador. They are, however, comparatively perma-
with some spots or streaks of dark brown usually nent, because they occur at approximately one
present, most commonly on chin, throat, and chest; "mule-day" intervals on the trail, and require a
where thin, midline streak may broaden into one good-sized flat area for a house and a corral. Only
or several spots in tandem. Triangular, diaper-like when highways are open to towns such as General
patch extending from anus onto ventral surface of Plaza, Macas, and others in the lowlands, will the
mule traffic stop, and these houses will revert to
thighs. Dorsal surfaces of feet similar to body and
the forest.
legs, but about half of specimens have first three
toes and first two fingers distinctly lighter than ETYMOLOGY.—The species name is derived from
rest of foot, in some becoming quite yellow. Ven- the Greek word nepios, meaning "infant," and
tral surfaces of feet usually unicolor light yellow, zoma, referring to a girded garment or the position
but in some individuals may be dark spotted. of the girdle, i.e., the girded garment of an infant,
as close as one can come to "diaper" in Greek. It
COLOR IN LIFE.—Dark greenish above with dark refers to the patch of dark pigment that encloses
brown spots. Ventrally orange on thighs, arm in- the anus, upper thighs, and the posteriormost part
sertions, and chin. Belly yellowish. Spots on throat of the venter of these frogs, which gives them the
and below anus very dark brown. look of a diapered baby.
 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

• •?< J I
SAN VICENTE^

! SAN, JUAN IBOSCOJ PLAN DE MILAGRO

—

FIGURE 24.—Route of collecting transect made in August 1962, by J. A. Peters and party,
showing localities along mule trails from whence material was obtained. Contour lines are
in meters. (Based on the Mapa Geografico de Ecuador, published by the Instituto Geografko
Militar, Quito, 1950.)

SPECIMENS EXAMINED.—The following are desig- Schmidt (1858) gave additional information con-
nated as paratypes: ECUADOR: Morona-Santiago cerning the type locality, stating it to be "Westen
Province; Cerro Negro, 3450 m (USNM 193544), von Neu-Granada, bei Bonaventura, in einer Hohe
near Pailas, circa 2150 m (USNM 193545), San von 5000'."
Juan Bosco, 2000 m (BMNH 1972.735, AMNH DESCRIPTION.—Snout from above blunt, rounded,
79463, USNM 193546-47), San Vicente, circa 3450 not projecting, distinct angle in canthus at nostril,
m (UKMNH 108946, USNM 193549-51), Suro canthi from nostril to corner of eye form straight
Rancho, 3400 m (USNM 193548). The following line, angled slightly outward. Canthus and outer
were examined but are not considered paratypes: eye margin slightly fleshy, with distinct ridge from
ECUADOR: Morona-Santiago Province; San Vi- posterior corner of eye to end of head. Dorsum
cente, circa 3450 m (USNM 193552-53, JAP of head flat, without ridges, tubercles, or fleshy
7972-73). prominences; eye does not protrude dorsally. Snout
from side slightly protrusive beyond end of lower
Atelopus pachydermus (Schmidt) jaw; nostril directed laterally, slightly below can-
thus; loreal region weakly concave; eye slightly
FIGURE 25
overhung by fleshy edge of eyelid; temporal region
Phirix pachydermus O. Schmidt, 1857:1") [type locality: "Neu- without tubercles or spines, but in some individuals
Granada"]; 1858:257, pi. 3: fig. 26. with glandular swellings. Dorsum of body broad,
A.[telopus] pachydermus.—Rivero, 1963:107; 1968:23. flat, with minute wrinkling and folding; in some
Atelopus pachydermus Rivero, 1965:138.
individuals scattered, rounded, glandular promi-
HOLOTYPE.—In the Zoological Museum of Cra- nences. Sides of body studded with rounded,
cow, Poland. Present condition not known. glandular swellings, which are usually covered with
NUMBER 145 35

FIGURE 25.—A representative individual, USNM 193495, of A. pachydermus Schmidt.

several to many whitish spinules. Chin, chest, and ried forward along body; heels do not overlap when
belly wrinkled or lined, but without pustules. tibiofibulae are parallel to femora at right angles
Forelimb short, stout in males, more slender in to body. Entire foot with somewhat fleshy web,
females, forearm more fleshy than upper arm in covering all but tips of first and second toes; ex-
both sexes; dorsum of arm with glandular swellings, tends to tips of other digits, but is deeply incised
rest smooth with wrinkles or folds. When arm is between them. No subarticular tubercles, no tuber-
pulled back along body, digits reach level of hind cles on sole, outer metatarsal tubercle low, rounded,
leg perpendicular to body in females, not so far somewhat indistinct, inner metatarsal tubercle
in males. Forefoot slightly fleshy, webbing very ovate and low.
reduced, with all digits distinct and free; first digit COLOR IN ALCOHOL.—The typical pattern for this
more swollen than others, wth horny excrescence species is shown in Figure 25. It may be described
from tip of digit to base in males, covering a raised in general as: ground color or entire body cream-
hump of flesh, second digit with cornified area near yellow, with irregular black vermiculations and
base of digit. Subarticular tubercles absent, but spots dorsally, which may be so heavy as to cover
palm with many raised glandular areas, often with head and back completely. Sides, venter, and legs
pair at base of each of second and third digits; also spotted, stippled, and vermiculated with black,
large palmar tubercle, less prominent plantar with belly and undersides of limbs less spotted
tubercle. Tips of digits slightly swollen. than elsewhere. Color patterns extremely variable.
Hind limb rather stout, occasionally glandular COLOR IN LIFE.—The lighter dorsal areas are
along femora, more often heavily wrinkled or bright yellow. Many of the warts and pustules have
folded. Tibiotarsal articulation reaches to points white tops. The ventral surfaces are white in al-
between end of head and arm insertion when car- most all individuals, but some may have orange
36 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

areas or spots, and some have red bellies, similar present in adults, there is a distinct ontogenetic
to those seen in specimens of ignescens from nearby color change. Individuals 27 millimeters or less in
localities. body length lack the strong contrast between the
HABITAT.—This high altitude species is ex- dark and light areas seen in adults. The dorsal areas
tremely common in the vicinity of Papallacta, that presumably will be yellow in adults are dark
where they were found along the borders of the Rio gray or light brown in juveniles, and the darker
Papallacta, under rocks and stones in moist, spongy, areas are not strongly differentiated from the
runoff areas. lighter. Some juveniles appear almost totally black
VARIATION.—All species of Atelopus tend to be dorsally. Ventral color in the juveniles is the same
variable in coloration, but the available material in adults.
of A. pachydermus shows a striking amount of REMARKS.—Rivero (1965:138; 1968:23) has dis-
variability. Figures 26 and 27 show a series of adults cussed the use of the name Atelopus pachydermus
from the immediate environs of Papallacta, ar- (Schmidt) for the population described here. He
ranged to show the decrease in amount of light examined the AMNH material from Oyacachi, and
pigment in the series. While no individuals were tentatively assigned it to this taxon, and I have
taken in the region of Papallacta which lack the followed him in this action. As Rivero pointed
dark pigment entirely, some of the specimens from out, however, there are problems involving the type
the region of Oyacachi are apparently entirely locality of Schmidt's taxon. The holotype of pachy-
free of any dark spotting and mottling, and in life dermus was said to have come from above Buena-
must present a striking golden appearance. In ventura, at 5000 feet, presumably in the province
addition to the variation in the amount of pigment of Valle, on the western slope in Colombia. All

