The distributions of the gladiator frogs (Hyla boans group) in Colombia, with comments on size variations and sympatry

17 Pages·2001·2.9 MB·English

by  Lynch, John Douglas

Caldasia 23(2): 491-507
THE DISTRIBUTIONS OF THE GLADIATOR FROGS
(HYLA BOANS GROUP) IN COLOMBIA, WITH
COMMENTS ON SIZE VARIATION AND SYMPATRY
JOHN D.LVNCH
Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Apartado 7495, Bogo-
tá, Colombia. [email protected]
ÁNGELA M. SlIÁREZ-MAVORGA
Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Apartado 7495, Bogo-
tá, Colombia. [email protected]
ABSTRACT
Fivespecies ofgladiator frogs(theHylaboansgroup) occur inColombia. Hyla wavrini
barely enters Colombia (itisconfined totheextreme eastern edge ofthe country) and
H. rosenbergi isendemic to the biogeographic Chocó. The other three species are
more widely distributed. Hyla boans isfound inthe Chocó, Middle Magdalena, and
Amazonas-Orinoquia below 1QOO m. "Hylacrepitans" isfound inthe Magdalena
drainage aswell asthe western edge ofthe Orinoquia and lesscommonly along the
Caribbean coast and inthe Maracaibo Basin. This species occurs from sea level to
more than 2000 m. Hyla pugnax occurs across the Caribbean lowlands and up the
Magdalena valley aswell as inthe Maracaibo Basin and atafew cis-Andean sites,
but always below 500 m. Sympatry isrelatively common between H. boans and H.
rosenbergi but, less common between H. boans and either "H. crepitans" or H.
pugnax. Maximal sympatry results in sympatry of H. boans, H. pugnax, and H.
rosenbergi - normally nomorethantwospecies aresympatric. Sexual dimorphism in
size is evident in the smallest species ("H. crepitans ") but not in the others. AH
exhibit geographic variation insize,perhaps related tothe combination ofsympatric
species ofthe group.
Key words. Amphibia, Hylidae, variability, new records, treefrogs.
RESUMEN
En Colombia se encuentran cinco especies de ranas gladiadoras (del grupo Hyla
boans). Hyla wavrini apenas entra aColombia (seencuentra enellímite oriental del
país) eH.rosenbergi esendémica del Chocó biogeográfico. Las otras tres especies
tienen distribuciones más amplias. Hyla boans se encuentra en el Chocó
biogeográfico, Magdalena Medio yen laregión cisandina por debajo de los 1000m.
"Hyla crepitans" se encuentra en el drenaje del Río Magdalena, así como en la
parte occidental de losLlanos Orientales ycon menor frecuencia en laregión Cari-
be; laespecie sedistribuye desde elnivel del mar hasta unpoco más de los2000 m.
Hyla pugnax seencuentra en lastierras bajas de laregión Caribe, Magdalena Me-
dioytambién enalgunos sitios de laOrinoquia, siempre por debajo de los500 m. Es
bastante común la simpatría entre H. boans e H. rosenbergi pero menos común
entre H. boans e "H. crepitans" oH.pugnax. Sehan registrado hasta tres especies
The distributions of the gladiator frogs
simpátricas (H. boans, H. pugnax e H. rosenbergi) pero en general no coexisten
más de dos especies. El dimorfismo sexual en tamaño es evidente en laespecie más
pequeña (uH. crepitans ") pero no en las otras. Cada especie muestra variación
geográfica en tamaño, tal vez relacionada con lacombinación de especies simpátricas.
Palabras clave. Amphibia, Hylidae, variabilidad, nuevos registros, ranas plataneras.
