International Society for the Study

and Conservation of Amphibians

(International Society of Batrachology)

SEAT

Laboratoire des Reptiles et Amphibiens, Muséum national d'Histoire naturelle, 25 rue Cuvier, 75005

Paris, France. — Tel.: (33).(0)1.40.79.34.87. — Fax: (33).(0)1.40.79.34.88. - E-mail: dubois@mnhn.fr.

BOARD

President: W. Ronald HEvEr (Washington, USA).

General Secretary: Alain Dupois (Paris, France).

Treasurer: Jean-Louis DENIAUD (Paris, France).

Deputy Secretaries: Britta GriLLrrscH (Wien, Austria); Stephen J. RICHARDS (Townsville, Autralia).

Deputy Treasurers: Julio Mario Hoyos (Paris, France); Mark L. WyGoDA (Lake Charles, USA).

Councillors: Rafael DE SA (Richmond, USA); C. Kenneth Dobp, Jr. (Gainesville, USA); Jon LoMAN

(Lund, Sweden).

TARIFF FOR 1999

Individuals Institutions

Subscription to Alyres alone 280FF/565 (regular) 560FF/112$

140 FF / 28 $ (student)

Subscription to Alyres + ISSCA + Circalytes 300 FF/60$ (regular) 600FF/120$

150 FF / 30 $ (student)

Back issues of Alyres: single issue T0 FF/14$ 140 FF/28$

Back issues of Alpes: one complete volume (4 issues) 225 FF/45$ 450 FF/90$

Back issues of Alytes: complete set volumes 1t015 2700 FF/540$ 5400 FF / 1080 $

Five-year (1999-2003) individual subscription to Alytes: 1120 FF / 224$.

Five-year (1999-2003) individual subscription to Afyres + ISSCA + Circalytes: 1500 FF / 300$.

Life individual subscription to Alytes from 1999 on: 5600 FF / 1120 $.

Life individual subscription to Alytes + ISSCA + Circalytes from 1999 on: 7000 FF / 1400 $.

Patron individual subscription to Alytes from 1999 on: 11200 FF / 2240 $ or more.

Patron individual subscription to Alyres + ISSCA + Circalytes from 1999 on: 14000 FF / 2800 $ or

more.

Important notice: from 1996 on, any new life or patron individual subscriber to A/ytes will be offered a free

complete collection of back issues of Alytes from the first issue (February 1982) until the start of her/his

subscription.

Circalytes is the internal information bulletin of ISSCA. Back issues of this bulletin are also available:

prices can be provided upon request by our Secretariat.

Inclusive Section or Group affiliation to ISSCA: 250 FF / 50 $.

Individual subscription to the ISSCA Board Circular Letters: 200 FF / 40 $.

MODES OF PAYMENT

— In French Francs, by cheques drawn on a French bank payable to “ISSCA”, sent to our Secretariat

(address above).

— In French Francs, by direct postal transfer to our postal account: “ISSCA”, Nr. 1-398-91 L, Paris:

if you use this mode of payment, add 15 FF to your payment for postal charges at our end.

— in US Dollars, by cheques payable to “ISSCA”, sent to Mark L. WyGoDa, Department of Biological

and Environmental Sciences, PO Box 92000, McNeese State University, Tale Charles, Louisiana

70609-2000, USA.

Source : MNHN, Paris

AIRNTTES

INTERNATIONAL JOURNAL OF BATRACHOLOGY

June 1999 Volume 16, N° 3-4

Alytes, 1999, 16 (3-4): 77-83. 77

Combat behavior in Centrolene bucklevi

and other centrolenid frogs

Wilmar BoLivaR-G., Taran GRANT ! & Luis A. OsORIO

Laboratorio de Herpetologia, Departamento de Biologia, Universidad del Valle, A. A. 25360,

Cali, Valle del Cauca, Colombia

Observations of fighting behavior in Centrolene buckleyi revealed that

males dangled by their feet and grappled venter-to-venter. One of the males

repeatediy uttered a soft, short squeak, inflating the vocal sac and prving off

the other combatant in the process. Sonagrams of this aggressive call and

the advertisement call reveal markedly different structures. Both frogs were

visibly injured, presumably in combat. Of the few species coded for combat

behavior (n = 7), those of the genera Centrolene and Cochranella exhibit

the derived state of dangling by the feet and grappling venter-to-venter,

whereas species of Hyalinobatrachium have primitive combat composed

of one male grasping the other in amplexus. We predict that the derived

behavior will be discovered in all Centrolene and in all or a large part of

Cochranella (representing a synapomorphy that unites the two groups),

and that no Hyalinobatrachium species will exhibit the apomorphic state.

INTRODUCTION

Although the past few decades have seen an unprecedented increase in our knowledge of

centrolenid frogs, most workers have concentrated on resolving taxonomic and phylogenetic

issues from a strictly morphological perspective. In so doing, they have inadvertently ignored

an abundance of characters that could provide invaluable clues as to the relationships of these

frogs. The purposes of this paper are to describe the physical combat of Centrolene buckleyi

(Boulenger, 1882) and to discuss the phylogenetic significance of combat behavior in the

family Centrolenidae.

1. Current address: Department of Herpetology, American Museum of Natural History, Central Park West at

T9th Street, New York, NY 10024, USA

Bibliothèque Centrale Muséum

(l : MNHN, Pas

78 ALYTES 16 (3-4)

METHODS

Our observations of Centrolene buckleyi combat were made on 2 April 1996, just below

El Boquerôn, near the border between Departamento del Valle del Cauca and Departamento

del Chocé at 2220 m in the Cordillera Occidental of Colombia (GPS coordinates: 4°4439"N,

76°18/16"W). The site is in relatively intact cloud forest along the road from El Cairo to El

Boquerén.

The frogs were found at the side of the road approximately 2.5 m directly above a trickle

of run-off water and 6 m from a fast flowing stream along which we heard many advertisement

calls of this species. It rained quite heavily throughout most of the day and night, before,

during and after these observations.

Calls were recorded using a Sony WM D6C Professional Walkman and a Sony ECM 909

microphone at 15.6°C air temperature. Sonagrams were generated on a Kay 5500 DSP

Sona-Graph. Data were also obtained using Computerized Speech Research Environment

(CSRE) 4.5 PC-based signal analysis software.

Preserved specimens are kept in the Universidad del Valle Colecciôn de Anfbios y

Reptiles (UVC).

RESULTS

Our observations began at 22.05 h. Frog A (UVC 12729; SVL 28.7 mm) was hanging

from a horizontal twig by its feet and was grasping frog B (UVC 12730; SVL 27.9 mm) with

its hands at the base of B’s arms. B was grasping À in the same way, but was hanging with its

feet free. The two frogs were facing the same direction.

B began swinging its body and grasped a leaf withits foot. After several minutes, itswung

its body up and wrapped its feet and legs around A’s body. At this point the two frogs were

oriented venter-to-venter, hanging head down (fig. 1).

A then began to emit a call sporadically, consisting of a single, short, soft squeak lasting

0.14-0.15 s (fig. 2A). The frequencies of this call fall between 4000 and 7100 Hz and are

strongly modulated. The emphasized frequency begins at around 4600 Hz and climbs to near

5000 Hz, at which point it jumps to about 7100 Hz and then rather abruptly falls to terminate

near 6000 Hz. The first part of the call is notably pulsatile (although the number and pattern

of pulsesis variable), while the second, higher, part is only weakly so. When A called, the vocal

sac expanded, which, in turn, pushed B from A, thereby loosening B's grip. After several

minutes of this behavior, at 22.42 h, B fell to a leaf below, at which time it was collected. A

climbed up onto the twig from which it had been hanging, and it was also collected. The

observed combat lasted 37 min.

Both frogs were visibly wounded, presumably in combat. B, the “loser” of the encounter,

had a red, swollen hematoma just dorsal and slightly anterior to the insertion of the right arm;

the location of the injury corresponds to the position of A’s humeral spine during combat.

While A did not show any wounds or marks directly attributable to B's humeral spine, the skin

Source : MNHN, Paris

BoLivar-G., GRANT & OSORIO 79

Fig. 1. - Illustration of combat in Centrolene buckleyi

on the dorsal surface of the outer edge of the right hand and fingers III and IV was torn. The

right hand was observed not to be used when climbing in the plastic collecting bag. Both

specimens were sluggish once collected.

DISCUSSION

The above description conforms well with descriptions of combat in Cochranella grif-

fithsi Goin, 1961 (DUELLMAN & SAVITZKY, 1976) and C. ignota Lynch, 1990 (RESTREPO-TORO,

1996), both of which lack the humeral spine in males (although C. griflithsi males exhibit “a

large bladelike ventral crest on the humerus”; LYNCH & RUIZ-CARRANZA, 1997: 529, fig. 3).

Similar fighting has also been observed in Centrolene prosoblepon (Boettger, 1892) by JACOB-

SON (1985) and C. acanthidiocephalum (Ruiz-Carranza & Lynch, 1989) by Pedro M. RUIZ-

CARRANZA (personal communication), two species which exhibit a humeral spine in males.

There are two differences between previous observations and ours. First, previous

reports have not mentioned any evidence of physical damage inflicted by the humeral spine

(although JACOBSoN observed seven combat encounters). It is common to find scars on the

head and body of males of Centrolene geckoideum Jiménez de la Espada, 1872 (personal

Source : MNHN, Paris

80 ALYTES 16 (3-4)

4 1 Î

2 1

1 A | B

5 —————— —— ————

8 0 02 0 0.2 04 0.6 0.8 1.0 12 14

Time in seconds

Fig. 2. - Audiospectrograms (graphed with wide-band 300-Hz filter) and waveforms of Centrolene

buckleyi vocalizations, both recorded at 15.6°C air temperature, produced from tape copy on

AMNH herpetology reel 276. (A) Combat call of UVC 12729. (B) Advertisement call of UVC

12589; C. buckleyi was observed to call sporadically, .e., couplets are not usually emitted in series.

observation), presumably inflicted in combat by the extremely large and sharp spines of this

species (as opposed to the blunt spine of C. buckleyi, see RUIZ-CARRANZA & LYNCH, 1991,

and RUEDA-ALMONACID, 1994, for spine shape and size in C. buckleyi and C. geckoideum,

respectively), but to date combat has not been observed to confirm this suspicion. Our

evidence of physical damage inflicted by the humeral spine is cireumstantial (i.e., we did not

examine the individual immediately prior to combat, so we cannot confirm the origin of the

hematoma) but is more convincing than any previously reported.

The second difference is the use of a call during combat. Our interpretation of this event

as non-accidental is based on the fact that we observed frog A to produce over 25 such calls

(including 14 in recordings TG 9604 and 9605), all with the same effect. Inasmuch as the call

or, more precisely, the inflation of the vocal sac - appeared to be used to physically loosen the

opponent’s grip during combat, we suspect that the acoustic qualities of the call are unim-

portant. Despite this conjecture, the call varies remarkably little; all of the calls recorded

exhibit essentially the same amplitude and frequency modulation as that shown in fig. 2A, i.e.,

itis not simply a random emission of sound made while inflating the vocal sac. As seen in fig.

2, this call differs markedly from the advertisement call (fig. 2B), which is a high-pitched,

Source : MNHN, Paris

BoLivAR-G., GRANT & OSORIO 81

Centrolene Cochranella Hyalinobatrachium

Humeral Spines Bulbous Liver

Combat Behavior

Fig. 3.- Conjectured phylogeny of centrolenid genera based exclusively on unique synapomorphies, i.e.,

character states that do not occur in any other anuran.

pulsed croak of 0.12-0.15 s duration consisting of two notes; the first note contains between

three and six well defined pulses, while the second note contains three. The two notes are

separated by 0.05-0.06 s intervals. The emphasized frequency lies at approximately 5200 Hz.

The aggressive call also differs acoustically from the short, sporadic bursts of random noise

that compose the encounter call of Centrolene buckleyi (fide John D. LYNCH, personal

communication). Although Bis an adult with vocal slits, it was not observed to call during the

event.

The physical combat of frogs of the genus Hyalinobatrachium differs from that of

Centrolene and Cochranella. MCDiARMID & ADLER (1974) described the combat behavior of

H. fleischmanni (Boettger, 1893) (as Centrolenella viridissima Taylor, 1942) and H. valerioi

(Dunn, 1931), in which one of the males grasps the other in amplexus; their description of H.

fleischmanni combat was corroborated by GREER & W 1980) and JAcOBsON (1985). Strict

outgroup comparison (sensu LyYNCH, 1997: 355, footnote 2) reveals that this is the primitive

behavior, while combat in which males dangle by their feet grappling venter-to-venter is

derived.

Source : MNHN, Paris

82 ALYTES 16 (3-4)

Although the data set is exceedingly small (data are available for only 6.1 % of the

family), it is sufficient to allow us to make a number of predictions based on a cladistic

interpretation of known character distribution and published phylogenetic hypotheses (pri-

marily RUIZ-CARRANZA & LyNCH, 1991). First, we predict that the derived combat will be

found in all 33 species of Centrolene for which combat remains to be observed. Similarly, we

predict that none of the 24 uncoded species of Hyalinobatrachium will exhibit this derived

state (i.e., they will exhibit either the plesiomorphic state or some other, unknown type of

combat). Convincing evidence of monophyly has not been put forth for the more than 50

species (or any sizeable portion thereof) placed in Cochranella. However, the expression of the

derived type of combat in two small but seemingly quite distantly related monophyletic

groups of Cochranella — viz., the ocellata group sensu stricto (i.e., sensu LYNCH, 1990) and the

griffithsi group (sensu LYNCH & RUIZ-CARRANZA, 1997: 529; named by RUIZ-CARRANZA &

LyNCH, 1995: 3) — is suggestive of a widespread distribution of this state throughout

Cochranella. Consequently, we postulate (fig. 3) that the derived combat behavior constitutes

a synapomorphy for Centrolene + (at least some part of) Cochranella, and therefore resolves

the polytomy reported by RUIZ-CARRANZA & LYNCH (1991). Data on the use of an aggressive

call in combat are too limited (7 = 1 species) to be phylogenetically informative at this

time.

RESUMEN

En nuestras observaciones del combate fisico de Centrolene buckleyi, los machos se

colgaron de los pies y pelearon vientre-a-vientre. Un macho emitié repetidamente un chillido

débil y corto, y asi inflaba la bolsa vocal y empujaba al otro combatiente en el proceso. Los

sonogramas de este canto agresivo y el canto de advertencia demuestran estructuras marca-

damente diferentes. Cuando las colectamos, ambas ranas estaban visiblemente heridas,

presumiblemente durante el combate. De las pocas especies codificadas por el comporta-

miento de combate (n = 7), Centrolene ÿ Cochranella exhiben el estado derivado de colgarse de

los pies y pelear vientre-a-vientre, mientras que Hyalinobatrachium presentan el combate

primitivo en el cual un macho agarra al otro en amplexus. Predecimos que se descubrirä el

comportamiento derivado en todas las especies de Centrolene y una gran parte de Cochranella

(y por ende representa una sinapomorfia para estos dos grupos), y que ninguna especie de

Hyalinobatrachium presentarä el estado apomérfico.

ACKNOWLEDGMENTS

We thank Janalee P. CALDWELL, Maureen A. DONNELLY, John D. LYNCH, Pedro M. Ruiz-

CaRRANZA and Erik R. WiLD for encouraging our work and improving the manuscript. We also thank

two anonymous reviewers for their constructive comments. The illustration of combat was rendered by

Santiago CasraRo. Charles W. Myers graciously allowed TG to use the American Museum of Natural

History equipment to generate and prepare sonagrams. The recording equipment and GPS were provided

by Idea Wild. Field work at El Boquerôn was made possible in part by the Departamento de Biologia,

Source : MNHN, Paris

BoLivar-G., GRANT & OSORIO 83

Universidad del Valle and donations to the Kelowna Museum by L. BELL and H. D. GRANT. Fernando

CasrRo’s efforts to promote herpetological investigations at the Universidad del Valle were the primary

stimulus for our work.

LITERATURE CITED

DUELLMAN, W. E. & SavrrzkY, A. H., 1976. - Aggressive behavior in a centrolenid frog, with comments

on territoriality in anurans. Herpetologica, 32: 401-404.

GREER, B. J. & WELLS, K. D., 1980. - Territorial and reproductive behavior of the tropical American frog

Centrolenella fleischmanni. Herpetologica, 36: 318-326.

JacosoN, $. K., 1985. - Reproductive behavior and male mating success in two species of glass frogs

(Centrolenidae). Herpetologica, 41: 396-404.

LyNCH, J. D., 1990. — A new ocellated frog (Centrolenidae) from western Colombia. Proc. biol. Soc.

Washington, 103: 35-38.

Ce 1997. — Intrageneric relationships of mainland Eleutherodactylus. I. A review of the Eleutherodac-

1vlus sulcatus group. Rev. Acad. colomb. Ciene., 21: 353-372.

LyNCH, J. D. & RuIZ-CARRANZA, P. M., 1997. — A remarkable new centrolenid frog from Colombia with

a review of nuptial excrescences in the family. Herpetologica, 52: 525-535.

McDiarmiD, R. W. & ADLER, K., 1974. - Notes on territorial and vocal behavior of Neotropical frogs of

the genus Centrolenella. Herpetologica, 30: 75-78.

RESTREPO-TORO, J. H., 1996. — Ecologia conductual de una rana arbérea Neotropical.Tesis de Magister,

Universidad del Valle, Cali.

RUEDA-ALMONACID, J. V., 1994. - Estudio anatémico y relaciones sistemäticas de Centrolene geckoideum

(Salientia: Anura: Centrolenidae). Trianea, 5: 133-187.

Ruiz-CaRRANZA, P. M. & LYNCH, J. D., 1991. - Ranas Centrolenidae de Colombia. I. Propuesta de una

nueva clasificaciôn genérica. Lozania, 57: 1-32.

-- 1995. - Ranas Centrolenidae. V. Cuatro nuevas especies de Cochranella de la Cordillera Central.

Lozania, 62: 1-23.

Corresponding editor: Janalee P. CALDWELL.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 84-96.

Anäâlisis trôfico en dos poblaciones

de Scinax nasicus (Anura, Hylidae)

de Argentina

Paola M. PELTZER & Rafael C. LAIMANOVICH

Instituto Nacional de Limnologia (INALI-CONICET),

José Maciä 1933, Santo Tomé (3016), Santa Fe, Argentina

A comparative study of diets and morphometric analyses were made in

Scinax nasicus in two localities of Santa Fe province (Argentina). À

discriminant analysis was_carried out to determine the. morphometric

variation of S. nasicus. The quantitative composition of diet for each

locality was studied through the quantification of the trophic spectrum,

niche trophic diversity and amplitude, prey size, as well as the Index of

Relative Importance. Trophic relationships were obtained using overlap

matrices based on Pianka’s index. The results showed that S. nasicus has a

strategic behavior for capture food between specialist and non-specialist

Csit-and-wait”).

INTRODUCCION

Scinax nasicus es un hilido que se distribuye en Argentina en las provincias de Corrientes,

Chaco, Formosa, Entre Rios, Jujuy, Salta, Santa Fe, Santiago del Estero y Tucumän; ademäs,

seencuentra en el sur de Bolivia, centro de Brasil, Paraguay y Uruguay (Cet, 1980; GALLARDO,

1987; LANGONE, 1994). Es frecuente hallarla asociada a ecosistemas periurbanos, en tanques

de agua, piletas, u otros lugares hümedos durante la estacién seca. En los ambientes del

litoral mesopotämico argentino su reproduccién ocurre generalmente en cuerpos de agua

temporarios desde octubre hasta abril, dependiendo del régimen pluviométrico. Pone huevos

que se encuentran sujetos, en forma de racimos gelatinosos, a plantas acuäticas (GALLARDO,

1987).

La informacién sobre datos bioecolégicos de S. nasicus es dispar. Una primera aproxi-

macién al conocimiento de su dieta, en häbitats naturales de la provincia de Corrientes

(Argentina), fue realizada por DURÉ & KEHR (1997). ORDANO et al. (1999) estudiaron sus

häbitos alimentarios en ambientes antrépicos de Santa Fe (Argentina). La dieta larval fue

descripta por LAJMANOVICH (1997) en ecosistemas del rio Paranä, provincia de Santa Fe.

La relacién entre caracteres biométricos y determinadas caracteristicas biolôgicas en

anuros ha sido tratada por diversos autores, e.g. EMERSON (1976, 1986), SALTHE & CRUMP

(1977), WiLeuR et al. (1978) y GATZ (1981).

Source : MNHN, Paris

PELTZER & LAJMANOVICH 85

Considerando que la utilizacién del alimento en los anfibios tiene un importante rol en la

dinämica poblacional y en las interrelaciones interespecificas, y que es un factor relevante

para la evoluciôn y organizaciôn de sus comunidades (CRUMP, 1974; DUELLMAN, 1978; TOFT

& DUELLMAN, 1979; JONES 1982), el presente trabajo tiene como objetivos analizar la dieta de

S. nasicus en dos localidades de la provincia de Santa Fe y comparar las caracteristicas

morfométricas de las ranas, que permitan establecer variaciones geogräficas en la especie.

MATERIALES Y MÉTODOS

ÂREAS DE ESTUDIO

Se colectaron manualmente un total de 50 adultos de Scinax nasicus durante el verano de

1996 en dos localidades de la provincia de Santa Fe, distantes aproximadamente a 300 km. La

fijacién de los ejemplares se realizé in situ con una soluciôn fijadora de formol al 10 %,

inyectändose fijador en la cavidad abdominal con el fin de detener los procesos digestivos,

tomando en cuenta que el tiempo que transcurre desde que los animales son capturados hasta

su preservaciôn puede afectar los resultados de los anälisis de dieta (CALDWELL, 1996).

Segün la clasificacién de las regiones batracolôgicas propuestas por Cr1 (1980), el sitio

Las Gamas se encuentra en la regién Chaqueña y el sitio Colastiné en la Litoral Mesopotä-

mica. En la colecciôn Herpetolégica del Museo Provincial de Ciencias Naturales “Florentino

Ameghino” de la ciudad de Santa Fe (Argentina) se conservan el total de ejemplares

utilizados en este estudio (ap. 1).

Sitio Las Gamas, Dpto. Vera, Santa Fe (29°27'S, 60°230)

Fitogeogräficamente se ubica en el Distrito Chaqueño Oriental (CABRERA, 1976). Se

caracteriza por presentar bosques semixerfilos de Schinopsis balansae alternando con pas-

tizales, esteros y bañados. Climäticamente, corresponde a una regién tropical con estacién

seca, temperatura media anual de 20°C, precipitaciones anuales medias entre 950 y 1000 mm,

y abundantes Iluvias en verano (aproximadamente 300 mm).

Sitio Colastiné, Dpto. La Capital, Santa Fe (30°40'S, 60°300)

El ärea se sitüa en la zona sur de la [lanura aluvial del rio Paranä. El ambiente se

caracteriza por presentar numerosos cursos de agua que forman una extensa red de drenaje

con gran nümero de islas y cuerpos lénticos de distinta importancia, tales como lagun:

bañados y pantanos. Fitogeogräficamente pertenece al Dominio Amazénico, Provincia

Paranaense, Distrito de Selvas Mixtas (CABRERA, 1976), destacändose especies vegetales

como Sulix humboldtiana, Acacia caven, Tessaria integrifolia, Croton urucurana y Sapium

haematospermum, entre otras. Climâticamente, el ârea corresponde a una Zona subhümeda-

hümeda mesotermal con temperaturas medias anuales de 18°C y precipitaciones anuales

medias de 1000 mm.

Source : MNHN, Paris

86 ALYTES 16 (3-4)

METODOLOGÏA DE LABORATORIO Y ANÂLISIS ESTADISTICO

En los ejemplares de ambos sitios se midieron longitudes de: hocico-cloaca (HC); ancho

cabeza (AC); distancia interocular (DIO); borde anterior del ojo a la narina (BAON); mano,

desde el tubérculo metacarpal externo al dedo mäs largo (LM); fémur (LF); tibia (LT); pie,

desde el tubérculo metatarsal al dedo mäs largo (LP). Las medidas se tomaron con un calibre

milimétrico de precisiôn 0,01 mm. A partir de estas medidas originales se determinaron sus

proporciones con respecto a la longitud hocico-cloaca. En el anälisis morfométrico, las

mediciones se transformaron a su logaritmo natural con el fin de asegurar su distribuciôn

normal y reducir la dispersiéôn de los datos (SOKAL & ROHLF, 1979).

La diferenciaciôn de las poblaciones se realizé a través de un anälisis discriminante entre

los 4 grupos (ejemplares colectados en Las Gamas y en Colastiné, hembras y machos).

Los cälculos estadisticos se realizaron con el programa STATGRAPHICS® Plus For

Windows (ANONIMO, 1994).

Con el fin de analizar la dieta, los estémagos fueron disecados y estudiados indivi-

dualmente. Para la determinacién y cuantificacién de los items alimentarios se consideraron

como individuos aquellas estructuras o piezas claves para la identificaciôn (cabezas, élitros

etc.).

Para calcular la diversidad trôfica de los contenidos estomacales se usé el método

propuesto por PIELOU (1966):

H= (IN) x (log, N!- Elog, Nj!),

donde N es el nümero total de organismos hallados en el estémago de cada individuo y Nes

el nümero total de organismos de la especie i en cada estémago.

Se calculé la diversidad media (H) y la diversidad tréfica acumulada (h,) que se utiliza

para determinar la muestra minima en estudios herpetolégicos (HURTUBIA, 1973), segün la

siguiente formula:

Die Ni Hi Ni Hi) / Ni = Nic)

donde H, y H4, son las diversidades tréficas acumuladas en k y k-1 estémagos, y N4 y

N4. son el nümero total de individuos de todas las especies presa en k y k-1 estémagos.

La amplitud tréfica del nicho se obtuvo mediante el indice de LEVINS (1968):

Nb= (EP),

donde P,;; es la probabilidad de la proporcién del item i en la muestra j.

Para establecer la contribuciôn de cada categoria de alimento a la dieta, se aplicé un

indice de importancia relativa segün PINKAS et al. (1971):

IRI = % FO (% N + % V),

donde % FO es la frecuencia de ocurrencia de las categorias de alimentos, % N es el

porcentaje numérico y % V el porcentaje volumétrico, calculado por desplazamiento de agua

con una precisiôn de 0,01 ml.

Source : MNHN, Paris

PELTZER & LAJMANOVICH 87

La comparacién de dieta de ambas poblaciones se elaboré en base al indice de solapa-

miento de PIANKA (1973):

= EP; Pix /(E P32 E Py2)”,

siendo P;; y P,, las proporciones en que los individuos j y k utilizan las diferentes clases que se

reconocen en el recurso i. Este indice presenta valores que varian entre 0 y 1 en sentido

creciente de coincidencia en la utilizacién de recursos.

RESULTADOS

Srrio Las GaMAS

Del total de 25 individuos colectados, 10 fueron hembras y 15 machos. La comparaciôn

entre la proporciôn de sexos no fue significativa: Ji? (con correcciôn de Yates) = 0,7; P > 0,05.

Los registros morfométricos del total de ejemplares colectados se presentan en la tab. 1.

Elespectro tréfico, basado en la identificaciôn de 56 presas, result integrado por 22 taxa

animales, restos de insectos no identificados y restos vegetales (tab. 2). La contribucién de

cada categoria de alimento a la dieta fue obtenida por la aplicaciôn del indice de importancia

relativa (IRI) (tab. 3, fig. 1), que presenté mayores valores en himenépteros y coleépteros. Las

presas con mayor porcentaje de presencia fueron himenépteros (Brachymirmex spp.) (40 %);

le siguieron las larvas de dipteros (36 %). Numéricamente, los himenépteros, con hormigas de

la familia Dorylidae, son los mäs importantes (17,8 %), seguidos por las larvas de dipteros

(5,8 %). Dentro de la fracciôn vegetal, el 40 % de los estémagos analizados presentaron resto

de tallos y hojas, que no fueron evaluados numéricamente.

La diversidad media (H) resulté 0,26 (s = 0,41). La diversidad trfica acumulada (h,) fue

de 3,26. Con la suma de las 25 muestras, la curva de diversidad tréfica tiende a la estabilizaciôn

(fig. 2). La amplitud del nicho (Nb) para el periodo estudiado presenté un valor de 11,5. La

distribuciôn de frecuencias del tamaño de presas (fig. 3) presenta una distribucin homogénea

en los intervalos considerados.

En el intestino medio y posterior de 10 hembras y 2 machos se encontraron un total de

12 paräsitos pertenecientes al phylum Nematoda.

Srrio COLASTINÉ

De los 25 ejemplares colectados, 12 fueron hembras y 13 machos. La comparaciôn entre

la proporcién de sexos no fue significativa: Ji? (con correcciôn de Yates) = 0,083; P > 0.05. Los

registros morfométricos del total de especimenes colectados se detallan en la tab. 1.

El espectro tréfico, resultado de la identificacién de 53 presas, estuvo integrado por

17 taxa animales, restos de insectos no identificados y restos vegetales (tab. 2). La contribu-

ciôn de cada categoria de alimento a la dieta (IRI) (tab. 3, fig. 1) mostré mayores valores en

Source : MNHN, Paris

88 ALYTES 16 (3-4)

100

xN LAS GAMAS COLASTINÉ

50 — HYMENOPTERA DIPTERA

DIPTERA OTROS HYMENOPTERA

INSECTA

]

%v ARACHNIDA ARACHNIDA

COLEOPTERA

100

29 27 12 23 9 EL 28 26 "12

% FO

Fig. 1. -Representaciôn gräfica del indice de importancia relativa (IRI) de los distintos componentes de

la dieta de Scinax nasicus en la provincia de Santa Fe. % N, porcentaje numérico; % V, porcentaje

volumétrico; % FO, porcentaje de ceurrencia.

Tabla 1. — Caracteristicas morfométricas (en mm) evaluadas en Scinax nasicus. HC,

longitud hocico-cloaca. Proporciones respecto de la longitud hocico cloaca: AC,

ancho cabeza; DIO, distancia interocular;, BAON, longitud desde el borde

anterior del ojo y la narina; LM, longitud de la mano, desde tubéreulo metacarpal

externo al dedo mäs largo; LF, longitud fémur; LT, longitud tibia; LP, longitud

del pie, desde el tubérculo metatarsal al dedo mäs largo. x, media; s,, error

eständar; s, desviacién eständar; , coeficiente de variaciôn. Diferencia de

medias: test f, x, > x,, Significaciôn: *, P < 0,01.

Las Gamas Colastiné

AC/HC

DIO/HC

BAON/HC

LM/HC

LF/HC

LT/HC

LP/HC

Source : MNHN, Paris

PELTZER & LAJMANOVICH

Tabla 2. — Dieta de Scinax nasicus en dos ambientes de la provincia de Santa Fe. n,

nümero total de los 25 contenidos estomacales; %, porcentaje de la categoria en

el total de las presas; f, frecuencia absoluta de la categoria en los estémagos; x,

no evaluado numéricamente; (ni), no identificado.

Sitio Las Gamas

Sitio Colastiné

ñ % fl % n

Categorias gi

Insecta

Coleoptera

Carabidae 1 LS 1 a a E

Coccinelidae 2 3,6 2 = +

Curculionidae 1 LE À = =

Elateridae 4 7,14 4 1 1,89 1

Scarabaeidae 1 LS 1 1 1,89 1

Hydrophilidae ï LE ï 3 5,66 2

Sylphidae I 1 L8 1 6 11,32 3

Dytiscoidea E = S 1 1,89 ï

Gi) = = = 2 3,77 2

Hymenopiera

Formicidae

AcromyrmEx Spp- Z 3,6 F 6 11,32 5

Brachymirmex spp._| 3 535 10 = a s

Dorylidae 10 17,85 3 £ 3,77 L

Mirmicinae .] 5,35 3 a BE

(ni) 3 3,35 2 4 7,54 3

Diptera

Tabanidae (larvae) 3 5,35 ] a E

Chironomidae = = = 5 9,43 2

Larvae (ni) 9 15,79 9 10 18,8 6

Pupae (ni). = 7 : 3 5,66 ï

(ni) = = 4 7,54 2

Orihoptera

Tridactyloidea 1 18 4 é =

Tettigonoidea 5 L D 139 ï

Homoptera

(ni) 1 18 1 = =:

Hemipiera

Corixidae 2 3,6 2 = = -

Lepidoptera

Tarvae (ni) 4 T4 4

Apierygota

Collembola 1 LS L -

Arachnida

Arancomorphae

Capontidae ï LE ï Z 377 2

Araneidae 1 LS 1 2

Dictynidae L S T KE T

Acariformes 1 LS 1 = =

Opiliôn (ni) = = = D T9 i

Restos de insectos (ni) x 16 x = 15

Restos vegetales x = 10 x 6

Total de presas 56 5

(4H) 0,26(+ 0,41) 0,22 (+ 0,32)

(Hk) 3,26 3,26

(Nb) 1,52 10,6

Famaño promedio de presa por estômago 3,97 mm (4 0,61) 2,48 mm (4 0,22)

Source : MNHN, Paris

90 ALYTES 16 (3-4)

Tabla 3. — fndice de importancia relativa (IRI) de los distintos componentes de la dieta

de Scinax nasicus. % FO, porcentaje de frecuencia de ocurrencia; % N,

porcentaje numérico; % V, porcentaje volumétrico.

Sitio Las Gamas

IRI = % FO

CN+% V)

Hymenoptera 1642

Coleoptera 1515

Diptera 554

Otros Insecta 488

Arachnida 147

IRI = % FO

CN+% V)

Diptera 2189

Coleoptera 1985

Hymenoptera 1159

Arachnida 120

13,2

dipteros y coleépteros. Las presas con mayores porcentajes de presencia fueron las larvas de

dipteros (24 %), y le siguen en orden de importancia las hormigas (Acromyrmex spp.) (20 %).

Numéricamente, las larvas de dipteros son las mäs representadas (18,8 %). El 24 % de los

estémagos contuvo restos de tallos y hojas.

La diversidad media (H) resulté 0,22 (s = 0,32). La diversidad tréfica acumulada (h,) fue

de 3,26 y con la suma de las 25 muestras la curva de diversidad trfica tiende a la estabilizaciôn

(fig. 2). La amplitud tréfica del nicho (Nb) en el periodo de estudio fue de 10,6. La

distribuciôn de frecuencia del tamaño de presas (fig. 3) presenté una mayor concentracién en

elintervalo 1,5-3 mm.

En el intestino medio de un ejemplar macho se encontraron un total de 2 nemätodos.

RELACION ENTRE LAS POBLACIONES

Las comparaciones morfométricas realizadas entre los individuos provenientes de

ambos sitios (tab. 1) arrojaron diferencias significativas en el 100 % de las medidas y relaciones

evaluadas; las medias de Las Gamas son mayores que las de Colastiné.

Source : MNHN, Paris

PELTZER & LAJMANOVICH 91

Diversidad tréfica acumulada (Hk)

0 T TT T Car

1234567 8 910111213141516171819202122232425

Némero de estémagos

—4- LAS GAMAS -æ- COLASTINE

Fig. 2. - Curva de diversidad tréfica acumulada (H,) versus nümero de estémagos analizados que

determinan la muestra minima para Scinax nasicus en las poblaciones Las Gamas y Colastiné. La

flecha sobre la curva indica aproximadamente el punto en donde se alcanza la estabilizaciôn.