FIGURE 26.—A scries of specimens from the environs of Papallacta selected to show the variability
in dorsal pattern and distribution of colors in A. pachydermus. Numbers refer to field series, and
all have been recataloged.
NUMBER 145 37
of my material is from the eastern slope in Ecuador, is certainly appropriate to my specimens, and I use
almost entirely at altitudes considerably above that it for them temporarily, until further evidence
of the type locality. I consider it most unlikely that accumulates on the taxon.
any Atelopus species is capable of maintaining a A series of specimens from the collections of
genetic continuum between these two areas, and if Gustavo Orces-V. (GOV 8698-700) are recorded
the existence of a species answering the description as collected at Loja, Ecuador, but I am sure this
of pachydermus in the region of the type locality is an error in the transcription of the data, per-
is verified, my material will probably need a new" haps made by me when transferring that collection
name. I continue to use the name pachydermus to the National Museum. The series is unicolor
here for two reasons. First, the species has never light golden, and is identical with material from
again been taken at or near the type locality, nor Oyacachi. I do not believe the same taxon, and
has anything resembling it shown up there. It is the same striking variation in pattern, also occurs
entirely possible that the type may have come
in Loja, and I have not included that locality in
from the eastern slope, not the western, of Colom-
the following list.
bia, with either an erroneous locality given in the
SPECIMENS EXAMINED.—ECUADOR: Napo Prov-
description by Schmidt, or, as Rivero (1968:23)
suggested, there is a village called Buenaventura ince; 2 km E of Chalpi, 2755 m (USNM 193520-
on the eastern slopes of Colombia. Second, my spe- 21), Oyacachi, 3300 m (AMNH 20505-21, BMNH
cimens do fit the original description fairly well, 1972.732-733, CAS-SU 10348-50, 134200, JAP 3791,
and the series includes specimens very similar to MCZ 84201-02, USNM 193494, 193522-33), Pueblo
the holotype in coloration. They are found at Viejo, Oyacachi (AMNH 20492-93), Papallacta
comparatively high altitudes in northern Ecuador, and environs, 3150 m (KU 117876-79, 142950-56,
and there seems to be a good likelihood that the UMMZ 131020-21, UPR-M 5088-89, USNM
same species could range into Colombia. The name 193495-518).

FIGURE 27.—Ventral view of same series as in Figure 26.

38 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Atelopus palmatus Andersson lateral ridge behind eye. Side of head shagreened,
without raised tubercles or glandular areas. Dorsum
FIGURE 28
and sides of body lightly shagreened; vertebral
Atelopus palmatus Andersson, 1945:12, fig. 4 [type locality: column forms indistinct ridge. Chin and belly
"Rio Pastaza, Ecuador. 1000 to"].—Rivera, 1963:110; 1968: finely shagreened or lightly wrinkled and folded.
21; 1969:142. Forelimb long, slender; forearm heavier than
SYNTYPES.—Stockholm Royal Museum number humeral region, dorsum lightly studded with small
1908, 5 individuals. pustules or finely shagreened, ventral surfaces
DESCRIPTION.—Snout from above with pointed, smooth or finely wrinkled. Forefoot with webbing
enclosing all but tip of first toe, weakly developed
projecting tip; canthus forms straight line from
on other digits. Subarticular tubercles absent or
snout tip to level of nostril and then forms inward
very weakly developed, palmar tubercle large,
curve from nostril to corner of eye. Snout, can-
prominent, no plantar tubercle. Edge of prepollex
thus, and eyelid raised, but weakly fleshy; bony with horny excrescence in males.
ridge from posterior corner of eye to corner of head, Hind limb slender, lightly pustulate along dor-
ending in raised boss. Dorsum of head finely sum of upper leg, smooth elsewhere. Tibiotarsal
shagreened, forming a plane surface, eyes slightly articulation reaches to points between nostril and
raised above plane. Snout from side protrudes well anterior border of eye when carried forward along
beyond both upper lip and lower jaw, may be body; heels touch or overlap when tibiofibulae are
slightly concave laterally, and may curve slightly parallel to fermora held at right angles to body.
downward at tip, giving "hooked" appearance. Webbing extends to tips of toes, with first and
Nostril lateral, slightly below canthus, loreal re- second digits almost totally obscured; webbing
gion slightly concave. Vertical ridge joins dorso- rather deeply excised between other toes, reaches

FIGUXE 28—A representative individual, FNHM 36657, of A. palmatus Andersson.