INTRODUCTION that males were smaller than females whereas
males and females ofthe other species (that he
Gladiator frogs are large, aggressive treefrogs recognized) were of equal sizes (an observa-
distributed from Costa Rica to southern Brasil tion consistent with sexual selection). So far
(Kluge 1979). Five species are known from as isknown, among frogs ofthe H.boans spe-
Colombia: H. boans, UH. crepitans ", H. cies group, there ismarked aggression arnong
pugnax, H. rosenbergi, and H. wavrini, but males for the possession of a breeding site.
non e is endemic. Lynch &Vargas (2000) re- Nevertheless, this does not translate into abias
ported the Colombian distribution of H. insize for males inthe H. boans species group.
wavrini, aGuyana element that barely enters Over the past thirty years, biologists frorn the
Colombia, being known from only Instituto de Ciencias Naturales of the
Departamentos Guainía and Vaupés. The most Universidad Nacional have descended into the
recent summary ofthe gladiator frogs and their lowlands on rnany occasions and, whenever the
distributions in Colombia isby Kluge (1979) opportunity presented itself, have collected
who concluded that H. rosenbergi was discon- gladiator frogs. These are rather large treefrogs
tinuously distributed in the biogeographic and therefore conspicuous (which does not
Chocó, H.pugnax was anear-Colombian en- always translate into vouchers). Both "H.
demic, found in the Magdalena valley and crepitans" and H. pugnax are known locally
along the Caribbean coast, H. boans was dis- as ranas plataneras, meaning that neither lo-
continuously distributed (biogeographic cal biologists nor campesinos distinguish the
Chocó as well as Amazonia and the Orinoco, two species. Beginning in 1980, efforts were
absent from the Magdalena valley), but was made to document distributions, ensuring at
unaware that H. wavrini is a species distinct least one voucher per locality (but one speci-
from H. boans, and that "H. crepitans" oc- men is not adequate to test the proposition of
curred in the Magdalena valley (and coastal sympatry). Given that local biologists did not
lowlands) as well as in the upper parts ofthe distinguish the species and that each species
Río Orinoco drainage. Because these are large is locally abundant, the database available to
conspicuous treefrogs of lowlands, Kluge us is diminished because most potential col-
(1979) presumed that absence of published lectors did not bring specimens to ICN (rea-
records reflected absence of species, aconclu- soning, perhaps, that enough specimens were
sion we regard as premature. We have little to available already and acting "ecologically").
add to the published record as concerns H. Relatively little ofthe available data have been
rosenbergi and H. wavrini; our study concerns published (excepting Lynch &Vargas, 2000)
primarily the other three species (H.boans, "H. and the data available require amore cornpli-
crepitans ".and H.pugnax). cated view that that presented by Kluge (1979),
The theory of sexual selection places a good offering many options to students who wish to
deal of attention on sexual dimorph ism (here confront avariety of issues inecology and re-
in terms of size). Kluge (1979) noted that his productive biology.
material of "H.crepitans" supported the view
492
John D.Lynch &Ángela M. Suárez-Mayorga
MATERIALS ANO METHODS other two species are srnaller and have re-
duced webbing of the fingers (not reaching
Our study ofthese frogs isbased on the study
the disks nor extending past the ultimate
ofmore than 700 specimens available from the
subarticular tuberc\e ofFingers III or IV). The
Instituto de Ciencias Naturales (ICN) and the
lower palpebrurn of H.pugnax lacks pigrnen-
Instituto Alexander von Hum boldt (formerly
tation whereas the anterior half of the
the collection of INDERENA) (lAvH). For
palpebrum of "H. crepitans" is pigmented.
each specimcn, we recorded sex, size, rnatu-
The two species can be distinguished as well
rity, afew morphological characters (pigmen-
because "H. crepitans "isasrnaller frog (see
tation ofthe lower eyelid, calcar developrnent,
species account), has less webbing ofthe fin-
degree ofwebbing of the hand, degree ofweb-
gers and toes, lacks barring on the anterior
bing of the foot, and color pattern data), and
face of the thighs, and has simple flank bar-
±
locality. Means are reported Istandard er- ring (Kluge 1979). Kluge (1979) suggested that
ror ofthe mean, webbing characters were unreliable when deal-
ing withjuveniles. Inour experience, the flank
RESULTS and thigh pigmentation patterns are unreli-
able when dealing withjuveniles butthe palpe-
Both sexes of these species exhibit angled
bral pigmentation patterns are not influenced
vomerine odontophores (a trait shared by few
by size or age.