%100

"Las Gamas" "Colastiné"

n=56 n=53

501

0+—

——

0 2 4 6 8 0 15 3 45 6

Tamaño de las presas (mm)

Fig. 3.- Distribuciôn de frecuencias de los tamaños de las presas consumidas por Scinax nasicus en Las

Gamas y Colastiné. Sobre los histogramas figura el nümero total de presas medidas (n)

Source : MNHN, Paris

92 ALYTES 16 (3-4)

El anälisis discriminante entre los 4 grupos (ejemplares de Las Gamas y Colastiné,

machos y hembras) explica en sus dos primeras funciones discriminantes (FD) el 96 % de la

variaciôn total. En la tab. 4 se representan los coeficientes de correlaciôn canénicos estanda-

rizados de cada variable. En la representacién gräfica (fig. 4) quedan separados por las dos

primeras FD los siguientes grupos: la mayoria de los individuos de la poblaciôn de Las Gamas

se sitüan en la parte positiva de las FD 1 y 2, y los de Colastiné en la negativa de la primera y

positiva de la segunda. Con respecto al dimorfismo sexual, a pesar de notarse cierto aleja-

miento entre los grupos, no se aprecia una clara separacién entre los sexos. La diferencia mâs

destacable seria en relaciôn a su longitud hocico-cloaca (hipôtesis no comprobada).

El anälisis de la dieta presenté una baja superposicién en los items alimentarios (valor

del indice de superposiciôn de Pianka = 0,55).

DISCUSION

Hasta el momento no se han descrito subespecies de Scinax nasicus; sin embargo, Scinax

es un género cuya taxonomia es compleja a causa de la importante variacién que presentan

sus especies (FAIVOvICH, 1997). Este autor encontré diferencias osteolégicas entre dos

poblaciones de Scinax fuscovarius y entre dos poblaciones de Scinax berthae. Un anälisis

cladistico del genero Scinax en las especies argentinas fue realizado por FAIVOvICH (1988),

estudios que ampliarän el numero de especies del género.

En las poblaciones investigadas, los resultados de los anälisis morfométricos realizados

no son evidencia suficiente para considerar que se trate de subespecies. Los efectos de la

temperatura en el tamaño del cuerpo de los animales ectotermos son de dificil interpretaciôn

(ATKINSON, 1996). La resoluciôn a la paradoja de “porque los organismos usualmente son de

mayor tamaño en ambientes mäs frios” estä fundamentalmente relacionada con el aumento

en el tamaño celular a bajas temperaturas (VAN VOORHIES, 1996; ATKINSON & SIBLY, 1997).

Este fenémeno explica el incremento en el tamaño del cuerpo de los ectotermos a bajas

temperaturas, independientemente de la ecologia especifica de las especies. Estudios sobre la

relaciôn entre la aridez y el tamaño corporal, en anuros, no han encontrado relaciôn entre las

variables (LEE, 1993).

Los ejemplares provenientes de Las Gamas presentaron una importante proporcién de

hormigas y mayor diversidad de coleépteros en su dieta. En general se puede observar en los

dos a un nümero de presas por estémago relativamente bajo en comparacién con otras

especies de anuros simpätricas estudiadas en la regiôn (ver LAIMANOVICH, 1995, 1996). Los

especimenes de Colastiné, con una menor amplitud de nicho, predaron preferentemente sobre

dipteros, coincidiendo con lo hallado por Duré & KEHR (1997) en la provincia de Corrientes,

donde los érdenes mejor representados fueron los dipteros, himenépteros y coleépteros. Es

menester aclarar que las caracteri: de los ambientes de la provincia de Corrientes

concuerdan con los de Colastiné. En coincidencia con DURÉ & KEHR (1997), se considera que

S. nasicus sigue una estrategia para capturar alimento intermedia entre forrajera y no

and-wait” (HUEY & PIANKA, 1981; Torr, 1981). Los especialistas son busca-

dores activos (por ejemplo, de hormigas), presentan gländulas venenosas y consumen muchas

Source : MNHN, Paris

PELTZER & LAJMANOVICH 93

Funciôn discriminante 2

-44 “24 0.4 16 36

Funcién discriminante 1

Fig. 4. - Resultado del anälisis discriminante: 1 y 2, machos y hembras de Las Gamas; 3 y 4, machos y

hembras de Colastiné.

Tabla 4. - Resultados del anälisis discriminante. Ordenacién de las variables segün los

dos primeros coeficientes de correlacién canénicos (C.C.). HC, longitud hocico-

cloaca. Proporciones respecto de la longitud hocico cloaca: AC, ancho cabeza;

DIO, distancia interocular; BAON, longitud desde el borde anterior del ojo y la

narina; LM, longitud de la mano, desde tubéreulo metacarpal externo al dedo

mäs largo; LF, longitud fémur; LT, longitud tibia; LP, longitud del pie, desde el

tubérculo metatarsal al dedo mäs largo.

AC/HC

DIO/HC

BAON/HC

LM/HC

LF/HC

LT/HC

LP/HC

Source : MNHN, Paris

94 ALYTES 16 (3-4)

pequeñas presas por dia; en contraposiciôn, los no especialistas son depredadores inmôviles

que esperan el paso de presas ocasionales (TorT, 1981). Los valores de amplitud tréfica

calculados se asemejan al hallado en otra especie generalista de la regiôn (Leptodactylus

ocellatus) (LAIMANOVICH, 1996).

RESUMEN

Se realizé un estudio comparativo de la dieta y de la morfometria de Scinax nasicus en

dos localidades de la provincia de Santa Fe (Argentina). Mediante un anälisis discriminante

se establecieron variaciones morfométricas en la especie. Ademäs, se cuantificé el espectro

tréfico, se calcularon la diversidad y la amplitud tréfica del nicho, el tamaño de presa y el

indice de importancia relativa. La comparacién de las dietas de ambas poblaciones se obtuvo

en base al indice de Pianka. Los resultados obtenidos muestran a S. nasicus con una estrategia

para capturar alimento intermedia entre forrajera (especialista) y no especialista “sit-and-

wait”.

AGRADECIMIENTOS

A Juliän FAIvovicH por la colaboraciôn en la colecta de los ejemplares de Vera y a Adolfo BELTZER

por la lectura critica del manuscrito. À Santiago RON y muy especialmente a Analia PUGENER por sus

sugerencias y aportes bibliogräficos.

LITERATURA CITADA

AN6NIMO, 1994. - STATGRAPHICS® Plus For Windows. Versiôn 1.11. Licensed to: INALI. Serial

ATkiNson, D. 1996. - Ectotherm life-history responses to developmental temperature. Ja: 1. A. JOHNS-

roN & À. F. BENNETT (ed.), Animals and temperature, New York, Cambridge University Press:

183-204

ATkIxsoN, D. & SiBLY, R. M, 1997. — Why are organisms usually bigger in colder environments? Making

sense of a life history puzzle. Trends in Ecology and Evolution, 12 (6): 235-239.

CasrerA, A. L., 1976. — Regiones fitogeogräfi inas. Enciclopedia Argentina de Agricultura y

Jardineria, 2 (1): 1-25.

CaLpweLz, J. P, 1996. — The evolution of myrmecophagy and correlates in poison frogs (family

Dendrobatidae). J. Zool., Lond, 240: 75-101

Ci, 1 M. 1980. - Amphibians of Argentina. Monit. zoo. it., (n.s.). Mon. 2: 1-609.

Cru, M. L., 1974. - Reproductive strategies in à tropical anuran community. Misc. Publ. Mus. nat

Hist. Univ. Kansas, 61: 1-68.

Duré, M. I. & KR, A. I, 1997. - Superposicién de nichos tréficos en dos especies sintépicas de hilidos

(Anura: Hylidae) en la provincia de Corrientes. Res. VI Jor. de Sc. Nat. del Lioral. 12

DUELLMAN, W. E., 1978. - The biology of an equatorial herpetofauna in Amazonian Ecuador. Misc.

Publ. Mus. nat. Hist. Univ. Kansas, 65: 1-352.

Source : MNHN, Paris

PELTZER & LAJMANOVICH 95

EMERSON, $. B., 1976. - Burrowing in frogs. / Morph., 149: 437-458.

dE 1986. — Heterochrony and frogs: the relationship of life history trait to morphological form. Am.

Nat., 127: 167-183.

Faivovicn, J., 1997, - Nuevas notas sobre la osteologia de las especies argentinas de Scinax (Anura:

Hylidae). Res. III Congr. Herp.: 33.

= 1988. - Nuevos caracteres para un anälisis del género Scinax (Anura, Hylidae). Res. XIII Reun. arg.

herp.: 15.

GaLLARDO, J. M., 1987. — Anfibios argentinos. Guia para su identificaciôn. Buenos Aires, Biblioteca

Mosaico: 1-98.

Garz, J, 1981. - Non-random mating by size in american toads, Bufo americanus. Anim. Behav., 29:

1004-1012.

Huy, R. B. & Pianka, E. R., 1981. - Ecological consequences of foraging mode. Ecology, 62 (4):

991-999.

Hurruta, J, 1973. — Trophic diversity measurement in sympatric predatory species. Ecology, 54:

885-890.

Jowrs, K. L., 1982. - Prey patterns and trophie niche overlap in four species of Caribbean frogs. Wild! Res.

Rep., 13: 49-55.

LAIMANOVICH, R. C., 1995. — Relaciones tréficas de bufénidos (Anura: Bufonidae) en ambientes del rio

Paranä, Argentina. Alytes, 13 (3): 87-103.

_ 1996. - Dinämica trôfica de juveniles de Leptodactylus ocellatus (Amphibia: Anura), en una isla del

Paranä, Argentina. Cuad. Herp., 10 (1-2): 11-23.

_— 1997. — Alimentaciôn de larvas de anuros en ambientes temporales del sistema del rio Paranä,

Argentina, Doñana, Acta Vertebrata, 24 (1-2): 191-202.

LANGONE, J. A. 1994. - Ranas y sapos del Uruguay. (Reconocimiento y aspectos biolôgicos). Publ. Museo

Damaso Antonio Larrañaga, Serie Divulgaciôn, 5: 1-123.

Lke, J,, 1993. - Geographic variation in size and shape of Neotropical frogs: a precipitation gradient

analysis. Occ. Pap. Mus. nat. Hist. Univ. Kansas, 163: 1-20.

LEVINS, R., 1968. — Evolution in changing environments. New Jersey, Princeton, Univ. Press: 1-120.

OrbANo, M., COLLINS, P. & LAIMANOVICH, R., 1999. - Suburban habitat uses by Scinax nasicus (Anura:

Hylidae) frog in Santa Fe, Argentina. Herp. Rev. en prensa.

Piaka, E. R., 1973. - The structure of lizard communities. Ann. Rev. Ecol. Syst., 4: 53-74.

PIELOU, E. C., 1966. - The measurement of diversity in different types of biological collections. J. theor.

Biol., 13: 131-144.

PinkAs, L., OLiPHANT, M. $. & IVERSON, Z. L., 1971. - Food habits of albacore bluefin, tuna and bonito

in California waters. Calif. Dep. Fish Game, Fish Bull, 152: 1-1105.

Saurur, S. N. & CrumP, M. L., 1977. - A Darwinian interpretation of hindlimb variability in frog

populations. Evolution, 31: 737-749.

SokaL, R. R. & ROLHF, F. J, 1979. - Biometria. Madrid, Blumé: 1-830.

Torr, C. A., 1981. - Feeding ecology of Panamanian litter anurans: patterns in diet and foraging mode.

J. Herp., 15 (2: 139-144.

Torr, C. À. & DUELLMAN, W. E., 1979. — Anurans of the lower rio LLullapichis, Amazonian Perü: a

preliminary analysis of community structure. Herpetologica, 35 (1): 71-77.

Van Voornirs, W. A., 1996. — Bergmann size clines: a simple explanation for their occurrence in

ectotherms. Evolution, 50 (3): 1259-1264.

Wizeur, H. M., RUBENSTEN, D. L. & FaIRCHILD, L., 1978. - Sexual selection in toads: the roles and male

body size. Evolution, 32: 264-270.

Corresponding editor: Janalee P. CALDWELL.

Source : MNHN, Paris

96 ALYTES 16 (3-4)

APÉNDICE I

MATERIAL ESTUDIADO

MFA-ZV-H: Colecciôn Museo Florentino Ameghino, Zoologia Vertebrados, Herpetologia, Santa Fe,

Argentina.

Scinax nasicus

Sitio Las Gamas, Dpto. Vera, Santa Fe (29°27'S, 60°230): MFA-ZV-H 507 (lote de 25 ejemplares).

Sitio Colastiné, Dpto. La Capital, Santa Fe (30°40'S, 60°30/0): MFA-ZV-H 508 (lote de 25 ejemplares).

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 97-110. 97

The life-history traits

of Eurvcea guttolineata

(Caudata, Plethodontidae),

with implications for life-history evolution

Jeremy L. MARSHALL !?

Department of Biolo,

P.O. Box 42451, L

E-ma

University of Southwestern Louisiana,

yette, Louisiana, 70504-2451, USA

marshal@usl.edu

Evaluating life-history traits allows for the assessment of local adapta-

tion and its correlated fitness consequences. The goal of this study was to

describe the life-history traits of a spring-dwelling population of Eurycea

guttolineata to gain a better understanding of life-history evolution in the

Plethodontidae. Size at first reproduction, z 50.00 mm SVL, was similar

between males and females and was attained at 22-24 months of age.

However, a larger variance in size of sexually mature females (about twice

male variance) may suggest that some females do not become sexually

mature until 34-36 months of age. The data suggest a period of sexual

activity from late summer to early winter (July-December), ovipositing

occurring in early winter (November-December), and egg hatching rebeble

occurring in January or February. During ontogeny, growth rates were high

during the first (2.48 mm SVL/mon) and second (1.70 mm SVL/mon) years,

but decreased (0.11 mm SVL/mon) once sexual maturity was reached. I

found that metamorphosis occurred typically in June, at a size of 23.08 mm

SVL, at 5-6 months of age. A coefficient of variation analysis revealed that

age at metamorphosis was significantly more variable than size. This, in

conjunction with the fast larval growth rates and short larval period of this

species, is consistent with a hypothesis based on larval adaptation to warm,

stable aquatic environments in which an optimal size at metamorphosis is

reached at an early age. This analysis does not support the hypothesis that

larvae of this species are adapted to uncertain environments.

INTRODUCTION

Evaluating life-history traits across the geographic distribution of a species is critical to

interpreting the influence of local environments on life-history variation (STEARNS, 1992;

TiLLEY & BERNARDO, 1993). Such variation in life-history traits may reflect phylogeny

1 we ork completed at Department of Biology, University of Mississippi, University, Mississippi 38677,

Us,

2 Dis paper is dedicated to the memory of Nick PIETROPAOLO.

Source : MNHN, Paris

98 ALYTES 16 (3-4)

CHARVEY & PAGEL, 1991) or may represent adaptation to local environments (LEVINS, 1968;

STEARNS, 1992). Life-history studies that address phylogenetic history and local adaptations

are now being conducted at the level of genus and species with comparative methods

(BAUWENS & Diaz-URIARTE, 1997; IRsCHICK & Losos, 1998). However, a fundamental

criterion for evaluating the evolution of life-history traits, with comparative methods, is that

such traits are known for each of the taxa or populations under consideration (HARVEY &

PAGEL, 1991).

Life-history traits of the genus Eurycea (Caudata, Plethodontidae) have been documen-

ted from across the eastern United States (E. longicauda, ANDERSON & MARTINO, 1966: E.

multiplicata, IRELAND, 1974; E. quadridigitata, SEMLITSCH & MCMILLAN, 1980; E. junaluska,

SEVER, 1983; E. wilderae, BRUCE, 1988; E. cirrigera, MARSHALL, 1997; E. lucifuga, CARLYLE et

al., 1998). Considerable intraspecific variation in life-history traits has been observed, espe-

cially within those species that inhabit a wide variety of habitats (TILLEY & BERNARDO, 1993;

Voss, 1993; MARSHALL, 1996, 1997; CARLYLE et al., 1998). Habitat differences are the impetus

for local adaptation and may lead to the evolution of novel life-history characteristics (e.g.,

BAHERT, 1996; MARSHALL, 1996). Therefore, the assessment of life-history traits among

closely related species or populations within different habitats and regions should illuminate

potential sources of life-history variation (BERVEN, 1982; TILLEY & BERNARDO, 1993).

The three-lined salamander, Eurycea guttolineata Holbrook, 1838, (formerly E. longi-

cauda guttolineata) was raised to specific status by CARLIN (1997). This species has a bi-phasic

life cycle (CONANT & COLLINS, 1991; DUELLMAN & TRUEB, 1994) and inhabits a wide variety

of seepage, spring, river swamp, and creek systems in the eastern United States (CONANT &

CoLLins, 1991). The life-history traits of Æ. guttolineata and E. longicauda have been studied

in a variety of geographic locations. The traits of larvae and just metamorphosed specimens

of E. guttolineata from a spring-fed marsh in North Carolina were described by BRUCE (1982),

while some of the developmental and reproductive characteristics from a flood plain pop-

ulation in Florida were described by GORDON (1953). A detailed life-history study of £.

longicauda inhabiting temporary ponds in New Jersey was conducted by ANDERSON &

MARTINO (1966). IRELAND (1974) described the life-history traits of Æ. L melanopleura from a

spring-fed pond in Arkansas.

The previously studied populations of E. guttolineata were located largely in ephemeral

habitats. I examined a population of E. guttolineata that inhabits an annually invariant,

stenothermic spring ecosystem in the coastal plain of northern Mississippi, USA. The goals

of my study were to describe the life-history traits of this spring-dwelling population of E.

guttolineata, compare the findings to the results from other populations, and evaluate the

influence of intra- and interspecific variation on the evolution of life-history characteristics

among members of the Eurycea longicauda complex.

MATERIALS AND METHODS

Thestudy site was Poplar Cove, an approximately 50 m° spring, located at The University

of Mississippi Biological Field Station in the North-Central Hills physiographic province of

Source : MNHN, Paris

MARSHALL 99

Lafayette County, Mississippi, USA. Year round, the spring was stenothermic (x + s = 16.9

+ 1.8°C), with dissolved oxygen levels ranging from 0.7 to 9.4 ppm (x + s = 7.28 + 1.79

ppm). Water depths ranged from 0.005 to 0.100 m (x + s = 0.033 + 0.02 m). The area

surrounding Poplar Cove Spring (PCS) was a mixed pine-hardwood forest comprised of

short-leaf pine (Pinus echinata), eastern red cedar (Juniperus virginiana), blackjack oak

(Quercus marilandica), southern red oak (Q. falcata), water oak (Q. nigra), white oak (Q. alba)

and sycamore (Platanus occidentalis). The immediate area of the spring had a canopy

dominated by tulip poplar (Liriodendron tulipifera), an understory of American beech (Fagus

grandifolia), red maple (Acer rubrum), giant cane ( Arundinaria gigantea), American holly

(Ilex opaca), and the herbaceous plants netted chain-fern (Woodwardia areolata) and lizard’s

tail (Saururus cernuus). At this site, E. guttolineata co-occurred with several other caudates,

including the southern two-lined salamander (E. cirrigera), the red-spotted newt (Norophthal-

mus viridescens), the Mississippi slimy salamander (Plethodon mississippi), the red salamander

(Pseudotriton ruber) and the lesser siren (Siren intermedia), although the latter two species

were rarely seen.

T'installed a 35 m long drift fence constructed of 0.61 m wide aluminum flashing. The

fence began at the point of emergence of the spring and lay approximately 4 m from the

spring’s margin. The drift fence bordered approximately three-fourths of the total margin but

did not impede water flow. The bottom of the fence was buried to a depth of 0.10 m. The fence

was supported at 3.0 m intervals with two 0.50 m lengths of 0.02 m diameter PVC piping

fastened by plastic electrical ties. I placed pitfall traps adjacent to, and on each side of, the

fence at approximately 3.0 m intervals, with single-ended funnel traps placed at the ends

because of soil saturation in those locations. Coverboards (0.62 x 0.19 x 0.025 m wooden

planks) were then placed in between pitfall traps at certain locations along the fence. The

pitfall traps were 944 ml plastic buckets (0.115 m in diameter) with a 0.025 m internal lip to

help prevent escape (sensu Dopp & Scorr, 1994).

Daily surveys of the drift fence were conducted from April 1995 to December 1996. As

this research was part of a larger life-history study on caudates, E. guttolineata measurements

were taken rarely during 1995. However, more thorough measurements were taken during

1996. In addition to the daily surveying along the drift fence at PCS, samples of aquatic and

terrestrial salamanders were collected in May, July, August and November 1996. The aquatic

samples were conducted with the aid of a dip net. The terrestrial samples were taken with the

aid of a potato rake for searching through ground litter. The time spent surveying the aquatic

(180 min) and terrestrial (60 min) habitats was relative to their total area (.e., the aquatic and

terrestrial habitats were 50 and 17 m°, respectively). This method was used to reduce the bias

of sampling any particular area unequally. AÏI survey data were used for determining activity,

sizes, ages and months of metamorphosis and sexual activity.

I determined mean body sizes (to the nearest 0.01 mm SVL) of larvae, juveniles and

adults on a monthly basis. Reproductive status of adults was determined by the presence of

yolked oocytes in females (seen through the venter) and secondary sexual characteristics, such

as nasal cirri and mental glands, in males (ARNOLD et al., 1993; DUELLMAN & TRUEB, 1994).

Icompared SVL of just metamorphosed and adult individuals within and between years with

the Mann-Whitney U and Kruskal-Wallis H tests (ZAR, 1984). Size classes of individuals were

established from the monthly data.

Source : MNHN, Paris

100 ALYTES 16 (3-4)

Based on the size class data from the monthly samples at PCS, ages were estimated and

then assigned for each individual. This was accomplished by utilizing the three size classes of

individuals (see fig. la and 2b, May-August) and assigning ages between 0 and 11 months for

the first, 12 and 23 months for the second, and 24 and 35 months for the third size class,

respectively. Larval hatching was assumed to occur in January based on the presence of a few

newly hatched larvae at PCS. Larvae found in January were assigned an age of zero month.

Although there may be some error in the estimates of older age classes (1.e., z 31 months of

age), this technique provides an adequate method for assigning respective ages of larval,

juvenile and subadult salamanders with non-overlapping size classes (BRUCE, 1988; STEARNS,

1992).

TLestimated growth rates by regressing month of capture versus size (SVL) for each size

class of individuals (ZAR, 1984). A general model of growth over the first 35 months of life

was estimated by regressing estimated age versus SVL. This approach allowed for the general

assessment of larval period, juvenile period, age and size at metamorphosis, and age and size

at sexual maturity. I then compared these life-history characteristics to those of other

populations of E. guttolineata and E. longicauda.

I utilized a Haldane coefficient of variation analysis for samples, ie, Va = (1 +

1/4n)(s/x), to determine differences in variation of age and size at metamorphosis among

populations of the £. longicauda complex. This analysis corrects for the bias of small sample

size and the effects of sampling (HALDANE, 1955; SOKAL & BRAUMANN, 1980; DELAUGERRE &

Dugois, 1985). A Fratio test was used to determine statistical significance between coefficient

of variation values (LEWONTIN, 1966). Assumptions of normality were met for all analyses.

Finally, the relationships between larval growth rate, size at metamorphosis, and age at

metamorphosis, were assessed intra- and interspecifically among populations of the E.

longicauda complex. Growth rates were estimated using BEACHY’s (19954) equations. Rela-

tionships were evaluated using regression and correlation coefficient analyses (ZAR, 1984).

Statistical significance was set at x = 0.05.

RESULTS

The population structure of the 1995 (n = 48) and 1996 (n = 61) samples of E.

guttolineata from PCS revealed two juvenile age classes prior to the first adult age class

(fig. 1b). Mean sizes for each age class and month are presented in tab. 1. A few newly hatched

larvae were collected in January 1995, although not measured. Larvae were present at PCS

through May (tab. 1). 1 found that metamorphosis occurred in June and July (tab. 1). This

corresponds to an age at metamorphosis of 5-6 months, assuming hatching occurred in

January. After metamorphosis, the juvenile period lasted 17-19 months.

The data on sexually mature individuals (tab. 1) indicated a late summer to late autumn

(July-October) period of sexual activity, with the smallest females becoming sexually active

during the latter part of the season (see tab. 1, October and December). The smallest female

at sexual maturity was 50.5 mm SVL. I estimated the age of this individual to be 23 months.

Therefore, age at first reproduction is reached at the end of the second year of life at 22-

Source : MNHN, Paris

MARSHALL 101

FREQUENCY

16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70

SVL (mm)

*] — #4 $—

[e)

SVL (mm)

MONTH

Fig. 1. (a) The frequency distribution of snout-vent lengths (SVL) of Eury cea guttolineata from the 1995

and 1996 pooled data from Poplar Cove Spring in Lafayette Co., Mississippi, USA. The three

designated size classes are 17-31 mm, 37-46 mm and 50-70 mm SVL. (b) The pooled population

structure of the 1995 (n = 48) and 1996 (n = 61) monthly samples of E. guttolineata. Growth rates

are based on the regression slope for first and second year juveniles, as well as the adult estimate

solid lines). The regression analyses for each growth period were as follows: first year growth, slope

= 2.48, r° = 0.84, df = 65, F = 326.79, P = 0.0001; second year growth, slope = 1.70, r?

” = 0.82, df =

Le 27.41, P = 0.0019; adult estimate, slope = 0.11, r? = 0.001, df = 37, F= 0.04, P = 0.8356.

Source : MNHN, Paris

102 ALYTES 16 (3-4)

metamorphosis (SVL = 23 mm)

0 D 5 1

0 5 10 15 20 25 30 35

MONTHS FROM HATCHING

Fig, 2. - A general growth model for E. guttolineata (n = 109), from the pooled data of 1995 and 1996.

This model incorporates timing of metamorphosis and sexual maturity. Individual salamanders

were assigned an age based on their size and month of capture. These age classed data were then

used to generate this growth model. À second order polynomial regression was used to generate

this model. The model is: size (SVL) = -0.023 age” + 2.121 age + 13.506, r° = 0.971.

24 months of age. However, the majority (85 %) of sexually mature individuals were probably

at least 30 months of age with a SVL > 55.00 mm. The grand mean, standard deviation,

range and coefficient of variation (V;,) for size (mm SVL) of sexually mature individuals are

as follows: males, n = 23, x + s = 56.99 + 2.12, range 52.60-60.80, V4, = 3.79; females, n = 14,

x + s= 58.29 + 5.35, range 50.50-69.80, V,, = 9.51. Females exhibited significantly more

variation in size than males (F,4,3 = 5.95, P < 0.001). Egg-laying probably occurred from

November to January based on the disappearance of females during late autumn and the

presence of new hatchlings in January (although I observed no egg masses).

There was no difference in size at metamorphosis between samples from June 1995 and

1996 (1995, n = s= 22.98 + 1.09, range 20.76-25.32; 1996, n = 3, x + s= 23.94 + 1.93,

range 21.89-2 Ti u= = 23, P = 0.2815). There was also no difference in grand mean female

SVL between 1995 and 1996 samples (1995, see tab. 1; 1996, 7 =9, x + 5 = 56.65 + 5.94, range

50.5-69.8; U = 8, P = 0.0532). Sexually active males in the 1996 sample exhibited no monthly

differences in mean SVL (H = 2.24, df =2, P = 0.327). The lack of significant differences

Source : MNHN, Paris

MARSHALL 103

Table 1. - SVL (in mm) for each age class of Eurycea guttolineata for each month of 1996. For

each sample, the table gives x + s, followed by n in parenthesis. Data are from Poplar

Cove Spring, Lafayette County, Mississippi, USA. Jralies, data from the 1995 sample

only. Bold, pooled data from 1995 and 1996.

October

August

Age class

Ist year juveniles

2nd year juveniles + 1.66 | + 4.05 +1.50 |

G) | @) @)

- - - - 56.12 | 56.97 | 58.05 | 54.74 |-

Adult males +1.79 | 42.09 | + 2.26 | + 0.00

(6)_| @) | @) | @)

- 55.00 |- 56.07 | 61.25 | 54.00 | 58.21 |- 50.50

+000 | + 2.33 | + 0.00 | + 6.67 + 0.00

@ | © | O | © @)

Adult females

between the 1995 and 1996 samples justified pooling these data for use in growth analyses

(SokaL & ROHLF, 1995).

The solid lines in fig. 1b represent growth rates for first and second year juveniles, and

adults. The first year rate of growth (2.48 mm SVL/mon) incorporates both larval and juvenile

growth (see fig. 1b and tab. 1 for data). The y-intercept of the first year growth equation

(11.21 mm) closely estimates published SVL data for hatchlings of E. gurtolineata from other

populations (GORDON, 1953; ANDERSON & MARTINO, 1966; BRUCE, 1970, 1982). Second year

growth (1.70 mm SVL/mon) decreased from the first year, while estimated adult growth was

minimal (0.11 mm SVL/mon). The combined rate of growth for juveniles, i.e., from metamor-

phosis to sexual maturity, was 1.49 mm SVL/mon. I generated a predictive growth model for

E. guttolineata, based on estimates of age and measures of size, that incorporates timing of

metamorphosis and maturity (fig. 2).

I summarized life-history data on size, age and timing of metamorphosis from 10

populations within the E. longicauda complex (tab. 2). Age at metamorphosis data from each

population incorporated potential variation from egg-laying dates. Using these data, I

calculated the Haldane coefficient of variation for both age and size at metamorphosis. The

mean data for size (tab. 2) were notdifferent from the raw data with respect to the mean or

variance (1 = 0.95, df = 127, P = 0.344; mean data, n = 10, V,ssize = 9.53; raw data, n = 119,

Vusize = 8.54; Fo 19 = 1.293, P > 0.50). I used data from the first year’s metamorphosing

Source : MNHN, Paris

104 ALYTES 16 (3-4)

Table 2. - Data on metamorphosis for populations of the Eurycea longicauda complex. E. g., E.

guttolineata; E. L L, E. L longicauda; E. L m., E. L melanopleura. SM, size at

metamorphosis (mm SVL). AM, age at metamorphosis (months). Rate, larval growth rate

(mm/month). MM, month(s) when metamorphosis occurs. Perm, permanent habitat type.

Ephl, ephemeral habitat type. Jialies, standard deviation from a larval sample with same

size range as just metamorphosed individuals. Bold, standard deviation estimated from

mean and range. Sources: (1) this study; (2) BRUCE, 1982 (Caney Fork); (3) BRUCE, 1970:

(3a) Cox Cove, (3b) Horse Cove; (4) GORDON, 1953; (5) SINCLAIR, 1951; (6) ANDERSON

& MARTINO, 1966; (7) FRANZ & HARRIS, 1965; (8) IRELAND, 1974; (9) RUDOLPH, 1978.

——

MS: Lafayette | 28 |23.08 + 1.19 (20.71-25.71)|5.50 (5.0-6.0) | 2.378 | Jun-Jul | Perm 1

NC: Jackson 9 |25.70 + 0.71 (25.00-27.00)|4.50 (4.0-5.0) | 3.044 |Jul-Aug| Perm 2

NC: Jackson 5 |24.40 + 0.89 (23.00-25.00)|3.75 (3.5-4.0) | 3.840! Jun | Ephl 3a

NC: Macon 9 |26.60 + 2.96 (23.00-32.00) | 4.50 (3.5-5.5) | 3.689| Aug ? 3b

FL: Jackson 1 [21.00 6.50 (6.0-7.0) | 1.692 | Jun-Jul | Ephl 4

TN: Haywood | 2 |23.25 + 1.06 (22.50-24.00) ? ? Jun ? 5

NI: Sussex 18 |22.50 + 1.15 (20.20-24.50)|3.50 (3.0-4.0) | 3.742 | Jun Ephl 6

MD: Garrett 15 |19.50 (18.00-21.00) LA ? Jul Perm 7

E. L.m. | AR: Washington| 24 |25.50 + 1.25 (23.00-28.00) | 6.00 (5.0-7.0) | 2.583 | Jun-Jul | Perm 8

E.l.m. |OK: Delaware | 25 |24.28 + 2.25 (19.00-29.00) | 6.50 (4.0-9.0) | 2.197 | Jul-Oct| Perm 9

Grand means,

136/23.58 + 2.19 (18.00-32.00) |5.09 (3.5-9.0) | 2.720 [Jun-Oct| - -

ranges, totals

populations only, as this was a more conservative measure of variation in age at metamor-

phosis. Including individuals that over-wintered, i.e., > 12 month larval period, increased the

coefficient of variation for age more than size. I found that age at metamorphosis had a

significantly greater coefficient of variation than size at metamorphosis within the complex

(Vaage = 24.19, V,ssize = 9.53, F3 10 = 5.81, P < 0.05). Moreover, this finding was consistent

when habitat type (i.e., populations occurring either in permanent or ephemeral habitats) was

included in the analysis (permanent, V,jage = 16.13, V,isize = 5.24, F,, = 10.09, P < 0.05;

ephemeral, Vage = 39.35, V,isize = 8.16, F, 3 = 18.87, P < 0.05). In contrast, populations of

E. quadridigitata, the dwarf salamander, which utilize ephemeral habitats, have significantly

greater variation in size than age at metamorphosis (V,jage = 7.37, Vijsize = 2241, F5 =

10.58, P < 0.05; data from: Bishop, 1947; HARRISON, 1973; SEMLITSCH, 1980; DUNDEE &

ROSSMAN, 1989).

Source : MNHN, Paris

MARSHALL 105

Finally, I analyzed the relationships between larval growth rate, size at metamorphosis,

and age at metamorphosis within and among species in the complex (fig. 3a-c). I found that

among populations there was not a significant relationship between larval growth rates and

size at metamorphosis (r = 0.46, P > 0.20; H,: b = 0, 1 = 1.28, P = 0.2489) and age at

metamorphosis and size at metamorphosis (r= 0.23, P > 0.50; H,: b = 0, 1= 0.57, P = 0.5869).

However, there was a significant relationship between larval growth rate and age at metamor-

phosis among populations (= 0.94, P <0.001; Ho: b=0,1= 6.74, P = 0.0005). When the data

were analyzed within species, only data from populations of £. guttolineata provided suffi-

cient sample sizes. Among populations of £. guttolineata, there was a significant correlation

between larval growth rates and size at metamorphosis (r = 0.85, P < 0.02; H,:b =0,1= 1.84,

P = 0.0701), larval growth rate and age at metamorphosis (r = 0.96, P < 0.001; H;:b=0,1=

6.23, P = 0.0084), and age at metamorphosis and size at metamorphosis (r = 0.80, P < 0.05;

H5:b=0,1= 2.33, P =0.1018). However, only the relationship between larval growth rate and

age at metamorphosis was significantly different from the null hypothesis b = 0 (see above).

The relationships between these traits for £. /. longicauda and E. I. melanopleura are shown in

fig. 3a-c.