NUMBER 145 39

tip of fourth toe as a fringe. Subarticular tubercles (UKMNH 122544), Rio Azuela, 1740 m (UKMNH
weak and occasionally indistinct, outer metatarsal 142957-64).
tubercle prominent, inner metatarsal tubercle
absent.
Atelopus planispinus Espada
COLOR IN ALCOHOL. — Dorsal ground color
chocolate-brown, dorsum of head spotted and varie- FIGURE 29
gated in light gray, usually with dorsolateral streak Atelopus planispina Espada, 1875:148, pi. 2: figs. 1-2 [type
or series of spots of same color from eye to groin, locality: "San Jose de Moti, Ecuador"].
scattered spots with irregular margins on back. Atelopus longirrostris [sic].—Espada, 1875: pi. 2: fig. 1 [see
All light gray areas bordered with narrow dark below, in "Remarks"].
brown to black margin. Ground color extends down Phryniscus cruciger.—Boulenger, 1882a: 154 [a questioned syn-
onymy].
side of head and body about halfway, with lips, A.[telopus] cruciger Nieden, 1926:84 [a questioned syn-
shoulder, and lower sides as well as all ventral onymy].
surfaces unicolor light brown to yellow. Limbs as A.[tetopus] planispinus.—Rivero, 1963:107.
body dorsally and ventrally; dorsum of first digits
on both fore and hind foot usually yellowish, with SYNTYPES.—Twenty-eight specimens, MM 212,
no brown pigment dorsally. Museo Nacional de Ciencias Naturales, Madrid,
Spain.
COLOR IN LIFE.—"Dorsal surfaces dark brown
with bright green spots. Flanks orange-yellow. DESCRIPTION.—Snout from above with rounded,
protrusive tip, with sharp angle at nostril; canthus
Ventral surfaces, including inner 2 digits of fore-
rostralis from nostril to angle of eye forms straight
and hindlimbs, deep yellow. Groin bright orange-
line angling outward from anterior to posterior;
yellow. Iris black with minute greenish yellow
no heavy fleshy margin on canthus. Dorsum of
flecks and ring around pupil." (Taken from field
head flat, in same plane as broad, flat dorsum of
notes on KU 122544, made by W. E. Duellman.)
body; bony ridge from corner of eye to end of
REMARKS.—Rivero (1969:142) has examined the
skull above tympanum, ending in boss; head
syntypes of A. palmatus, and he indicated that the
shagreened with tiny spines, heavier on eyelids.
species was valid, but closely resembled A. planis- Snout from side overhangs lower jaw, ending in
pinus, a larger species from the slopes of Mt. heavy point, which may be bent somewhat down-
Sumaco. I agree with this conclusion and, as indi- ward, i.e., "hooked," nostril laterally directed,
cated by the list below, have found a good series below canthus, loreal region vertical, does not form
of specimens in the various museums. Rivero concave area. Eye slightly overhung by moderately
mentioned that the posterior margin of the fronto- fleshy margin of eyelid. Upper lip and head behind
parietal and the upper margin of the suprascapulas eye spinose and pustulose. Entire body finely
are raised and prominent in some of the syntypes, speckled with tiny spinules visible under high
but this is not evident in the series I have examined power of microscope; on sides and in groin spines
except in individuals either dried or preserved in are much more prominent and pustules are also
a solution with too high an alcohol content. present.
Little is known about the ecology of this species.
Forelimb slender, elongate, shagreened with
A specimen collected by W. E. Duellman was taken
spicules and spines. When forelimb is pulled back
on the margin of a small stream during daylight
along body, palm reaches vent. Hand fleshy; first
hours.
digit enclosed in fleshiness except for its tip; other
SPECIMENS EXAMINED.—ECUADOR: Napo Prov- digits free or with slightly fleshy border, not
ince; El Chaco, Upper Rio Quijos, 1454 m (USNM webbed. Second digit shorter than fourth. Subar-
193541-42). Pastaza Province; Abitagua, 1200 m ticular tubercles absent or very ill defined, larger
(UMMZ 90367), Baiios (FMNH 28079), Hacienda palmar tubercle, no tubercle at all at base of first
San Francisco, on Bafios-Mera trail (FMNH 36657, digit.
152136), Mapoto (FMNH 36687), Puyo (FMNH Hind limb slender, covered with spicules and
28078), Sarajacu (AMNH 33891-92), southern slope scattered spines, not pustulate. Tibiotarsal articu-
of Cordillera Du£ above Rfo Coca, 1150 m lation reaches to points between nostril and pos-
40 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 29. A representative individual, AMNH 22198, of A. planispinus Espada.

terior margin o£ eye when carried forward along SYNTYPES.—The 28 syntypes in the Madrid Mu-
body. Heels overlap when tibiofibulae are parallel sum are badly faded, and a few appear to have
to femora at right angles to body. Entire foot been dried. Most of them have lost all signs of
fleshy, webbing almost entirely encloses first and pattern, but one small male still retains most of
second toes; webbing extends to tip of other toes its color, and it is possible to see enough parts of
as fringe, but is deeply incised between them. Sub- the pattern on other individuals to be sure that
articular tubercles weakly defined, outer metatarsal the more recent specimens identified here as plani-
tubercle forms rounded hump, inner metatarsal spinus belong to the same species. Insofar as they
tubercle obscured byfleshinessoffirsttoe. can be determined on the type series, the charac-
COLOR IN ALCOHOL.—Dorsal ground color light ters given above for the species as a whole are
tan, with highly vermiculated markings of light to accurate for the types. Characters such as spicules
dark brown. Dorsolateral region usually solid and spines, or degree of fleshiness, could not be
ground color with scattered small dark spots, rest determined on the syntype.
of dorsum very irregular and sometimes so heavily REMARKS.—The type locality for this species,
marked with dark brown that ground color forms given as "San Jos£ de Mod" by Espada, is appar-
isolated spots. Lower sides, chin, chest, and venter ently the town in Napo Province now called San
usually without spots, yellowish white (one juve- Jose Nuevo. The United States Board on Geo-
nile has brown markings on chin and throat). graphic Names in their volume on Ecuador (1957:
Ground color of limbs tan, with dark brown bars 155) gave a reference to San Jose de Mote, with a
or vermiculations which do not extend onto cross reference to San Jose" Nuevo, at 0°26'S,
yellowish white lower surfaces of limb. Outer 77°20/W, which is on the lower northwestern slopes
margins of soles of feet dark brown, inner parts of Mount Sumaco, the only place from whence I
yellowish white. have seen specimens. The town itself is quite low,
NUMBER 145 41

and it is unlikely that the type series originated the new taxon the name "planispina" rather than
there. I suspect the 28 specimens came from the to continue to regard it as a variety of longirostris.
slopes of Mount Sumaco, as do the AMNH speci- The species shown in Plate 2; figure 2, in Espada,
mens. I doubt also the reference to the "top" of is probably actually A. longirostris, although the
Sumaco for AMNH specimens, since that is at 3900 figure is not detailed enough to be certain. Espada
meters, an altitude higher than any other species did have specimens of longirostris from Nanegal in
of Atelopus in Ecuador except ignescens, and much his collection (1875:159).
higher than any other species in the longirostris This species and A. halihelos are apparently re-
group, which, as pointed out above, tends to be stricted tc isolated mountain masses to the east
found at lower, warmer altitudes. of the main Andean chain in Ecuador. It is unlikely
Espada (1875, pi. 2: figs. 1, 2) illustrated what that either of them are found in the lowland areas
he identified in the legends for the Plate 2 varieties surrounding these masses, because other wide-
of A. longirrostris (sic), but did not present a figure ranging species of Atelopus are found there.
of his own new taxon, "A. planispina." If one reads SPECIMENS EXAMINED.—ECUADOR: Napo Prov-
the section termed "Observ." on page 154 of ince; San Jose de Moti (MM 212-syntypes), Top of
Espada's text, one discovers that Espada consid- Sumaco Mountain (AMNH 22193, 22195-97,
ered longirostris and planispinus to be closely re- 22199-200, USNM 193960), side of Sumaco Moun-
lated, and in fact he thought of the two samples tain (AMNH 22300-301).
he had as variants of longirostris for some time,
until he decided that "son verdadermente dos
Atelopus pulcher pulcher (Boulenger)
especes, muy afines si, pero distantas." Since his
figure 1 on Plate 2 is the same taxon I am calling FIGURE 30
planispinus on the basis of the AMNH specimens
Phryniscus pulcher Boulenger, 1882:154, pi. IS: fig. 2 [type
from Mount Sumaco, I believe that the plate legend locality: "Chyavetas, Peru" perhaps the same as Chaya-
did not get changed when Espada decided to give huitas, Peru, according to Rivero, 1968:19)].