other Hyla,but also by some exc\uded genera,
such as Osteocephalus) and adult males pos- Hylaboans (Linnaeus)
sess aprotuberant prepollex (shared by afew Kluge (1979) was perplexed by the variance
other groups of Hyla, sorne ofwhich also ex- of size of this species. However, Lynch &
hibit angled vornerine odontophores). AII spe- Vargas (2000) asserted that cis-Andean Colom-
cies, for which tadpoles are known, possess a bian populations differed markedly in adult
larval tooth row formula of 2/4. Although size (populations from the Orinoco are signifi-
Kluge (1979) thought that this cornbination cantly srnaller frogs than those frorn
might be sufficient, we are not sanguine that Amazonia). Kluge (1979) also opined that the
the Hyla boans group ismonophyletic (in the absence of records frorn the Magdalena drain-
face ofthe available data and arguments). age did not reflect collecting effort. We beg to
Only H. boans exhibits an obvious calcar. That disagree but are uncertain when the first H.
species as well as H. wavrini exhibits areticu- boans was collected inthe Middle Magdalena,
lum inthe lower eyelid (a feature shared by H. although JDL collected specirnens in eastern
geographica and H. microdcrma of the H. Caldas in 1981. One might quibble that the
geographica group). Each ofthe species ofthe report frorn southern Córdoba (Renj ifo &
H. boans group exhibits geographic variation Lundberg 1999) hardly represents adistribu-
in size ofadults (and, presumably, age at rna- tion outside of the biogeographic Chocó (a
turity) but this isminimal in H.pugnelX. Hyla view with which we agree) butthe records frorn
wavrini has significantly reduced manual web- eastern Antioquia, eastern Caldas, and west-
bing (but cannot be confused with the small ern Boyacá, Cundinamarca, and Santander re-
species ofthe group, whose manual webbing flect aMiddle Magdalena distribution (Fig. 1),
is very reduced). Hyla rosenbergi is an only a pattern reinforced by such taxa as Bufo
slightly smaller treefrog, also with extensive haemaiiticus, Hyalinobatrachium co/ymbiphy-
webbing of the hands. Unlike the other two llum, and Hylapalmeri. Minimally, H. boans
large species (frorn aColombian perspective), isdistributed inthe biogeographic Chocó and
it lacks a reticulurn on the lower palpebrum Nechí districts (following Müller 1973) ofCo-
(rather pigment isdistributed unifonnly). The lombia, as well as the connecting terrain.
493
The distributions ofthe gladiator frogs
78 67
8 - 8
4
o
200
KILOMfTE1I:S
~ 3000 METERS & ABOVE
3 1000 METER CONTOUR ,é/ 3
,
r
~---78L-------------------~~~----------------~--------6~8------~
Figure 1. Distribution of Hyla boans in Colombia. Solid symbols represent localities for
specimens examined. Open symbols represent literature records (Kluge 1979).
494
.101mD. Lynch &Ángela M. Suárez-Mayorga
In the collections of ICN, there are only 12 graphic Chocó (Costa Rica to western Ecua-
males from the biogeographic Chocó (76.9- dor, Boulenger 1898, Cochran & Goin 1970,
107.6 [mean 87.8 ±2.3] mm SVL and three Duellman 1970 and Kluge 1979). At least in
females [82.5-97.1, mean 89.7] mm SVL). the Middle Magdalena, H. boans co-occurs
These are notably smalIer than the frogs re- with either "H.crepitans" orH.pugnax, each
ported by Duellman (1970) from eastern of which isa markedly smaller frog than H.
Panamá. Thesamp1efromthe interior (Middle boans (even considering that the Middle
Magdalena, excluding the headwaters ofthe Magdalena H.boans aresmall incomparison
Río Sinú) consists of20 males, five females, with populations in western Colombia,
andthreejuveniles. The males are68.9to95.3 Panama, orEcuador).