DISCUSSION

Thelife-history traits of this population of E. guttolineata were similar to other taxa and

populations in the E. longicauda complex. The larval period of this population was compa-

rable to North Carolina and Florida populations of E. guttolineata and a population of E. !.

melanopleura in Arkansas, but longer than that of E. /. longicauda from New Jersey (tab. 2).

Metamorphosis also appeared to take place at a similar time regardless of the population

(tab. 2). This semi-consistent pattern of timing of metamorphosis may be a function of

phylogenetic history among these closely related populations, i.e., a relatedness constraint.

However, there was variation in age at metamorphosis among populations, which was

significantly more variable than size at metamorphosis. Therefore, variation in age at meta-

morphosis could result from plasticity in growth rates, as a function of the habitat, to reach an

optimal size at metamorphosis (WiLBur & COLLINS, 1973) and/or genetically based diffe-

rences in age at metamorphosis among populations (BERVEN, 1982).

Previous studies suggest that the short larval period of members of the Æ. longicauda

complex reflects an adaptation to uncertain/ephemeral aquatic environments (ANDERSON &

MaRTINO, 1966; BRUCE, 1982). To evaluate this hypothesis, some theoretical predictions

should be considered. WizBur & CoLLiNs (1973) stated that species that exploit

certain/permanent environments should have a narrow range of sizes at metamorphosis (.e.,

around an optimum) and a greater range in age at metamorphosis. This pattern should result

in increased variation in age at metamorphosis (e.g., from a few months to a year). In contrast,

those species which exploit uncertain/ephemeral habitats should exhibit the opposite trend

(Wizgur & CoLLINS, 1973). Moreover, if selection is favoring an optimal size at metamor-

phosis, then growth rates should only influence the time it takes to reach an optimal size.

BRUCE (1982) elaborated on WiLBur & COLLINS’ (1973) model by stating that in uncertain

environments slower growing larvae should metamorphose at a smaller size, as opposed to

delaying metamorphosis until the optimal size is reached. These theoretical predictions

Source : MNHN, Paris

106 ALYTES 16 (3-4)

SIZE AT METAMORPHOSIS

(mm SVL)

1 15 2 2.5 3 3.5 4

LARVAL GROWTH RATE (mm/month)

TAMORPHOSIS

(months)

AGE AT ME

1 15 2 2.5 3 3.5 4

LARVAL GROWTH RATE (mm/month)

SIZE AT METAMORPHOSIS

(mm SVL)

3 4 5 6 F4

AGE AT METAMORPHOSIS (months)

Fig.3. - Relationships between larval growth rate, age at metamorphosis, and size at metamorphosis for

each member of the £. longicauda complex. Open squares and dashed lines, £. L melanopleura;

solid circles and lines, E. guttolineata; open triangles, E. L. longicauda. (a) Relationship between

larval growth rate and size at metamorphosis for each species: £. guttolineata, size (SVL) = 2.075

rate + 18.078, r? = 0.718; E. L. melanopleura, size (SVL) = 3.161 rate + 17.336. (b) Relationship

between larval growth rate and age at metamorphosis: £. gurtolineata, age (months) =-1.139 rate

+8.256, 1° = 0.928; E. l. melanopleura, age (months) = -1.29$ rate + 9,346. (c) Relationship between

age and size at metamorphosis: £. guttolineata, size (SVL) = -1.664 age + 32.393, r° = 0.645 E. L.

melanopleura, (SVL) = -2.440 age + 40.140.

Source : MNHN, Paris

MARSHALL 107

provide the basis for my evaluation of the hypothesis of an adaptation to uncertain environ-

ments for this complex.

The findings of this study, that age at metamorphosis is significantly more variable than

size at metamorphosis, do not support the hypothesis of adaptation to uncertain environ-

ments. Instead, the data support the alternative prediction of WiLBur & COLLINS’ (1973)

model, which states that in stable environments individuals should remain in the aquatic

environment until an optimal size at metamorphosis is reached. The significant relationship

(e., correlation coefficient and b) between larval growth rate and age at metamorphosis, but

not larval growth rate and size at metamorphosis (i.e., b = 0), supports the latter prediction.

Both within and among species in this complex, the relationship between age at metamor-

phosis and size at metamorphosis was not significantly different from the null hypothesis b =

0. In addition, there is corroborating evidence that several populations within the £. longi-

cauda complex have fast growing larvae that metamorphose within months of hatching and

slow growing larvae that metamorphose more than 12 months after hatching (FRANZ, 1967;

RUDOLF, 1978; BRUCE, 1982). Moreover, populations of E. quadridigitata that inhabit ephe-

meral habitats exhibited the opposite trend (.e., significantly greater variation in size than age

at metamorphosis). Therefore, populations of the £Æ. longicauda complex meet the predictions

of Wizgur & CoLLINs’s (1973) model and support, at least in part, the hypothesis of selection

for an optimal size at metamorphosis.

Although the data do not support the hypothesis of an adaptation to uncertain environ-

ments, the hypothesis of selection for an optimal size at metamorphosis does not address

directly why members of this complex have shorter larval periods and smaller sizes at

metamorphosis relative to other semi-aquatic plethodontids. One evolutionary explanation is

that larvae are adapted to stable, warmer aquatic environments with increased food regimes

(e.g., food availability), resulting in increased growth rates and smaller sizes at metamorphosis

(BEACHY, 1995b).

Several studies have shown that increases in temperature and food result in increased

larval growth rates (WiLDER, 1924; STEWART, 1956; BIZER, 1978; SEXTON & B1ZER, 1978;

CHY, 1995b). However, a conflict, over the influence that increased temperature has on size

at metamorphosis, has arisen between alternative models of metamorphosis. SEXTON & B1ZER

(1978) stated that increases in temperature should result in shorter larval periods and smaller

sizes at metamorphosis. However, JUTERBOCK (1990) stated that temperature influences on

growth are not consistent among plethodontids (e.g., that sometimes decreases in temperature

result in smaller sizes at metamorphosis). BEACHY (1995b) stated that the discrepancies could

be accounted for by the complex relationship between increased temperatures and food

regimes (i.e., that increases in temperature are accompanied by increases in food regimes).

This complex temperature-food interaction can allow for increased larval growth rates,

shorter larval periods, and a range of sizes at metamorphosis. This reconciles the question of

how an optimal size at metamorphosis, facilitated by a stable environment, can be accompa-

nied by a shorter larval period. À warmer, more stable aquatic environment would allow an

Optimal size at metamorphosis to be reached at an earlier age through an increased growth

rate. Therefore, the data support the notion that habitat parameters (such as temperature and

food) directly influence larval growth rates, which then influence the age at which an optimal

size at metamorphosis is reached.

Source : MNHN, Paris

108 ALYTES 16 (3-4)

The majority of plethodontid life-history theory has centered on the genus Desmogna-

thus (for a review, see TILLEY & BERNARDO, 1993). However, the dominant theory for the

desmognathines, that increased adult body sizes are due to increased ages at maturation, does

not hold for salamanders in the genus Eurycea. Eurycea guttolineata and its close relatives are

at least 20 mm SVL larger (BRUCE, 1982; CONANT & COLLINS, 1991; this study) and become

sexually mature sooner than or at the same age as other salamanders in the genus (i.e., £.

bislineata complex). This suggests that age at maturity could not account for the differences in

adult body size. Moreover, it appears that juvenile growth rate, juvenile period, and/or size at

maturation, account for the differences in adult body size within this genus (MARSHALL,

unpublished data). Although different taxa in the family Plethodontidae appear to be

utilizing different strategies to attain larger body sizes, the influence of aquatic habitats on

larval development may be consistent among genera (i.e., increases in temperature result in

increase in larval growth rates). Moreover, this analysis provides evidence that intra- and

interspecific variation in life-history traits is influenced by local environments, which play a

critical role in shaping life-history evolution.

RESUMEN

La evaluaciôn de caracteristicas de la historia de vida nos permiten estimar la adaptaciôn

local y sus consecuencias correlacionadas de ajuste. El objetivo de este estudio fue describir

las caracteristicas de la historia de vida de una poblaciôn de manantial, Eurycea guttolineata

(Plethodontidae), para obtener un mejor entendimiento en la evoluciôn de la historia de vida

de Plethodontidae. Se encontré que la metamorfosis tipicamente ocurre en junio, con un

tamaño de 23.08 mm SVL, a una edad de 4-6 meses. El tamaño en la primera reproduccién,

2 50.00 mm SVL, fue similar entre machos y hembras a una edad de 22-24 meses. Sin

embargo, un gran variabilidad en tamaño en hembras sexulamente maduras (2 veces la

variabilidad en machos) sugiere que algunas no Ileguen a su madurez sexual hasta los 34-36

meses de edad. Los datos sugieren un periodo de actividad sexual del final del verano hasta el

comienzo del invierno (julio a diciembre), con deposicién de huevos al comienzo del invierno

(noviembre-diciembre), y su eclosién en enero o febrero. Tasas de crecimiento fueron altas

durante el primer (2.48 mm SVL/mes) y segundo (1.70 mm SVL/mes) años de vida, mientras

que decrecieron (0.31 mm SVL/mes) una vez alacanzada la madurez sexual.

ACKNOWLEDGMENTS

Lam grateful to E. D. Keiser, Jr. for his support and comments during this research. This

manuscript was improved greatly by the long discussions with C. Camp and E. D. KEISER, Jr, on

plethodontid life-history evolution. 1 thank C. Cap, P. CHIPPINDALE, R. JARGER, S. MARSHALL,

V. TownsEND, Jr, and an anonymous reviewer, for critically reviewing this manuscrit. This work was

supported by the Mississippi Wildlife Heritage Program and an University Doctoral Fellowship from The

University of Southwestern Louisiana. 1 am also indebted to S. MARSHALL and M. MARSHALL for their

support and encouragement

Source : MNHN, Paris

MARSHALL 109

LITERATURE CITED

ANDERSON, J. D. & MaRTINO, P. J., 1966. - The life history of Eurycea L. longicauda associated with ponds.

Am. Midl. Natur., 75 (2): 257-279.

ARNOLD, S. I, RFAGAN, N. L. & VERRELL, P. A., 1993. — Reproductive isolation and speciation in

plethodontid salamanders. Herpetologica, 49 (2): 216-228.

BaHErT, R., 1996. - Ecology of lake dwelling Eurycea bislineata in the Shawangunk Mountains, New

York. J Herp., 30 (3): 399-401.

BAUWENS, D. & Diaz-URIARTE, R., 1997. — Covariation of life history traits in lacertid lizards: a

comparative study. Am. Natur., 149 (1): 91-111.

Bracuy, C. K., 1995a. — Age at maturation, body size, and life-history evolution in the salamander family

Plethodontidae. Herp. Rev., 26 (4): 179-181.

— 1995b. — Effects of larval growth history on metamorphosis in a stream-dwelling salamander

(Desmognathus ochrophaeus). J. Herp., 29 (3): 375-382.

BERVEN, K. A., 1982. - The genetic basis of attitudinal variation in the wood frog, Rana sylvatica. 1. An

experimental analysis of life history traits. Evolution, 36 (5): 962-983.

Biskor, $. C., 1947. - Handbook of salamanders. Ithaca, Comstock Publ. Co. i-xiv + 1-555.

BizR, JR. 1978. - Growth rates and size at metamorphosis of high elevation populations of Ambystoma

tigrinum. Oecologia, 34 (2): 175-184.

BRUCE, R. C., 1970. — The larval life of the three-lined salamander, Eurycea longicauda guttolineata.

Copeia, 1970 (4): 776-779.

— 1982. — Larval period and metamorphosis in two species of salamanders of the genus Eurycea.

Copeia, 1982 (1): 117-127.

= 1988. — An ecological life table for the salamander Eurycea wilderae. Copeia, 1988 (1): 15-26.

Cain, J L., 1997. - Genetic and morphological differentiation between Eurycea longicauda longicauda

and £. guttolineata (Caudata: Plethodontidae). Herpetologica, 53 (2): 206-217.

CarLvLe, JC. SaNDERS, D. E. & PLUMMER, M. V., 1998. — Eurycea lucifuga (Cave Salamander).

Reproduction. Herp. Rev., 29 (1): 37-38.

ConaNT, R. & CoLLINS, J. T., 1991. — Repriles and amphibians of eastern and central North America. 3rd

edition. Boston, Houghton Mifflin Co. i-xx + 1-450.

DELAUGERRE, M. & Dunois, A., 1985. - La variation géographique et la variabilité intrapopulationnelle

chez Phyllodactylus europaeus (Reptilia, Sauria, Gekkonidae). Bull. Mus. natn. Hist. nat., (4), 7 (A)

(3): 709-736.

Doop, C. K., Jr. & Scorr, D. E., 1994. - Drift fences encircling breeding sites. Jn: W. R. HEvER, M. A.

DONNELLY, R. W. MCDiARMI, L. A. HAYEK & M. S. FOSTER (ed.), Measuring and monitoring

biological diversity: standard methods for amphibians, Washington, Smithsonian Institution Press:

125-130.

DuLLMAN, W. E. & TRUEB, L., 1994. - Biology of amphibians. 2nd edition. Baltimore & London, The

John Hopkins Univ. Press: i-xix + 1-670.

Dunber, H. A. & RossmaN, D. A., 1989. - The amphibians and reptiles of Louisiana. Baton Rouge,

Louisiana State University Press: i-xi + 1-300.

Franz, R., Jr, 1967. — Notes on the long-tailed salamander, Eurycea longicauda longicauda (Green), in

Maryland. Bull. Maryland herp. Soc., 3 (1): 1-6

FRANZ, R. Jr. & HARRIS, H., 1965. - Mass transformation and movement of larval long-tailed salaman-

ders, Eurycea longicauda longicauda (Green). JL. Ohio herp. Soc., $ (1): 32.

Gorpow, R. E., 1953. — A population of Holbrook's salamander, Eurycea longicauda guttolineara

(Holbrook). Tulane Stud. Zool., 1 (4): 55-60.

HALDANE, J. B. $., 1955. - The measurement of variation. Evolution, 9 (3): 484.

HaRRISON, J. R., 1973. — Observations on the life history and ecology of Eurycea quadridigitata

(Holbrook). Herp. Inf. Search Syst. News J., 1: 57-58.

HaRvEY, PH. & PaGez, M. D., 1991. - The comparative method in evolutionary biology. Oxford, Oxford

Univ. Press: i-viii + 1-239.

Source : MNHN, Paris

110 ALYTES 16 (3-4)

IRELAND, P. H., 1974. - Reproduction and larval development of the dark-sided salamander, Eurycea

longicauda melanopleura (Green). Herpetologica, 30 (4): 338-343.

IrscHick, D. J. & Losos, J. B., 1998. - À comparative analysis of the ecological significance of maximal

locomotor performance in Caribbean Anolis lizards. Evolution, 52 (1): 219-226.

JurerocK, JE, 1990.- Variation in larval growth and metamorphosis in the salamander Desmognathus

fuscus. Herpetologica, 46 (3): 291-303.

LEvINs, R., 1968. - Evolution in changing environments. New Jersey, Princeton Univ. Press: 1-120.

LEWoNTiN, R. C., 1966. — On the measure of relative variability. Syst. Zool., 15 (2): 141-142.

MARSHALL, J. L., 1996. - Eurycea cirrigera (Southern Two-lined Salamander). Nest Site. Herp. Rev., 27

(2): 75-76.

997. The life history of the southern two-lined salamander, Eurycea cirrigera (Caudata: Plethodon-

tidae), with interspecific comparisons to the members of the Eurycea bislineata complex. Masters

Thesis, University, The University of Mississippi: 1-94.

Ruporrs, D. C. 1978. Aspects of the larval ecology of five plethodontid salamanders of the western

Ozarks. Am. Midl. Natur., 100 (1): 141-159.

SeuLrrscH, R. D., 1980. - Growth and metamorphosis of larval dwarf salamanders (Eurycea quadridi-

gitata). Herpetologica, 32 (2): 138-140.

SemLiTscH, R. D. & MCMILLAN, M. À. 1980. — Breeding migrations, population size structure, and

reproduction of the dwarf salamander, Eurycea quadridigitata, in South Carolina. Brimleyana, 7

(3): 97-105.

Sever, D. M. 1983. - Observations on the distribution and reproduction of the salamander Eurycea

“junaluska in Tennessee. J. Tennessee Acad. Sei., 58 (3): 48-50.

Sexron, O. J. & B1ER, J. R., 1978. — Life history patterns of Ambystoma tigrinum in montane Colorado.

Am, Midl. Nat., 99 (1): 101-118.

SincLAIR, R. M., 1951. - Notes on recently transformed larvae of the salamander Eurycea longicauda

guttolineata. Herpetologica, 7 (2): 68.

SokaL, R. R. & BRAUMANN, C. A. 1980. - Significance test for coefficient of variation and variability

profiles. Syst. Zool., 29 (1): 50-66.

SokaL, R. R. & RonLr, EI, 1995.- Biometry. New York, W. H. Freeman.

STEARNS, S. C., 1992. - The evolution of life histories. Oxford, Oxford Univ. Press: i-xii + 1-249.

STEWART, M. M., 1956. - The separate effects of food and temperature differences on the development of

marbled salamander larvae. / Elisha Mitchell Sci. Soc., 72 (1): 47-56.

Tue, S. G. & BERNARDO, J,, 1993. — Life history evolution in plethodontid salamanders. Herpetologica,

49 (2): 154-163.

Voss, $. R., 1993. - Relationship between stream order and length of larval period in the salamander

Eurycea wilderae. Copeia, 1993 (3): 736-742.

Wizeur, H. M. & COLLINS, J. P, 1973. - Ecological aspects of amphibian metamorphosis. Science, 182:

1305-1314.

WiLoer, L W, 1924. - The relations of growth to metamorphosis in Eurycea bislineata (Green). J. exp.

Zool., 40 (1): 1-112.

ZaR, JH., 1984. - Biostatistical analysis. 2nd edition. New Jersey, Prentice Hall, Englewood Cliff: i-xiv

+717

Corresponding editor: Janalee P. CALDWELL.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 111-122. 111

Adaptation aux particularités climatiques

du cycle biologique d’un anoure tropical,

Nectophrynoides occidentalis

Angel, 1943 (Bufonidae)

Maxime LAMOTTE & Coralia SANCHEZ-LAMOTTE

Laboratoire des Reptiles & Amphibiens, Muséum national d'Histoire naturelle,

25 rue Cuvier, 75005 Paris, France

In the distribution area of the viviparous toad Nectophrynoides occi-

dentalis Angel, 1943, restricted to a few square kilometers of a low grass

savanna above 1200 m on the crests of the Mount Nimba, a very dry season

of about 5 months alternates with a rainy season of 7 months. The life and

breeding cycles of this species are closely linked with this seasonal cycle.

During the dry season, the toads burrow underground and become dor-

mant. They emerge as soon as the rains start, between February and the end

of March, in the following order : first the gravid females, then the virgin

females and finally the males. All births of young take place in June.

Fertilization takes place from September for females older than one year, to

the end of October for females of that year. Al females burrow immediately

after fertilization. The cycle does not seem to be modified by the amount of

water available in the year, which may vary by twice as much according to

the place in the chain or to the year. However, monitoring of the climatic

cycle and of toad populations over several years have shown that the dates

of burrowing and of dormancy are closely linked to the beginning and above

all to the end of the rainy season, that may vary more than one month from

year to year. These variations result in important differences in the propor-

tion of young females that are virgin before their first burrowing for the dry

season. They have therefore consequences for the reproduction rate of the

population.

Nectophrynoides occidentalis Angel, 1943 est un petit amphibien anoure de la famille des

Bufonidae dont la longueur museau-anus dépasse rarement 24 mm chez les mâles et 27 mm

chez les femelles (fig.1). La coloration est d’un brun ocre chez les mâles, nettement plus claire

chez les femelles (ANGEL, 1943; ANGEL & LAMOTTE, 1944, 1948).

L'espèce N. occidentalis ne vit que sur les quelques kilomètres carrés de la prairie

d'altitude (savane à herbes basses) couvrant les crêtes de la chaîne du Nimba dans sa partie

située en Guinée et Côte d'Ivoire près de la frontière du Libéria (fig. 2). Présente jusqu’au

sommet à 1750 m, elle ne descend pratiquement pas au-dessous de 1200 m d’altitude. Cette

localisation très stricte est liée à deux caractéristiques très particulières du milieu.

La première est un relief abrupt (LAMOTTE & ROUGERIE, 195$). Celui-ci exclut presque

totalement la présence de mares permettant la vie de têtards et élimine ainsi la concurrence de

la presque totalité des autres amphibiens. Nectophrynoides occidentalis, en revanche, a pu

Source : MNHN, Paris

112 ALYTES 16 (3-4)

Fig. 1. - Aspect d’une femelle de Nectophrynoides occidentalis mettant bas un nouveau-né. Dessin de

Y. ScHacH-Duc d’après une photographie de F. XAVIER.

répondre à ce défi par un développement direct dans les oviductes conduisant à la naissance

de jeunes entièrement métamorphosés, longs de moins de 8 mm. La figure 3 représente les

principaux stades de ce développement, qui dure près de 9 mois (ANGEL & LAMOTTE, 1944,

1948; LAMOTTE, 1959; LAMOTTE & XAVIER, 1972).

En second lieu, les conditions climatiques font alterner une saison des pluies très

favorable durant laquelle règne en permanence une forte humidité de l'atmosphère et une

saison sèche particulièrement rigoureuse où le degré hygrométrique s’abaisse souvent au

dessous de 30 % (RICHARD-MOLARD et al., 1955) (fig. 4). L'espèce répond à ce contraste

climatique accentué par un cycle biologique déterminé lui-même avec rigueur.

LE CYCLE SAISONNIER MOYEN DES POPULATIONS

Durant la saison pluvieuse, l'humidité persistante du milieu liée aux précipitations, aux

bruines et aux brouillards lui permet de maintenir son activité de façon ininterrompue, tandis

que pendant la saison sèche aucun amphibien ne peut survivre autrement qu’enfoui dans le

sol, ce que fait effectivement Necrophrynoides en mettant à profit des fissures de la roche

sous-jacente.

Des prélèvements quantitatifs effectués sur des surfaces de 25 m° en divers sites de la

chaîne et au cours des mois successifs de plusieurs années ont permis de suivre les variations

de la densité et de la composition des populations. Ils étaient complétés par une étude de la

Source : MNHN, Paris

LAMOTTE & SANCHEZ-LAMOTTE 113

Fig. 2.- Situation et relief de la chaîne du Nimba. On a figuré les courbes de niveau 600, 1000 et 1400 m;

la zone située-au dessus de 1000 m est représentée avec un figuré de tirets horizontaux Les nombres

indiquent les emplacements où ont été réalisées des collectes du peuplement animal.

taille des individus (liée de façon directe à leur âge) et par la détermination de l'étape de la vie

sexuelle des femelles (vierges, gravides, après l'accouchement) qu’indique l’état des oviductes

et des ovaires.

Une caractéristique essentielle du cycle biologique est le fait que toutes les naissances se

produisent durant le mois de juin, en pleine période de vie active. Encore nettement distincte

par sa taille plus petite (de 7 à 13 mm), une nouvelle cohorte vient alors se joindre aux deux

plus anciennes (fig. 5). Ces femelles adultes, alors âgées de 12-15 mois à 2 ou 3 ans, renferment

des individus encore vierges et d’autres qui viennent d’accoucher ; leur taille est de 17 à 28 mm

tandis que les mâles des mêmes cohortes mesurent de 14 à 21-22 mm. La population renferme

alors un nombre sensiblement égal de mâles et de femelles, et cette égalité plus ou moins

Source : MNHN, Paris

114 ALYTES 16 (3-4)

Fig. 3.- Quelques stades du développement embryonnaire de M. occidentalis (d’après LAMOTTE & XAVIER,

1972). Dessins de Y. ScHACH-Duc.

Source : MNHN, Paris

LAMOTTE & SANCHEZ-LAMOTTE 115

Lundi 16 Margt 17

; Li pires 16 [y ;ieuer 19, ; J Nendreas 20 Saseai 21 | JyDiansene 22 février

Lion va

MYGROUETRIE Mu 16 ou “5 Cévrier 2

ë Ÿ x FETE

II 1 Perret 0 pp Peu D ji Vendredi lo, ,,, Mmtél LU, inanghe Li aveils

F 1 Ï (

= si 1 nt :

HEROMETRIE \ (| du 6 au ds nsril

te à AE Ÿ

Lundi 3 Haras à ; ENT fig pimanene v stat y

11:

= f 1 Î

= ; 1 LECELL

F ITET

routiers Eh rermnnent

fl ô TT

T Hi IT

1 ( T d 130 TT

Î AUOT

RAA PH mt (ET

LS a)

1 NAT I ET RUN EREE RER ds

111 TT fi

Ê f 11 [II

HULL l j

fl Run! JYirri | F

anti THOIL TOOL

Î Î Î NTI

EN ï DEMAIN +

i MYGROMETRIE du 15 Toi

Fig. 4. — Les variations journalières du degré hygrométrique de l'air dans la prairie d'altitude du Nimba

à 1600 m. Le degré hygrométrique apporte l'indication la plus adéquate sur les conditions plus où

moins favorables du milieu pour un amphibien. De haut en bas: du 16 au 22 février (saison sèche), du

6 au 12 avril (première saison des tornades), du 3 au 9 août (pleine saison des pluies) et du 15 au

21 octobre (seconde saison des tornades).

Source : MNHN, Paris

116 ALYTES 16 (3-4)

début de juin

8 10 12 4 16 18 20 22 24 26 28 8 10 12 14 16 18 20 22

millimètres millimètres

femelle juvénile

femelle vierge

femelle gravide

femelle après

accouchement

8 10 #2 14 6 8 10 12 14 16 18 20 22

Fig. 5. — Composition de la population (femelles et mâles) au début du mois de septembre. Après la

période des naissances (en juin), les mois de juillet, août et septembre voient une croissance active de

tous les individus. La cohorte des jeunes de l’année, alors âgés de 1 à 4 mois, se distingue par sa taille

nettement plus petite (entre 7 et 14 mm). Il apparaît en outre chez les femelles plus vieilles une

coexistence de deux cohortes (respectivement âgées d'environ 16 et 28 mois).

Source : MNHN, Paris

LAMOTTE & SANCHEZ-LAMOTTE 117

Œ& femelle après accouchement M femelle vierge

femelle à œufs utérins

Q [3] (premières heures de la FA femelle gravide

A gravidité)

oct. avril

8 10 12 14 16 18 28 B 10 12 14 16 18 20 22 24 26 28

millimètres

juv.

} V

novembre mai Vi

8 10 12 14 16 18 20 22 24 265 28 8 10 12 14 16 18 20 24 265 28

Fig. 6. - Évolution apparente de la population des femelles durant la période d'enfouissement (à gauche)

et durant la période d’émergence (à droite).

complète des effectifs des deux sexes persiste durant toute la saison des pluies jusqu’en

septembre, tant chez les juvéniles de l’année que chez les individus plus âgés (voir fig. 5).

En fin septembre, avant la fin des pluies, commence la période d'enfouissement qui va

permettre à l’espèce de résister à la saison sèche ; cet enfouissement s'étale sur plusieurs

semaines. Dès le mois de septembre, période pourtant encore très pluvieuse, les femelles de

grande taille, dont c’est la seconde ou la troisième gravidité, s’enfouissent dès qu'elles sont

fécondées. Les mâles, au contraire, et surtout les plus jeunes, attendront la fin de la saison des

pluies, qui survient en général durant le mois d’octobre. C’est le cas aussi des jeunes femelles

vierges, nées 4 mois auparavant : elles quittent la vie active épigée au fur et à mesure de leur

fécondation, elle-même liée à leur degré de développement (fig. 6, à gauche). Les jeunes

femelles qui, lorsque cessent les pluies, n’ont pas atteint une maturité suffisante, s’enfouissent

encore vierges. Il résulte de ces décalages qu’en octobre et novembre la population de N.

occidentalis ne comprend plus que des mâles et des femelles vierges nés dans l’année.

La fin de la période de vie enfouie coïncide avec l’arrivée des pluies qui se produit

généralement en fin mars, plus exceptionnellement en avril ou en février. Les divers individus

Source : MNHN, Paris

118 ALYTES 16 (3-4)

de la population ne sortent toutefois pas tous en même temps, mais avec un décalage qui

s'étale sur près d’un mois. Les femelles gravides émergent en premier, puis les mâles et les

femelles vierges nées 9 mois auparavant (fig. 6, à droite). Au tout début de la saison des pluies,

la population active ne comprend que des femelles gravides (ANGEL & LAMOTTE, 1944 ;

LAMOTTE, 1959) ; elles ne sont rejointes qu’ensuite par des femelles vierges et les mâles.

LES CONSÉQUENCES DES VARIATIONS INTERANNUELLES DU CLIMAT

SUR LE CYCLE DES POPULATIONS

La présentation des traits généraux du cycle des populations en a fait apparaître la liaison

étroite avec les variations saisonnières de la pluviosité. Cette dépendance très stricte de la vie

de N. occidentalis vis-à-vis des facteurs climatiques donne à penser que toute variation de ces

facteurs se traduira sur la biologie de l’espèce et notamment sur son cycle de reproduction. Or

de telles variations du climat se produisent inévitablement au cours des années successives et

des différences existent aussi dans l’espace entre les divers sites de la chaîne où l'espèce est

présente.

La hauteur totale des précipitations annuelles est sans doute un facteur important de la

localisation de l'espèce puisque celle-ci est absente dans la partie septentrionale de la chaîne

où les pluies sont inférieures à 1500 mm. Elle est aussi très variable au sein de l’aire de

répartition puisqu'il tombe plus de 3000 mm d’eau au sud du mont Richard-Molard et

seulement 2000 mm dans la région septentrionale du Signal Sempéré et du mont Tô. Les

différences interannuelles de la pluviosité en un même site de la chaîne sont également très

fortes : à la station météorologique de Ziéla, la pluviosité annuelle a varié entre 1099 mm et

1757 mm durant les années 1949 à 1957. Il est toutefois difficile de détecter une influence de

cette hauteur annuelle des pluies sur la fécondité de l’espèce qui reste apparemment semblable

d’un bout à l’autre de son aire de répartition. Elle est masquée en effet par les variations

considérables liées à la taille de la femelle. De fait, les jeunes femelles de moins de

21 mm de longueur museau-anus -— fécondées à l’âge de 4 mois — ne portent généralement que

de 2 à 8 embryons, tandis que les femelles plus âgées, dont la taille dépasse 22 mm, en ont

généralement plus de 10 (fig. 7). Cette relation entre le nombre d’embryons et la taille de la

mère se retrouve dans tous les sites de la montagne et toutes les années.

Si la pluviosité annuelle ne semble pas être un facteur majeur du cycle biologique, tout

autre est le rôle du calendrier des pluies.

Au cours d’une même année, les dates d’émergence et d'enfouissement sont, comme

celles de l’arrivée et de la fin des pluies, sensiblement les mêmes dans toute l’aire de répartition

de l’espèce, depuis le Signal Sempéré jusqu’au sud du mont Richard-Molard. Au contraire,

ces dates du commencement et de la fin de la période pluvieuse sont très variables d’une année

à l’autre et elles déterminent toujours avec rigueur celles de l'émergence et de l’enfouissement

des Nectophrynoides. I] est ainsi des années où les pluies précoces provoquent une sortie des

crapauds dès la fin de février et d’autres où les pluies, et avec elles l'émergence, n'arrivent que

fin avril ou début mai. Inversement, la fin de la période des pluies et donc celle de la vie active

des derniers individus — jeunes femelles non fécondées et mâles parmi lesquels dominent des

jeunes de l’année — peuvent se produire dès le début du mois d'octobre ou au contraire au

début novembre, voire en décembre.

Source : MNHN, Paris

LAMOTTE & SANCHEZ-LAMOTTE 119

nombre d'embryons

22 .

18 x

16 e.

14 x ye

Ca * *

12 ee e. *

* x # *

10 * #4 *

8 PE *# #

x e x +

6 DC e

ea nt e +

4 { e

._. + *x

2 * *x

15 16 17 18 19 20 21 22 23 24 25 26 27

taille des femelles

Fig. 7. - Nombre d’embryons en fonction de la taille de la mère (longueur museau-anus). Les points et les

étoiles correspondent respectivement aux années 1981 et 1991.

Une analyse plus précise de la structure démographique des populations poursuivie

durant plusieurs années conjointement avec des enregistrements pluviométriques mensuels a

permis de pousser plus loin l’étude de l'influence du cycle saisonnier des pluies (LAMOTTE,

1959). Elle a fait apparaître une corrélation nette entre la pluviosité des mois d'octobre et

novembre et la proportion dans la population de jeunes femelles restées vierges parce

qu'immatures lors de l’enfouissement à l’arrivée de la saison sèche (fig. 8). Une venue précoce

de la saison sèche, dès le début octobre, diminue ainsi la participation de la cohorte de jeunes

femelles de l’année au renouvellement de la population, tandis que le prolongement de la

saison des pluies permet le développement jusqu’à leur maturité de la majorité de ces

individus.

Les femelles plus âgées, elles, sont toutes fécondées dès le mois de septembre et fournis-

sent donc toutes, quelle que soit la date de la fin des pluies, le même contingent d'embryons.

Durant les années à saison sèche précoce, la contribution à la natalité de la cohorte des jeunes

de l’année peut ainsi tomber à 7 % seulement, alors qu'elle représente jusqu’à 25 % quand la

saison des pluies se prolonge jusqu'en fin novembre. C'est dire l'influence considérable

qu’auraient plusieurs années défavorables consécutives sur la démographie de l’espèce.

Source : MNHN, Paris

120 ALYTES 16 (3-4)

—>

& 60%,

ÿ 50%

x 40%

30%

20%]

150 "200 250 300 "350 400 50

Pluies octobre + novembre

Fig. 8.- Variation interannuelle du pourcentage des femelles restées vierges lors de leur enfouissement en

fonction des quantités de pluie tombées en octobre et novembre de l'année de leur naissance (d'après

LAMOTTE, 1959).

Comme le montre la fig. 9, la pluviosité annuelle totale est, contrairement à la fin plus ou

moins précoce de la saison des pluies, sans action sur la proportion de femelles restant vierges

avant l’enfouissement.

CONCLUSIONS

Les études menées sur le terrain entre 1942 et 1991 ont fait apparaître l’étroite corrélation

qui existe entre les populations du petit bufonidé vivipare orobionte Nectophrynoides occi-

dentalis et le cycle climatique de la prairie d’altitude où il est localisé. La corrélation, qui se

manifeste déjà avec rigueur à l'échelle de l’année climatique moyenne, est coroborée et

précisée par la comparaison de plusieurs années différant par leur cycle saisonnier. Celle-ci

fait ressortir le rôle prépondérant du calendrier des pluies et plus particulièrement de l’arrivée

plus ou moins précoce de la saison sèche qui influe sur le pourcentage de jeunes femelles de

l’année fécondées avant de s'enfouir. La fécondité globale de l'espèce peut être ainsi considé-

rablement modifiée.