I -•'

FIGURE 30.—A syntype, BMNH 1947.2.14.80, of A. pulcher Boulenger.

42 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Atelopus pulcher.—Boulenger, 1894:375.—Peracca, 1904:22.— of feet bright red, with area on venter that is
Rivero, 1963:107. flesh colored in alcohol also bright red. In some
Atetopus spumarius.—Rivero, 1968:19. individuals the entire ventral surface is red.
Atelopus spumarius spumarius.—Duellman and Lynch, 1969:
232. REMARKS.—Rivero (1968:19) discussed at length
the two taxa A. pulcher Boulenger and A. spum-
SYNTYPES.—Four individuals in the British Mu- arius Cope, and finally concluded, with doubts
seum (Natural History), and USNM 193574. based on Cope's inadequate description of his type,
DESCRIPTION.—Snout from above rounded from that they were synonymous. I have not accepted
nostril to nostril, canthus forms a straight line from this action in my work for the following reasons.
nostril to eye, little or no fleshy area on snout First, A. pulcher Boulenger is a well-defined, easily
region or on canthus. Dorsum of head flat, smooth, recognized species, with a good figure accompany-
slight knob at corner of skull. Snout from side very ing the original description, and with type material
slightly protrusive beyond upper lip and lower available for comparisons. As a consequence, the
jaw; nostril directed laterally. Loreal region verti- name has been used and the species recognized by
cal; upper eyelid slightly fleshy, overhangs eye; all workers in eastern Peru and Ecuador since
upper lip not fleshy; temporal and tympanic areas Boulenger's Catalogues were published. Second,
smooth or very lightly wrinkled and shagreened. Cope's species is inadequately described; no figure
No tubercular areas on head, body, or limbs. was published to show what in the world he meant
Dorsum and sides of body lightly wrinkled, belly when he wrote about "aggregated annuli of green-
wrinkling somewhat more pronounced. ish yellow . . . the appearance of green foam," and
Forelimb slender and slightly wrinkled, fore- the holotype has never been located. The substitu-
arm usually thicker than humeral region. When tion of a poorly described, inadequately character-
lorelimb is drawn back along body, palm reaches ized taxonomic name for a well-described, clearly
thigh. Palm of hand fleshy; first digit distinct, with defined one with type material available, and which
tip of digit free of fleshy webbing; other digits en- has been in wide use for 80 years, seems to me to
tirely free of webbing; subarticular tubercles very be an inadvisable taxonomic practice, and I hope
indistinct; palmar tubercle very large, rounded. we can return herewith to the widely recognized
Hind limb slender, smooth or finely wrinkled, name with only a minimum of use of the substitu-
under surfaces of thighs below anus with markedly tion. This does not seem to me to require more
wrinkled and somewhat pustulate area, usually than a recognition by herpetologists that A.
distinctively colored. Tibiotarsal articulation spumarius is clearly a dubious name until such a
reaches points between anterior and posterior mar- time as the holotype is rediscovered, and, as such,
gins of eye; heels touch or overlap when tibio- should not be applied to a known species until
fibulae are parallel to femora at right angles to the doubts are resolved. This would require that
body. Entire foot somewhat fleshy, with first digit the subspecies described by Rivero (1968:23) be
entirely hidden in webbing, second with only its known as Atelopus pulcher andinus Rivero.
tip protrusive, other three digits webbed to tip, on Rivero has also discussed (1968:20) my views
fourth chiefly as fringe. Subarticular swellings pres- that high altitude records for this species are un-
ent; outer metatarsal tubercle present but small. doubtedly incorrect. He pointed out that he could
COLOR IN ALCOHOL.—Dorsal ground color green- recognize differences between specimens suppos-
ish yellow, with dark brown vermiculations and edly from Riobamba, a city in the inter-Andean
spotting more or less symmetrically distributed on highlands at 2800 meters, and the lowland material
head, back and sides; upper lip yellow, with a fine from the Upano Basin. Savage (1960:14) has dis-
brown line along margin. Belly and chin yellow, cussed the validity of Riobamba as the source of
with small rounded brown spots primarily located certain AMNH specimens, and I follow him in
posteriorly. Limbs barred above with dark brown, not accepting Riobamba as a valid locality for any
yellow below with brown spotting. Soles of feet and lowland species until verification in the form of
area just below anus flesh colored. newly collected specimens has been made. The
COLOR IN LIFE.—Dorsal ground color yellowish AMNH specimens were received from H. Feyer,
green, with jet black spotting and striping. Soles and his "Riobamba" records of lowland species
NUMBER 145 43
have been causing problems in Ecuadorian herpe- Present Position of Taxa Formerly Assigned to
tology for years. It seems clear that Feyer, with Atelopus In Ecuador
his residence in Riobamba, purchased specimens
from any seller, including mule drivers coming up There is a single species originally described in
from the Amazonian lowlands on their regular the genus Atelopus from Ecuador that is no longer
trips, carrying a few frogs and snakes along to sell considered to belong to that genus. Atelopus festae
to the "gringo," who then used "Riobamba" as Peracca is now considered to belong to the genus
the origin of the specimens. Rhamphophryne (Trueb, 1971:22). A second spe-
The specimens mentioned by Rivero (1968:20) cies, Atelopus minutus Melin, although not origin-
from Chancha, Chimborazo Province, are from ally described from Ecuador, has been collected
Normandia, in Morona-Santiago Province. This in the country several times. It was recently trans-
locality is discussed by Savage (1960:13), who lo- ferred to the genus Dendrophryniscus by McDiar-
cated it at 1300 meters on the trail between Macas mid (1971:40).
and Riobamba. This is within the known range of
the species.
Duellman and Lynch (1969:232) described and Incertae Sedis
figured a single presumed tadpole of this species,
which was collected in the Rio Puyo in 1969. They Atelopus longirostris marmorata Werner
followed Rivero in the use of spumarius for the
species, as did McDiarmid (1971:39). This taxon was described by Werner in 1910,
The species Atelopus tricolor Boulenger (1902: page 600, from "Ecuador," without more detailed
397) is close to A. pulcher in its characters. I have information on the type locality. He noted that the
had an opportunity to examine the syntypes, how- females in his collection appeared different from
ever, and they appear to me to represent a distinct the males of longirostris, and used the trinomen
taxon. One of the syntypes is now in the USNM above for them. Nieden (1926:81) included the
collection, through the kindness of the British taxon in the synonymy of Atelopus longirostris, and
Museum (Natural History) (USNM 193575). the few citations since that time (Rivero, 1963;
Rivero (1968:19) has discussed the spelling of Cochran and Goin, 1970) accepted Nieden's action
the type locality in detail, and concluded that it as having done away with the taxon. Careful exam-
was very likely that the Peruvian village called ination of all of Nieden's work in this volume of
Chayahuitas, at 5°5'S, 76°10'W, was the original Das Tierreich, however, shows that he included
source of the type material. Since Chyavetas has all trinomials in existence prior to his work as
never been located, his argument has considerable
synonyms of a binomial, and then mentioned those
validity, and I have indicated that the village he
he considered to be recognizable in the text below
located is probably the legitimate type locality.
the synonymy I find only three subspecies men-
SPECIMENS EXAMINED.—ECUADOR: (AMNH
17433-38, 17447). Morona-Santiago Province; Chan- tioned in the entire volume, all in the text material.
cha, Normandia (AMNH 16695-712), Cutucii Cor- One of these is marmorata. It is, therefore, not par-
dillera (AMNH 33913-15), Macas (AMNH 21492- ticularly valid to cite Nieden as the authority for the
94, USNM 65487-90), Miazal (GOV S940, JAP synonymy of marmorata with longirostris. I think
3843), Limon (CAS 134202-03, JAP 6966, 7520, it entirely possible that Werner had specimens of
UMMZ 131022-23, USNM 193479-90). Pastaza a completely different species in hand. Since ele-
Province; "Bafios and Canelos" (AMNH 16713- gans occurs on the western slopes in Ecuador, and
15), Canelos (MCZ 17927-28), Puyo (FMNH may actually be sympatric with longirostris in part
28078), 2l/2 km southeast of Puyo (JAP 1937-38), of its range, I would suggest that Werner may not
Rio Oglan (GOV S939), Veracruz (JAP 6105, have recognized specimens of elegans among his
UKMNH 120480-85). longirostris material. Unfortunately, the type
PERU: Chayahuitas (BMNH 1947.2.14.80, specimens were destroyed during World War II,
USNM 193574, syntypes of A. pulcher). and marmorata becomes a nomen dubium.
44 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Zoogeography and Evolution rious types of barriers are common, and are most
often encountered between the traditionally used
Rivero (1963:112) has remarked, "Frogs of the trails, one would expect each transect to produce
genus Atelopus seem to be very sensitive to en- different taxa, each evolving in an isolation pro-
vironmental factors." This is certainly true of duced by the barrier of altitude on one hand, and
Ecuadorian species, particularly at altitudes above the many horizontal barriers mentioned above on
the subtropical zone. The Ecuadorian terrain the other hand. This proved to be particularly true
makes exploration by herpetologists difficult, and of frogs of the genus Atelopus and lizards of the
not too much work has been done on the outward genus Proctoporus. Since representatives of both
facing slopes of the Andes as a consequence. Access genera are fairly often found even if the total col-
to areas between 1500 and 3300 meters has always lecting time available is slight, they become good
been and still generally is available only on foot indicators of evolutionary patterns, effects of isola-
or horseback, although road construction on both tion, adaptive radiation, and other phenomena
east and west slopes continues to make more areas often associated with the study of insular faunas. In
accessible by car. Most of my personal collecting point of fact, mountains such as the Andes of
efforts in Ecuador have been focussed on these Ecuador can be considered "continental islands,"
slopes. I have spent comparatively little time in and many of the phenomena long discussed by
the Pacific lowlands, and even less in the Amazon- students of insular biology are equally observable
ian rain forest. Herpetologically, the latter has been there, with the added advantage that the mountains
the most popular and thoroughly explored area in are not dependent upon waifs and drifters for re-
the country, and, while it continues to produce plenishment of the supply of evolutionary ma-
both novelties and surprises, it still does not hold terials.
the same promise for investigation of ecological and
The taxa in the genus Atelopus lend themselves
evolutionary phenomena as do the Andean slopes.
to the study of these phenomena because of the sen-
In 1958-1959 I made a series of seven transects sitivity to environmental factors already noted by
from the inter-Andean valleys to the Pacific low- Rivero. But, although they are sensitive to, and
lands, and in 1962 my group made five transects on often limited by, very slight environmental
the eastern slopes, usually, although not always, changes, they are also capable of adaptation and di-
penetrating to the Amazonian lowlands. Such tran- vergence, which permits them to conquer new
sects perforce follow available routes, and of the environments. Suitable habitat for the taxa in the
total of twelve transects, five were by mule and lowlands o.f Amazonian Ecuador as well as in the
on foot, six were by road in truck or Landrover, lower altitudes on the Pacific side appears to be
and one was by rail. Since the routes taken were continuous, with few barriers to gene flow, and
either mule trails several hundred years old, or lowland species tend to have a considerable lati-
roads that follow such old trails and replace them, tudinal range. A. pulcher ranges throughout the
it is to be expected that each trail would be sepa- southern half of Amazonian Ecuador and a con-
rated by some distance and usually some fairly sub- siderable distance in Peru, from whence it was orig-
stantial physical or physiographic barrier from the inally described. It stops at about the Rio Pastaza,
next trail, since arrieros follow paths of least re- and is replaced by A. palmatus, the northern limit
sistance, and do not develop and follow duplicate of which is currently unknown. A. elegans is a
trails ending at the same place very often (except lowland species on the west side of the Andes in
as smugglers). The physiographic barriers are deep Ecuador, and is found north well into Colombia
river valleys, sheer cliffs, or too heavily dissected and south to the point where the wet areas are
slopes, all of which can and do act as biological replaced by the dry coastal scrub, under the in-
barriers to species movement equally well. In addi- fluence of the Humboldt Current. Figure 31 is an
tion, rain shadow slopes, where the vegetation and attempt to show how the situation changes when
moisture available change radically within a few the species found on the slopes at slightly higher
thousand feet, occur on both east and west slopes, altitudes are examined, and the opportunities for
and represent a major biotic barrier to horizontal horizontal migration are diminished by the multi-
migration of mountainside species. Since these va- plicity of barriers. The extremely narrow distribu-
NUMBER 145 45