(mean 81.6±1.7)mm SVL, theovigerous fe-
"Hyla crepitalls"Wied
males are 70.8 to 84.2 (mean 77.0) mm SVL
Like Kluge (1979), we cannot assert that this
and thejuveniles 45.6 to 58.6 mm SVL. The
species isthe same asthat described by Wied
samples areminimal, but itwould appear that
(1824) from extreme eastern Brasil (Bahía).
specimens of H. boans from the Middle
That caveat given, wewi11refertothesmallest
Magdalena are smaller than those from the
Colombian species ofthe H. boans group as
Chocó(andPanama,following Duellman 1970,
"H.crepitans ".InColombia, "H.crepitans"
and Kluge 1(79)-adding to the geographic
occurs ashigh as2450 m(western Santander)
variation insizesofadult ofH.boans reported
and there are many records aboye 1000 mon
byLynch&Vargas (2000).
bothflanksofthe Cordillera Oriental. The low-
In Colombia, H. boans occurs at elevations est altitudinal record is 5 m (Puerto Colom-
below 1000mandwesuspect that itisdistrib- bia, Atlántico). The species isabundant inthe
uted continuously along the biogeographic middle andsouthern portions ofthe Magdalena
Chocó (Panamá to northwestern Ecuador), drainage andnormally scarce(meaning thatwe
across the northern fringes of the cordilleras have few records as vouchers) towards the
Occidental and Central into the Middle Caribbean coast. The distributional disjunc-
Magdalena and then discontinuously (due to tion (absence ineastern Panamá and western
the elevations ofthe Cordillera Oriental) into Colombia) noted by Kluge (1979) appears to
the eastern lowlands (Fig. 1).Combining our bereal (we areaware ofnorecords tothewest
datawiththatofLynch &Vargas (2000), large ofthe Golfo deMorrosquillo inColombia, an
frogs occur inthe biogeographic Chocó and areawithrelatively fewvouchers).
in the drainage of the Río Amazonas but Kluge (1979) emphasized that "H.crepitans"
smaller H. boans occur in the Middle was rarely sympatric with H.pugnax, andthat
Magdalena andwestern Orinoco. itnever co-occurred with both H.pugnax and
Inthe biogeographic Chocó, H.boans issym- H.rosenbergi (something that would require
patrie with the slightly smaller H.rosenbergi, observation in central Panama, where he
afact that may explain itslargesizethere (un- worked, under our assumption that "H.
derthe assumption thatthese two compete for crepitans" is discontinuously distributed in
some resource, perhaps afoolish assumption, northwestern Colombia and Panamá). InCo-
because this argument does not account for lombia, three gladiator frogs (H. boans, H.
equally largespecimensfromAmazonia, where pugnax, and H. rosenbergi) co-occur in the
H. boans isnot sympatric with another large headwaters ofthe Río Sinú and inthe Parque
speciesofthe group). Hylarosenbergi ispoorly Nacional Natural de Katios in the northern
represented inthe ICNcollections but occurs Chocó, otherwise, nomore than two aresym-
in southern Córdoba as well as the biogeo- patrie (although localities forthree, H.boans,
495
The distributions ofthe gladiator frogs
"H crepitans ". and H pugnax, are spaced (2-3 weeks ofobservation in 1998and 1999) as
very closely ineastern Caldas). InColombia, did 10days offieldwork atElCeibal (Municipio
"H crepitans" is never sympatric with H Santa Catalina) innorthern Bolivar inOctober
rosenbergi and appears to rarely be sympatric 2000 and one night of fieldwork in southern
with H boans (two occurrences, Municipio Depto. Atlántico inOctober 2000 - and attwo
Yacopí, Cundinarnarca, and at Orocué, localities in cis-Andean Arauca (but see be-
Casan are, aswell asKluge' s[1979] report from low). Hundreds (or thousands) of H pugnax
either extreme eastern Boyacá or adjacent were heardl observed but not one "H
Meta, at Guaicaramo). Kluge (1979) reported crepitans " during fieldwork at El Ceibal,
"H crepitans" from 28 trans-Andean locali- Polonuevo, and just south of Santa Marta.
ties in Colombia and, for two ofthese, noted These observations do not contradict the claim
sympatry with H pugnax. (Kluge, 1979) that the two species arel were
Our observations (hundreds ofhours of casual sympatric, intheearly 1920's,atFinca ElAranar,
observation on the part of JDL during the past near Bonda or at Fundación (both in Depto.