Source : MNHN, Paris

LAMOTTE & SANCHEZ-LAMOTTE 121

% de femelles vierges

60 %

50 %

4%| ©

30 %

20 % e e

1099 1224 1244 1501 1506 1576 1730 1757

pluies de l’année

Fig, 9. - Pourcentage de femelles restées vierges lors de leur enfouissement en fonction de la pluviosité

iotale (en mm) de l'année de leur naissance.

RÉSUMÉ

Dans l'aire de répartition du crapaud vivipare Nectophrynoides occidentalis Angel, 1943,

limitée à quelques kilomètres carrés d’une savane à herbes courtes couvrant les crêtes du mont

Nimba au-dessus de 1200 m d'altitude, le climat est caractérisé par l’alternance d’une saison

très sèche de l’ordre de 5 mois contrastant avec une saison de 7 mois de pluies et de bruines.

Les cycles de vie et de reproduction de l'espèce sont étroitement liés à ce cycle saisonnier.

Durant la saison sèche, les crapauds sont enfouis dans le sol en état de vie ralentie. Ils sortent

dès l’apparition des pluies, entre février et fin mars, avec un décalage entre les femelles

gravides, qui sortent les premières, les femelles vierges et enfin les mâles. Toutes les mises-bas

ont lieu en juin. Les fécondations se font en septembre pour les femelles âgées de plus d’un an,

jusqu’en fin octobre pour les femelles de l’année. Toutes s’enfouissent aussitôt fécondées.

Le cycle ne semble pas modifié par la quantité d'eau tombée annuellement, pourtant

variable du simple au double selon l'emplacement dans la chaîne et selon l’année. En

revanche, le suivi du cycle climatique et celui des populations au cours de plusieurs années a

montré que les dates de l’enfouissement et de la sortie de la vie ralentie sont liées étroitement

à l’arrivée et surtout à la fin de la saison des pluie qui peuvent différer de plus d’un mois selon

les années. Ces variations se traduisent par des différences importantes de la proportion de

jeunes femelles restées vierges avant de s’enfouir pour leur première saison sèche. Elles se

répercutent ainsi sur la fécondité globale de la population.

Source : MNHN, Paris

122 ALYTES 16 (3-4)

RÉFÉRENCES BIBLIOGRAPHIQUES

ANGEL, F, 1943. Description d’un nouvel amphibien anoure ovovivipare de la Haute Guinée française.

Bull. Mus. natn. Hist. nat., 15: 167-169.

ANGEL, F. & LaMorTE, M., 1944. — Un crapaud vivipare d'Afrique Occidentale, Nectophrynoides

occidentalis Angel. Ann. Sci. nat. Zool., 6: 63-89.

_— 1948. — Nouvelles observations sur Nectophrynoides occidentalis Angel. Remarques sur le genre

Nectophrynoides. Ann. Sci nat. Zool., 10: 115-147.

LamorTE, M., 1959. - Observations écologiques sur les populations naturelles de Nectophrynoïdes

occidentalis. Bull. biol. Fr. Belg., 93: 355-413.

Laworrr, M. & ROUGERIE, G., 1955. - Description régionale de la chaîne du Nimba. /n: LECLERC et al.

(1955): 23-58.

Laorrr, M. & XAVIER, F, 1972. - Recherches sur le développement embryonnaire de Nectophrynoides

occidentalis Angel, amphibien anoure vivipare. Ann. Embr. Morph., 5: 315-340.

Lecerc, J. C., RicHARD-MOLARD, J., LAMOTTE, M., ROUGERIE, G. & PORTÈRES, R., 1955. — La chaîne du

Nimba, essai géographique, Mém. IFAN, 43: 1-271.

RicHaRD-MOLARD, J,, LAMOTTE, M. & PORTÈRES, R., 1955. - Les conditions climatiques. In: LECLERC et

al. (1955): 59-115.

Corresponding editor: Alain DuBois.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 123-129. 123

Notes on morphological variation

and the biology

of Nototriton guanacaste

Good & Wake, 1993

(Caudata, Plethodontidae)

Michael FRANZEN

Zoologische Staatssammlung, Münchhausenstr. 21, 81247 München, Germany

The variation in body size, body proportions, and coloration of Noto-

triton guanacaste Good & Wake, 1993 is greater than documented

previously. Data from seven newly collected specimens suggest that the

character “snout-gular length”, previously considered to be diagnostic,

widely overlaps with that of other Costa Rican Nototriton species. The

variation of some aspects of coloration is considerably greater than in the

type series. Regarding the habitat, N. guanacaste seems to prefer locations

among roots of epiphytes growing in moss mats.

INTRODUCTION

Despite the comprehensive study of Goop & Wake (1993), the diminutive and incon-

spicuous plethodontid salamanders of the genus Nototriton are among the least known

species of the Costa Rican amphibian fauna. I collected specimens of the recently described

Nototriton guanacaste Good & Wake, 1993, which is endemic to two isolated peaks in

northwestern Costa Rica. This material provides new information on morphological varia-

tion with respect to body size, body proportions, coloration and on habitat and biology.

MATERIAL AND METHODS

Specimens of Nototriton guanacaste here studied are deposited in the collection of the

Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn, Germany (ZFMK).

Seven specimens and two clutches were collected at the type locality in Guanacaste National

Park, Costa Rica, in the western summit area of Cerro Cacao (1450-1550 m) on 7 September

1993 (ZFMK 67726), 13 September 1993 (ZFMK 67728) and 23 September 1993 (ZFMK

67727, 57729-57732). Furthermore, the material consists of a clutch of eight eggs deposited

on 7 September 1993 (three preserved on 7 September 1993: ZFMK 57733; five preserved on

12 October 1993: ZFMK 57734) and a clutch of five eggs found on 23 September 1993 (one

Source : MNHN, Paris

124 ALYTES 16 (3-4)

preserved on 23 September 1993: ZFMK 57735; four preserved on 12 October 1993: ZFMK

57736). Measurements follow the standards defined by BRAME (1968), and were made under a

dissecting microscope fitted with an ocular micrometer.

RESULTS

MORPHOLOGY

Measurements and morphometric dimensions of the newly collected material are shown

in tab. 1. Most of the newly ascertained body measurements and proportions (axilla-groin

length, head width, nostril diameter, forelimb length, hindlimb length, foot width, third and

fith toe length) agree well or are at least very near to those of the type series. However,

maximum length and variation of some body proportions are larger than previously docu-

mented. Three of the seven newly collected specimens have larger snout-vent lengths than the

largest specimens of the type series (holotype: 29.7 mm). Furthermore relative trunk width is

consistently larger in the newly collected material, with no overlap with the range of the type

series. In contrast, relative tail length is shorter in the newly collected material, likewise with

no overlap with the range of the type series. With respect to the relative snout-gular length

(‘head length”), only the two smallest specimens are near to measurements of the type

specimens, with head lengths of 20.3 and 21.3 percent of snout-vent length. Relative snout-

gular length is strongly negatively correlated with snout-vent length (r = -0.88; P = 0.004).

Regarding the development of the parotoid glands, only a somewhat physically enlarged

parotoid region is discernible in all specimens of the new material.

The preserved specimens are shown in fig. 1. Two specimens (ZFMK 57726 and 57732)

show a conspicuous bright dorsal ground coloration, which was light brownish-orange in life.

Within the entire series, the lateral and dorsal ground color varied from light brownish-orange

to dark brown in life. One of the seven new specimens (ZFMK 57731) has a bright lateral

coloration. In two specimens (ZFMK 57727 and 57729), the flanks are slightly brighter than

the dark brown dorsum, whereas the four remaining specimens have a lateral coloration which

is identical (ZFMK 57728) or darker than the dorsal ground color. The bright coloration of

the parotoid region is evident in all the new specimens. However, a bright elongate blotch on

the parotoids is indistinct and very small in ZFMK 57728 and 57730. In ZFMK 57726 and

57732, parotoid glands cannot be discerned by their color due to an overall bright dorsal

coloration.

NOTES ON BIOLOGY

AI specimens of Nototriton guanacaste Were observed in 10-20 cm thick dripping wet

moss mats growing on trees in “lower montane rain forest” (sensu Tosi, 1969: common

names, “cloud forest”, “elfin forest”) near the summit of Cerro Cacao. During 12 hours of

searching, three salamanders were taken from moss clumps hanging from twigs and branches,

whereas four specimens were found within 30 minutes on horizontal branches among the

Source : MNHN, Paris

Table 1. - Measurements (mm), followed in parentheses by morphometric ratios (percent of snout-vent length), of the seven newly collected

Nototriton guanacaste specimens compared with the range of the type series (after Goo & WAKE 1993). SVL: snout-vent length.

57727 57729 57726 57728 57731 57730 57732 (hoc loco)

Sex male male female | female | female | cf. female | cf female =

Snout-vent length 30.9 26.8 Fi.) 330 27.5 22.2 22.1 _ _

Axilla-groin length] 17.7 (57.3) | 14.0(52.2) | 20.0 (60.6) | 20.0(59.7) | 15.7(57.1) | 12.0(54.1) | 12.3 (55.7 | 522-606 | 545-563

Trunk width 4.7 (15.2) 4.0 (14.9) 4.5 (13.6) 5.0 (14.9) 4.0 (14.5) 3.2(144) 3.3 (14.9) 13.6-—15.2 | 110-122 eu

Tail length “ + 39.0 (116.4) - 28.9 (105.1) | 24.6 (110.8) | 23.9 (108.1) |105.1 — 116.4/121.0 — 133.7| £

Snout-gular length | 5.8 (18.8) 5.2 (19.4) 6.1 (18.5) 5.9 (17.6) 5.0 (18.2) 4.5 (20.3) 4.7 (21.3) 17.6-—21.3 | 21.6-224 2

Head width 41033 | 30040 | 42a2n | 44G31) | 38038 | 33049) | 35058 [127-158 | 145-157

Nostril diameter 0.24 (0.78) | 0.19(0.71) | 0.12 (0.36) | 0.19 (0.57) | 0.17 (0.62) | 0.21 (0.95) | 0.21 (0.95) | 0.36-0.95 04-09

Forelimb length 5.8 (18.8) 4.6 (17.2) 5.6(17.0) 5.3 (15.8) 42(153) 3.8 (17.1) 3.9 (17.6) 15.3-18.8 | 17.0-17.9

Hindlimb length 6.4 (20.7) 5.3 (19.8) 5.8 (17.6) 5.8 (17.3) 4.8 (17.5) 4.4 (19.8) 4.4 (19.9) 17.3-20.7 | 18.5-20.1

Foot width 2.5 (8.1) 2.0 (7.5) 2.3 (7.0) 2.1 (6.3) 1.9 (6.9) 1.3 (5.9) 1.5 (6.8) 59-81 66-72

Third toe length 1.0 (3.2) 0.9 (34) 1.2 (3.6) 0.9 (2.7) 0.8 (2.9) 0.7 (32) 0.6(2.7) 2.7-3.6 2.8-3.1

Fifth toe length 06(19) | 04(1.5) | 0618) | 05{:5) | 050.8) | 034 | 03G4) | 14-18 | 11-17

Source : MNHN, Paris

126 ALYTES 16 (3-4)

Fig. 1.- Variation in coloration in Nototriton guanacaste from Cerro Cacao (Guanacaste, Costa Rica).

Photo: Juliane Pop.

roots of epiphytes growing in the moss mats. In such microhabitats, the habitus of the

salamanders was remarkably similar to that of epiphytic roots. AI animals were found 0.5 to

5 meters above the ground. Substrate temperatures ranged from 18.6 to 21.5°C.

When grasped, all salamanders showed “coil-uncoil flip” and “running flip” defensive

behaviors (DopD & BRODIE, 1976). Flips were observed as far as 50 cm.

On 7 September 1993, female ZFMK 57726 and two single eggs were taken from a moss

clump growing on an air root two meters above the ground. The night after capture, the female

deposited six eggs in the moss of the transportation container. An unguarded clutch of five

eggs, containing well developed embryos, was taken on 23 September 1993 from a moss mat

Source : MNHN, Paris

FRANZEN 127

growing on a vertical tree trunk about two meters above the ground. Both clutches were stored

in wet moss at room temperature in the laboratory for a time. Most eggs of both clutches

developed well until they were perserved.

DISCUSSION

Most of the here ascertained differences in morphometric dimensions are not surprising,

since only five individuals of N. guanacaste were analyzed in the original description by Goon

& Wake (1993). The data on the snout-vent lengths of the new specimens suggest that those

of the type series are not fully grown, though obviously mature (according to Goop & WAKE,

1993: 138, one male specimen of the type series has a “rather flat and inconspicuous mental

gland”). The new ascertained maximum snout-vent length of 33.5 mm (female, ZFMK

57726) makes N. guanacaste the second largest among the Costa Rican Nototriton species.

Only the single known specimen of N. major Good & Wake, 1993 has a larger size with a

snout-vent length of 37.9 mm. Differences in relative tail lengths of the new specimens

compared to the types may be caused by slightly different measurements. I measured snout-

vent length from the anterior tip of the snout to the posterior angle of vent. If measured to the

anterior angle of vent (and subsequently tail length from anterior angle to the tip of the tail),

the new specimens have relative tail lengths of 112.8 to 128.9 (mean 122.5 + 7.12) percent of

snout-vent length. This is well within the range of the data given in the original description.

Relative snout-gular length is a major diagnostic feature which separates N. guanacaste from

all other Costa Rican Nototriton (Goo & Wake, 1993). The revised range of 17.6 to 22.4

percent in this character (including data from Goop & Wake, 1993) widely overlaps with N.

picadoi (Stejneger, 1911), N. richardi (Taylor, 1949), N. tapanti Good & Wake, 1993, N. major

Good & Wake, 1993 and various populations of N. abscondens (Taylor, 1948). The differences

between my own data and those of Goop & WAKkE (1993) can be explained by the smaller size

of the type specimens (see above): relative snout-gular length is significantly negatively

correlated with snout-vent length. In other words, smaller animals have longer heads and

head length shows a changing relationship to body size as animals grow. Another diagnostic

feature which separates N. guanacaste from N. abscondens according to Goop & WaAKkE (1993)

is its prominent parotoid glands. I ascertained only rather flat and inconspicuous parotoid

regions in the new material. However, it is relative to some degree to regard a character as

“prominent” or “indistinct”, and the difference may be caused by my limited experience with

other Nototriton species. A single specimen of N. abscondens (EI Angel Waterfall, Provincia de

Alajuela, Costa Rica, in my private collection) indeed shows much more reduced, almost

invisible parotoid glands.

The robust habitus (as measured by “trunk width”) of N guanacaste, that makes it

unmistakable among Costa Rican species, is confirmed by the newly collected material. The

revised range with a maximum of 15.2 percent of snout-vent length even emphasizes dif-

ferences to the other species. However, one should keep in mind that differences between the

new material and the type series may be caused by different methods of conservation.

AI in all the robust habitus and the confirmed small nostril diameter (which is a

Major character separating N. guanacaste from the geographically nearest population of

Source : MNHN, Paris

128 ALYTES 16 (3-4)

N. abscondens at Monteverde) support the specific status of N. guanacaste from the morpho-

logical point of view.

Variation in coloration of the newly collected specimens is considerably greater than in

the type series. À bright lateral coloration, as reported in previously collected specimens, is

evident in only one specimen. Bright parotoid blotches are indistinct and very small in two

specimens. GooD & WAKE (1993) mentioned that these markings were less evident in their

smaller specimens. However, markings are inconspicuous among the new material in one

large (ZFMK 57728) and one small specimen (ZFMK 57730).

The observations regarding the biology agree well with data known for N. guanacaste and

other Nototriton species. Like all previously observed specimens, the new material was found

in moss mats on trees above the ground. Regarding the microhabitat, the new specimens were

observed with different success in two different structures: in moss clumps hanging from air

roots or growing on vertical branches (0.25 specimen/hour) and in moss mats among roots of

epiphytic ferns and bromeliads on horizontal branches (8 specimens/hour). Due to the small

number of observed specimens, these results may be accidental. Nevertheless, it can be

considered that humidity conditions are more stable in the latter microhabitat due to a higher

proportion of humus and an overall thicker and more compact substrate cover.

The defensive behaviors “coil-uncoil flip” and “running flip” were previously reported

by Dopp & BRODIE (1976) for other neotropical plethodontids, including “Chiropterotriton

picadoi” (i.e., Nototriton richardi or N. abscondens sensu Goop & WAkE, 1993). I observed

that juvenile and adult N. picadoi and N. abscondens show the same behaviors in the field and

in captivity.

The clutch sizes of five and eight eggs observed during the present study correspond to

the data given by Goop & WaKkE (1993): two clutches with four and seven eggs. In other

Nototriton species, clutch size may be as high as 17 eggs (JOKUSCH & GARCIA-PARIS, 1998).

Nototriton and Oedipina are presumed to be the only bolitoglossines which abandon

their clutches (Goo & Wake, 1993). The finding of another unguarded clutch of M.

guanacaste supports this to some degree. It should be noted that I also found two further

unguarded clutches (with two and three eggs) of unidentified Nototriton on 3 and 4 October

1993 at Tapanti, Costa Rica.

Though considerably different to the type series in some aspects, I regard the newly

collected material as belonging to a single species. Differences in morphometric dimensions

are consistent among the newly collected material (tail length) or vary gradually (snout-gular

length). Furthermore, differences in coloration (parotoids and flanks) are not associated with

differences in body proportions or snout-vent length. Nototriton species can be highly specific

to microhabitats (see CAMPBELL & SMITH, 1998), so the different microhabitats observed in the

present study may give a hint for a specific differentiation. However, the occurrence of color

morphs (e.8., animals with dark flanks or animals with a bright overall coloration) did not

correspond to a certain type of microhabitat.

Source : MNHN, Paris

FRANZEN 129

ACKNOWLEDGMENTS

I thank the “Servicio de Parques Nacionales of Costa Rica” for research and collection permits and

Roger BLANCO for his invaluable help during my fieldwork at Maritza. 1 am grateful to Juliane Popr for

the photograph in fig. 1, to Kirsten Tirr for polishing the English and to Kenneth Dopp, Jr. and four

anonymous reviewers for valuable comments on earlier drafts of the manuscript.

LITERATURE CITED

BRAME, À. H., Jr., 1968. - Systematics and evolution of the Mesoamerican salamander genus Oedipina.

J. Herp., 2: 1-64.

CamPBELL, J. A. & SMITH, E. N., 1998. - New species of Nototriton (Caudata: Plethodontidae) from

eastern Guatemala. Sci. Pap. nat. Hist. Mus. Univ. Kansas, 6: 1-8.

Don», C.K., Jr. & BRODIE, E. D., 1976. - Defensive mechanisms of neotropical salamanders with an

experimental analysis of immobility and the effect of temperature on immobility. Herpetologica,

32: 269-290.

Goop, D. A. & Wake, D. B., 1993. - Systematic studies of the Costa Rican moss salamanders, genus

Nototriton, with descriptions of three new species, Herp. Mon., 7: 131-159.

JokuscH, E. L. & Garcia-PaRis, M., 1998. — Nororriton abscondens (Cordilleran Moss Salamander).

Reproduction. Herp. Rev, 29 (1): 38.

Tost, J. A., 1969. - Repiblica de Costa Rica. Mapa Ecolôgico. San José, Centro Cientifico Tropical.

Corresponding editor: Franco ANDREONE.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 130-138.

Limits of the morphometric method

for field identification of water frogs

Alain PAGANO & Pierre JOLY

ESA CNRS 5023 Ecologie des eaux douces et des grands fleuves,

Université C. Bernard Lyon 1, 69622 Villeurbanne Cédex, France

e-mail: Pagano@univ-lyonl.fr, pjoly@biomservuniv-lyonl.fr

Taxonomic identification of the water frogs has evolved since hybrido-

genesis has been revealed within the Rana esculenta complex. Although

the study of protein polymorphism has proved robust in taxonomic infor-

mation, morphometric measurements are currently used despite of some

limitations of the method. By comparing results obtained with these two

techniques, this study shows that morphometry is not always decisive for

field identification. In the three populations studied, in the mid-Rhône

floodplain, the morphs of Rana ridibunda and the hybrid Rana kl. escu-

lenta greatly overlap in morphometric characters.

INTRODUCTION

The Palearctic water frog group is composed of several species (for a review see DUBoIS

& OHLER, 1995) and is characterized by three hybridogenetic complexes (synkleptons sensu

PoLLs-PELAZ, 1989). The Rana esculenta complex, which is widespread in central Europe, is

the more studied of these complexes. The three taxa of this synklepton (Rana ridibunda, Rana

lessonae and the hybridogenetic hybrid Rana kl. esculenta) have been distinguished by several

morphological characters for a long time (e.g. CAMERANO, 1884), but the systematics of water

frogs remained confused until the existence of a hybrid complex was demonstrated (BERGER,

1968). In this context, the morphometric indices proposed by BERGER (1966) to discriminate

three morphs among the hybridogenetic complex strongly contributed to the systematics of

the group, and this method is still commonly used (for a recent review, see OGIELSKA, 1995).

Nevertheless, several morphometric investigations showed an overlap among the charac-

teristic morphs of several taxa (e.g. GÜNTHER et al., 1991; PoLLs-PELAZ, 1991; RYBACKI,

1995). Besides using the morphological indices proposed by BERGER (1966), some authors

applied sophisticated analysis (discriminant analysis, multivariate analysis) to maximize the

morphological differences between taxa (e.g. UZZELL & Horz, 1979; PLÔTNER et al., 1994).

Despite the increasing complexity of taxonomic identification on the basis of morphometric

variables, this morphometric method still remains. On the other hand, the analysis of protein

polymorphism proves robust in taxonomic identification.

Although the use of quantitative morphological traits fails in identification of water

frogs in eastern France (JoLY et al., 1995; TUNNER, personal communication), some studies

Source : MNHN, Paris

PAGANO & JoLY 131

only used the morphometric method in frog taxonomy. Because of large number of indi-

viduals to be identified, field studies need simple methods. In this context, the aim of this

paper was to compare the simplest morphometric measurements currently used (e.g. Dp/Cint)

with the analysis of allozymic markers.

MATERIAL AND METHODS

SITES AND SAMPLE SIZES

Three populations (Morte-de-la-Barre, Jons, Pierre-Bénite) were investigated in sites

located near the active channel of the Rhône river. The former two ponds are gravel-pits while

the last one is a regularly overflowed side arm of the Rhône. The sample size is the following:

Pierre-Bénite, n = 28 (15 males and 13 females); Jons, n = 31 (19 males and 12 females);

Morte-de-la-Barre, 7 = 33 (25 males and 8 females). Voucher numbers are: Jo26-33, Jo35,

Jo37-38, Jo40-47, JoSS, Jo92-102, PBS50-54, PB103-125, MB56-63, MB65-71, MB74-91, all

deep-frozen carcasses, kept in our laboratory (Université Lyon 1, France).

PROTEIN ELECTROPHORESIS

Electrophoresis was performed on skeletal muscles. Tissue samples were crushed in a

1.2 g Tris + 0.37 g EDTA + 1 1 H,0 + 50 ml NADP 1 % solution. Migration was performed

in a Tris citrate gel at pH 6 during 3 to 5 hours under 180 Volts. Tris citrate gel composition

was: 48 gstarch (12 %), 1.4 ml buffer 1 * (composition of the 10 x buffer: Tris 270 g, citric acid

181 g, H,0 1000 ml), 398.6 ml H,0. Staining solutions were prepared using modifications of

standard procedures (PASTEUR et al., 1987; HoTz, unpublished).

Four loci were analyzed for somatic tissues: lactate deshydrogenase (LDH-1, Enzyme

Commission 1.1.1.27), mannose-phosphate-isomerase (MPI, E.C. 5.3.1.8), phosphogluco-

mutase (PGM-2, E.C. 2.7.5.1) and creatine kinase (CK, E.C. 2.7.3.2). These enzymes were

chosen because they are known to be efficient for taxonomic identification of several species

and hybrids of water frogs (for review, see HOTZ, 1983 and BEERLI, 1994).

Reference specimens from the collection of the Zürich University (H. HoTz) were used as

control samples (2 specimens for each of the following taxa): Rana perezi (Elvo Delta, Spain),

Rana kl. grafi (Pouzolles, France), Rana ridibunda (Mosina, Poland), Rana kl. esculenta

(Hellberg, Switzerland) and Rana lessonae (Poznan, Poland and Hellberg, Switzerland).

Respective voucher numbers are: 17027, 17030, 17570, 17572, 18095, 18096, 18011, 18109,

18094, 18102, all deep-frozen tissues (no carcasses), kept in the Zürich University (Switzer-

land).

MORPHOMETRY

The method of SAGNES (1995) was used in collecting morphometric data. Demedulated

animals were disposed on a box, near a scale. A photograph taken using a video camera was

numerized by the computer. Using the “Image © software”, we scaled the photographs and

the variables were measured (fig. 1). Because this software allows to zoom a part of the

photograph for measuring variables of small size (the metatarsal tubercle in our study), the

Source : MNHN, Paris

132 ALYTES 16 (3-4)

Fig. 1. A specimen of water frog numerized and measured by computer software.

errors in measuring parameters were minimized (SAGNES, 199$). Five variables were measured

on computerized frog photographs: Le (body length), Ti (Tibia length), Dp (First toe length)

Cint (Metatarsal tubercle length) and Cint-a (Metatarsal tubercle height). These measure-

ments were used to calculate morphometric indices (Dp/Cint, Ti/Cint, Ti/Cint-a) that are

known to discriminate the three forms of the esculenta synklepton (BERGER, 1966). Male and

female analyses were done separately. Measurements were made before freezing the animals.

RESULTS

ELECTROPHORETIC IDENTIFICATION

The analysis of specific markers in the loci studied established the presence of Rana

ridibunda and R. kl. esculenta, and the absence of R. lessonae, R. perezi and R. kl. grafi in the

sites studied (tab. 1).

Whereas the Jons population was exclusively composed of R. ridibunda, the others were

mixed populations of R. ridibunda and R. KI. esculenta with 12% and 19 % of hybrids in

Morte-de-la-Barre and Pierre-Bénite, respectively.

Source : MNHN, Paris

PAGANO & JoLY 133

Table 1. - Specific allozymes or specific genotypes which allow taxonomic identification of

water frogs.

Allozymes or genotypes Number of frogs per site

Species

Pierre Morte

Bénite PS | Barre

Allozyme Rana

aorc ridibunda

Genotype Rana kl.

ah esculenta

Allozyme Rana

lorm perezi

(1) No specific marker between R. lessonae and R. ridibunda. The identification of R.

kl. esculenta is not possible with only this locus.

(2) No specific marker between R. perezi and R. ridibunda.

MORPHOMETRIC IDENTIFICATION

The graph Dp/Cint versus Ti/Cint usually discriminates the different forms of the

esculenta synklepton (BERGER, 1966). However, in the populations studied and with the

morphometric method used (based on computerized photographs), these morphological

indices did not clearly separate the different morphotypes neither for males nor for females

(fig. 2). Thus, for males, the use of genetic taxonomic markers revealed that the morphological

indices of R. kl. esculenta widely overlapped those of R. ridibunda in the populations studied

(fig. 2), and most of the hybrids could not be distinguished from R. ridibunda using these

indices. Whereas an overlapping was also evidenced for females, the small sample size does not

allow a decisive conclusion.

DISCUSSION

In central and eastern Europe, each taxon of the R. esculenta synklepton can be

identified by several morphological indices (BERGER, 1966; BLANKENHORN et al., 1971;

Source : MNHN, Paris

134

Ti/Cint a

ALYTES 16 (3-4)

[=]

1 | 3 4

Dp/Cint

100 +

80 +

Ti/Cint

Fig. 2. - The usual graphs Ti/Cint versus Dp/Cint and Ti/Cint versus Ti/Cint-a reveal a great overlapping

between the morphotypes of R. kl. esculenta (black) and R. ridibunda (white). Squares symbolize

males and circles females. Several thresholds are represented. These limits discriminate R. kl.

esculenta from R. ridibunda in the following respective references:

CEANU & TEsio, 1993; (3) PoLLs-PELAZ, 1991; (4) RÉGNIER & NEVEU, 198

GELDER, 1976. None of these references make it possible to identify the fro,

: (1) BERGER, 1966; (2) COLGANI-

5) WunaNDs & VAN

of the present sample.

Source : MNHN, Paris

PAGANO & JOLY 135

WuNANDS & VAN GELDER, 1976). However, in several studies, morphological identification

did not correspond with genetic identification (e.g. GÜNTHER et al., 1991; POLLS-PELAZ, 1991;

RyBacKki, 1995). In our study, morphological features of esculenta males greatly overlapped

with those of R. ridibunda and no clear morphotype (as currently described) was detected.

Thus the morphometric indices are not always valid for taxonomic identification in the field.

Morphometric identification is far from being secure, at least in the studied region and using

our method (photographs of non-fixed animals). Other studies evidenced similar problems of

taxonomic identification (JoLy et al., 1995; KorLic & SULOVA, 1995; LaDa et al., 1995;

RyBACKI, 1995; MoRAND et al., in preparation). Thus, the limitations of identification using

these indices are striking when we report the values of Ti/Cint given by several authors as

discriminating values for the three morphs of the R. esculenta synklepton. Thresholds vary

between studies (see tab. 2 for a review and fig. 2). Though it may be argued that there are

artefactual differences linked to differences in methods (fixed specimens or living frogs,

differences in measurement methods, investigations with or without taking care of morpho-

metric differences between males and females), such a variation in morphological traits

suggests several other hypotheses or questions:

(1) Are morphological traits more representative of adaptation than of phylogenetic

relationships? Some ecological variables in relation to a gradient of flood disturbance lead to

this hypothesis (MORAND et al., in preparation). The sites we studied were within a floodplain

where ecological successions are rapid and different habitats patchily distributed. In tadpoles,

variation in size is greater in unpredictable environments than in predictable ones (WILBUR &

CoLLis, 1973). Morphology is probably determined on the one hand by phylogenetic

constraints and on the other hand by environmental conditions. The absence of distinct

morphotypes can be explained by the expression of phenotypic diversity in the context of

unpredictable and heterogeneous environments. So, we hypothesize that morphological

discrimination found in several studies in stable environments is perhaps more an effect of

different, separate and stable habitats than the result of phylogenetic lineage. However, there

is no evidence in the literature to support this statement because of a lack of ecological

description of sites (PAGANO et al., in preparation). Morphometric method was more used as

a taxonomic tool than for ecological investigations. In a same taxon, the morphological

variation between populations of different biogeographic regions (tab. 2) can be the result of

genetic structurations. Several studies have shown that R. ridibunda is highly variable (HOTZ

et al., 1985; BEERLI, 1994; PAGANO et al., 1997). Besides, the genetic distance between R. kl.

esculenta of France and central Europe is unknown. The hypothesis of genetic structuration

Within a taxon remains to be tested.

(2) According to GROSSENBACHER (1988), the presence of R. ridibunda in the upper-

Rhône river is recent and due to introductions. In this respect, we can hypothesize that, for a

long time, R. kl. esculenta lived alone in habitats favorable for R. ridibunda. So its morphology

may reflect its adaptation to these habitats. The absence of distinct morphotypes for R.

ridibunda and R. kI. esculenta could be explained by convergence.

(3) Does temperature influence morphological variation? REPA (1977) showed that tibia

length was related to the mean water temperature of the ponds. The epigenetic origin of

morphological variation has to be studied. Such an idea has been suggested to explain the

high values of indices in water frogs from western France (RÉGNIER & NEVEU, 1986).

Source : MNHN, Paris

136 ALYTES 16 (3-4)

Table 2. - Differences in the discriminating values of the index Ti/Cint for the identification of

water frogs in some countries of Europe.

Rana Rana kl. Rana

References lessonae | esculenta | ridibunda

RÉGNIER & NEVEU, France (Bretagne,

1986 4 North-East)

France (Paris

POLLS-PELAZ, 1991 É _- É

region)

GÜNTHER, 1975 - | Germany

+ T

WILINANDS & ë >8.5 Netherlands

VAN GELDER, 1976

BERGER, 1966 x Poland

COGALNICEANU &

Romania

TESIO, 1993

In several studies, investigations were performed on the basis of the sole morphometric

identification, but we assert that such an identification is far from being secure. For the

moment, only genetic identification provides decisive criteria for taxonomic identification.

Because several studies (experimentation, field studies, etc.) need identification of living

animals, we may recommend the use of electrophoresis. It is possible to perform such an

analysis on a small piece of tissue (a cut toe or blood; HOTZ, personal communication;

PAGANO, unpublished data), so that data collection is easy in the field. However, other

morphological criteria allowing identification may be found, such as the shape of the

vomerine teeth (CROCHET et al., 1995), though the pertinence of such methods has to be

checked by extensive comparison with electrophoretic data.

RÉSUMÉ

Pour des raisons historiques, la morphométrie est couramment utilisée pour la détermi-

nation taxinomique des grenouilles vertes du complexe Rana esculenta. L'utilisation de

l'éléctrophorèse de protéines est souvent utilisée à des fins identiques. Dans cette étude, la

détermination des spécimens a été effectuée à la fois par l’analyse du polymorphisme enzy-

matique et par la morphométrie en analyse d'images, contribuant à montrer que cette dernière

technique n’est pas totalement fiable pour des déterminations sur le terrain.

Source : MNHN, Paris

PAGANO & JOLY 137

ACKNOWLEDGEMENTS

Thanks to Hansjürg Horz who made electrophoretic investigation possible by giving check samples.

We also wish to thank Pierre SAGNES (his protocol was helpful for our morphometric investigations),

Hansjürg HOTZ, Alain MORAND and Pierre-André CROCHET for comments, and all those who helped us

in the field: Dr. C. P. HENRY, Eric PAGANO, Nadège PERRET, Laurie MAUCLAIRE, Wojteck MAILLE and

Severine MESNAGER. Two anonymous reviewers, Dr. M. MATsUI and Dr. A. DuBois provided fruitful

comments.

LITERATURE CITED

B&erLi, P, 1994. - Genetic isolation and calibration of an average protein clock in western Palearctic water

Jrogs of the Aegean region. Inaugural dissertation, Universität Zürich: 1-92.

BERGER, L., 1966. - Biometrical studies on the population of green frogs from the environs of Poznan.

Ann. Zool. pol. Akad., 23: 303-323.

ee 1968. - Morphology of the F1 generation of various crosses within Rana esculenta complex. Acta

Zool. Cracov., 12: 301-321.

BLANKENHORN, H. J., HEUSSER, H. & VOGe., P., 1971. - Drei Phänotypen von Grünfroschen aus dem

Rana esculenta Komplex in der Schwei 1242-1247.

CAMERANO, L., 1884. - Recherches sur les variations de la Rana esculenta et du Bufo viridis dans le bassin

de la Méditerranée, Zoologie et Zootechnie, séance du 16 avril 1884: 680-692.

COGALNICEANU, D. & Tesio, C., 1993. — On the presence of Rana lessonae in Roumania. Amphibia-

Reptilia, 14: 90-93.