tion of the species found at altitudes above 2000 ubiquitous over a considerable range. It is the only
meters is in part an artifact of the collecting oppor- species that may in several cases be sympatric with
tunities, which are restricted to a few hundred those species with restricted ranges on the outer
yards laterally to the trails, b u t it is fairly certain slopes. I have found A. ignescens within 2 to 3
that there is a high degree of endemism on the kilometers of populations of A. pachydermus, and
higher slopes, and it is very likely that the endemic it appears that the same is true for A. bomolochos,
species are totally allopatric. Finally, after an al- much farther to the south. A. ignescens is, of course,
titude sufficient to permit the inter-Andean high- sympatric with A. bufoniformis, which has been
lands to be reached, the picture is reversed. T h e taken as high as 3400 meters, and is known from
highly successful species A. ignescens is capable of both slopes of the Andes, but only in the most
living at very high altitudes under most inhospi- northern provinces of Ecuador. A. bufoniformis is
table conditions. As a result, it is the most wide- the end point in the evolutionary sequence dis-
spread amphibian species in the Andean highlands, cussed below.
ranging from well into Colombia practically to the While the evolutionary sequences are not en-
Peruvian border. T h e ability to live in paramo tirely clear, and the evidence used here is primarily
and subparamo climates has eliminated all phys- morphological and zoogeographic and thus could
ical barriers to movement by the species, and it is be easily disproven by data from cytological, ge-