22 years, supported by the recalled observa- Magdalena), based on the fieldwork of
tions ofthe late Pedro M. Ruiz,pers. comm., Alexander Ruthven, butwe suspect that ineach
and our consultation offieldnotes available in case a macrogeographic perspective has ob-
the Laboratorio de Anfibios, ICN) ofthese two scured the actual distributions ofthe two spe-
species suggest that they differ ecologically in cies. Inthe 191Osand 1920s, afinca (El Aranar)
that "H crepitans "nearly always calls from or atown (Fundación) presumably covered an
the ground (like H rosenbergi, based onJDL's area with a diameter measured in kilometers
recalled observations in western Ecuador) (not allowing atest ofour suggestion that these
whereas H pugnax calls from shrubs ortrees two species are never microsympatric). Taking
(like H boans and H wavrini, although once locality data atface value, without reference to
a nest is made or selected, H boans and H the collector's fieldnotes, also provides the il-
wavrini can be found calling from the ground lusion of sympatry at two localities in Depto.
aswell). We must emphasize that these obser- Arauca (Orinoquian Colombia). Kluge (1979)
vations were casual, and are perhaps mistaken, reported H pugnax from Girardot (Depto.
but suggest some behavioral difference, not Cundinamarca) and from Espinal and Honda
noted previously, between these two species. (Depto. Tolima) inthe Magdalena valley.just
The trans-Andean records (from Colombia) west of Bogotá. We have experience at Nilo
available to us reinforce Kluges assertion - (Depto. Cundinamarca) andatVenadillo (Depto.
that sympatry of "H crepitans" and H pugnax Tolima), geographically intermediate sites,
israre or non-existent. At acoarse-grain level, where we observedl heard onIy "H crepitans".
the two are sympatric in the Parque Tayrona The record sfrom Depto. Sucre involve speci-
(Depto. Magdalena) ~ but the park isa large mens taken by different collectors indifferent
geographic unit and JDL's experience in 1983 years. The only locality for which sympatry
revealed only "H crepitans" inthree weeks of seems certain is in Dept. Caldas (lCNMHN
fieldwork (and see below)- and Kluge's (1979) 34533-38, H pugnax, and 34548-51, "H
reports for two localities inDepto. Magdalena crepitans ")where both species were collected
(and five records reported here, two inDepto. inapasture on23 March 1994.
Arauca one inDepto. Caldas, and two inDepto. Hyla crepitans occurs from the Caribbean
Sucre). Collecting/ observation on the part of coast tothe upper Magdalena valley ateleva-
JDL in and around the campus of the tions between sea level and more than 2400 m.
Universidad de Magdalena (just S of Santa In the Parque Nacional Natural Tayrona
Marta) revealed only the presence ofH pugnax (Depto. Magdalena), "H crepitans" builds
496
John D.Lynch &Ángela M. Suárez-Mayorga
nests made of small diameter gravels (Lynch males). Our measurements of228 adult males
& Vargas, 2000) but otherwise itdeposits its and 54 adult females (from Colombia) ingen-
floating egg film on the surface of ponds (pers. eral confirm Kluge's (1979) claim that "H.
observations at sites aboye Cañaverales, crepitans" from the Llanos are larger than
Magdalena, aswell asatVenadillo, Tolima, and those frorn the Caribbean/ Magdalena (Table
in the vicinity of Villavicencio, Meta). 1).That said, there are significant differences
ICNMHN 33115 (H. pugnax) isrecorded from in size arnong adults from Meta as contrasted
"Cañaverales," a site a few meters aboye sea with those frorn slightly higher elevations in
leve\. In 1983, .rOL did not collect at Boyacá and Cundinamarca or those from the
Cañaverales (the staging site for the Segunda lowlands of Arauca and Casan are. Ingeneral,
Expedición Botánica) but did collect at the frogs frorn trans-Andean Colombia are smaller
slightly higher El Cedro, where only "H. than those from cis-Andean Colombia but the
crepitans" were found/ observed. These 10- small sample available from southern Tolima
calities are separated by only a few kilorne- consists of frogs as large as those frorn cis-
terso Maps ofthe scale used here and by Kluge Andean Colombia.