CROCHET, P. A., DUBOIS, A., OHLER, A. & TUNNER, H., 1995. Rana ( Pelophylax) ridibunda Pallas, 1771,

Rana (Pelophylax) perezi Seoane, 1885 and their associated klepton (Amphibia, Anura): morpho-

logical diagnoses and description of à new taxon. Bull. Mus. natn. Hist. nat., (4), 17: 11-30.

Dusois, A. & OHLER, A., 1995. — Frogs of the subgenus Pelophylax (Amphibia, Anura, genus Rana): a

catalogue of available and valid scientific names, with comments on name-bearing types, complete

synonymies, proposed common names, and maps showing all type localities. Zool. Pol., 39:

139-204.

GROSSENBACHER, K., 1988. — Atlas de distribution des amphibiens de Suisse. Documenta Faunistica

Helvetiae,

GÜNTHER, R., 1975. - Zum natürlichen Vorkommen und zur Morphologie triploider Teichfrôsche, * Rana

esculenta” L., in der DDR (Anura, Ranidae). Mitt. zool. Mus. Berlin, 51: 155-157.

GÜNTHER, R., PLÔTNER, J. & TETZLAFF, L., 1991. - Zu einigen Merkmalen der Wasserfrôsche (Rana synkl.

esculenta) des Donau-Deltas. Salamandra; 27: 246-265.

Horz, H., 1983. - Genic diversity among water frog genomes inherited with and without recombination. Ph.

Dr. inaugural dissertation, Universität Zürich.: 1-136.

Horz, H., MANCINO, G., BUCCI-INNOCENTI, S., RAGGHIANTI, M., BERGER, L. & UZZELL, T., 1985.- Rana

ridibunda varies geographically in inducing clonal gametogenesis in interspecies hybrids. J. exp.

Zool., 236: 199-210.

JOLY, P., PAGANO, A. & MORAND, A., 1995. - Biometrical investigations of water frogs in an alluvial valley

and a plateau in eastern France. Zool. Pol., 39: 493-499.

Koric, P. & SuLova, K., 1995. - Syntopic occurrence of three taxa of water frogs in Czech Republic.

Zool. Pol., 39: 417-424.

Laba, G. A. BoRki, L. J. & VinoGrapov A. E., 1995. — Distribution, populations systems and

reproductive behavior of green frogs (hybridogenetic Rana esculenta complex) in the central

Chernozem territory of Russia. Russian J. Herp., 2: 46-57

MORAND, A. JOLY, P. & PAGANO, A. in preparation. — Morphological variation along a gradient of river

influence in water frogs (Rana esculenta complex).

Source : MNHN, Paris

138 ALYTES 16 (3-4)

Ocuetska, M., 1995. II International symposium on ecology and genetics of European water frogs.

Zool. Pol., 39: 117-528.

PAGANO, A. JoLy, P. & HoTz, H., 1997. - Taxon composition and genetic variation of water frogs in the

mid-Rhône floodplain. C. r Acad. Sci., Sci. Vie, 320: 759-766.

PAGANO, A., PLENET, S. & JoLY, P. in preparation. - Ecology of the hybridogenetic water frog complex.

I. Breeding habitat partitioning.

PASTEUR, N., PASTEUR, G., BONHOMME, F, CATALAN, J. & BRITTON-DAVIDIAN, J., 1987. —- Manuel

technique de génétique par électrophorèse de protéines. Paris, Lavoisier: 1-217.

PLÔTNER, J., BECKER, C. & PLÔTNER, K., 1994. - Morphometric and DNA investigations into European

water frogs (Rana ki. esculenta synklepton) (Anura, Ranidae) from different population systems.

J. zool. Syst. evol. Res., 32: 193-210.

PoLLs-PELAZ, M., 1989. - The biological klepton concept (BKC). Alytes, 8: 75-89.

es 1991. — Unisexualité mâle dans un peuplement hybridogénétique de grenouilles vertes: génétique,

éthologie, évolution. Thèse, Muséum National d'Histoire Naturelle de Paris: 1-182.

RÉGNIER, V. & NEVEU, A., 1986. — Structures spécifiques des peuplements en grenouilles du complexe

Rana esculenta de divers milieux de l’ouest de la France. Acta Œcologica, 7: 3-26.

Repa, P., 1977. - Biometrische Analyse der Grünfrôsche aus dem Ceskyles Gebirge und der Tachovka

Brazda Senke (Sudwestbôhmen). Vest. cs. Spolec. zool., 61: 121-134.

RyBacKki, M., 1995. - Water frogs (Rana esculenta complex) of the Bornholm Island, Denmark. Zoo!.

Pol., 39: 331-344.

SAGNES, P., 1995. — Un outil de prise de données sur une image numérisée et son utilité dans les études

relatives aux poissons: exemple d’une application concrète en morphométrie. Bull. fr Pêche Piscie.,

337-339: 131-137.

Uzzeui, T. & Horz, H., 1979. - Electrophoretic and morphological evidence for two forms of green frogs

(Rana esculenta complex) in peninsular Italy. (Amphibia, Salientia). Mitt. zool. Mus. Berlin, 55:

13-27.

WuNaANDs, H. E. J. & VAN GELDER, J. J., 1976. - Biometrical and serological evidence for the occurrence

of three phenotypes of green frogs (Rana esculenta complex) in the Netherlands. Neth. J. Zool., 26:

414-424.

WiLBur, H. M., & COLLINS, J. P., 1973 - Ecological aspects of amphibian metamorphosis. Science, 182:

1305-1314.

Corresponding editors: Masafumi MaTsuI & Alain DuBois.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 139-147. 139

Estructura del condrocräneo

y esqueleto visceral de larvas

de Pseudis minuta (Anura, Pseudidae)

E. O. LaviLLA* & Rafael O. DE SA**

* Fundacién Miguel Lilo, Miguel Lillo 251, 4000 Tucumän, Argentina

** Department of Biology, University of Richmond. Richmond, VA 23173, Estados Unidos

The chondrocranium and visceral skeleton of Pseudis minuta tadpoles

are described, based on a series of five larvae in stages 31-35 of Gosner

(1960). Among their striking characters are the presence of peculiar

articular surfaces between cornua trabeculae and suprarostral cartilage, the

incomplete development of the orbital cartilage, the high fenestration of the

floor of the cavum cranii, the fusion of posterior foramina, and the fusion of

spicules 3 and 4 in the hyobranchial skeleton.

INTRODUCCION

Diversos autores han tratado la morfologia externa de las larvas de Pseudidae. En

tiempos recientes KENNY (1969), DUELLMAN & TRUEB (1985), EMERSON (1988) y Cais &

Vizorro (1993) se ocuparon de Pseudis paradoxa, haciendo hincapié en el gigantismo

corporal, en tanto que Dixon et al. (1995) analizaron los cambios ontogenéticos en el patrôn

de coloracién. FERNANDEZ & FERNANDEZ (1921) y DE SÀ & LAviLLA (1997) caracterizaron la

larva de Pseudis minuta, y KEHR & Basso (1990) hicieron lo propio con la de Lysapsus

mantidactylus.

Si bien contamos con informacién adecuada sobre la morfologia externa, es muy poco lo

que sabemos sobre otros aspectos larvales. Por ejemplo, la informaciôn relativa a caracteres

anatémicos es escasa y antigua, y estä restringida a las descripciones del condrocräneo

(PARKER, 1882) y esqueleto visceral (PARKER, 1882; RIDEWOOD, 1898) de Pseudis paradoxa.

Sabemos que la familia Pseudidae es uno de los agrupamientos enigmäticos en el

conjunto de anuros neotropicales. Sus relaciones filogenéticas no han sido satisfactoriamente

dilucidadas, dado que no se han identificado apomorfias que la separen claramente de

Hylidae, su aparente grupo hermano (DUELLMAN & TRUEB, 1985; FORD & CANNATELLA,

1993; Hay et al., 1995), y la estructura interna de la familia también es materia de discusiôn:

aunque sélo se han reconocido dos géneros, su composicién especifica necesita revision

Tomando en cuenta lo que se acaba de decir, y considerando que la suma de caracteres

larvales puede ayudar a solucionar algunos de los problemas enunciados, es el objectivo de

Source : MNHN, Paris

140 ALYTES 16 (3-4)

este trabajo describir el condrocräneo y esqueleto visceral de larvas de Pseudis minuta, un

taxon cuya atribuciôn genérica fuera materia de disputa hasta no hace mucho tiempo.

MATERIAL Y MÉTODOS

La descripcién del condrocräneo y esqueleto visceral estä basada en cinco larvas en

estadios comparables a 31-35 de GOsnER (1960) (una por estadio), que forman parte de un

lote mayor depositado en la coleccién herpetolégica del Museo Nacional de Historia Natural,

Smithsonian Institution, bajo los nümeros USNM 497619 a 497639. El material fue obtenido

en laboratorio a partir de puestas naturalmente inducidas de una pareja proveniente de

Laguna del Cisne, Salinas, Departamento Canelones, Uruguay (27.X.94, A. OLMos y R. DE

SA col.). La hembra estä depositada como ejemplar de referencia como USNM 498369. Los

renacuajos fueron criados en acuarios de 40 litros con una densidad de 25 ejemplares por

acuario para estandarizar variables dependientes de la densidad, y alimentados ad libitum

con comida para peces carnivoros.

El material estudiado fue fijado en formol 10 % y teñido diferencialmente para hueso y

cartilago, y posteriormente diafanizado, segün la técnica de DINGERKUS & UHLER (1977). Las

observaciones se realizaron bajo glicerina en una lupa binocular Wild M3C.

RESULTADOS

NEUROCRÂNEO (FIG. la-c)

El cartilago suprarrostral es una estructura ünica, fuerte y completamente condrificada,

que se dirige hacia adelante y hacia abajo a partir del extremo distal de los cuernos trabecu-

lares. El cuerpo presenta una profunda escotadura dorsal en forma de V, y se une a las alas por

medio de una banda de cartilago relativamente ancha, que deja una escotadura ventral a cada

lado, de märgenes irregulares. Las alas son cuadrangulares; el proceso dorsal posterior es

proporcionalmente delgado y con extremo romo, y estä proyectado hacia afuera y hacia aträs.

El proceso ventral posterior no estä definido. En el margen proximal de cada ala, préximo a

su uniôn con el cuerpo, existe un ârea engrosada que actüa como superficie articular con el

cuerno trabecular respectivo.

Los cuernos trabeculares corresponden aproximadamente al 17 % de la longitud del

neurocräneo. Son estructuras fuertes, completa y uniformemente condrificadas y divergen

hacia adelante. El extremo distal estä levemente expandido, y los märgenes interno y anterior

son irregulares. Ventralmente, en el ângulo externo de cada cuerno se observa un ärea

cartilaginosa engrosada, que se corresponde con la superficie articular descripta para el

Ssuprarrostral. Préximo a la regin basal de cada cuerno, sobre su margen externo, se insinüa

el proceso lateral; la lamina cartilaginosa y el proceso prenasal estän ausentes.

Los cuernos trabeculares se continüan hacia aträs con la porciôn trabecular del piso del

neurocräneo, y en esa regiôn no se han diferenciado aüûn estructuras tales como la placa

etmoidal, el septo nasal, el techo nasal ni la lâmina orbitonasal.

Source : MNHN, Paris

FCP

CM PM AS FO

Fig. 1. - Condrocräneo de Pseudis minuta, estadio 33 de GOsnER (1960). (a) Vista dorsal. (b) Vista ventral. (c) Vista lateral. AS, arco subocular;

C, cuadrado: CaO, câpsulas éticas: CCCA, comisura cuadrado-craneal anterior; CM, cartilago de Meckel: CO, cartilago orbital; COc, céndilos

occipitales; CT, cuerno trabecular; FBC, fenestra basicraneal; FCP, foramen carotideo primario; FO, fenestra oval + operculum; FY, foramen

yugular: IR, cartilago infrarrostral: N, notocorda; PA, proceso ascendente; PM, proceso muscular; SR, cartilago suprarrostral; TS, techo sinético.

YS 4Q # VTIAVT

[LA

Source : MNHN, Paris

142 ALYTES 16 (3-4)

Los cartilagos orbitales son vestigiales, estando limitados a un par de proyecciones

cartilaginosas estrechas, oblicuas, una anterior y otra posterior, y que serian homélogas a las

pilas metoptica y antotica respectivamente. El espacio comprendido entre estas dos estructu-

ras estä abierto y no se reconocen foramenes de manera individual. La proyecciôn posterior

del cartilago orbital no tiene contacto con la câpsula ética, de modo que el foramen proôtico

estä abierto dorsalmente. No existen tenia tecti marginalis ni tenia tecti transversa, y en la

parte media del techo sinôtico se observa una proyecciôn triangular hacia adelante, que

corresponderia a un esbozo de tenia tecti medialis.

El piso de la cavidad craneal està poco condrificado y muestra la fenestra basicraneal

abierta y proporcionalmente muy grande, correspondiendo al 40 % de la longitud del neuro-

crâneo. Los fordmenes craneopalatinos estarian incluidos en dicha fenestra, mientras que los

forâmenes carotideos primarios son circulares, pequeños y estän claramente definidos. En la

regiéôn posterior, el arco occipital estä bien desarrollado y fusionado a las câpsulas éticas, los

céndilos occipitales estän esbozados pero aün no osificados, los foramenes yugulares estän

definidos y la notocorda penetra por un distancia equivalente al 25 % de la longitud del piso

de la cavidad craneal.

Las câpsulas 6ticas son cuadrangulares, oblongas y corresponden a aproximadamente el

35 % de la longitud del neurocräneo. La fenestra ovales grande (equivale al 1/3 de la longitud

de la câpsula ética) y el operculum, diferenciado como un elemento cartilaginoso, subcircular

y pequeño, ocupa aproximadamente 1/6 de la abertura. La cresta parôtica no se reconoce

como una estructura discreta, aunque desde el ängulo anterior externo, e inmediatamente por

delante de la fenestra oval, surge el proceso ético larval (en el sentido de DE BEER, 1937),

dirigido hacia adelante y hacia abajo, formando un arco. En la regiôn posterior ventral de

cada câpsula ética se observa un sélo foramen, de aproximadamente la mitad del tamaño de

la fenestra oval, que corresponderia a la fusiôn de los foramenes acüstico, perilinfaticos y

endolinfätico.

Dorsalmente las câpsulas ôticas estân unidas por el recho sinético, en forma de una

banda cartilaginosa que presenta en la regiôn media del margen anterior la proyecciôn

subtriangular ya mencionada.

SUSPENSORIO

En el suspensorio, el proceso ascendente tiene un desarrollo similar al del proceso ético y

se une al piso del neurocräneo (uni6n baja).

El arco subocular se presenta como una lämina delgada, claramente ensanchada en los

tercios medio y posterior y curvada hacia abajo.

En el cuadrado, el proceso muscular es subtriangular, de märgenes irregulares y extremo

romo. Està fuertemente curvado hacia adentro, de modo que su porcién distal se presenta casi

paralela al cuadrado. El margen posterior del proceso coincide con el margen posterior de la

comisura cuadrado craneal anterior. Esta comisura, con âreas de condrificaciôn débil, Ileva

en su margen anterior el proceso cuadrado etmoidal, y en el posterior el proceso pseudopte-

rigoideo. Ambos procesos tienen desarrollo similar, son subtriangulares y de vértice agudo.

La fosa hiocuadrado es poco notable, y el proceso articular, condilar, se muestra como un

Source : MNHN, Paris

LaviLLA & DE SÂ 143

engrosamiento cartilaginoso subtriangular y romo, ubicado en el margen lateral externo del

cuadrado, a nivel de la base del proceso muscular. El tünel muscular es abierto, y estä limitado

por abajo por la base del cuadrado y la comisura cuadrado-craneal anterior, y por su margen

externo y dorsalmente por el proceso muscular.

MANDIBULA INFERIOR

Los cartilagos de Meckel son subcilindricos y contorneados, con el proceso retroarticular

protruido y romo, mäs un pequeño proceso, también romo, ubicado en el margen interno, a

nivel del ängulo. Se unen a los infrarrostrales por medio de cépulas intermandibulares

ligamentosas.

Los cartilagos infrarrostrales, pares, son oblongos y curvados, y Ilevan una proyecciôn

posterior por la que articulan con los cartilagos de Meckel. La copula intramandibular es

conectiva.

ESQUELETO VISCERAL (FIG. 2a-b)

En el esqueleto hiobranquial no se reconoce la copula I. Los ceratohiales estän mejor

condrificados distal que proximalmente, muestran el proceso hiocuadrado oblongo, pro-

truido y bien desarrollado y los procesos anterior y lateral subtriangulares y notables.

La pars reuniens està muy débilmente condrificada y es de contorno aproximadamente

rectangular y mäs ancha que larga.

La copula II, aunque poco definida, estä mejor desarrollada que la estructura anterior:; es

aproximadamente dos veces mäs larga que ancha, con el extremo distal angular, y Ileva un

proceso urobranquial corto y romo.

La copula II estä relacionada con las placas hipobranquiales, muy poco condrificadas,

por tejido conectivo.

Los ceratobranquiales I a IV constituyen las estructuras mejor desarrolladas del esque-

leto hiobranquial y distalmente estän unidos entre si por comisuras terminales, mientras que

la uni6n con las placas hipobranquiales se realiza a través de bandas de tejido escasamente

condrificadas. Los ceratobranquiales I y III, por su parte, se unen entre si por medio de un

proceso branquial fuerte. Ventralmente existen dos espiculas delgadas y poco condrificadas

(que corresponden a los ceratobranquiales I y IT), mas una placa irregular, poco condrificada,

cribosa y continua con las placas hipobranquiales, que continüan los ceratobranquiales III y

IV (fig. 2b).

DISCUSION Y CONCLUSIONES

La ausencia de informacién sobre la estructura del condrocräneo en miembros del

género Lysapsus nos impide, por el momento, señalar el conjunto de caracteres derivados

compartidos por los Pseudidae y que podrian emplearse para dilucidar sus relaciones con

Source : MNHN, Paris

144 ALYTES 16 (3-4)

—1 PH

. 2. Esqueleto visceral de Pseudis minuta, estadio 33 de GosneR (1960). (a) Vista general. (b) Detalle

de placa hipobranquial. CB, ceratobranquiales; CBI, ceratobranquial 1; CBII, ceratobranquial II:

CBIII, ceratobranquial III; CBIV, ceratobranquial IV; CH, ceratohi , copula II; E, espiculas:

El, espicula 1; E2, espicula 2; E3+4, espicula 3 y 4 fusionadas: PH, placa hipobranquial; PR, pars

reuniens; U, proceso urobranquial.

Source : MNHN, Paris

LaviLLA & DE SÂ 145

Hylidae. Hasta obtener dicha informaciôn creemos conveniente analizar un conjunto de

caracteres del condrocräneo y esqueleto visceral de las larvas de Pseudis minuta que Ilaman la

atencién por no haber sido reportados previamente entre los anuros conocidos, o por ser

comunes a lo reportado para Pseudis paradoxa pero muy poco frecuentes en larvas de tipo IV.

(1) Llama la atenciôn los variados patrones de condrificaciôn del esqueleto cefälico.

Existen regiones fuertemente condrificadas (i.a., cartilagos supra e infrarrostral, cuernos

trabeculares), otras donde sélo se hacen evidentes las paredes de los condrocitos y otras mâs

donde el tejido condrogénico muestra una estructura irregular y difusa (como ciertas regiones

del piso del crâneo).

(2) La presencia de dos superficies articulares engrosadas en el margen proximal del

cartilago suprarrostral, en la regién de unién de cuerpo y ala, es también un caräcter

particular. Estas estructuras se corresponden con superficies articulares de caracteristicas

similares ubicada en la regiôn ventral del margen anterior de cada cuerno trabecular. Las dos

superficies articulares son planas, y se mantienen en posiciéôn y se flexionan por medio de

ligamentos. La ilustraciôn brindada por PARKER (1882: lâm. 2 fig. 1) muestra una estructura

aparentemente similar a la que aqui se describe. Por otra parte, la estructura del cartilago

suprarrostral seria derivada, considerando las discusiones de FABREZI & LAVILLA (1992),

PLASOTA (1974) y SOKkoL (1981).

(3) Los cuernos trabeculares son continuos hacia aträs con la porci6n trabecular del piso

del neurocräneo, sin que se hayan desarrollado aüûn (estadio 35 de GosnEr, 1960) las

estructuras caracteristicas de la regiôn etmoidal. PARKER (1882) reporté la presencia de una

estructura equivalente al septo nasal en Pseudis paradoxa.

(4) Los cartilagos orbitales, escasamente desarrollados, estän representados por un par

de pilares, uno en el extremo anterior (asimilado tentativamente a la pila metoptica) y otroen

el posterior ({pila antotica?) del piso del neurocräneo, dejando un gran espacio vacio entre

ellos. JACOBSON (1968) y SokoL (1981) consideraron a la ausencia de cartilago orbital (tal

como se observa en algunos Microhylidae) como derivada; la presencia de los pilares

extremos mostraria una condiciôn intermedia.

(5) En correlaciôn a la ausencia de un cartilago orbital continuo, el proceso ascendente se

une directamente al piso del neurocräneo, una condicién considerada como altamente

derivada por FABREZI & LAVILLA (1992), y que los asemeja a algunos hilidos (i.a., Phyllome-

dusa sauvagii, P. boliviana, Phasmahyla guttata, Hyla nana, Scinax acuminatus).

(6) La gran fenestracién del condrocräneo larval de Pseudis minuta se acentüa al

considerar el notable desarrollo de la fenestra basicraneal, equivalente, como dijéramos, al

40 % de la longitud total del crâneo. La mencionada fenestra se obtura en estadios tempranos

de desarrollo en Pseudis paradoxa, tal como se desprende de la descripciôn de PARKER (1882).

Es conveniente resaltar que los érganos del sistema nervioso central estän rodeados por una

fascia conectiva muy resistente y firmemente adherida a los elementos esqueléticos de la

regiôn.

(7) En la regién posterior del crâneo se destaca la fusiôn de los foramenes acüstico,

perilinfäticos y endolinfätico. El foramen resultante, de gran tamaño, se ubica en la regiôn

posterior ventral de cada cäpsula ética.

Source : MNHN, Paris

146 ALYTES 16 (3-4)

(8) El palatocuadrado muestra al menos dos caracteres notables. Uno es la expansién

proporcionalmente grande de las regiones media y posterior del arco subocular, y otro es el

notable desarrollo del proceso ôtico larval, que alcanza proporciones similares a la del

proceso ascendente. Ambos muestran condiciones equivalentes en Pseudis paradoxa.

(9) La posicién del proceso muscular del palatocuadrado, ubicado de modo que forma

parte del techo del tünel muscular, es un estado de caräcter derivado, si se tiene en cuenta el

anälisis de FABREZI & LAVILLA (1992).

(10) Pseudis minuta y P. paradoxa comparten la presencia de los procesos cuadrado

etmoidal y pseudopterigoideo en los märgenes anterior y posterior de la comisura cuadrado-

craneal anterior, respectivamente. En P paradoxa el proceso cuadrado-etmoidal de cada lado

estä en contacto con el proceso lateral de la base del cuerno trabecular, limitando completa-

mente a la coana; en P minuta los procesos laterales estân poco desarrollados y la coana estä

abierta anteriormente.

(11) La estructura de las espiculas en el esqueleto hiobranquial es también peculiar. Los

ceratobranquiales I y II estän seguidos por espiculas de estructura clâsica, en tanto que los

ceratobranquiales III y IV se continüan en una placa cuadrangular, poco condrificada y

cribada, formada por la fusién de las espiculas 3 y 4 (fig. 2b). Dichas placas son continuas con

la placa hipobranquial respectiva. Una condicién similar fue reportada por PARKER (1882) y

RIDEWOOD (1898) para Pseudis paradoxa.

RESUMEN

Se describe el condrocräneo y esqueleto visceral de las larvas de Pseudis minuta en base

a 5 ejemplares en estadios 31 a 35 de Gosner (1960). Entre los caracteres peculiares

observados se encuentran la presencia de una superficie articular particular entre los cuernos

trabeculares y el cartilago suprarrostral, el desarrollo incompleto del cartilago orbital, la gran

fenestracién del piso del crâneo, la fusién de los foramenes posteriores de la regiôn ética y la

fusién de la espiculas 3 y 4 en el esqueleto hiobranquial.

AGRADECIMIENTOS

Este trabajo fue parcialmente financiado con fondos del proyecto NSF BIR-9510228 otorgado a

R. DE S4. El estudio fue iniciado durante la estadia de E. O. LAVILLA como Visiting Research Associate

en la Universidad de Richmond, Virginia.

LITERATURA CITADA

Caïs, A. & Vizorro, L. D., 1993. - Biologia e ontogênese de Pseudis paradoxus (Amphibia, Anura,

Pseudidae). Livro de Resumos, 3° Congresso Latino-Americano de Herpetologia: 105.

De BeR, G. R., 1937. - The development of the vertebrate skull. Oxford, Clarendon Press: i-xxiti + 1-543,

143 läm.

Source : MNHN, Paris

LaviLLa & DE SÂ 147

DE Si, R. O. & LavicLA, E. O., 1997. - The tadpole of Pseudis minuta (Anura: Pseudidae), an apparent

case of heterochrony. Amphibia-Reptilia, 18 (3): 229-240.

DiNGERKUS, G. & UnLer L. D., 1977. - Enzyme clearing of alcian blue stained whole small vertebrates for

demonstration of cartilage. Stain Technol., 52: 229-232.

Dixon, J. R., MERCOLLI, C. & YANOSKY, A. À. 1995. - Some aspects of the ecology of Pseudis paradoxa

from northeastern Argentina. Herp. Rev, 26 (4): 183-185.

DUELLMAN, W. E. & TRUEB L., 1985. - Biology of amphibians. New York, MeGraw-Hill: i-xix + 1-670.

EMERSON, S. B., 1988. — The giant tadpole of Pseudis paradoxa. Biol. J. Linn. Soc., 34: 93-104.

FABREzI, M. & LAVILLA, E. O., 1992. - Estructura del condrocräneo y esqueleto hiobranquial en larvas

de algunos hilidos neotropicales (Anura: Hylidae). Acta Zool. Lilloana, 41:155 — 164.

FERNANDEZ, K. & FERNANDEZ, M., 1921. — Sobre la biologia y reproduccién de algunos batracios

argentinos. I. Cystignathidae. An. Soc. Ci. Arg., 91: 97-140, 3 lâm.

Forp, L.S. & CANNATELLA, D. C., 1993. - The major clades of frogs. Herp. Mon., 7: 94-117.

Goswer, K. L., 1960. — A simplified table for staging anuran embryos and larvae, with notes on

identification. Herpetologica, 16: 183-190.

Hay, I M., RuviNskY, L., HEDGES, S. B. & MAxsON, L., 1995. - Phylogenetic relationships of amphibian

families inferred from DNA sequences of mitochondrial 12S and 165$ ribosomal RNA genes. Mol.

Biol. Evol., 12 (5): 928-937.

Jaconson, C. M., 1968. - The development of the chondrocranium in two species of the Australian

anuran genus Pseudophryne Fitzinger. Aust. J. Zool., 16 (1): 1-15.

KENNY, JL S., 1969. - The Amphibia of Trinidad. Stud. Fauna Curaçao Caribb. Isl., 108: 1-78, 15 läm.

KEHR, A. L. & BASso, N. G., 1990. - Description of the tadpole of Lysapsus limellus (Anura: Pseudidae)

and some considerations on its biology. Copeia, 1990 (2): 573-575.

PARKER, W. K., 1882. - On the structure and development of the skull in the Batrachia. Part LIL. Phil.

Trans. r. Soc. London, “1881”, 172: 1-266, 44 läm.

PLAsoTA, K., 1974. - The devlopment of the chondrocranium (neurocranium and visceral arches) in

Rana temporaria L. and Pelobates fuscus (Laur.). Zool. Pol., 24 (1): 99-168.

RipewooD, W. G., 1898. — On the larval hyobranchial skeleton of the anurous batrachians, with special

reference to the axial parts. Z Linn. Soc. London, Zool., 26: 474-487, 1 läm.

Sokoz, O. M., 1981. — Larval chondrocranium of Pelodytes punctatus, With a review of tadpole

chondrocrania. J Morph., 169: 161-193.

Corresponding editor: Marvalee H. Wake.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 148-164.

The adult skeleton of Spea multiplicata

and a comparison of the osteology

of the pelobatid frogs (Anura, Pelobatidae)

Anne M. MAGLIA

Division of Herpetology, Natural History Museum and Biodiversity Research Center,

and Department of Ecology and Evolutionary Biology,

The University of Kansas, Lawrence, Kansas 66045-2454, USA

E-mail: magliaa@ukans.edu

Among the pelobatids (Anura, Pelobatidae), the skeletal anatomy of the

North American genera Spea and Scaphiopus is poorly known. Based on

dry-skeletal and cleared and double-stained specimens, 1 describe the

osteology of Spea multiplicata and compare it to that of all other pelobatid

taxa (Spea, Scaphiopus, Pelobates). Several anatomical structures are

shared by Spea and Scaphiopus, including the absence of a quadratojugal

bone, the presence of a palatine process of the facial process ofthe maxilla,

a long postchoanal process of the vomer, and a completely cartilaginous

sternum. Spea is characterized by a poorly developed maxillary process of

the nasal, the lack of a well-developed posteromedial process of the

parasphenoid, and possibly a well-developed pars ascendens plectri of the

auditory apparatus. Most other diagnostic features of Spea relate to the

limited cranial ossification of this genus relative to other members of the

family.

INTRODUCTION

Among “basal” frogs, the largest and arguably the most poorly known group is the

Pelobatoidea. These frogs comprise about 95 extant species (FROST, 1985) in three families

(Pelobatidae, Megophryidae and Pelodytidae), and are distributed throughout the Holarctic

Region extending into the Old World tropics (DUELLMAN & TRUrB, 1994). Among the

Pelobatidae are frogs in the genera Pelobates, Scaphiopus and Spea. Although the skeletal

anatomy of frogs in the genus Pelobates has been considered by several authors (e.g.,

ANDERSEN, 1978; ROCEK, 1981; RODRIGUEZ TALAVERA, 1990), the adult osteology of the

North American genera Spea and Scaphiopus remain poorly understood.

Of the few authors who have considered the skeleton of the North American pelobatids,

JURGENS (1971) included Spea intermontana in his description of the nasal cartilages of

anurans, RAMASWAMI (1939) described the cranial osteology of Scaphiopus holbrookii, and

FABREZI (1992) described the carpus of Scaphiopus couchii. The only thorough description of

the anatomy of these frogs is that by WIENS (1989) on the osteological development of Spea

bombifrons. K is in part because of the lack of detailed morphological descriptions of Spea

Source : MNHN, Paris

MAGLIA 149

and Scaphiopus that the phylogenetic relationships within the family Pelobatidae are unre-

solved (FORD & CANNATELLA, 1993). Therefore, I provide a detailed description of the adult

skeleton of Spea multiplicata, a species for which the anatomy is relatively unknown, and

compare its skeleton to that of other frogs in the family Pelobatidae, with the hope of

attaining information that may be phylogenetically useful.

MATERIALS AND METHODS

Osteological descriptions of Spea multiplicata were made from male and female dried

skeletons and cleared and double-stained specimens. Dry-skeletal and cleared and double-

stained specimens of Spea bombifrons, S. hammondii, S. intermontana, Scaphiopus couchii,

S. holbrookiü, S. hurteri, Pelobates cultripes, P. fuscus, P. syriacus and P. varaldii also were

examined (app. 1). Osteological terminology is that of DE SA & TRUEB (1991), TRUEB (1993),

DUELLMAN & TRUEB (1994) and FABREZI & ALBERCH (1996, for manus and pes). Descriptions

and illustrations were made with the aid of a stereo microscope equipped with a camera

lucida.

RESULTS

CRANIUM

The cranium is square and well ossified, but lacks dermal ornamentation (fig. 1). Both

the neopalatine and quadratojugal are absent in this species. The frontoparietal fontanelle is

exposed as a moderate-sized fenestra, and the maxillae and premaxillae bear teeth.

Nasal cartilages

The septum nasi is extensively ossified, synostotically fused to the sphenethmoid, and

extends forward anterior to the nasal roofing bones. The tectum nasi also is ossified and is

invested by the medial margins of the nasals. The oblique cartilages, which form the antero-

dorsal roof of the nasal capsule, are confluent anteromedially with the septum and tectum

nasi and posterolaterally with the commissura lateralis (fig. 2). À minute and blunt anterior

maxillary process projects forward from the anteroventral border of the planum antorbitale

toward the posterior half of the facial process of the maxilla. The posterior maxillary process

projects posteriorly from the posteroventral margin of the planum antorbitale, and is fused

synchondrotically to the pterygoid process of the palatoquadrate cartilage. The anterolateral

margin of the oblique cartilage unites with the robust crista subnasalis, which extends

ventrally to abut the anterior margin of the facial process of the maxilla. Posteriorly, the crista

subnasalis fuses with the solum nasi, the horizontal sheet of cartilage extending medially from

the septum nasi that forms the floor of the nasal capsule. A small, bifurcate process extends

posteriorly from the solum nasi to articulate with the sphenethmoid and the dorsal surface of

the vomer. The cup-shaped alary cartilage lies above the anterior margin of the solum nasi,

providing support for the anterior margin of the nares. The alary cartilage is united synchon-

Source : MNHN, Paris

vomer

OST

premaxilla

maxilla

spheneth

frontopar

pterygoid

prootic

exoccipital

optic f

lam perp fp

(t-£) 91 SALATV

max p nasal

2mm

septomaxilla

tymp ann

Fig. 1.- Cranium of Spea multiplicata (KU 86662) in (a) dorsal, (b) ventral, and (c) lateral view. Gray denotes cartilage, black denotes

foramina. Abbreviations: cultrif p, cultriform process of parasphenoïd: f, foramen; frontopar, frontoparietal: lam perp fp, lamina

perpendicularis of frontoparietal; max p nasal, maxillary process of nasal; p acen pl, pars acendens plectri; p ext pl, pars externa

plectri; p int pl, pars interna plectri; p med pl, pars media plectri; spheneth, sphenethmoid; tymp ann, tympanic annulus. Source : MNHN, Paris

MAGLIA 151

ant max p

— commis lateralis

crista subnasalis

sup prenasal c

inf prenasal c

Fig. 2. - Nasal cartilages of Spea multiplicata (KU 86664) in oblique anterior view. Gray denotes

cartilage, white denotes bone. Abbreviations: ant max p, anterior maxillary process; c, cartilage;

commis, commissura: inf, inferior; plan ant, planum antorbitale: sup, superior.

drotically with the superior prenasal cartilage, which extends ventromedially to the alary

process of the premaxilla. The inferior prenasal cartilage extends anteroventrally from the

solum to the premaxilla.

Septomaxillae

Each of these small bones lies medial to the fusion of the oblique cartilages and the crista

subnasalis, and posterior to the alary cartilage (fig. 2). Although complex in shape, only the

medial and lateral rami are exposed in dorsal view (appearing to be U-shaped).