4000
-•INTO
3500 K -r

3000 V
cc
Ul
2500 -• COLOMBIA

2000
G
1500
1
1000 f- INTO^- .1 I

500 - PERU **- INTO

4--I7
0 - • COLOMBIA

5°S 4°S 3°S 2°S 0*

LATITUDE

A ARTHURI F ELEGANS K NEPIOZOMUS

B BALIOS G HALIHELOS L PACHYDERMUS
C BOMOLOCHOS H IGNESCENS M PALMATUS
D BOULENGERI I LONGIROSTRIS N PLANISPINUS
E BUFONIFORMIS J MINOOENSIS 0 PULCHER

FIGURE 31.—A diagrammatic representation of the known distribution patterns of the species of
Atelopus in Ecuador. Each square (or line) is based on known extremes, and therefore may
well include areas from whence a particular species is not known. In each species a complete
rectangle is shown, although this is not easy to discern where two species have common limits.
Dashed lines indicate uncertainty, usually explained in the text.
46 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

netic, biochemical, and other more modern tests, although the horizontal distribution of each is
it still seems legitimate to construct hypotheses certainly greater than my data show. The vertical
concerning what has happened to Atelopus phy- restrictions are for the most part real, and I think
logeny in Ecuador. The potential absurdity and es- the various taxa show sharp response to and are
sential sterility of such hypotheses can only become closely confined by the changing conditions which
evident after considerable research on the taxa accompany increasing altitude. All of these species
concerned, research which I do not anticipate be- are living in a comparatively young area, geologi-
ing done for many years—and which might even cally speaking. The surrounding lowlands are
verify my guesses here. much older, and it seems plausible that the an-
In any event, the evidence derived from an ex- cestral species in the genus Atelopus are the low-
amination of the distribution and morphology of land taxa, and the highland taxa have been re-
the two groups of species indicates the direction peatedly derived from the lowland stocks. I would
that evolution has taken. With regard to distribu- further suggest that each of the highland taxa
tion, the taxa of the longirostris group are almost (with the probable exception of A. ignescens and
entirely low altitude species. The only exception the possible exception of A. bufoniformis) is the
may be A. planispinus, which is found on Mount product of a separate, successful invasion of the
Sumaco, which reaches an altitude of 3900 meters. highlands by lowland stock. This, of course, means
The type locality for the species, at the foot of the that the ignescens group is clearly not monophy-
mountain, is at 500 meters, and specimens in the letic, but simply represents a series of separately
American Museum of Natural History are recorded derived species all differentiating from the ancestral
as coming from the top of Sumaco, which gives the stock along similar morphological lines.
species a possible altitudinal range of 500 to 3900 The morphology of the two groups of species
meters, a most unlikely situation. Since this taxon tends to verify the hypothesis. The adaptive
is part of the longirostris group, and since it has changes evolved in response to higher altitudes and
not been taken elsewhere in the Amazonian low- colder temperature include shorter, more stocky
lands, as would be expected if it actually occurred bodies and a reduction of the limbs from long and
at San Jose Nuevo (the modern name for San slender to short and stout. This suggests, of course,
Jose de Moti, or Mote), I predict it will even- the development of mechanisms to reduce total sur-
tually be found to occupy the slopes of Sumaco face and concomitant heat loss. The highland spe-
at lower to intermediate altitudes (perhaps 1000 to cies are much less likely than the lowland taxa
2000 meters) and probably on the eastern slope to be found out and actively moving about except
only (because of habitat change on the back or during breeding periods, when they become abun-
western side of the mountain). Other than A. dant.
planispinus, the taxa in the longirostris group do While not as sharply distinctive between groups
not reach altitudes above 2000 meters, and enjoy as the body and limb configuration, the thickening
their greatest abundance below 1000 meters.
of the skin in the webbing between the digits, and
The taxa of the ignescens group, on the other the overall fleshiness of the highland species in
hand, are almost all high altitude forms. The one general, may well be another modification toward
exception, A. mindoensis, is an enigma. As I point
heat conservation. It reaches maximum develop-
out frequently in this paper, this taxon is inter-
ment in A. bufoniformis, the taxon which has both
mediate between the two groups in many ways.
For example, it tends to fall halfway between the the maximum development of the characteristics
groups in almost all of the analyses of proportional discussed here and reaches, with A. ignescens, maxi-
measurements. Leaving that species to one side, mum altitudes.
the remaining members of the igenscens group all McDiarmid (1971:58) has discussed the evolu-
show a common zoogeographic pattern. They are tionary relationships of Atelopus with other genera
very restricted horizontally, and appear to be of the family Bufonidae, and demonstrated that
equally restricted vertically to a fairly narrow al- the genus is quite advanced rather than primitive
titudinal band. They have not, and I believe they in a large number of its characteristics. He pointed
are unable to, spread widely on a horizontal plane, out several phenomena concerning the genus Ate-
NUMBER 145 47