(1979) provide the illusion of syrnpatry. No Other than sizes of adults, Kluge (1979) dis-
records are available for this species from the tinguished "H. crepitans" and H.pugnax us-
drainage ofthe Río Cauca (a pattern repeated ing webbing ofthe hands (the membrane does
by most species ofthe Caribe/ Río Magdalena not reach the distal subarticular tubercles in
pattern). The species also is found along the "H. crepitans" and does in H.pugnaxs, lower
piedmont of the departments of Arauca, palpebral pigmentation (anterior half in "H.
Casanare, and Meta to the east of the Andes crepitans ",only the dorsal edge inH.pugna,),
and inthe southwestern edge ofthe drainages two characters ofpigmentation (anteriorthigh
toLago Maracaibo inVenezuela (Fig. 2). Given with bars in H. pugnax but not in "H.
that the species can occur aboye 2000 m, the crepitans",flanks with prominent vertical bars,
Táchira depression ofwestern Venezuela isnot often doubled, inH. pugnax, whereas the bar-
a barrier (a second pass is available just N of ring istenuous and simple in "H. crepitans")
Picachos [Caquetá= Huila], enabling contact and calls (a feature obvious to anyone with
between the populations ofthe Magdalena and experience with each species). Inaddition, we
Orinoco) . The eastern limits of distribution noticed that "H. crepitans" has generally less
remain to be described. Personnel frorn ICN webbing of the interior toes than does H.
have rarely collected the llanos atany distance pugnax (Fig. 3), anobservation expected, given
from the Cordillera Oriental and the most east- than webbing of the hand is correlated with
ern localities of gladiator frogs suggest the webbing ofthe foot.
presence of H.pugnax, not H. crepitans (see
below), a further issue for study. HylapUgl111X Schmidt
Kluge (1979) noted that "H. crepitans" isthe Kluge (1979) pointed out that Duellrnan (1970)
smallest of the species ofthe H.boans group erred in considering this a synonym of "H.
and noted that some geographic variation in crepitans" and declared ita near-endemic of
size is apparent. Our data suggest that geo- Colombia. He reported itfrom 30 trans-Andean
graphic variation in size is much more corn- Colornbian localities and one in the Colom-
plex than suggested by Kluge (1979), who bian Amazon (not rnapped, nor commented
found that the trans-Andean Colombian popu- upon), saw a few (6) Panamanian specimens
lations (59 males, 14 females) were smaller (including the holotype), and disputed arecord
frogs than those from Panamá (31males, 9fe- from relatively high altitudes (Garagoa,
males) and the Orinoco (27 males and 14fe- Boyacá) inthe Amazonian versant. We agree
497
The distributions ofthe gladiator frogs
78 67
8 8
200
KIlOMETER5
~ 3000 METE RS&..•IO.VE
·3 1000 METER CON TOUR 3
L----7~8----------------------74L-----------------~---------6~8-------J
Figure 2. Distribution of "H. crepitans" inColombia. Solidsymbols represent localities for
specimens examined. Open symbols represent literature records (Kluge 1979).
498
JohnD.Lynch&ÁngelaM.Suárez-Mayorga
with his skepticism concerning the Garagoa The available data require that we view H.
record, in part because we have taken "H. pugnax asalow-elevation treefrog (below 500
crepitans" inthe Garagoa area and doubt that m)but weare unable toexplain why this spe-
H. pugnax occurs at such elevations. Hyla ciesshould berestricted tovery lowelevations.
pugnax iswidely distributed inthe lowlands It may be that H. pugnax (in contrast to H.