Sphenethmoids

These extensively ossified elements are fused dorso- and ventromedially to form a single

bone, exposed dorsally in a diamond-shaped area between the frontoparietals and the nasals

(fig. la). Anteriorly, the ossified septum nasi is synostotically fused to the sphenethmoid,

forms the shared medial wall of the nasal capsules, and extends forward beyond the length of

the nasal roofing bones. Ventrally, the sphenethmoid floors the nasal capsules, extending

anteriorly to the level of the pars palatina of the premaxilla (fig. 1b). The sphenethmoid

extends laterally for most of the planum antorbitale to articulate with the pars facialis of the

maxilla. Ventromedially, the orbitonasal foramen opens posteriorly at the level of the anterior

margin of the cultriform process. Dorsomedial ossification includes the anterior margin of

the frontoparietal fontanelle; posteroventral and lateral ossification continues to the level of

the anterior margin of the optic foramen, thereby forming the anterolateral wall of the

neurocranium.

Source : MNHN, Paris

152 ALYTES 16 (3-4)

Prootics and exoccipitals

The prootics and exoccipitals are synostotically united with one another. The prootics

form the anterior and ventrolateral parts of the otic capsule, and are invested dorsomedially

by the frontoparietals (fig. 1a). The posterolateral margin of the frontoparietal extends to the

eminentia epiotica. Each prootic forms the posterior margin of the optic foramen; anterior

ossification extends only to the posterior three fourths of the prootic foramen. Laterally, the

prootic narrows to form an extensively ossified crista parotica, with only the distal tip being

cartilaginous. The lateralmost tip of the prootic, ventral to the crista parotica, is mineralized

and articulates with the basal process (sensu REIss, 1997) of the pterygoid.

The exoccipitals form the posteromedial part of the otic capsule, as well as the margin of

the foramen magnum and the occipital condyles. The hyal of the hyoid attaches to the

posterolateral margin of the exoccipital, and possibly to the posterior margin of the basal

process, via a small ligament (or other unstained connective tissue). The margin of the

foramen magnum is incompletely ossified dorsomedially and dorsoventrally (fig. 1b). The

occipital condyles, lateral to the foramen magnum, are well developed. Internal and slightly

lateral to the occipital condyles are the jugular foramina.

Plectral apparatus

The plectral apparatus is ventral to the crista parotica, oriented horizontally (fig. 1b-c).

The fully ossified pars interna plectri is expanded but separate from the fenestra ovalis and

operculum. The operculum is robust and completely ossified, except for the posterolateral

margin. The pars media plectri is columnar, slightly sigmoidal, and expanded medially to

articulate with the pars interna plectri. Distally, the pars externa plectri forms a flat carti-

laginous plate that fills about one-third of the tympanic annulus. A well-developed pars

ascendens plectri extends from the medial portion of the pars externa plectri to the crista

parotica. The tympanic annulus attaches dorsally to the cartilaginous tip of the crista

parotica, and except for a slight separation at this articulation, forms a complete ring.

Nasals

The rhomboidal, paired nasals overlie the nasal capsule (fig. la). Medially, they overlap

the septum nasi of the sphenethmoid, although this element is clearly visible between them.

Posteriorly, the nasals overlap the planum antorbitale, but do not articulate with the fronto-

parietals. Laterally, the poorly-developed maxillary process of each nasal narrows to extend

to the level of the pars facialis of the maxilla, but does not articulate with it.

Frontoparietals

These paired, dorsal elements form the lateral and posterior margins of the frontoparie-

tal fenestra (fig. la). Anteriorly, they invest the sphenethmoid to the level of the tectum

anterior; anterolaterally, each narrows away from the anterior margin of the fenestra and

lacks a supraorbital flange. Laterally, each forms the lamina perpendicularis, which extends

Source : MNHN, Paris

MAGLIA 153

ventrally about one third of the height of the braincase and posteriorly to the anterior margin

of the optic foramen (fig. 1c). Posterodorsally, these elements overlap the prootic to the

eminentia epiotica. À narrow ridge, the occipital crest, forms anterior to the eminentia

epiotica. Anterior to this crest, the occipital foramen opens posteriorly. Although completely

covered, the occipital canal is visible through the bone, traversing obliquely from the lateral

margin of the frontoparietal to open at the level of the posterior margin of the frontoparietal

fenestra. In some specimens, a smaller foramen opens dorsally at the midpoint of the occipital

canal.

Parasphenoid

The parasphenoid is broad, smooth, and lacks bony ornamentation. The anterior half of

the broad cultriform process overlaps the sphenethmoid, and narrows to a point just posterior

to the level of the planum antorbitale (fig. 1b). The parasphenoid alae are broad, anterolat-

erally oriented, and ventrally invest the otic capsule. A distinct posteromedial process is

absent; however, the posterior margins of the alae converge to form the posteromedial

margin. This part underlies the ventral cartilaginous margin of the foramen magnum

(between the exoccipitals).

Vomers

The vomers are large, bear about five teeth each, and contribute to the floor of the nasal

capsules (fig. 1b). The anterior process is rectangular and extends obliquely from its anterior

margin just posterior to the maxilla-premaxilla articulation toward the midline of the body.

At the level of the dentigerous process, a small prechoanal process extends laterally, Medial to

this process is a small foramen for the palatine ramus of the facial nerve. The dentigerous

process is rounded, and narrowly separated from its counterpart. The postchoanal process is

long and slender, and invests the planum antorbitale. This process extends beyond the planum

antorbitale to articulate at its most dorsolateral end with the anterior ramus of the pterygoid

via the posterior maxillary process of the planum antorbitale and may articulate weakly with

the pars facialis of the maxilla.

Premaxillae

The premaxillae are narrowly separated from one another; each has a well-developed

alary process that is inclined anteriorly, curved slightly laterally, and ends dorsally in a

bifurcated, rounded tip. The pars dentalis curves dorsolaterally; its anteroventral surface

appears wavy because of the presence of approximately 12 teeth (fig. 1b). The palatine process

of the pars palatina is a short, flat plate that forms a right triangle. À small posterolateral

process of the pars palatina also is present.

Maxillae

Each maxilla possesses approximately 36 teeth and lacks pre- and postorbital processes.

The pars facialis of each is well developed and reaches its maximum height at the level of

Source : MNHN, Paris

154 ALYTES 16 (3-4)

Pan "se

Meckel's c mener en

Meckel's c

Fig. 3.- Mandible of Spea multiplicara (KU 86664) in dorsal (left), lateral (right, top), and lingual (right,

bottom) view. Gray denotes cartilage. Abbreviations: angulospl, angulosplenial; c, cartilage.

the planum antorbitale (fig. 1c). The anterior tip of the pars facialis articulates with the lateral

margin of the premaxilla and the medial margin articulates with the planum antorbitale of the

sphenethmoid. A small foramen, possibly a foramen for a ramus of the maxillaris superior

vessels, traverses vertically through the pars facialis.

Mandible

The angulosplenial, dentary, and mentomeckelian bones comprise the mandible (fig. 3).

The mentomeckelian bones form the anterior margin of the mandibles; they are small and

relatively well ossified, and are fused to one another medially. The thin dentary articulates

with the posterior portion of the mentomeckelian, and extends posteriorly for more than half

the length of the mandible, investing the lateral margin of Meckel's cartilage. The angulo-

splenial forms the posterior portion of the mandible and serves as the attachment point for

the mandible to the cranium. The angulosplenial extends anteriorly to invest most of the

lingual margin of Meckel’s cartilage. Posteromedially, the angulosplenial possesses a well-

developed coronoid process.

Squamosals

The zygomatic ramus of the squamosal is short and projects anteriorly (fig. 1e). The otic

ramus of the squamosal invests the anterolateral tip of the crista parotica. The ventral ramus

extends posteroventrally at a 45° angle relative to the horizontal axis of the skull and invests

the ossified portion of the palatoquadrate cartilage. A thin, sheetlike process extends antero-

medially from the ventral ramus, ventral to the zygomatic ramus, and invests the palatoqua-

drate cartilage.

The triradiate pterygoids are well developed, with robust anterior and medial rami

(fig. la-b). The anterior ramus projects anterodorsally, invests the pterygoid process of the

palatoquadrate, and articulates with the pars palatina of the maxilla. The anterior ramus

Source : MNHN, Paris

MAGLIA 155

synchondrotically fuses to the lateral margin of the postchoanal process of the vomers. The

posterior ramus invests the ventrolateral surface of the pars articularis of the palatoquadrate.

The medial ramus invests the pterygoid process of the palatoquadrate and articulates with the

basal process.

Palatoquadrates

The pars articularis of the palatoquadrate (quadrate process) is ossified to the level of

midheight of the ventral process of the squamosal. The basal process extends medially to

articulate with the prootic, and is invested by the medial ramus of the pterygoid (fig. 1c).

HYOID APPARATUS

Hyoid apparatus

There is little sexual dimorphism in the hyoid apparatus. The hyoid plate shows no

mineralization and is narrow; the length along the longitudinal axis (midlength) is about

two-thirds the length along the transverse axis (fig. 4). The hyoglossal sinus is U-shaped.

Separate anterolateral processes are not present; they are fused to the hyoid plate in devel-

opment, creating oval lateral foramina (WiENs, 1989), which are larger in males. As in other

pelobatoids (CANNATELLA, 1985), the hyals are disassociated from the hyoid plate, with each

ventrally investing the lateral margin of the hyoid plate, posterior to the lateral foramen,

narrowing posterolaterally, and extending forward to articulate with the exoccipital (or basal

process of the palatoquadrate; see Exoccipitals above).

The slender posterolateral processes project from the posterior margin of the hyoid plate

at approximately a 45° angle to the transverse axis of the hyoid plate. These processes are

about equal in length to the midlength of the hyoid plate. The ossified posteromedial

processes project posterolaterally from the posteromedial margin of the hyoid plate at

approximately a 60° angle to the transverse axis of the hyoid plate. In males, the shaft of each

posteromedial process is one-third the width of the proximal and distal heads; in females, the

shaft is half the width of either head.

Laryngeal cartilages

There is sexual dimorphism in both the size and shape of the laryngeal cartilages. In

males, the laryngeal apparatus nearly fills the entire space between the posteromedial process-

es; in females, only half of this space is filled. In ventral view, the paired arytenoid cartilages,

which are much larger in males, lie within the cricoid ring. As each of these cartilages extends

dorsomedially, it becomes more narrow and less concave, and appears to form discrete dorsal

and ventral parts (fig. 4). In males, the dorsal portion extends almost the full length of the

ventral portion. In females, the dorsal portion is only half the length of the ventral part, and

the anterodorsal margin is acuminate. The elongate, paired bronchial processes project

ventrolaterally from the cricoid ring at the level of the distal heads of the posteromedial

processes. The distal portion of each bronchial process terminates in a head with three

Source : MNHN, Paris

156 ALYTES 16 (3-4)

hyoglossal sinus

lat foram

hyal

hyoid plate.

posterolateral p

posteromedial p

EQ arytenoid c = FN ]

À Pa

— bronchial p cricoid ring A

esophageal p

Fig. 4. — Ventral view of hyoid apparatus (top) and lateral view of cricoid ring (bottom) of male (left:

KU 86664) and female (right; KU 86662) Spea multiplicata. Stippled pattern denotes bone, gray

denotes cartilage. Abbreviations: c, cartilage; lat foram, lateral foramen; p, process.

fingerlike projections. In males, the bronchial processes extend to the level of the posterior

margin of the arytenoid cartilages; in females, these processes extend to the level of the

posterior margin of the cricoid ring. Slightly posterior to the origin of the bronchial processes,

shelf-like expansions extend medially from the cricoid ring. In males, a small square esopha-

geal process extends ventrally from the posterior margin of the cricoid ring; in females, this

process is less distinct.

AXIAL SKELETON

The vertebral column is composed of eight notochordal presacral vertebrae, the sacrum

and the urostyle (fig. 5a). The vertebrae are slightly imbricate, and ossified intervertebral

bodies are present between the centra. Each neural arch bears a low neural ridge with two

small, posterior projecting spinous processes; the articular facets of the pre- and postzyga-

pophyses are simple. The relative lengths of transverse processes and sacral diapophyses are as

follows: III > sacrum = IV >> V=VI=VII= VIIL. Transverse processes of presacrals II-V

are almost perpendicular to the notochordal axis, whereas those of presacrals II, VI, VIT and

VIII are directed anteriorly. Small, posteriorly directed uncinate processes are present on the

transverse processes of vertebrae II-IV.

The cervical cotyles of the atlas are Type II (LYNCH, 1973) and are nearly contiguous.

The urostyle is rounded in cross section, fuses with the sacrum, and bears a dorsal ridge

Source : MNHN, Paris

omosternum

cleithrum

clavicle

suprascap coracoid

sternum

sacral diap C £

urostyle

Fig. 5. (a) Dorsal view of axial skeleton of Spea multiplicata (KU 86664). (b) Ventral view of pectoral girdle (KU 86662), with the scapula

and suprascapula deflected ventrally. (c) Ventral (left) and lateral view of pelvic girdle (KU 86664). Gray denotes cartilage. Abbrevia- S

tions: c, cartilage; f, fossa; procor, procoracoid; sacral diap, sacral diapophysis; suprascap, suprascapula.

Source : MNHN, Paris

158 ALYTES 16 (3-4)

throughout its anterior two thirds. The sacrum consists of vertebrae IX and X, and the

slightly expanded sacral diapophyses (expanded transverse processes of vertebrae IX; WIENS,

1989) are oriented perpendicular to the midline of the body. À bony webbing, which has been

mistaken for postsacral transverse processes (discussed by WIENS, 1989), is present on the

posterior margin of the sacrum, between vertebrae IX and X.

APPENDICULAR SKELETON

Pectoral girdle

The sternum is a spade-shaped plate of cartilage that floats between the epicoracoid

cartilages, typical of the arciferal arrangement of the girdle (fig. 5b). À completely cartilagi-

nous, knob-shaped omosternum articulates with the epicoracoid bridge of the epicoracoid

cartilages. The anterior margins of the paired procoracoid cartilages are completely invested

by the clavicles and are synchondrotically contiguous posteromedially with the epicoracoid

cartilages. The pectoral fenestrae are large and tear-shaped; each is bordered anteriorly by the

procoracoid cartilage, medially by the epicoracoid cartilage, posteriorly by the coracoid, and

laterally by the glenoid fossa.

The relatively long clavicles (one-third longer than the coracoids) are posteriorly

concave; the glenoidal end of each is flared anteriorly, forming a wedge-shaped process that

abuts the pars acromialis of the scapula. The clavicles do not reach the midline and are

separated medially by the epicoracoid bridge. The long axes of the coracoids are slightly

arcuate; each of these robust bones is narrowly separated anterolaterally from the clavicle and

articulates with the pars glenoidalis of the scapula. The sternal end of each coracoid is

moderately broad (twice the width of the shaft), but narrower than the glenoidal end

(approximately 80 % of width of glenoidal end). The scapular end of the coracoid is also

broad (almost three times the width of the shaft) and its distal concavity articulates with the

pars acromialis, forming the posterior surface of the relatively deep glenoid fossa.

The scapula is about three times the length of the glenoid fossa, with its greatest width

being half of its total length. The pars glenoidalis is a thin, concave plate, and the pars

acromialis is a robust, rounded process; both form the remaining portion of the glenoid fossa.

The shaft of the scapula is short and constricted (width about one-fourth total length of

scapula). The distal head of the scapula is expanded to articulate with the cleithrum; its width

is twice the width of the shaft and half the total length of the scapula. The cleithrum invests

most of the anterior two-thirds of the suprascapular cartilage. It is narrow anterodistally and

broadens at the scapular end to form the shape of a cleaver. The suprascapular cartilage

extends posteriorly as a broad, flat blade.

Forelimb

The humerus has a large, flangelike crista ventralis, a slightly smaller, well-developed

crista medialis, and a low crista lateralis. The glenoid head of the humerus (caput humeri) is

cartilaginous, whereas the distal head (eminentia capitata) is completely ossified. The flat-

tened radioulna is about two-thirds the length of the humerus and its distal head is wider than

Source : MNHN, Paris

MAGLIA 159

radius

sulc int

capit-

ulnare

olecranon

Fig. 6. — Right forelimb of Spea multiplicata (KU 86664); top of figure is anterior. (a) Dorsal view of

radioulna (left) and lateral view of humerus. (b) Ventral view of hand, Gray denotes cartilage.

Abbreviatio: apit, capitulum; cr lat, crista lateralis; cr med, crista medialis; cris vent, crista

ventralis; dist prep, distal prepollex; sulc int, sulcus intermedius; u, ulnare; Y, element Y.

its proximal head (fig. 6a). A distinct groove, the sulcus intermedius, distinguishes the radius

and ulna, although they are fused to one another medially. A small flange is present along the

proximolateral edge of the ulna.

The manus resembles that of Scaphiopus couchit (morphology A) as described by

FABREZ1 (1992). Proximally, there is a small ulnare, and a slightly larger radiale; distally, there

is a large irregular-shaped element Y and a smaller carpal 5 (fig. 6b). Carpals 2, 3 and 4 lie

proximal to metacarpals IT, II and IV, respectively. Carpals 3 and 4 are partially fused to one

another, and lie on the ventral surface of the manus: carpal 2 is smaller, and lies dorsal to

element Y. AIl carpal elements are well ossified. À moderate-sized, ossified prepollex lies

distomedial to element Y. One completely ossified distal prepollical element and at least one

cartilaginous distal prepollex also are present. Relative lengths of the digits are IV > II > HIT

= V. The phalangeal formula is 2-2-3-3. There is apparent sexual dimorphism in the size and

shape of digit IT; in male specimens, the metacarpal and phalanges are thickened, with a small

protuberance on the medial border of the metacarpal.

Source : MNHN, Paris

160 ALYTES 16 (3-4)

Pelvic girdle

In dorsal view, the internal margins of the ilia form a narrow U-shape (fig. 5c). The ilial

shafts are simple, with no obvious crests, but have a small dorsal prominence. The preaceta-

bulum is moderate and the preacetabular angle (i.e., the angle between the ilial shaft and the

preacetabular margin) is slightly obtuse. The ilia are separated from one another medially and

from the ischia posteriorly by cartilage. The ischia are approximately D-shaped, and are fused

to one another to form the posterior margin of the acetabulum. The completely cartilaginous

pubis forms the ventral border of the acetabulum.

Hind limb

The femur is long and thin (length about 12 times width); both the distal and acetabular

heads are cartilaginous. The femur possesses a small ridge on the lateral margin. The

tibiofibula is about three-fourths the length of the femur, and its distal and proximal heads are

of similar size. Although the tibia and fibula are fused, a distinct groove separates them. The

tibiale and fibulare are short and robust (length less than half that of the tibiofibula), and are

fused to one another at their proximal and distal heads (fig. 7a).

The pes has a single ossified tarsal element proximal to digit II, and a large element Y

(FABREZI, 1993) proximal to metatarsal I (fig. 7b). An ossified prehallux and a large spadelike

distal prehallical element are present medial to element Y. Relative lengths of the digits are

IV>V>IN > I > I. The phalangeal formula is 2-2-3-3-3.

DISCUSSION

Although the anatomy of frogs in the genus Pelobates is relatively well known, the

phylogenetic relationships within the family Pelobatidae are unresolved (FORD & CANNA-

TELLA, 1993), in part because of a lack of detailed morphological descriptions of the other

members of the family, Spea and Scaphiopus. The description provided herein should

facilitate a more detailed comparison among pelobatid taxa. What follows is both a summary

of the most recent works on pelobatid osteology as well as my own observations. The

preliminary comparisons of Spea multiplicata to all other pelobatid taxa presented here were

incorporated as part of a phylogenetic analysis of the pelobatids (MAGLIA, 1998).

Most recent authors (e.g., FORD & CANNATELLA, 1993; DUELLMAN & TRUEB, 1994) agree

that pelobatids (Pelobates, Scaphiopus and Spea) form a monophyletic assemblage. However,

although there are several diagnostic characters for these frogs (including broad sacral

diapophyses and sculpturing of dermal cranial bones: ROCEK, 1981), few osteological features

have been proposed to be shared derived characters uniting Pelobates, Scaphiopus and Spea.

CANNATELLA (1985) proposed that the presence of cranial exostosis and a long zygomatic

ramus of the squamosal were synapomorphies for the pelobatids; however, both of these

features are absent in the genus Spea. He also cited the presence of a supraorbital flange of the

frontoparietal in all pelobatids; however, I have not seen evidence of this structure in any

Source : MNHN, Paris

MAGLIA 161

tibia Îl

fibula à

2mm

distal prehallux

Léa {| prehallux

Fig. 7. Right hind limb of Spea multiplicata (KU 86664). (a) Lateral view of femur (left) and ventral

view of tibiofbula. (b) Ventral view of right foot. Gray denotes cartilage. Abbreviation: Y,

element Y.

Spea. The other osteological characters uniting the pelobatids in CANNATELLA’S (1985)

analysis, the complete ossification of septum nasi and the fused articulation of the urostyle

and sacrum, are found in several non-pelobatid taxa.

In comparing the osteology of Spea multiplicata With that for all other pelobatid taxa, I

found several features shared by the pelobatids. AII of these taxa possess an occipital canal

that is roofed completely by bone. This feature is not present in non-pelobatid pelobatoids

(e.g., Megophrys, Pelodytes); however, it is present in other taxa (e.g., some neobatrachians:

LynCH, 1969; MENDELSON et al., in press). The presence of bony webbing on the posterior

margin of the sacral diapophyses is shared among all pelobatids. This has been identified by

some authors (e.g., LYNCH, 1973; DUELLMAN & TRUEB, 1994) as post-sacral transverse

processes, but was shown to originate in development from the sacral diapophyses (WIENS,

Source : MNHN, Paris

162 ALYTES 16 (3-4)

1989). Other morphologies shared by all pelobatids are the presence of relatively elongate,

convex clavicles and well-developed facial and preorbital processes of the maxilla.

The North American pelobatids Spea and Scaphiopus have several morphological

features that are unique to them, including the lack of a quadratojugal bone and the presence

of a palatine process of the facial process of the maxilla (CANNATELLA, 1985). These taxa also

possess a postchoanal process of the vomer that subtends the planum antorbitale (discussed

in CANNATELLA, 1985) and a completely cartilaginous sternum. ROEK (1981: 151) provided a

detailed comparison of the cranial differences between Pelobates and the North American

pelobatids, and included a discussion of several features common to Scaphiopus and Spea

(e.g., well-developed stapes, ossified operculum).

Several morphologies are unique to the genus Spea. For example, Spea lacks the exostosis

of the dermal cranial and suspensorium elements found in all other pelobatids. Also, the otic

ramus of the squamosal barely overlaps the crista parotica, whereas it forms an otic plate

investing nearly half the otic capsule in other pelobatids. Also in Spea, the frontoparietals do

not come into contact with the nasals; they lack supraorbital flanges, and they are in contact

posteromedially only, exposing the frontoparietal fontanelle. These features most likely relate

to the degree of ossification of the cranium of Spea: these frogs are much less ossified than

other pelobatids. If Scaphiopus and Spea share a most common ancestor, which seems to be

of little doubt (FORD & CANNATELLA, 1993; DUELLMAN & TRUEB, 1994), and if the clade

LSpea + Scaphiopus] is the sister group to Pelobates (also well supported; CANNATELLA, 1985),

then the limited ossification and small body size of Spea may be a reversal of the hyperossi-

fication present in Pelobates and Scaphiopus. However, it is just as likely that the common

ancestor shared by the Pelobatidae resembled Spea in amount of ossification, and that the

hyperossification present in Pelobates and Scaphiopus evolved separately in these taxa.

Morphologies are thought to be highly conserved among species of Spea, and primarily

one morphological feature, the frontoparietal boss, has been the subject of much discussion

CWiens & Tirus, 1991). Therefore, the only major works attempting to analyze the rela-

tionships within the genus Spea have relied on biochemical data (e.g., SAGE et al., 1982; WiENs

& Titus, 1991). However, comparing S. multiplicata to other members of the genus, I found

several features that vary to some degree among these frogs. For example, the maxillary

process of the nasal is poorly developed in S. multiplicata and S. bombifrons, but is more

extensive in the other taxa. Similarly, S. multiplicata and S. bombifrons lack a well-developed

posteromedial process of the parasphenoïd, whereas the other taxa possess this feature.

A small but striking anatomical feature present in Spea multiplicata is a well-developed

pars ascendens plectri of the auditory apparatus. Because this feature can only be observed on

cleared and double-stained specimens with well-developed plectral apparatuses, I was able

only to compare it among a few taxa in this sample (S. bombifrons, Scaphiopus couchii and

Pelobates varaldii). Of these, the pars ascendens plectri was only present in S. bombifrons

(although not described by WiEns, 1989). This structure may be unique to Spea, or may vary

among pelobatids. It is hoped that further comparisons of this feature and others discussed

here will help in resolving the relationships among the pelobatids.

Source : MNHN, Paris

MAGLIA 163

RESUMEN

Entre los pelobätidos (Anura, Pelobatidae), la morfologia esqueletal de los géneros

norteamericanos Spea y Scaphiopus es pobremente conocida. La osteologia de Spea multipli-

cata se describe en base a esqueletos secos y a especimenes diafanizados y doblemente teñidos,

y se la compara con todos los otros taxones de pelobâtidos (Spea, Scaphiopus, Pelobates).

Spea y Scaphiopus comparten varias morfologias, incluyendo la ausencia de cuadradojugal, la

presencia de un proceso palatino del proceso facial de la maxila, proceso postcoanal del

vomer largo, y esternôn completamente cartilaginoso. Spea se caracteriza por un proceso

maxilar del nasal pobremente desarrollado, falta de un proceso posteromedial de parasfenoi-

des bien desarrollado y posiblemente una pars ascendens plectri del aparato auditivo bien

desarrollada. La mayoria del resto de los caraceres diagnôsticos de Spea estän relacionados a

la limitada osificacién craneal de este género en relacién a otros miembros de la familia.

ACKNOWLEDGEMENTS

T'thank Linda TRUEB, L. Analia PÜGENER, Hélio R. DA SILVA and Christopher SHrIL, Alexander

Haas, and Borja SANCHIZ for their comments on this work. I also thank L. Analia PÜGENER for providing

the Spanish translation of the abstract. I am grateful to Charles MYERS at the American Museum of

Natural History, José RosADoO at the Harvard Museum of Comparative Zoology, and José E. GONZALEZ

from the Museo Nacional de Ciencias Naturales (Madrid) for the loan of specimens. This research was

supported by NSF Grant DEB 95-21691 to Linda TRUEB.

LITERATURE CITED

ANDERSEN, M. L., 1978. - The comparative myology and osteology of the carpus and tarsus of selected

anurans. PhD Thesis, Lawrence, The University of Kansas: 1-302.

CaNNATELLA, D. C., 1985.- À phylogeny of primitive frogs (Archaeobatrachians). PhD Thesis, Lawrence,

The University of Kansas: 1-404.

DUELLMAN, WE. & TRUEB, L., 1994. Biology of amphibians. Second edition. Baltimore, Johns Hopkins

University Press: i-xix + 1-670.

FABrEzi, M., 1992. — El carpo de los anuros. A/ytes, 10: 1-29.

= 1993. — The anuran tarsus. 4/ 11: 47-63.

FaBrezi, M. & ALBERCH, P., 1996. — The carpal elements of anurans. Herpetologica, 52: 188-204.

Forp, L. S. & CANNATELLA, D. C., 1993. - The major clades of frogs. Herp. Mon., 7: 94-117.

Frosr, D. R., (ed.), 1985. - Amphibian species of the world. Lawrence, Allen Press & Assoc. Syst. Coll.:

Liv] + iv + 1-732.

NS, J. D., 1971. — The morphology of the nasal region of the Amphibia and its bearing on the

phylogeny of the group. Ann. Univ. Stellenbosch, 46: 1-146.

Lynou, J. D. 1969. - Evolutionary relationships and osteology of the frog family Leptodactylidae. PhD

Thesis, Lawrence, The University of Kansas: 1-800.

= 1973. - The transition from archaic to advanced frogs. /n: J. L. ViaL (ed.), Evolutionary biology of the

anurans: contemporary research on major problems, Columbia, Univ. of Missouri Press: 133-182.

MaGLia, A. M., 1998.- Phylogenetic relationships of the pelobatid frogs (Anura, Pelobatoidea): skeletal

evidence. Sci. Pap. nat. Hist Mus. Univ Kansas, 10: 1-19.

Source : MNHN, Paris

164 ALYTES 16 (3-4)

MENDELSON, LR, III, Da SILVA, H. R. & MaGLIA, À. M. in press. — Phylogenetic relationships among

genera of marsupial frogs (Anura: Hylidae: Hemiphractinae), based on evidence from morphol-

ogy: reassessment of character and life history evolution. Zool. J. Linn. Soc, in press.

RAMASWAMI, L. $., 1939. - The cranial morphology of some examples of Pelobatidae (Anura). Anat.

Anz., 81: 65-96

Retss, J. O., 1997. — Early development of chondrocranium in the tailed frog Ascaphus truei (Amphibia:

Anura): implications for anuran palatoquadrate homologies. J. Morph., 231: 63-100.

Roërk, Z., 1981. - Cranial anatomy of frogs in the family Pelobatidae Stannius, 1856, with outlines of

their phylogeny and systematics. Acta Univ. Carol. Biol., 1980: 1-164.

RODRIGUEZ TALAVERA, MR, 1990. — Evoluciôn de peloditidos: morphologia y desarollo del sistema

esquelético. PhD Thesis, Madrid, Universidad Complutense de Madrid: 1-282.

S4, R. O. Dr & TRurB, L., 1991. — Osteology, skeletal development, and chondrocranial structure of

Hamptophryne boliviana (Anura: Microhylidae). J. Morph., 209: 311-330.

SAGE, R. D., PRAGER, E. M. & Wake, D. B., 1982. - A Cretaceous divergence time between pelobatid frogs

(Pelobates and Scaphiopus): immunological studies of albumin serum. J. Zool., London, 198:

481-494.

Tru, L., 1993. Patterns of cranial diversity among the Lissamphibia. In: J. HANKEN & B. K. HALL

(ed.), The skull, 2, Patterns of structural and systematie diversity, Chicago, The University of

Chicago Press: 255-341.

Wans, J.J., 1989. - Ontogeny of the skeleton of Spea bombifrons (Anura: Pelobatidae). J. Morph., 202:

29-51.

Was, J J. & Tirus, T. A., 1991. - A phylogenetic analysis of Spea (Anura: Pelobatidae). Herpetologica,

47: 21-28.

APPENDIX 1

SPECIMENS EXAMINED

INSTITUTIONS

AMNH: American Museum of Natural History, New York, USA.

KU: The University of Kansas, Lawrence, USA.

MCZ: Harvard University Museum of Comparative Zoology, Cambridge, USA.

MNCN: Museo Nacional de Ciencias Naturales, Madrid, Spain.

SPECIMENS EXAMINED

Pelobates cultripes: KU 148619, MNCN 20041.

Pelobates fuscus: KU 68819, 129240.

Pelobates syriacus: KU 146856.

Pelobates varaldii: AMNH 62935, MCZ 31970

Scaphiopus couchi: KU 20444, 73384, 209575.

Scaphiopus holbrookii: KU 20439, 145413.

Scaphiopus hurterit: KU 20472, 60173, 90096.

Spea bombifrons: KU 5405, 73382.

Spea hammondii: KÜ 176016.

Spea intermontana: KU 79436, 204563

Spea multiplicata: KU 27622, 39776A-B, 49468, 84888, 86662, 86664, 97355, 106225, 117347.

Corresponding editor: Masafumi MATSUI.

Source : MNHN, Paris

Alytes, 1999, 16 (3-4): 165-189. 165

Geographic variation of Hyla rubicundula

and Hvla anataliasiasi,

with the description of a new species

(Anura, Hylidae)

Marcelo Felgueiras NAPOLI & Ulisses CARAMASCHI

Departamento de Vertebrados, Museu Nacional do Rio de Janeiro,

Quinta da Boa Vista, 20940-040, Rio de Janeiro, RJ, Brasil

E-mail: napoli@acd.ufrj. br

Analyses ofintra- and interpopulation variation of the external morphol-

ogy of Hyla rubicundula Reinhardt & Lütken, 1862 and Hyla anatalia-

siasi Bokermann, 1972 indicate that four morphospecies are represented.

Hyla rubicundula comprises three of the four morphospecies. Its northern

morphospecies is described as a new species characterized by an immacu-

late dorsum and a pointed snout. Redescriptions of H. rubicundula and

H. anataliasiasi are provided.

INTROD UCTION

The species currently included in the Hyla rubicundula group share the following charac-

teristics: small size (SVL: males 16.0-25.5 mm, females 16.6-25.9 mm), thighs immaculate,

dorsum consistently green in life, and dorsal surfaces pink to violet in preservative. This group

occurs in northern, central, northeastern and southeastern Brazil (FROST, 1985), in open

habitats, mainly in “cerrado” formations, but also in transitional areas between cerrado and

rainforests.

According to BOKERMANN (1968) and FROST (1985), the Hyla rubicundula group is

composed of Hyla rubicundula Reinhardt & Lütken, 1862, Hyla tritaeniata Bokermann, 1965

and Hyla anataliasiasi Bokermann, 1972. Hyla elongata À. Lutz, 1925 was synonymized with

H. rubicundula by BOKERMANN (1968) but treated as a valid species by HADDAD et al. (1988);

the latter authors compared vocalizations of specimens from Serra da Canastra, Minas

Gerais, with the vocalizations of topotypic populations of H. rubicundula described by

CaRDOSO & VIELLIARD (1985), and considered A. rubicundula and H. elongata as distinct

species. However, our examination of the external morphology of the specimens from Serra

da Canastra revealed that they must be associated to the A. tritaeniata complex, and were

wrongly identified as H. elongata by HADDAP et al. (1988). Thus, the synonymization of H.

elongata With H. rubicundula proposed by BOKERMANN (1968) is valid.

Source : MNHN, Paris

166 ALYTES 16 (3-4)

Hyla tritaeniata, originally included in the A. rubicundula group, is not treated in this

paper because it has (1) a distinctive dorsal pattern (a single sacral stripe, instead of two in the

other species) and (2) different habitat preferences: this species is found in springs and streams,

whereas the rest of the group inhabits permanent or temporary ponds (BOKERMANN, 1965;

Jim, 1980). Also, (3) the large intra- and interpopulation variations of H. tritaeniata suggest a

species complex that must be analyzed separately.

The purposes of this paper are (1) to study the degree of intra- and interpopulation

variation in A. rubicundula and H. anataliasiasi, and (2) to describe a new species of the A.

rubicundula species group.

MATERIAL AND METHODS

Specimens used for description or examined for comparisons were previously deposited

in the collections of the Museu Nacional, Rio de Janeiro (MNRIJ), of the Museu de Zoologia,

Universidade de Säo Paulo (MZUSP), of the Naturhistorisches Museums, Vienna (NMW),

of the Werner C. A. BOKERMANN collection, deposited in the Museu de Zoologia, Universi-

dade de Säo Paulo, SP, Brazil (WCAB), of the Kobenhavns Universitet, Zoologisk Museum,

Copenhagen (ZMUC), and of the Museu de Historia Natural, Universidade Estadual de

Campinas (ZUEC). The analysis of the material was similar to that used by VANZOLINI (1970)

and HEYER (1984). Initially, large samples from each locality were analyzed (“basic samples”)

to determine the patterns of variation within samples. Specimens were sorted into morpho-

species (i.e., categories thought to represent different species). Subsequently, samples from

poorly represented localities were analyzed (“small samples”), and these specimens, when

possible, were associated to a morphospecies by similar morphology and proximity among

localities. The last step of the analysis corresponds to a careful examination of the patterns of

variation among morphospecies.