lopus that are of interest in understanding evolu- they represent repeated successful invasions of the
tionary events in Ecuador. He stated: outer slopes by A. ignescens, adapting and differ-
With the initial uplifts of the Andes in Peru and Ecuador in entiating in different areas. A. ignescens definitely
the Late Cretaceous . . . . entire new areas became available occurs on the eastern slope within a few kilometers
to and were exploited by Atelopus. The ancestral stock prob- of A. pachydermus, and the relationship between
ably came from southeastern portions of the Guiana shield them seems to be close. The same is true for A.
where the more primitive Atelopus flavescens is found today. bomolochos, which is equally close to A. ignescens
However, it is possible that Atelopus initially invaded the
montane habitat in the southern Andes and moved north both geographically and morphologically. Why,
along their eastern slope. It appears also that the higher then, is it not possible that A. ignescens is the di-
elevations have been invaded several times. rect ancestor to these restricted populations?
I would agree completely with the concept of Perhaps the strongest argument against this hy-
multiple invasion of the higher elevations, but pothetical ancestry is the fact that A. ignescens
think it unlikely that there has been very much tends to maintain its specific identity throughout
movement north along the eastern slope of the a considerable distance in the inter-Andean pla-
Andes from the south. The evidence presented teau. Although quite variable, and perhaps with
above seems to indicate fairly strongly that lateral recognizeable subspecies in several places, it still
or horizontal movement is not a common phe- remains one species throughout, in my estimation.
nomenon in Ecuador, probably as a consequence It is uncertain how much contact there is between
of multiple barriers. McDiarmid (1971:58) has populations separated by nudos, but I suspect there
himself provided an additional argument against is less contact between them than within such
horizontal movement, for he said: populations, certainly a situation to encourage
local populational differentiation, if the species is
The apparent close association of species of Atelopus with plastic enough to respond. I suggest that A. ignes-
mountain streams has restricted species to particular drainage cens is a species well advanced evolutionary, and
systems. The resulting isolation (drainage system and moun-
tain top) probably accounts for the striking differences among highly adapted to a very inhospitable environment,
many populations and probably has been an important factor and therefore has comparatively little evolutionary
contributing to speciation. potential. The likelihood of such a species invad-
ing a series of isolated drainages on both sides of
The only species of the ignescens group which
the Andes of Ecuador and giving rise to such a
has been able to break out of the limitations noted
large number of very distinct species as are found
above for other species is A. ignescens itself. It has there seems very slight to me.
been a highly successful invader of the inter-
On the other hand, I must reverse that argument
Andean valleys in Ecuador, and its vertical range
with regard to A. bufoniformis. This species has
is such that the nudos or knots between the valleys
gone beyond A. ignescens in every one of the adap-
are not a barrier to horizontal movement. The tive characteristics discussed here. It is clearly the
species ranges from the southernmost provinces in terminal species in the differentiation sequence.
Ecuador well into Colombia, a horizontal range It is difficult for me to visualize any other ancestor
almost as great as that of any lowland species. for this species than A. ignescens, and this causes
Only A. bufoniformis shows greater differentiation some considerable discomfort. At one time I had
than A. ignescens along the morphological se- thought A. bufoniformis to be an inhabitant of
quences to high altitude adaptation, as can be seen the high eastern slope, found slightly below the
in the graphs of proportional measurements. A. range of A. ignescens. But material in the
ignescens represents an explosive movement taking UKMNH collections show it from 3400 meters, and
place, because its ancestral stock continued the up- well into the higher parts of the inter-Andean Pla-
ward expansion of its range until it spilled over the teau. It does not seem to occur in the inter-Andean
wall of the Andes, opening up the whole inter- valleys, although it has been found low enough on
Andean plateau to settlement. the outer slopes that altitude would not be a bar-
The success and widespread distribution of A. rier. It appears to have been a comparatively re-
ignescens leads one to a second look at the other cent derivative of A. ignescens stock, and is perhaps
members of its group. It is entirely possible that even more thoroughly adapted to the cold, wind-
48 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

swept, high altitude paramo than the parent stock. Duellman, W. E., and J. D. Lynch
1969. Descriptions of Atelopus Tadpoles and Their Rele-
It is probably in the process of expanding its range vance to Atelopodid Classification. Herpetologica,
at those high altitudes, but may face problems in 25:231-240.
the fact that the habitat it utilizes is not sufficiently Dunn, Emmett R.
continuous to permit it to reach all existent pa- 1931. New Frogs from Panama and Costa Rica. Occasion-
ramo. al Papers of the Boston Society of Natural History,
5:385-401.
Espada, Marcos Jimenez de la
Literature Cited 1875. Vertebrados del Viaje al Pacifico verificado de 1862
a 1865 por una Comision de Naturalistas enviada
Acosta Sol is, M. pox el GoHerno Espanol. 208 pages, 6 plates.
1944. Nuevas Contribuciones al Conocimiento de la Pro- Batracios: Madrid.
vincia de Esmeraldas. 606 pages. Quito: Editorial Guibe, Jean
Ecuador. 1948. Catalogue des Types d'Amphibiens du Museum
Andersson, Lars G. National d'Histoire Naturelle. 71 pages. Paris: Im-
1945. Batrachians from East Ecuador Collected 1937, primerie National.
1938 by Wm. Clarke-Maclntyre and Rolf Blomberg. Gunther, Albert
Arkiv for Zoologi, 37A (2) :l-88. 1858. Catalogue of the Batrachia Salientia in the Collec-
Boettger, O. tion of the British Museum. 160 pages. London:
1892. Katalog der Batrachier-Sammlung in Museum der British Museum.
Senckenbergischen Naturforschenden Gesellschaft in 1859. Second List of Cold-blooded Vertebrates Collected
Frankfurt am Main. 73 pages. Frankfurt a. M., Ger- by Mr. Fraser in the Andes of Western Ecuador.
many. Proceedings of the Zoological Society of London.
Boulenger, George A. 1859:402-420.
McDiarmid, Roy W.
1880. Reptiles et Batrachiens recueillis par M. £mile de
1971. Comparative Morphology and Evolution of Frogs
Ville dans les Andes de l'Equateur. Bulletin de la
of the Neotropical Genera Atelopus, Dendrophry-
Societe Zoologique de France, 1880:41-48.
niscus, Melanophryniscus, and Oreophrynella. Bul-
1882a. Catalogue of the Batrachia Salientia s. Ecaudata in
letin of the Los Angeles County Museum of Natural
the Collection of the British Museum. Second Edi-
History, Science, 12:1-66.
tion, 495 pages, 30 plates. London: British Museum.
Nieden, F.
1882b. Account of the Reptiles and Batrachians Collected
1926. Amphibia. Anura II. Engystomatidae. Das Tier-
by Mr. Edward Whymper in Ecuador in 1879-80.
reich, 49:1-110.
Annals and Magazines of Natural History, series 5,
Noble, G. K.
9:457^67.
1921. Five New Species of Salientia from South America.
1894. On the Genus Phryniscus of Wiegmann. Annals
American Museum Novitates, 29:1-7.
and Magazine of Natural History, series 6, 14:374-
Orton, James
375.
1871. Contributions to the Natural History of the Val-
1898. An Account of the Reptiles and Batrachians Col-
ley of Quito-II. Reptiles. American Naturalist, 5:
lected by Mr. W. F. H. Rosenberg in Western
693.
Ecuador. Proceedings of the Zoological Society of
Parker, H. W.
London, 1898:107-126.
1934. Reptiles and Amphibians from Southern Ecuador.
Cochran, Doris M., and Coleman J. Goin
Annals and Magazine of Natural History, series 10,
1970. Frogs of Colombia. United States National Museum 14:264-273.
Bulletin, 288: xii -f 655 pages. 1938. The Vertical Distribution of Some Reptiles and
Cope, Edward D. Amphibians in Southern Ecuador. Annals and
1867. On the Families of Raniform Anura. Journal of Magazine of Natural History, series 11, 2:438-450.
the Academy of Natural Sciences of Philadelphia, Peracca, M. G.
series 2, 6:189-206. 1904. Viaggio del Dr. Enrico Festa nell'Ecuador a regioni
1868. An Examination of the Reptilia and Batrachia vicine. Rettili ed Amfibii. Bollettino dei Musei di
Obtained by the Orton Expedition to Equador and Zoologia ed Anatomia comparata della R. Univer-
the Upper Amazon, with Notes on other Species. sitd di Torino, 19 (465) : 1-41.
Proceedings of the Academy of Natural Sciences of Rivero, Juan A.
Philadelphia, 20:96-140. 1963. Five New Species of Atelopus from Colombia, with
Cornalia, Emilio Notes on Other Forms from Colombia and Ecuador.
1849. Vertebratorum Synopsis in Museo Mediolanense Caribbean Journal of Science, 3 (2,3): 103-124.
Extantium quae per Novam Orbem Cajetanus Os- 1965. Notes on the Andean Salientian (Amphibia) Atel-
culati Collegit Annuis 1846-47-48. Museo Medio- opus ignescens (Cornalia). Caribbean Journal of
lanense, 1849:304-315. Science, 5:137-139.
NUMBER 145 49

1968. More on the Atelopus (Amphibia, Salientia) from schaftlichen Classe der Kaiserlichen Akademie der
Western South America. Caribbean Journal of Sci- Wissenschaften, Vienna, 14:237-258.
ence, 8:19-29. Taylor, Edward H.
1969. Nota sobre Atelopus carinatus Anudersson y Atel- 1952. A Review of the Frogs and Toads of Costa Rica.
opus palmatus Andersson (Amphibia, Salientia) . University of Kansas Science Bulletin, 35:577-942.
Memorias de la Sociedad de Ciencias Naturales La Thominot, Alexandre
Salle, Caracas, 29 (83) : 142-145. 1889. Observations sur quelques Reptiles et Batraciens de
Savage, Jay M. la Collection du Museum d'Histoire Naturelle de
1960. A Revision of the Ecuadorian Snakes of the Colu- Paris. Bulletin de la Socie'te Philomathique, Paris,
brid Genus Atractus. Miscellaneous Publications of series 8, 1:21-30.
the Museum of Zoology University of Michigan, Trueb, Linda
112:1-86. 1971. Phylogenetic Relationships of Certain Neotropical
Schmidt. Oscar Toads with the Description of a New Genus. Los
1857. Diagnosen neuer Frosche des zoologischen Cabinets Angeles County Museum, Contributions to Science,
zu Krakau. Sitzungberichte der Mathematisch- 216:1HH).
Naturwissenschaftlichen Classe der Kaiserlichen Werner, Franz
Akademie der Wissenschaften, Vienna, 24:10-15. 1901. Ueber Reptilien und Batrachier aus Ecuador und
1858. Deliciae Herpetologicae Musei Zoologici Cracovien- NeuGuinea. Verhandlungen der k. k. Zoologische-
sis. Denkschriften der Mathematisch-Natunvissen- Botanische Gesellschaft im Wien, 1901:593-614.

U. * . GOVERNMENT PRINTING OFFICE: 1073 516-120/0

 Publication in Smithsonian Contributions to Zoology
Manuscripts for serial publications are accepted by the Smithsonian Institution Press, sub-
ject to substantive review, only through departments of the various Smithsonian museums. Non-
Smithsonian authors should address inquiries to the appropriate department. If submission is
invited, the following format requirements of the Press will govern the preparation of copy.
Copy must be typewritten, double-spaced, on one side of standard white bond paper, with
IVJ" top and left margins, submitted in ribbon copy with a carbon or duplicate, and accompa-
nied by the original artwork. Duplicate copies of all material, including illustrations, should be
retained by the author. There may be several paragraphs to a page, but each page should begin
with a new paragraph. Number consecutively all pages, including title page, abstract, text, litera-
ture cited, legends, and tables. The minimum length is 30 pages, including typescript and illus-
trations.
The title should be complete and clear for easy indexing by abstracting services. Taxonomic
titles will carry a final line indicating the higher categories to which the taxon is referable:
"(Hymenoptera: Sphecidae)." Include an abstract as an introductory part of the text. Identify
the author on the first page of text with an unnumbered footnote that includes his professional
mailing address. A table of contents is optional. An index, if required, may be supplied by the
author when he returns page proof.
Two headings are used: (1) text heads (boldface in print) for major sections and chap-
ters and (2) paragraph sideheads (caps and small caps in print) for subdivisions. Further
headings may be worked out with the editor.
In taxonomic keys, number only the first item of each couplet; if there is only one couplet,
omit the number. For easy reference, number also the taxa and their corresponding headings
throughout the text; do not incorporate page references in the key.
In synonymy, use the short form (taxon, author, date:page) with a full reference at the
end of the paper under "Literature Cited." Begin each taxon at the left margin with subse-
quent lines indented about three spaces. Within an entry, use a period-dash (.—) to separate
each reference. Enclose with square brackets any annotation in, or at the end of, the entry.
For references within the text, use the author-date system: "(Jones, 1910)" and "Jones
(1910)." If the reference is expanded, abbreviate the data: "Jones (1910:122, pi. 20: fig. 1)."
Simple tabulations in the text (e.g., columns of data) may carry headings or not, but they
should not contain rules. Formal tables must be submitted as pages separate from the text, and
each table, no matter how large, should be pasted up as a single sheet of copy.
Use the metric system instead of, or in addition to, the English system.
Illustrations (line drawings, maps, photographs, shaded drawings) can be intermixed
throughout the printed text. They will be termed Figures and should be numbered consecu-
tively; however, if a group of figures is treated as a single figure, the components should be
indicated by lowercase italic letters on the illustration, in the legend, and in text references:
"Figure 9b." If illustrations (usually tone photographs) are printed separately from the text as
full pages on a different stock of paper, they will be termed Plates, and individual components
should be lettered (Plate 9b) but may be numbered (Plate 9: figure 2 ) . Never combine the
numbering system of text illustrations with that of plate illustrations. Submit all legends on
pages separate from the text and not attached to the artwork. An instruction booklet for the
preparation of illustrations is available from the Press on request.
In the bibliography (usually called "Literature Cited"), spell out book, journal, and
article titles, using initial caps with all words except minor terms such as "and, of, the." For
capitalization of titles in foreign languages, follow the national practice of each language.
Underscore (for italics) book and journal titles. Use the colon-parentheses system for volume,
number, and page citations: "10(2) :5-9." Spell out such words as "figures," "plates," "pages."
For free copies of his own paper, a Smithsonian author should indicate his requirements
on "Form 36" (submitted to the Press with the manuscript). A non-Smithsonian author will
receive 50 free copies; order forms for quantities above this amount with instructions for pay-
ment will be supplied when page proof is forwarded.
 K
1

"iHr

^ ^MBI