(below 500m)ofthe Caribbean region ofCo- crepitans) isunable to breed inthe cooler en-
lombia and the Magdalena valley but isalso vironments aboye 500 m(aproposition easily
found inwestern Venezuela andintheOrinoco tested under laboratory conditions). Review-
of Colombia (Fig. 4). The records from the ing fieldnotes (deposited inthe laboratory of
Caribbean coastandtheMagdalena valleypro- amphibians, ICN)suggests thatH.pugnax may
videanaltitudinal range ofH.pugnax between be more inclined to utilize waters associated
5 and 450 m aside from one record (Minca, with streams that H. crepitans (mostIy adeni-
Depto. Magdalena, aboye 1500 m, that we zen ofponds, but not inthe Sierra Nevada de
question/ reject). The present elevations ofthe Santa Marta). Ifwe are correct in accepting
Táchira depression greatly exceed the altitu- this supposition, then the difference inaltitu-
dinal range of H. pugnax (see below) but we dinal ranges ofthe two species isdifficult to
presume that Pleistocene effects allowed com- explain (one might expect H.pugnax toenjoy
munication of populations between cis and agreater altiudinal distribution/ range than H.
trans-Andean South America. crepitans, stream habitats being more normal
Table 1.Geographic variation insizes of adult "Hylacrepitans" inColombia. The first line
reports the range and sample size,the second linethe mean ± 1standard error ofthe mean.
Males Females
51.3- 64.6 (29) 50.5- 66.4 (10)
Arau ca -Cas anare
55.9 ± 0.5 61.1±1.4
47.6-60.5 (23) 58.2-61.0 (3)
Eastern Boyacá/ Cundinamarca
54.0 ± 0.7 60.0
45.5-56.0 (55) 52.3-62.4 (8)
Meta
51.6±0.4 57.9 ± 1.4
47.7-54.5 (4) 53.7-56.7 (2)
Cesar
50.8
46.4-57.3 (29) 54.1-61.5 (15)
Magdalena
52.3± 0.5 58.1 ± 0.6
43.0-52.7 (13) 51.0-54.1 (3)
Chara lá/ San Gil, Santander
47.4 ± 0.9 53.0
43.2-50.6 (30) 54.0-57.0 (4)
San .losé de Suaita, Santander
47.2±0.3 55.6
46.3-51.7 (lO) 53.2-62.6 (6)
Western Boyacá/ Cundinamarca
48.6 ± 0.6 57.6
40.8- 52.3 (17)
Caldas
49.7 ± 0.7
Northern To lirna 48.8- 56.4 (2) 52.8-55.5 (2)
50.7-58.7 (11) 62.6
Southern Tolima
54.7 ± 0.8
45.5-48.1 (5) 57.8
Huila
47.0
499
The distributions ofthe gladiator frogs
at elevation than are pond habitats) - as a The differences inaltitudinal distributions (H.
result of this observation, we suspect that crepitans at5-2450 mandH.pugnax at5-450
temperature plays agreater role than altitude m) mean that, at rnany localities, only H.
(even though the two are grossly correlated, crepitans might be found but, that alllocali-
asuggestion that merits testing under labora- ties for H. pugnax potentially allow for sym-
tory control). patry (within the biogeographic unit west of
In contrast to "H. crepitans ", H.pugnax is the Cordillera Oriental, as well as the
nearly invariate in size (between sexes and Maracaibo Basin). Inspite ofthis observation,
between localities). Kluge (1979) asserted that we remain impressed bythe near absence of
mean sizes ofmales andfemales were notdis- sympatry (the absence, based on our experi-
tinguishable statistically and our study of ences) between "H.crepitans" andH.pugnax
samples from Bolivar(two localities), Boyacá, (recall ourcomments aboye). Reinforcing this
Cesar, and Córdoba, agree fully withhisposi- position isthe discovery of different species
tion, even though there are significant differ- at localities very near one another inthe ab-
ences (perhaps artifacts of sample sizes) be- sence ofsome obvious difference inclimate or
tween sizes atdifferent localities (Table 2). vegetation. Inthe upper part of the valley of
iO
\
\¿··:IO '\)":~:
O,,"
'! .::.: r;,
·7~
\. (:.
Figure 3. Plantar views of feet of "Hyla crepitans" (left, ICNMHN 40835) and H.pugnax
(right, ICNMHN 44440) showing difference inwebbing between innertwotoes. The difference
isbest described as "H.crepitans" hasmore deeply incised webbing between toes 1and Il.
500