Only adult males were examined because females and juveniles were rare in the samples.

We developed a series of standards for the general dorsal pattern, mid-dorsal pin stripe,

dorsolateral stripes, lateral limits of dorsum, upper surface of tibia, loreal and canthal stripes,

and dorsal head shape (fig. 1-3). Nine measurements (mm) were taken following DUELLMAN

(1970): SVL (snout-vent length), HL (head length), HW (head width), ED (eye diameter),

UEW (upper eyelid width), IOD (interorbital distance), IND (internarial distance), TD

(tympanum diameter) and TL (tibia length). Four measurements were made following HEYER

et al. (1990): UAR (upper arm), FAR (forearm), HAL (hand length) and THL (thigh length).

Five other measurements were END (eye to nostril distance: straight line distance between

anterior corner of orbital opening and posterior margin of external nare), NSD (nostril to tip

of snout distance: straight line distance between anterior corner of nostril to tip of snout), FL

(foot length: distance from heel to tip of fourth toe), 3FD (third finger disk diameter: greatest

horizontal distance between outer edges of third finger disk) and 4TD (fourth toe disk

diameter: greatest horizontal distance between outer edges of fourth toe disk). Webbing

formula notations followed SAVAGE & HEYER (1967).

Discriminant function analyses compared inter- and intra-morphospecies variation

(Marcus, 1990) without removing the size effect in the groups (Reis et al., 1990), and groups

Source : MNHN, Paris

NAPoLI & CARAMASCHI 167

Fig. 1.- Standards for dorsal and mid-dorsal pin stripe patterns. Patterns A3 (dorsum immaculate), A7

(one to few dots distributed irregularly) and B3 (absence of mid-dorsal pin stripe) are not figured.

Source : MNHN, Paris

168 ALYTES 16 (3-4)

Fig. 2. — Standards for dorsolateral stripes: C1-C2, thin and regular; C3-C4, thick and irregular; C5,

vestigial; C6, absent, is not figured; C7, thick and well marked. Lateral limits of dorsum: DI, above

the tympanum; D2, under the lower border of tympanum. Upper surface of tibia patterns: El,

white stripe over dark stripe; E2, white stripe absent; E3, white and dark stripes vestigial or absent:

E4, presence of a mid-dorsal pin stripe. Loreal and canthal stripes patterns: F1, thin white stripe

over dark stripe; F2-F3, thick clear band over dark stripe.

were defined a priori. Eigenvectors and associated eigenvalues were obtained from a variance-

covariance matrix, and the loadings were the correlations between the original variables and

the scores. We used /-tests to compare mean values from different measurement variables of

the same species. For character analyses, we used the chi-square test (4?) to compare patterns

among samples of the same morphospecies (SOKAL & ROHLF, 1981).

Vocalizations were recorded by Rogério P. BasTos with a Uher Report Monitor and a

Uher M 518 A microphone at a tape speed of 19 cm/s. Tapes were analyzed on a Macintosh

Classic coupled to a MacRecord Sound System 2.0.5.

Source : MNHN, Paris

NAPOLI & CARAMASCHI 169

VCII (5.92 %)

@, (85.29 %)

œs, vi (

Fig. 3.- Standards for the dorsal head shape patterns (G1-G7), and projection of centroids resulted from

the multiple discriminant function analysis for 18 morphometric characters of the combined

samples of morphospecies RU, PRU, CBO and ANA, in the first three canonical axes, À minimum

spanning tree connects the closest means, and the Mahalanobis distance is given for each link of

the tree; this procedure corrects the distortion caused by the three-dimensional projection

Source : MNHN, Paris

170 ALYTES 16 (3-4)

RESULTS AND DISCUSSION

MORPHOSPECIES

The four morphospecies were named and coded as follows (code, code name, number of

specimens analyzed, localities):

RU, Hyla rubicundula, n = 144. BAHIA: Barreiras and Jupaguä. MiNAS GERAIS: Alfenas,

Andrequicé, Arinos, Baräo de Cocais, Belo Horizonte, Buritis, Buritizeiro, Esmeraldas,

Jaboticatubas, Januäria, Lagoa Formosa, Lagoa Santa, Manga, Pirapora, Três Marias, Unai

and Vespasiano. Gols: Cristalina.

PRU, Hyla “pseudorubicundula”, n = 54. MiNas GERAIS: Uberlândia. GoÂs: Aragarças,

Cavalcante, Goïânia, laciara, Monte Alegre de Goiäs, Nova Roma, Porangatu, Santa Rita do

Araguaia, Säo Domingos and escarpa da Serra Dourada. PrAU‘: Uruçui.

CBO, “Cachimbo”, n = 15. PARA: Cachimbo.

ANA, Hyla anataliasiasi, n = 85. MATO Grosso: Posto Leonardo and Posto Diauarum.

COMPARISONS AMONG MORPHOSPECIES

Results from the analysis of the seven coloration patterns indicate two categories of

characters (tab. 1). In the first category, frequencies of character states differed among

morphospecies, but no states (e.g., mid-dorsal pin stripe or loreal and canthal stripes patterns)

were diagnostic. The second category was defined by states unique to certain morphospecies,

and specimens having such unique states were easily diagnosed from the other morphospecies

(e.g., any specimen that presented pattern A11 for general dorsal pattern was automatically

assigned to morphospecies ANA). General dorsal patterns, dorsolateral stripes, lateral limits

of dorsum, upper surface of tibia, and dorsal head shape patterns belonged to this category.

Taken in combination, pattern characteristics distinguished most but not all individuals of

the four morphospecies; that is, a specimen that had only character states common to all

morphospecies was not assigned to one of them.

MEASUREMENT VARIABLES

Multiple discriminant function analysis was used to analyze morphological variation

among the four morphospecies. We found three significant axes (Wilks à = 0.0753, F = 16.86,

df = 54 and 659.3; Bonferroni corrected, P < 0.01) (fig. 3). Morphospecies ANA and CBO

were easily discriminated from morphospecies RU and PRU, but the last two were only

partially discriminated from each other (tab. 2). The standardized discriminant function

coefficients and the loadings are presented in tab. 3.

Source : MNHN, Paris

NAPOLI & CARAMASCHI 171

Table 1. - Distributions and percentage (in parentheses) of patterns (fig. 1-3) among the four

morphospecies. A blank indicates no specimen had that state; a zero indicates that at

least one specimen with that state was examined, but the rate of occurrence per 100

specimens rounds off to zero. n = number of specimens for which data are available.

General dorsal patterns

Morphospecies _# AI A2 A3 A4 AS A6 A7 A8 A9 A10

RU 5340) 14) 15Q1) 1702) 4G) 64) 204) 9(@ 1541) 10)

PRU 16(33) 4(8) 1(25) 2(4) 244) 1(2) 1(2) 2(4) 7(14) 10)

CBO 563) 11 (66)

ANA 45(55) 143) 56

Mid-dorsal pin stripe Dorsolateral stripes

Morphospecies x | B1 B2 B3 C2 C3 c4

64(3) 3201)

48 (3)

49 7)

2(4) 10(20) 36(75) 8(7) 15(32) 10(21) 3(6) 5(10)

15 (100) 14(93) 1(6)

36(43) 20(2) 26(31) 50 (61) 102) 14) 5@

Lateral limits of dorsum Upper surface of tibia Loreal and canthal stripes

Morphospecies DI D2 EI E2 E3 E4 n F1 F2 F3 F4

134 (100) 91(67) 37026) 6) 140 |119(84) 120) 6(4) 30)

22(42) 26(54) 12 (26) 27 (60) 6(13) 46 | 4(8) 22(47) 14(30) 6(13)

15 (100) 10 (66) 5(33) 15 960) 6(40)

82 (100) 135) 66(80)] 82 |[40(46) 464) 38 (46)

Dorsal head shape

G2 G3 G4 Gs Gé G7 G8 G9

Morphospecies

RU 40 (28) 9@ 1H 1

58 (41)

20)

50) 264) 1@ 1@) 38 (80)

15 (100)

82 (100)

Source : MNHN, Paris

172 ALYTES 16 (3-4)

VARIATION WITHIN MORPHOSPECIES RU

The analysis examined the samples from Minas Gerais and Bahia. These samples were

grouped into four areas equidistantly distributed along a transect (fig. 4A) linking Barreiras

(Bahia) and Alfenas (Minas Gerais) that represented, respectively, the distribution limits

north and south for morphospecies RU. Distributions of pattern states were determined for

each of the four areas, and the observed occurrences were tested against expected occurrences

(based on frequency of distribution for entire sample RU) with a chi-square test. Some

character states were combined to avoid violating minimum cell-size requirements for ;?

analysis (app. 1; SOKAL & RoHLF, 1981).

Three directional clines were observed (fig. 4A). The first direction (shading “A”)

denoted a cline for general dorsal pattern and upper surface of tibia pattern (fig. SA). These

specimens showed an increase in dorsal melanization and a decrease of the dorsolateral white

stripe on the edges of tibia from southeastern to northeastern Minas Gerais. The second

direction (shading “B”) denoted a cline for dorsal head shape (fig. SA) involving areas I, II

and IV. We did not consider area IIT because it is not representative (the two geographical

samples in the direction “B” included only two specimens and neither were well preserved);

thus, there is a hiatus between areas II and IV. The third cline followed the transect line. It was

characterized by a decrease in occurrence of a mid-dorsal pin stripe (fig. SA) from south to

north (i.e., from area I/II to IV). The patterns of loreal and canthal stripes and dorsolateral

stripes did not show statistically significant level variation.

The similarity among these areas depended on each particular character, and there was

no specific pattern discriminating an area from the others. However, differentiation may be

computed in the degree of occurrence for a certain state. The similarity and dissimilarity

among areas shown by each character obtained from the }? test was as follows: general dorsal

pattern (1 = IV; II = III), mid-dorsal pin stripe pattern (I = I; III = IV), dorsolateral stripes

pattern (1 = II = III; IV), upper surface of tibia pattern (I = II = IV; ID), loreal and canthal

stripes pattern (1 = IV = II = INT), and dorsal head shape (1 = HI: I = IV).

MEASUREMENT VARIABLES

Multiple discriminant function analysis was used to analyze morphological variation

among nine samples previously combined. To increase the number of specimens analyzed,

samples from Três Marias and Andrequicé, Pirapora and Lagoa Formosa, and Vespasiano

and Baräo de Cocais were combined because of their proximity. Three significant canonical

axes (Wilks à = 0.02385, F = 3.274, df = 144 and 712.6; Bonferroni corrected, P < 0.0006)

resulting from this analysis represented 79 % of the total variation. The projection of the

individual scores in the first three axes (not figured) did not support additional discrimination

and made a mosaic of superpositions among the geographic samples. This result may be

interpreted as intraspecific variation. All samples were considered to belong to H. rubicun-

dula.

Source : MNHN, Paris

NapoLi & CARAMASCHI 173

VARIATION WITHIN MORPHOSPECIES PRU

This analysis examined samples from Goiäs. These were grouped into three areas (fig.

4B) with the same criteria as for morphospecies RU, but the small number of specimens in

each sample, mainly in areas I and III, made the use of the }? test (pattern analysis) impossible

in most comparisons. The discriminant function analysis used to analyze morphological

variation (measurement variables) among five previously combined samples furnished only

one significant canonical vector (Bonferroni corrected) without any relevant discrimination

result.

Frogs from areas I and II were similar to each other in the majority of characters but were

different from those from area III. A cline, characterized by the straight line between Santa

Rita do Araguaia and Säo Domingos (fig. 4B), was observed for (1) dorsolateral stripes (a

progressive disappearance of the dorsolateral white stripe from northern to southern Goiäs)

and (2) dorsal head shape patterns (a decrease of diversity of dorsal head shape patterns from

northern to southern Goiäs; fig. 5B). The similarity among areas shown for each character,

obtained for certain characters by the ;? test, is as follows: general dorsal pattern (1 = IL; III),

mid-dorsal pin stripe pattern (1 = IL; ID), dorsolateral stripes pattern (1= IL; IT), lateral limits

of dorsum pattern (1 = I; III), upper surface of tibia pattern (1 = II; HIT), loreal and canthal

stripes pattern (1; I; III) and dorsal head shape (I = II; III). Differences between areas I and

Il were mainly by degree of occurrence of some states, rather than kind; area III differed from

the others by degree and kind.

TAXONOMIC CONCLUSIONS

Morphospecies RU and PRU were not well discriminated from each other. Pattern

Standards denoted variation in degree between these morphospecies but not in kind. Such

variation occurred for all character similarity between area III of Minas Gerais (fig. 4A) and

area I of Goiäs (fig. 4B). The discrimination obtained by the discriminant function analysis

was not robust (tab. 2). Also, the comparisons between advertisement calls of topotypic Hyla

rubicundula (CARDOSO & VIELLIARD, 1985) (morphospecies RU) and a sample from Silvä

Goïäs (morphospecies PRU; see Vocalization in Hyla rubicundula redescription below) failed

to provide additional support for discrimination.

The distribution of morphospecies PRU in Goiäs (central Brazil) deserves consideration.

The Serra do ap, Serra Dourada, Serra dos Pirineus and heterogeneous vegetation

separate the examined population samples in three areas in northern, southern and eastern

Goiïäs (Goïânia). The vegetation (ANONYMOUS, 1989) is mainly represented by seasonal

semi-deciduous forest, seasonal deciduous forest and transitional areas (“ecological stress

areas”). Because these frogs never cross tropical rainforests, the discontinuity of cerrado

formation in central Brazil, where different kinds of relief and vegetation are found, may

reduce or obstruct genetic flow among local populations and favor the formation of hetero-

geneous morphotypes.

The “Espigäo Mestre” (scarps, 1200-3000 m), with tropical rainforests, between Goiäs

and Bah: s well as the semi-deciduous seasonal forest of southern Goïäs (ANONYMOUS,

1989) adjacent to Minas Gerais, may function as ecological barriers between populations of

Source : MNHN, Paris

174 ALYTES 16 (3-4)

+is

Santa Rita

do Armguai

+ 35e

- 50°

Fig. 4. — Geographic distribution of (circles) Hyla rubicundula, (squares) H. anataliasiasi and (stars) H.

cachimbo. Each plot may represent more than one sample. Closed symbols show the localities of

examined samples, and open symbols the localities of samples of H. anataliasiasi not examined in

this paper. (A) Distribution of morphospecies RU in Minas Gerais and Bahia. A transect line links

Barreiras and Alfenas, the distribution limits north and south for RU. Shading areas À and B show

directions of morphological variation explained in text (see Variation within morphospecies RU).

(B) Distribution of morphospecies PRU in Goiäs, A transect line links Säo Domingos and Santa

Rita do Araguaia, the distribution limits north and south for PRU. For detailed explanation of

each character involved, see Variation within morphospecies PRU. BA, Bahia; ES, Espirito Santo:

GO, Goiäs; MG, Minas Gerais; MS, Mato Grosso do Sul, MT, Mato Grosso; SP, Säo Paulo; TO,

Tocantins. Roman numerals indicate areas equidistantly distributed throughout the transect.

RU and PRU which occur only in cerrado habitats. The greatest morphological similarity

between these two morphospecies occurs right in the cerrado corridors that allow interactions

between populations of RU in Minas Gerais and Bahia and PRU in Goiäs. We conclude that

both morphospecies RU and PRU belong to Hyla rubicundula.

Source : MNHN, Paris

NAPOLI & CARAMASCHI 175

General dorsal patterns

Mid-dorsal pin stripe

100 DAFAIEAT 100} (m5? 1585 |

80! 80

co! 60 |

| D |

40! 40

| È e |

| |

AL n k o Lu

1 ü ui w 1 u on W

Upper surface of tibia Dorsal head shape

100 (mET OEZ m6 | 100: SHORSS |

à +05

| G+67

æ] 80 LRES

6 | ; &! ,

| | | | |

40 | 40!

| 7

20! ] ] ë | 20 | ë

NL En n°7 Rex

Dorsal head shape

RO DONS |

(e67

[ nl on

Fig. 5. - Frequency (in percentage) of patterns obtained in morphospecies (A) RU and (B) PRU for areas

I-IV (fig. 4A) and areas I-HI (fig, 4B) respectively. Patterns were combined (For criteria, see app. 1)

in order not to violate minimum cell-size requirements for chi-square analysis.

Morphospecies ANA (Hyla anataliasiasi) and CBO are well discriminated from each

other and from the other two morphospecies (Hyla rubicundula) by the analyses of external

morphology and morphometries. Morphospecies CBO is restricted to an isolated savanna

which is separated from cerrado by 200 km of tropical rainforest and was probably connected

to the cerrado during periods of drier climate (Pleistocene; PRANCE, 1996). As we stated, these

frogs never cross tropical rainforests; thus, this geographic isolation obstructs genetic flow and

Source : MNHN, Paris

176 ALYTES 16 (3-4)

Table 2. - Classification table for specimens based on the results of the discriminant function

analysis for the combined samples RU, PRU, CBO, and ANA; Results presented

graphically in fig. 5. # = number of specimens.

Morphospecies RU PRU ANA

RU 96(7742%) | 23(18.55%) 4(3.23%) 1 (0.81%)

5 (12.20%) 33 (80.49%) 3(7.32%) 0

0 12 (100%) 0

0 0 65 (100%)

suggests a speciation mechanism. Morphospecies CBO and ANA may be considered full

species, and we assigned the following morphospecies to these species: morphospecies RU and

PRU to Hyla rubicundula Reinhardt & Lütken, 1862; morphospecies ANA to Hyla anatalia-

siasi Bokermann, 1972; and morphospecies CBO to a new species described below.

SPECIES DESCRIPTIONS

Hyla cachimbo sp. nov.

(fig. 6A, 7A, 8A)

Holotype. - MZUSP 21912, adult male, collected at Cachimbo (about 09°21’S, 54°57/W),

Parä, Brazil, between 200 and 400 m, 18 October - 9 November 1955, by E. DENTE, FE S.

PEREIRA and W. BOKERMANN.

Paratopotypes. — Thirteen adult males (MNRJ 17298-17299; MZUSP 21911, 21913-21918,

21920-21926) and an adult female (MZUSP 21910), collected with the holotype.

Diagnosis. - Species characterized by the following combination of traits: (1) small size (SVL:

males 19.8-21.0 mm; female 24.2 mm); (2) lateral limits of dorsum above the tympanum

(pattern D2; fig. 2); (3) head as long as wide, width contained about 3.1 times in the snout-vent

length; and (4) dorsal snout profile acuminate (fig. 6A, 7A).

No specimen of H. cachimbo has two divergent dorsal brown stripes from the anterior

section of head to near the middle of the body nor two parallel sacral stripes, but many

individuals of H. rubicundula have such a pattern (patterns A1, A2, A4-A6 and A8-10; fig. 1).

No specimen of H. cachimbo has a mid-dorsal pin stripe, but many individuals of H.

rubicundula have such a pattern (fig. 1). No specimen of H. cachimbo has the lateral limits of

dorsum under the lower border of tympanum (pattern D2; fig. 2), but many individuals of H.

rubicundula from Goïäs have such a pattern. No specimen of H. cachimbo has a light pinkish

to white stripe above a brown stripe on the edges of the tibia (pattern El; fig. 2), but many

individuals of H. rubicundula have such a pattern; also, no specimen of the former has a thin

Source : MNHN, Paris

NAPOLI & CARAMASCHI 177

Fig. 6.- Dorsal views of adult males. (A) Hyla cachimbo, holotype, MZUSP 21912, Cachimbo, Parä; (B)

H. rubicundula, MNRIJ 17294, Lagoa Santa, Minas Gerais; (C) Æ. rubicundula, MNRJ 17295,

Goiänia, Goiäs; (D) H. anataliasiasi, MZUSP 49610, Posto Diauarum, Mato Grosso.

Source : MNHN, Paris

178 ALYTES 16 (3-4)

Table 3. - Standardized discriminant function coefficients for 18 morphometric characters of

the combined samples of morphospecies RU, PRU, CBO and ANA. r, correlation

coefficient (Pearson) of the original data with the scores resulted from the discriminant

function analysis; *, not significant; * P < 0.05; ** P < 0.02; *** P < 0.01.

Characters

longitudinal central brown stripe composed of small dots, whereas many individuals of H.

anataliasiasi have such a pattern (pattern E4). The presence in H. cachimbo of a broad pinkish

stripe above a canthal brown stripe (patterns F2-F3; fig. 2) distinguishes it from H. anatalia-

siasi Which presents a canthus well delimited by a thin white stripe above a brown stripe

(pattern F1). A pointed snout (fig. 6A, 7A) differentiates H. cachimbo from H. rubicundula

(fig. 6B-C, 7B-C). The head of the former is as long as wide, about 3.1 times into the

snout-vent length, and this feature distinguishes it from H. anataliasiasi which has a head

longer than wide, its width being contained about 3.6 times in the snout-vent length.

Description. — Descriptive statistics are provided in tab. 4. Head as long as wide, its width

contained about 3.1 times in snout-vent length; internarial distance greater than eye-nostril

distance (n = 15, 1 = 2.76, P = 0.01) and smaller than eye diameter (n = 15, 1 = 20.66, P = 0);

eye diameter greater than eye-nostril distance (n = 15, 1 = 19.68, P = 0); snout acuminate in

Source : MNHN, Paris

Table 4. - Descriptive statistical tables of morphometric characters for Hyla cachimbo sp. nov. (morphospecies CBO) and H. anataliasiasi (morphospecies

ANA). n = number of specimens for which data are available;

= mean; s = standard deviation; CV = coefficient of variation.

Morphospecies CBO Morphospecies ANA

Characters Males Females Males Females (1 = 4)

no x min max s CV |(@=1)| n x min max C2 x min max S CA

SvL 15 20.74 198 210 064 311] 242 | 80 1885 160 1.51 803 | 1970 166 216 224 1139

HW 15 639 60 68 025 394| 77 80 528 44 218 140 7.72 | 546 46 Gl 065 1201

HL 15 649 62 68 021 323| 77 80 568 47 GI 040 705 | 607 54 68 059 984 Z

ED 15 238 22 26 012 523] 25 80 219 19 65 010 456 | 227 20 24 015 683 ë

UEW 14 141 12 17 O14 1012] 17 77 122 09 24 O16 1321 | 115 10 12 00 793 É

10D 14 229 20 26 016 736| 25 78 184 14 22 016 903 | 195 17 22 022 1165 È

END 15 154 14 18 010 702| 17 80 121 10 22 O1 966 | 133 11 16 O18 l4il >

IND 15 163 15 17 006 695| 18 80 127 10 16 011 882 | 132 12 14 00 72 È

THL 15 990 93 105 035 359| 125 | 80 838 7. LS 069 832 | 888 73 100 124 1401 &

TL 15 1001 94 106 033 334| 124 | 80 880 75 101 078 886 | 925 78 105 113 1231 ä

TD 14 100 08 1 009 96 | 11 73 091 06 106 012 1344 | 087 06 10 016 18.95

NSD 15 119 10 13 008 714! 13 80 093 O7 14 008 9.10 | 092 08 10 008 936

UAR 15 600 57 64 02 371| 70 80 527 44 11 042 802 | 535 48 58 045 846

FAR 15 387 36 42 019 495] 5.1 80 337 28 63 029 868 | 343 30 39 039 115

HAL 15 591 55 62 02 37| 75 80 492 40 40 043 892 | 513 45 57 054 1070

3FD 15 088 07 10 007 810| 10 80 065 05 59 008 1283 | 071 05 O8 OI 16.58

FI 15 1410 131 151 056 399 189 | 80 1231 103 O8 112 909 |1343 110 151 181 13.53

Lam 15 081 07 09 006 847| 10 74 059 04 149 008 1450 | 058 05 06 007 1276

Source : MNHNI, Paris

180 ALYTES 16 (3-4)

dorsal outline and protruding or rounded in lateral outline; loreal region slightly oblique; eyes

moderately prominent; tympanum distinct and nearly circular; a supratympanic fold being

sometimes present, partially covering tympanum; nostrils dorsolateral; internarial region flat;

vomerine teeth often present in two patches between choanae; tongue cordiform or ovoid:

vocal sac single and subgular.

Forearm more robust and shorter than upper arm (7 = 15, 1 = 28.09, P = 0); hands with

a distinct palmar tubercle; subarticular tubercles rounded:; distal tubercle of third finger bifid

or rounded; distal tubercle of fourth finger always bifid; supernumerary tubercles present;

third finger disk diameter greater than fourth toe disk (n = 15,1= 5.72, P = 0); modal webbing

formula, I 2.50-2.50 II 2-2.25 III 2.75-2.25 IV. Legs slender; femur and tibia with about the

same stoutness and length (n = 15, 1 = 0.87, P= 0.39); sum of thigh and tibia lengths smaller

than snout-vent length (n = 15, = 3.42, P = 0). Foot with robust toes; subarticular tubercles

always rounded; supernumerary tubercles not distinct; prehallux distinct; plantar tubercle

distinct; modal webbing formula, I 2--2.25 II 1.25-2.25 III 1.25-2.75 IV 3-1.75 V.

Color in preservative. — Dorsum reddish, immaculate, with occasional dark brown dots;

mid-dorsal pin stripe absent; canthus rostralis delimited by a subcanthal brown stripe

(patterns F2-F3; fig. 2); lorus with variable melanization; a slender lateral brown stripe

sometimes present on flanks from posterior corner of orbit to near groin, sometimes bordered

by a light pinkish stripe (patterns C2-C3; fig. 2); thigh light brown, immaculate; a brown stripe

sometimes present on anterior and posterior edges of upper surface of tibia in addition to

dorsal random dots (patterns E2-E3; fig. 2); ventral surfaces immaculate buff. Color in life

unknown.

Measurements of holotype. — SVL 21.3; HW 6.8; HL 6.8; ED 2.4; UEW 1.4; IOD 2.6; END

1.7; IND 1.5; THL 10.5; TL 10.6; TD 1.0; NSD 1.1; UAR 6.4; FAR 4.2; HAL 6.2; 3FD 0.9;

FL 15.1; 4TD 0.8.

Etymology. - The specific name, a noun in apposition, refers to the type-locality, Cachimbo.

Geographic distribution. - Known only from the type-locality (fig. 4). This area is character-

ized as an “ecological stress area” (ANONYMOUS, 1991) or a transitional area between the

Cerrado Domain and the Amazon Equatorial Domain (AB’ SABER, 1977).

Hyla rubicundula Reinhardt & Lütken, 1862

(fig. 6B-C, 7B-C, 8B-C)

Hyla rubicundula Reinhardt & Lütken, 1862; BOKERMANN, 1968, 1972.

Specimens examined. —- BRAZIL. BaHiaA: Barreiras (MNRJ 0934, 0946, 0935-0940, 0933,

6145-6154); Jupaguä (MNRJ 0943-0944). Minas GErais: Alfenas (MNRJ 17126-17128,

17129-17133, 17134); Andrequicé (MNRIJ 17110); Arinos (MZUSP 64500-64504); Baräo de

Cocais (MNRJ 17210-17212); Belo Horizonte (MNRJ 17214-17220; MZUSP 519, 34647):

Buritis (MZUSP 64449-64452, 64455-64458, 64460-64464); Buritizeiro (MNRJ 17111-17112,

17113-17116); Esmeraldas (ZUEC 4023); Jaboticatubas (MZUSP 57712-57713); Januäria

(MNRJ 0942); Lagoa Formosa (MNRJ 17123); Lagoa Santa (topotypes, MNRJ 17117-

Source : MNHN, Paris

NAPOLI & CARAMASCHI 181

Fig. 7. - Dorsal and lateral views of the heads of adult males. (A) Hyla cachimbo, holotype, MZUSP

21912, Cachimbo, Parä; (B) H. rubicundula, topotype, morphospecies RU, MNRJ 17294, Lagoa

Santa, Minas Gerais; (C) H. rubicundula, morphospecies PRU, MNRJ 17295, Goiânia, Goiäs; (D)

H. anataliasiasi, MZUSP 49610, Posto Diauarum, Mato Grosso.

17121, 17124-17125, 3081, 13287, 0947, 6155-6177; MZUSP 34012-34023; ZUEC 4150);

Manga (MNRIJ 0941); Pimenta (MNRIJ 17319-17321); Pirapora (MNRJ 0928-0932, 0945,

0923-0927); Santa Luzia (MNRJ 17322-17323); Três Marias (MNRJ 17101-17109); Uberlân-

dia (MNRJ 17305-17308); Unai (MZUSP 64398-64402, 64386, 64389-64392, 64396; MNRJ

17135); Vespasiano (MNRJ 17221-17223; MZUSP 12691-12693). Golâs: Aragarças

(MZUSP 20983); Cavalcante (MZUSP 66543, 66570, 66574, 66576); Cristalina (MZUSP

64522); Goiânia (MNRIJ 17136-17155, 17300); laciara (MZUSP 66527-66528); Monte Alegre

de Goiäs (MZUSP 66403-66407, 66450, 66456); Nova Roma (MZUSP 66358-66360); Poran-

gatu (MNRJ 17167-17168); Santa Rita do Araguaia (MZUSP 66650-66654); Säo Domingos

(MZUSP 66597-66601, 66602, 66603); escarpa da Serra Dourada (ZUEC 7505). Praui:

Uruçui (MNRIJ 17224).

Syntypes. - NMW 16511, ZMUC 1440-1441, Lagoa Santa (about 19°37S, 43°53/W), Minas

Gerais, Brazil, 760 m (BOKERMANN, 1968; FROsT, 1985); specimens not examined by us.

Diagnosis. - Species characterized by the following combination of traits: (1) small size (SVL:

males 18.0-23.4 mm; females 21.6-25.1 mm); (2) in preservative, dorsum with two divergent

brown stripes from anterior section of head to sacral region, and two sacral stripes of same

color and orientation extending to cloacal region (pattern Al; fig. 1); (3) a thin brown

dorsolateral stripe bordered by a thin light stripe from posterior corner of orbit to near groin

(pattern CI; fig. 2); and (4) head as long as wide, its width contained about 3.3 times in

snout-vent length (fig. 6B-C, 7B-C).

Source : MNHN, Paris

182 ALYTES 16 (3-4)

The presence of dorsal brown stripes (patterns A1-A2, A4-A6 and A8-A10; fig. 1) in

many individuals of Hyla rubicundula differentiate them from H. cachimbo which never has

such a pattern. The presence in many specimens of the former of two divergent dorsal brown

stripes, from the anterior section of the head to nearly the middle of the body, together with

two sacral brown stripes (patterns A1 and A4; fig. 1), with or without additional brown stripes

(patterns AS and A8-A10), distinguish them from H. anataliasiasi, which do not have such

patterns. No specimen of Æ. rubicundula has the two anterior divergent dorsal brown stripes

fused to the sacral ones (pattern A11), whereas many individuals of H. anataliasiasi have such

a pattern. À mid-dorsal pin stripe (patterns B1-B2; fig. 1) in many specimens of Æ. rubicun-

dula distinguish them from AH. cachimbo, in which it is often absent. A broad and irregular

dorsolateral stripe, with or without an upper white to pinkish stripe (patterns C3-C4; fig. 2) in

many specimens of A. rubicundula distinguishes them from H. anataliasiasi, which never has

such a pattern. The lateral limits of the dorsal coloration in many specimens of A. rubicundula

are under the lower border of the tympanum (pattern D2; fig. 2), whereas Æ. cachimbo and H.

anataliasiasi often have this limit above the tympanum (pattern D1), a pattern common to the

three species. The presence of a thin white to pinkish stripe on the edges of the tibia above a

thin brown stripe (pattern El; fig. 2) in many specimens of H. rubicundula distinguishes them

from H. cachimbo, which never has such a pattern; also, no specimen of H. rubicundula has a

thin longitudinal central brown stripe on the upper surface of tibia composed of thin dots

(pattern E4), whereas many individuals of H. anataliasiasi have such a pattern. The presence

in A. rubicundula of a thin pinkish to white canthal stripe above a brown loreal stripe (pattern

F1; fig, 2) distinguishes it from Æ. cachimbo which lacks such a pattern; also, the presence in

many specimens of the former of a broad canthal pinkish stripe above a brown loreal stripe

(patterns F2-F3) distinguishes them from A. anataliasiasi, which never has such a pattern.

Hyla rubicundula has a truncate or rounded snout (fig. 6B-C, 7B-C), whereas Æ. cachimbo has

an acuminate snout (fig. 6A, 7A); also, the former has a head as long as wide, its width being

contained about 3.3 times in the snout-vent length, and H. anataliasiasi has a head longer

than wide, its width being contained about 3.6 times in the snout-vent length.

Description. — The following description is based on topotypes and other geographic samples

from Minas Gerais and Bahia (morphospecies RU). The morphotype located in central Brazil

(morphospecies PRU) is characterized in the geographic variation section.

Descriptive statistics are provided in tab. 5. Head as long as wide (7 = 140, 1 = 1.65, P =

0.09), its width contained about 3.3 times in snout-vent length; internarial distance greater

than eye-nostril distance (n = 139, = 4.61, P = 0) and much smaller than eye diameter (7 =

139, 1 = 50.29, P = 0); eye diameter greater than eye nostril distance (n = 139, 1 = 53.66, P =

0); canthus rostralis distinct, slightly rounded; lorus slightly oblique, sometimes perpendicu-

lar to canthus rostralis; eyes slightly to very prominent; tympanum distinct and nearly

circular; supratympanic fold poorly developed; nostrils dorsolateral, slightly protuberant,

directed laterally or slightly forward; internarial region furrowed or not; vomerine teeth in two

patches between choanae, with irregular shape and position; tongue cordiform or rounded;

vocal sac single and subgular.

Forearm more robust and shorter than upper arm (7 = 139, s = 40.64, P = 0); hands with

a distinct palmar tubercle; subarticular tubercles rounded; distal tubercle of fourth finger

bifid, that of third finger bifid or rounded; supernumerary tubercles present; prepollex

Source : MNHN, Paris

Table 5. - Descriptive statistical tables of morphometric characters for Hyla rubicundula (morphospecies RU and PRU). # = number of specimens for which

data are available; x = mean; s = standard deviation, CW = coefficient of variation.

Morphospecies RU Morphospecies PRU

Characters Males Females (2 = 4) Males Females (n = 6)

mo x min max S$ CV | x min max os CV] nm x min max s CV | x min max s CV

svL 140 2127 180 234 097 4.582375 216 25.1 1.52 643| 47 2167 181 238 109 507|2393 222 254 143 598

HW 140 631 54 70 028 454657 62 69 033 535] 47 649 56 72 031 486|700 65 42 026 3.80

HL 140 637 55 71 027 436/681 GS 71 033 488| 47 645 57 70 026 415|705 65 74 030 437 Z

ED 139 233 20 27 O14 636|251 23 26 O14 594| 47 245 21 28 O14 585|258 23 27 016 620 &

UEW 136 1.56 12 20 0.15 966|1.57 15 1.7 O11 7.55] 46 1.56 10 18 O14 935|170 14 19 016 951 Ë

10D 129 216 17 26 019 912236 20 26 027 11.51| 46 218 18 25 O14 682|227 22 24 (009 4.34 a

END 139 148 11 18 010 685/158 15 17 008 537| 47 153 13 17 010 670|160 15 17 009 5.70 Ë

IND 139 155 LI 18 010 685/158 15 17 O11 698| 47 1.54 13 18 009 6o8|170 15 18 008 526 È

THL 137 981 80 121 056 577|1061 94 111 078 737| 47 1008 85 113 060 608|1115 103 118 0.51 4.66 a

TL 140 999 83 111 048 488/1076 97 113 073 685| 47 1005 83 111 058 585/1104 102 115 059 5.36 À

TD 138 097 06 14 O11 1217|112 10 12 O18 1326| 44 103 08 12 008 824|127 09 18 031 2501

NSD 139, 111 09 18 011 1041|1.15 09 13 0.18 1626| 47 113 09 12 008 725 | 121 11 13 007 5.12

UAR 139 583 44 72 047 807|628 60 65 022 357| 47 599 51 67 040 642/658 60 69 033 5.12

FAR 139 390 31 49 030 737|431 38 46 035 816| 47 392 34 45 025 642|414 37 46 030 745

HAL 139 584 44 70 043 738628 60 64 019 307| 47 Gil 53 74 039 653|660 60 69 034 522

3FD 139 086 06 11 007 927|096 O8 10 008 887| 46 08 06 10 008 902|097 08 11 009 9.59

FI 139 1461 117 163 076 526[1572 143 165 096 G.14| 47 1489 122 174 1489 1.001620 150 174 101 626

4TD 139 080 0.5 10 009 1168/087 08 09 006 737| 47 084 06 10 009 11.57|088 O7 10 0.10 11.69

Source : MNHN, Paris

184 ALYTES 16 (3-4)

Fig. 8. — Hands and feet of adult males. (A) Hyla cachimbo, holotype, MZUSP 21912, Cachimbo, Parä;

(B) H. rubicundula, topotype, morphospecies RU, MNRJ 17294, Lagoa Santa, Minas Gerais; (C)

H. rubicundula, morphospecies PRU, MNRJ 17295, Goïänia, Goiäs; (D) H. anataliasiasi, MZUSP

49610, Posto Diauarum, Mato Grosso.

distinct; third finger disk diameter greater than fourth toe disk (n = 139, = 5.72, P = 0); modal

webbing formula, 1 2.75-2.75 II 2-3.25 III 3-2.25 IV. Legs slender; femur and tibia with about

the same stoutness; femur length shorter than tibia length (n = 137, 1 = 2.88, P =0); sum of

femur and tibia lengths smaller than snout-vent length (7 = 137, 1 = 12.20, P = 0); toes not

robust; subarticular tubercles rounded; supernumerary tubercles variable in shape and num-

ber; prehallux distinct; modal webbing formula, I 2--2.25 II 1*-2.25 III 1*-2.25 IV 2.25-1* V.

Color. — In life, the analysis of four topotypic specimens from Lagoa Santa (Minas Gerais)

revealed that in the same specimen the dorsal surfaces vary from dark green to dark brown,

with an intermediate yellow phase; dots and dark brown stripes are not visible on the dorsum:

a dark brown stripe, bordered by a white stripe, is visible on the flanks and canthus rostralis;

thigh light brown and immaculate; vocal sac yellowish, belly white; finger and toe disks

reddish.

Source : MNHN, Paris

NAPOLI & CARAMASCHI 185

FREQUENCY (kHz)

105

Fig. 9. - Sonogram and oscillogram of advertisement call of Hyla rubicundula (morphospecies PRU)

from Silvänia, Goiäs. Calls are given sporadically. The intervals between the notes are not natural.

In preservative, dorsum reddish, with occasional dark brown stripes and dots (patterns

AI-A10; fig. 1); a mid-dorsal pin-stripe sometimes present on dorsum (patterns B1-B2; fig. 1);

canthus rostralis delimited by a dark subcanthal brown stripe bordered above by a light pink

to white stripe (pattern F1; fig. 2); lorus with a variable degree of melanization; dorsolateral

region delimited by a dark brown stripe bordered above or not by a light pink to white stripe

from posterior corner of orbit to near groin (patterns C1-C2 and C5; fig. 2), both often above

tympanum (pattern DI; fig. 2); thigh light brown, immaculate; a brown stripe sometimes

present on anterior and posterior edges of tibia in addition to random dots (patterns E1-E3;

fig. 2); ventral surfaces immaculate buff.

Geographic variation. - Samples from central Brazil (morphospecies PRU) have the following

differences when compared to samples from Minas Gerais and Bahia (morphospecies RU):

dorsal head shape pattern with pattern A7 (fig. 3, 6C, 7A); internarial distance and eye-nostril

distance nearly equal (n = 47, 1= 0.26, P = 0.79); lorus slightly to strongly concave;: tympanum

covered or not by a supratympanic fold; distal tubercle of fourth finger bifid or not; femur and

tibia the same length (1 = 47, 1= 0.22, P = 0.82); dorsolateral stripes pattern corresponding to

patterns C3-C4 (fig. 2); lateral limits of dorsal pattern corresponding to pattern D2 (fig. 2).

The other variations are of a matter of degree (tab. 1) and descriptive statistics are presented

in tab. 5.

Vocalization. — The advertisement calls studied are from one specimen from Silvânia, Gois

{morphospecies PRU: fig. 9). Each note composed of three pulses had a duration of nearly

0.03 s, and each note was composed of four pulses about 0.04 s. Broadcast frequencies range

between 3.5 and 4.8 kHz. Air temperature was 21.5°C. CARDOSO & VIELLIARD (1985) gave a

Source : MNHN, Paris

186 ALYTES 16 (3-4)

detailed description of the call of Hyla rubicundula from Lagoa Santa, the type-locality of

morphospecies RU. Comparisons between the two vocalizations reveal that they are very

similar and that both belong to H. rubicundula.

Geographic distribution. — Hyla rubicundula occurs in Minas Gerais, Goiäs, Bahia and Piaui

(fig. 4), mainly in the Cerrado Domain (AB’ SABER, 1977), and never crosses tropical rain

forests.

Hyla anataliasiasi Bokermann, 1972

(ig. 6D, 7D, 8D)

Specimens examined. - BRAZIL. MATO Grosso: Posto Diauarum (MZUSP 49588-49617);

Posto Leonardo (MZUSP 49339-49393).

Holotype. - WCAB 45272, adult male, collected at Belém-Brasilia highroad, 80 km before

Paraiso do Norte, Brejinho de Nazaré (about 11°00'S, 48°33/W), Goiäs [Tocantins], Brazil,

247 m, 17 January 1970, by C. A. BOKERMANN, Ladislau A. DEUTSCH and Milton S. CAROLLO.

Paratypes. — Four adult males: WCAB 45273, collected with the holotype; WCAB 45256-

45258, collected at Paranä (about 12°36'S, 47°52’W), Goiïäs [Tocantins], Brazil, 274 m,

December 1969, by Anatalias J. RODRIGUES.

Diagnosis. - Species characterized by the following combination of traits: (1) small size (SVL:

males 16.0-21.8 mm; females 16.6-21.6 mm); (2) dorsum with nearly parallel dark brown

stripes, the two anterior ones very near each other, joined with the two sacral ones (pattern

AIl:; fig. 1); and (3) head longer than wide, its width being contained about 3.6 times in

snout-vent length (fig. 6D, 7D).

The presence of two anterior dorsal brown stripes fused to the sacral ones in some

specimens of H. anataliasiasi (pattern A1 1; fig. 1) distinguishes them from A. rubicundula and

H. cachimbo, which lack such a pattern; also, the absence in the former of two divergent dorsal

brown stripes, from the anterior section of head to nearly half of the dorsum, barely separated

from two sacral brown stripes (patterns A1 and A4), with or without additional dorsolateral

stripes (patterns AS and A8-A10), distinguishes it from A. rubicundula, which has many

individuals with such patterns. A mid-dorsal pin stripe (patterns Bl and B6; fig. 1) in many

specimens of H. anataliasiasi distinguishes them from H. cachimbo in which stripes are

absent. A well-marked dark brown to black dorsolateral stripe under a thin white stripe

(pattern C7; fig. 2) in some specimens of H. anataliasiasi distinguishes them from H.

rubicundula and H. cachimbo which never possess such a pattern; also, the absence in the

former of a broad and irregular brown dorsolateral stripe, with or without an upper white to

pinkish stripe (patterns C3-C4), distinguishes it from many individuals of A. rubicundula with

such patterns. No specimen of A. cachimbo has the lateral limits of the dorsal coloration

below the lower border of the tympanum (pattern D2; fig. 2), but many individuals of H.

rubicundula from Goiïäs have such a pattern. The presence in some specimens of A. anatalia-

siasi of a thin white to pinkish stripe on the edges of tibia, above a thin brown stripe (pattern

Source : MNHN, Paris

NaApoLi & CARAMASCHI 187

El; fig. 2), distinguishes them from H. cachimbo, which never has such a pattern; also, the

presence in the former of a thin longitudinal central brown stripe on the upper surface of

tibia, composed of small dots (pattern Ed), distinguishes it from AH. rubicundula and H.

cachimbo which never possess such a pattern. No specimen of H. anataliasiasi has a broad

canthal pinkish stripe above a brown loreal stripe (patterns F2-F3; fig. 2), but many indi-

viduals of H. rubicundula and H. cachimbo have such a pattern. The snout in H. anataliasiasi

is acuminate in many individuals (fig. 6D, 7D), but it is rounded or truncate in H. rubicundula

(fig. 6B-C, 7B-C). In the former the head is longer than wide, its width being contained about

3.6 times in snout-vent length, whereas in H. rubicundula and H. cachimbo the head is as long

as wide, its width being contained, respectively, about 3.3 and 3.1 times in snout-vent length.

Description. — Descriptive statistics are provided in tab. 4. Head longer than wide (n = 80, 1=

6.23, P = 0), its width being contained about 3.6 times in snout-vent length; internarial

distance greater than eye-nostril distance (7 = 80, : = 3.09, P = 0) and much smaller than eye

diameter (n = 80, : = 54.51, P = 0); eye diameter greater than eye-nostril distance (7 = 80, : =

56.35, P = 0); snout truncate, rounded or acuminate in dorsal outline, and slightly protruding,

truncate or rounded in lateral outline; canthus rostralis distinct, especially when bordered by

loreal and canthal stripes, rounded or straight; lorus slightly concave; eyes moderately

prominent; tympanum distinct, nearly circular; a supratympanic fold sometimes covering

upper surface of tympanum; nostrils dorsolateral, slightly protuberant, directed laterally or

slightly anteriorly; internarial region furrowed; vomerine teeth in two patches with irregular

shapes and positions between choanae; tongue cordiform or rounded; vocal sac single,

subgular, not well developed.

Forearm shorter and more robust than upper arm (n = 80, : = 33.04, P = 0); hands with

a distinct palmar tubercle; subarticular tubercles distinct, rounded; distal tubercle of third

and fourth fingers bifid or not; supernumerary tubercles present; palmar tubercle distinct;

prepollex distinct; third finger disk diameter greater than fourth toe disk (n = 74, 1 = 4.92, P

= 0); modal webbing formula, I 2.50-2.75 II 2.25-3.25 III 2.75-2.25 IV. Legs slender; femur and

tibia with the same stoutness; femur longer than tibia (n = 80, 1 = 3.60, P = 0); sum of femur

and tibia lengths smaller than snout-vent length (7 = 80, : = 8.57, P = 0); foot with rounded

subarticular tubercles; supernumerary tubercles not very distinct; prehallux distinct; plantar

tubercle present or not; modal webbing formula, 1 1.75-2.25 II 1*-2.25 III 1.25-2.25 IV 3-1*

Color. — In life, dorsal surfaces green (BOKERMANN, 1972). In preservative, dorsum reddish

with occasional dark brown stripes and dots (patterns A2, A6 and All: fig. 1); a mid-dorsal

pin-stripe present or not (patterns B1-B2; fig. 2); canthus rostralis delimited, or not, by a

subcanthal dark brown stripe bordered above by a light pink to white stripe (patterns F1-F3;

fig. 2); lorus with a variable degree of melanization; a lateral brown stripe sometimes present

on flanks from posterior corner of orbit to near groin, sometimes bordered by a light pinkish

stripe (patterns C1-C2, CS and C7; fig. 2), both often above tympanum (pattern BI; fig. 2);

thigh light brown with numerous widespread light brown dots; a brown stripe sometimes

present on anterior and posterior edges of upper surface of tibia, bordered by a light pink to

white stripe, in a addition to dorsal random dots (patterns El and E3; fig. 2), or with a thin

longitudinal central stripe composed of small dots (pattern E4); ventral surfaces immaculate

buff.

Source : MNHN, Paris

188 ALYTES 16 (3-4)

Geographic distribution. - Recorded from Tocantins (Brejinho do Nazaré and Paranä; BOKER-

MANN, 1972) and northern Mato Grosso (Posto Diauarum and Posto Leonardo; fig. 4), both

in the Cerrado Domain (AB’ SABER, 1977) at elevations between 247 and 274 m.

RÉSUMÉ

Le groupe d'espèces de Hyla rubicundula, composé de H. rubicundula Reinhardt &

Lütken, 1862 et H. anataliasiasi Bokermann, 1972, est subdivisé en quatre morpho-espèces.

La variation intra- et inter-populationnelle de la morphologie externe de chaque morpho-

espèce est analysée. Hyla rubicundula renferme trois des quatre morpho-espèces. Celle située

au nord de sa répartition est décrite comme une espèce nouvelle, caractérisée principalement

par un dos immaculé et un museau pointu. Une redescription est présentée pour les espèces H.

rubicundula et H. anataliasiasi.

ACKNOWLEDGEMENTS

We thank Mauro CAVALCANTI for aid in the statistical procedures, Ronald ALTIG, W. Ronald HEYER,

Ronaldo FERNANDES, José P. POMBAL Jr. and Hussam ZAHER for their helpful comments on the manu-

script, Paulo R. NASCIMENTO for the drawings, Rogério P. Basros for the recording of Hyla rubicundula

from Goiäs, and Célio F. B. HappaD for the bioacoustic analysis. Adäo J. CARDOSO and Paulo E.

VANZOLINI permitted study of material under their care. For financial support, we are grateful to the

CAPES and CNPq, and FAPESP (Proc. 94/5634-5) for the aid to the Bioacoustic Laboratory of UNESP

— Rio Claro, SP.

LITERATURE CITED

ANONYMOUS [Fundaçäo Instituto Brasileiro de Geografa e Estatistica], 1989. — Geografia do Brasil

Regiäo Centro-Oeste. Rio de Janeiro, IBGE, 1: 1-267.

—— 1991. — Geografia do Brasil: Regio Norte. Rio de Janeiro, IBGE, 3: 1-307.

An° Sam, A. N., 1977. — Os dominios morfoclimäticos na América do Sul. Primeira aproximaçäo.

Geomorfologia, 52: 1-23.

BOKERMANN, W. C. A. 1965. — Très novos baträquios da regiäo central de Mato Grosso, Brasil

(Amphibia, Salientia). Rev. brasil. Biol., 28 (3): 257-264.

1968. — Notas sobre alguns anfibios descritos por Reinhardt & Lütken em 1862 (Amphibia). Rev.

brasil. Biol., 28 (3): 327-329.

— 1972. — Uma nova espécie de Hyla de Goiäs, Brasil (Anura, Hylidae). Rev brasil. Biol, 32 (4):

593-504.

CarDoso, À. J. & VIELLIARD, J. M. E., 1985. - Caracterizaçäo bio-acüstica da populaçäo topotipica de

Hyla rubicundula (Amphibia, Anura). Rex. brasil. Zoo. Säo Paulo, 2 (7): 423-426.

DuELLMAN, W. E., 1970. - The hylid frogs of Middle America. Mon. Mus. nat. Hist. Univ. Kansas (1):

1-753.

FRosT, D. R., (ed.), 1985.— Amphibian species of the world. Lawrence, Allen Press & Ass. Syst. Coll.: [i-iv]

+ iv + 1-732.

Happap, C. F. B., ANDRADE, G. V. & CARDOSO, A. J., 1988. - Anfibios anuros no Parque Nacional da

Serra da Canastra, Estado de Minas Gerais. Brasil florestal, 64: 9-20.

Source : MNHN, Paris

NAPOLI & CARAMASCHI 189

Heyer, W. R., 1984. — Variation, systematics, and zoogeography of Eleutherodactylus guentheri and

closely related species (Amphibia, Anura, Leptodactylidae). Smithsonian Contrib. Zool., 402: 1-42.

HEYER, W. R., RAND, A. S., CRUZ, C. A. G., PEIXOTO, O. L. & NELSON, C. E., 1990. — Frogs of Boracéia.

Arq. Zool., Säo Paulo, 31: 231-410.

Jim, J,, 1980. — Aspectos ecolôgicos dos anfibios registrados na regiäo de Botucatu, Sao Paulo ( Amphibia,

Anura). Doctoral Dissertation, Säo Paulo, Instituto de Biologia, Universidade de Säo Paulo:

1-332.

LuTz, A., 1925. - Batraciens du Brésil. Cr Soc. Biol., Paris, 93: 137-139.

MarCus, L. F., 1990. - Traditional morphometrics. /n: F. J. ROHLF & F. L. BOOKSTEIN (ed.), Proceedings

of the Michigan morphometries workshop, The University of Michigan Museum of Zoology,

Special Publ., 2: 77-122.

PRANCE, G. T., 1996. - Islands in Amazonia. Phil. Trans. r. Soc. Lond., 351: 823-833.

REINHARDT, J. & LÜTKEN, C. F, 1862. - Bidrag til kundskab om Brasiliens Padder og Krybdyr. Vid.

Medd. Naturh. Foren Kjobenhavn, 3: 143-242.

Res, S. F; PÊssoa, L. M. & STRAUSs, R. E., 1990. - Application of size-free canonical discriminant

analysis to studies of geographic differentiation. Rev. brasil. Genet., 13: 509-520.

SAVAGE, J. M. & HEYER, W. R., 1967. - Variation and distribution in the tree-frog genus Phyllomedusa in

Costa Rica, Central America. Beitr. neotrop. Fauna, 5: 111-131.

SokaL, R. R. & ROHLF, F. J., 1981. — Biometry. 2nd ed. New York, Freeman & Co.: 1-859.

VANZOLINI, P. E., 1970. - Zoologia sistemätica, geografia e a origem das espécies. Universidade de Säo

Paulo, Instituto de Geografia, Teses e Monografias: 1-56.

APPENDIX |

Criteria for combination of patterns

in analyses of interpopulation variation

of morphospecies RU and PRU

Patterns were joined by similarity and geographic distribution.

General dorsal patterns. A1 and A4 are typical from topotypic samples for Hyla rubicundula.

Compared to patterns AI and A4, A2, A3, A6 and A7 are incomplete, vestigial or absent,

whereas A5, A8, A9 and A10 have additional melanization.

Mid-dorsal pin stripe patterns. — B1 and B2, presence; B3, absence.

Dorsolateral stripes. — C1 and C2, typical from Lagoa Santa, Minas Gerais; C3 and C4,

typical from Goiäs; C5 and C6, vestigial or absent; C7, only for H. anataliasiasi.

Dorsal head shape patterns. - G1-G3, typical from Lagoa Santa, Minas Gerais; G4-GS,

typical from Barreiras, Bahia; G6-G7, typical from central Minas Gerais.

Corresponding editors: Ronald G. ALTIG & Alain DUBOIS.

©ISSCA 1999

Source : MNHN, Paris

190 ALYTES 16 (3-4)

Dumerilia

Publication de l'Association des Amis du Laboratoire des Reptiles et Amphibiens

du Muséum national d'Histoire naturelle de Paris

Le nom Dumerilia est un hommage à la mémoire d'André-Marie-Constant DUMÉRIL

(1774-1860), l'un des fondateurs de l'herpétologie, premier auteur du magistral traité intitulé

Erpétologie générale (1834-1854).

Le périodique Dumerilia a pour objectif de faire connaître les résultats de recherches

originales concernant tous les aspects de l'étude des Amphibiens et des Reptiles, c' dire aussi bien

la biologie que la zoogéographie, la systématique, l'évolution, la protection, où même les problèmes

purement nomenclaturaux. Dumerilia sera une publication ouverte aux articles en français ou en

anglais. Des articles d'importance très variable, de la courte note à la volumineuse monographie,

pourront être acceptés pour publication.

Dumerilia accepte pour publication des articles rédigés en français ou en anglais. Les

manuscrits seront adressés en trois exemplaires au Rédacteur en Chef. Ils seront soumis pour examen

par le Comité de Rédaction du joumal à deux lecteurs au moins, appartenant à la communauté

scientifique intemationale.

Si possible, après acceptation, une copie du manuscrit final sur une disquette (34) sera

adressée au Rédacteur en Chef. Les logiciels de traitement de texte suivants sont acceptés: (1) de

préférence, MS Word Windows, WordPerfect ou WordStar, (2) à défaut, format DOS ASCII où MS

Word pour le Macintosh (disquette haute densité). Les tableaux présent ur disquette seront de

préférence composés à l'aide de tabulations plutôt que de successions d'espaces.

La publication d'un article dans Dumerilia est gratuite, mais celle de photographies en

couleurs est payante. Dumerilia a le plaisir d'offrir aux auteurs 25 tirés à part gratuits de chaque

article publié. Des tirés à part supplémentaires payants peuvent être commandés

Source : MNHN, Paris

Dumerilia 191

Dumerilia: NUMÉROS PARUS

Volume 1. - Patrick Davip. Liste des reptiles actuels du monde. I. Chelonii.

Le corps de l'ouvrage présente les genres, espèces et sous-espèces de l’ordre des Chéloniens, classés par

familles et sous-familles. Au sein de chaque catégorie, à partir du niveau de la super-famille, les taxons de rang

immédiatement inférieur sont disposés selon l'ordre alphabétique, à l'exception du taxon-type d'une catégorie, quelle

qu'elle soit, qui est toujours cité en premier. Ainsi, dans chaque genre, l'espèce-type est toujours placée en tête, et, au

sein d'une espèce polytypique, il en est de même pour la sous-espèce nominative. En regard de chaque taxon sont

indiqués d'abord le nom du descripteur suivi de l'année de sa publication, puis les noms vernaculaires français et

anglais, s'ils sont connus, et enfin plusieurs (rarement une seule) références bibliographiques.

This work is intended to be a list of all currently valid taxa of living chelonians. All taxa, whatever their

systematic level, are cüted along with bibliographic references, given under the form of numbers which refer 10

complete citations listed in bibliographies. Taxa are arranged by families and subfamilies, then genera, species and

subspecies. Inside categories, taxa are listed in alphabetical order, wüth the exception of the type-taxon, whatever its

level, which is always given first inside the category. For every taxon, name(s) of describer(s), year of description

and common names in French and in English are given, followed by numbers refering to bibliographical references.

Volume 2. - 6 articles, par 7 auteurs:

Georges PASTEUR - Biodiversité et Reptiles: diagnoses de sept espèces nouvelles fossiles et actuelles du genre de

lézards Lygodactylus (Sauria, Gekkonidac).

Roger BOUR - Une nouvelle espèce de tortue terrestre dans le Péloponnèse (Grèce).

Alain DUBOIS - The valid scientific name of the Italian treefrog, with comments on the status of some early scientific

names of Amphibia Anura, and of some articles ofthe Code conceming secondary homonyms.

Roger BOUR, Alain DUBOIS & Robert G, WEBB. - Types of recent trionychid turtles in the Muséum national d'Histoire

naturelle, Pari

Edouard-Raoul BRYG00. - Constant DUMÉRIL était aussi entomologiste.

Roger BOUR, Jean-Michel PROBST & Sonia RIBÉS. - Phelsuma inexpectata Mertens, 1966, le lézard vert de Manapany-

les-Bains (La Réunion): données chorologiques et écologiques (Reptitia, Gekkonidae)

Volume 3. - Patrick Davin & Ivan Inecn. Les serpents venimeux du monde: systématique et répartition

Cet ouvrage fournit une liste à jour commentée des serpents venimeux du monde, avec leur répartition par

pays, établie d'après la littérature récente. Les principales références taxinomiques sont données pour tous les genres et

espèces valides, ainsi que leurs noms communs en Français et en Anglais. Les sous-espèces valides sont mentionnées.

La répartition de chaque espèce est donnée par l'énumération de tous les pays habités par le taxon, accompagnée de

références bibliographiques taxinomiques et biologiques. Enfin, une liste des espèces venimeuses est présentée pour

chaque pays du globe.

This work gives an up-t0-date, annotated list of venomous snake species ofthe world with their distribution

by country, derived from the recent literature. Essential taxonomie data and references are also provided for all valid

genera and species, along with common French and English names. Valid subspecies are mentioned. The distribution

is provided for each species by enumerating all countries in which it occurs, along with selected systematic,

distributional, and biological bibliographic references. The work also provides a list of venomous snake species for

each country in the world.

Volume 4, à paraître en plusieurs fascicules, fascicule (1) paru

Alain DUBOIS & Annemarie OHLER. - À new species of Leptobrachium (Vibrissaphora) from northern Vietnamvith a

review ofhe taxonomy ofthe genus Leptobrachium (Pelobatidae, Megophryinae)

Source : MNHN, Paris

192 ALYTES 16 (3-4)

COMITÉ DE RÉDACTION POUR 1999

Rédacteur en Chef. Roger Bour (Laboratoire des Reptiles et Amphibiens, Muséum national d'Histoire naturelle,

25 rue Cuvier, 75005 Paris, France).

Rédacteur Adjoint. Alain Duvois (Laboratoire des Reptiles et Amphibiens, Muséum national dflistoire naturelle,

25 rue Cuvier, 75005 Paris, France).

Comité de Rédaction: Robert BaRBAULT (Paris, France), Aaron M. BauER (Villanova, U.S.A.), Michel BREUIL

(Paris, France), Edouard-Raoul BRYG00 (Paris, France), James R. BuskiRk (Oakland, U.S.A.), José M. Cet

(Cascais, Portugal), Mare CHEvLan (Montpellier, France), Heinz GriLuirscH (Wien, Austria), France DE

LaPPaRENT (Paris, France), Raymond F. LAURENT (Tucumän, Argentine), Colin J. McCarniy (London,

United Kingdom), Annemarie OuLER (Paris, France), Hidetoshi Ora (Okinawa, Japan), Gcorges PASTEUR

(Montpellier, France), Olivier S. G, PauweLs (Bruxelles, Belgique), Peter C. H. Prrrciarp (Orlando, U.S.A.),

Jean-Claude RAGE (Paris, France), Hussam ZAHER (Paris, France), George R. Zu (Washington, U.S.A.).

Secrétariat de Rédaction (Paris, France): Roger BoUR; Alain DuBoIs, Annemarie OHLER

TARIFS POUR 1999

Francs us

Français Dollars

Dumerilia

Volume 1, juin 1994, 128 pages.

Patrick DAVID. Liste des reptiles actuels du monde. 1. Chelonii

Volume 2, octobre 1995, 124 pages.

6 articles par 7 auteurs (Roger BOUR, Edouard Raoul BRYG00,

Alain DUBOIS, Georges PASTEUR, Jean-Michel PROBST,

Sonia RIBES, Robert G. WEB)

Volume 3, février 1999, 500 pages. avant le 01.06.99

Patrick DAVID & Ivan INEICH, Les serpents Venimeux du monde: après le 01.06.99

systématique et répartition

Volume 4, à paraître en plusieurs fascicules en 1998 avant le 01.06.99

et 1999, fascicule (1) paru en octobre 1998. après le 01.06.99

Plusieurs articles par différents auteurs (dont Herman A.

IN DEN BOSCH, Roger BOUR, Patrick DAVID, Annemarie OHLER,

Olivier PAUWELS, Hussam ZAHER); au total environ 200 pages

MODES DE PAIEMENT

- Paiement en Francs Français: par chèque à l'ordre de “AALRAM”, adressé à: AALRAM, clo Alain DUBOIS,

Laboratoire des Reptiles et Amphibiens, Muséum national d'Histoire naturelle, 25 rue Cuvier, 75005 Paris,

France,

- Payment in U. S. Dollars: by cheque payable to “AALRAM”, sent to: Dr. John B. IVERSON, Department of Biology,

Earlham College, Richmond IN 47374-4095, U.S.A.

Source : MNHN, Paris

AINTTES

International Journal of Batrachology

published by ISSCA.

EDITORIAL BOARD

Chief Editor: Alain Dupois (Laboratoire des Reptiles et Amphibiens, Muséum national d'Histoire naturelle,

25 rue Cuvier, 75005 Paris, France).

Deputy Editor: Janalee P. CALDWELL (Oklahoma Museum of Natural History, University of Oklahoma,

Norman, Oklahoma 73019, USA).

Editorial Board! Jean-Louis ALBARET (Paris, France); Franco ANDREONE (Torino, Italy); Günter GOLLMANN

(Wien, Austria); Tim HALLIDAY (Milton Keynes, United Kingdom); W. Ronald HEYER (Washington,

USA); Esteban O. LAVILLA (Tucumän, Argentina): Karen R. Lirs (Canton, USA); Thierry LODÉ

(Angers, France); Masafumi Marsut (Kyoto, Japan); John C. Poynron (London, England); Erik R.

WiLD (Dubuque, USA).

Technical Editorial Team (Paris, France): Alain Dugois (texts); Roger Bour (tables); Annemarie OHLER

(figures).

Book Review Éditor: Annemarie OuLer (Paris, France).

SHORT GUIDE FOR AUTHORS

(for more detailed Instructions to Authors, see Alytes, 1997, 14: 175-200)

Alytes publishes original papers in English, French or Spanish, in any discipline dealing with amphibians.

Beside articles and notes reporting results of original research, consideration is given for publication to synthetic

review articles, book reviews, comments and replies, and to papers based upon original high quality illustrations

(such as colour or black and white photographs), showing beautiful or rare species, interesting behaviours, etc.

The title should be followed by the name(s) and address(es) of the author(s). The text should be

typewritten or printed double-spaced on one side of the paper. The manuscript should be organized as follows:

Énglish abstract, introduction, material and methods, results, discussion, conclusion, French or Spanish

abstract, acknowledgements, literature cited, appendix.

Figures and tables should be mentioned in the text as follows: fig. 4 or tab. 4. Figures should not exceed

16 * 24 cm. The size of the lettering should ensure its legibility after reduction. The legends of figures and tables

should be assembled on a separate sheet, Each figure should be numbered using a pencil.

References in the text are to be written in capital letters (BOURRET, 1942; GRAF & POLLS PELAZ, 1989;

INGER et al., 1974). References in the Literature Cited section should be presented as follows:

BouRker, R., 1942. — Les batraciens de l'Indochine. Hanoï, Institut Océanographique de l’Indochine:i-x+ 1-547,

1. 1-4.

Grar, L.-D. & Pouus PELAz, M., 1989. - Evolutionary genetics of the Rana esculenta complex. In: R. M. DAWLEY

& I. P. BoGarT (ed.), Evolution and ecology of unisexual vertebrates, Albany, The New York State

Museum: 289-302.

INGER, RE, Vors, H. K. & Voris, H. H., 1974. - Genetic variation and population ecology of some Southeast

Asian frogs of the genera Bufo and Rana. Biochem. Genet., 12: 121-145.

Manuscripts should be submitted in triplicate either to Alain DUBOIS (address above) if dealing with

amphibian morphology, systematics, biogeography, evolution, genetics or developmental biology, or to Janalee

P. CALDWELL (address above) if dealing with amphibian population genetics, ecology, ethology or life history.

Acceptance for HS will be decided by the editors following review by at least two referees.

If possible, after acceptance, a copy of the final manuscript on a floppy disk (3 4 or 5 ) should be sent

to the Chief Editor. We welcome the following formats of text processing: (1) preferably, MS Word (1.1 to 6.0,

DOS or Windows), WordPerfect (4.1 to 51, DOS or Windows) or WordStar (3.3 to 7.0); (2) less preferably,

DR (ASCII) or DOS-formated MS Word for the Macintosh (on a 3 high density 1.44 Mo floppy

isk only).

Page charges are requested only from authors having institutional support for this pi se. The

publication of colour photographs is charged. For each published paper, 25 free reprints are offered by ISSCA

Lo the author(s). Additional reprints may be purchased.

Publish with the support of AALRAM

(Association des Amis du Laboratoire des Reptiles et Amphibiens

du Muséum National d'Histoire Naturelle, Paris, France).

Directeur de la Publication: Alain DuBois.

Numéro de Commission Paritaire: 64851.

Alytes, 1999, 16 (3-4): 77-192.

Contents

Wilmar BoLivar-G., Taran GRANT & Luis A. OsORIO

Combat behavior in Centrolene buckleyi and other centrolenid frogs ... 71-83

Paola M. PELTZER & Rafael C. LAJMANOVICH

Anälisis tréfico en dos poblaciones de Scinax nasicus

(Anura Hylhidae)de ATSENtTa 2... eh te res 84-96

Jeremy L. MARSHALL

The life-history traits of Eurycea guttolineata (Caudata,

Plethodontidae), with implications for life-history evolution ........... 97-110

Maxime LAMOTTE & Coralia SANCHEZ-LAMOTTE

Adaptation aux particularités climatiques du cycle biologique d’un

anoure tropical, Nectophrynoides occidentalis Angel, 1943 (Bufonidae) . 111-122

Michael FRANZEN

Notes on morphological variation and the biology of Nototriton

guanacaste Good & Wake, 1993 (Caudata, Plethodontidae) ........... 123-129

Alain PAGANO & Pierre JOLY

Limits of the morphometric method

for field identification of water frogs ......,:.....:1.......MMe mn 130-138

E. O. LavicLA & Rafael O. DE SÂ

Estructura del condrocräneo y esqueleto visceral

de larvas de Pseudis minuta (Anura, Pseudidae) ....................... 139-147

Anne M. MAGLIA

The adult skeleton of Spea multiplicata and a comparison of the

osteology of the pelobatid frogs (Anura, Pelobatidae) .… 148-164

Marcelo Felgueiras Napoi & Ulisses CARAMASCHI

Geographic variation of Hyla rubicundula and Hyla anataliasiasi,

with the description of a new species (Anura, Hylidae) ................ 165-189

ANNOUNCEMENT

TO OT MR ee MO RO en ne ur rt ot ec ee 190-192

Alytes is printed on acid-free paper.

Alytes is indexed in Biosis, Cambridge Scientific Abstracts, Current Awareness in Biological

Sciences, Pascal, Referativny Zhurnal and The Zoological Record.

Imprimerie F. Paillart, Abbeville, France.

Dépôt légal: 2°"° trimestre 1999.

Source : MNHN, Paris: