Analysis of the Intcrgencric Relationships of the Australian Frog Family Myobatrachidae

W. RONALD HEYER and DAVID S. LIEM

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 233 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti- tution, Joseph Henry articulated a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge." This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, com- mencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of professional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. These pub- lications are distributed by mailing lists to libraries, laboratories, and other interested institutions and specialists throughout the world. Individual copies may be obtained from the Smithsonian Institution Press as long as stocks are available.

S. DILLON RIPLEY Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 233

Analysis of the Intergeneric Relationships of the Australian Frog Family Myobatrachidae

W. Ronald Heyer and David S. Liem

SMITHSONIAN INSTITUTION PRESS City of Washington 1976 ABSTRACT Heyer, W. Ronald, and David S. Liem. Analysis of the Intergeneric Relation- ships of the Australian Frog Family Myobatrachidae. Smithsonian Contributions to Zoology, number 233, 29 pages, 28 figures, 2 tables, 1976.—Forty-two characters of external morphology, myology, osteology, and life history are studied. For each character, the evolutionary directions of changes of states are inferred. This information is used to construct a phylogenetic hypothesis of the intergeneric relationships of the primarily Australian frog family Myobatrachidae. Three major groupings of myobatrachids are proposed; these are accorded subfamilial status. One of the subfamilies is defined as new. In addition, three genera are proposed as new. It is proposed that the ancestral myobatrachid stock was associated with the Antarcto-Tertiary Geoflora. Several genera still demonstrate a distribution asso- ciated with this beech-podocarp vegetation. Major evolutionary events in the family have been associated with stream adaptations, trends toward and includ- ing direct development and fossorial adaptations. Convergence in these events with the leptodactylid frog fauna of southern South America is commented upon.

OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESICN: The coral Montastrea cavernosa (Linnaeus).

Library of Congress Cataloging in Publication Data Heyer, W Ronald. Analysis of the intergeneric relationships of the Australian frog family Myobatrachidae. (Smithsonian contributions to zoology ; 233) Bibliography: p. 1. Myobatrachidae. 2. Amphibians—Evolution. 3. Amphibians—Classification. 4. Amphibians —Australia. I. Liem, David S., joint author. II. Title. III. Series: Smithsonian Institution. Smithsonian contributions to zoology ; 233. QL1.S54 no. 233 [QL668.E2615] 591'.08s [597'.8] 76-608054 Contents

Page Introduction 1 Methods and Materials 1 Acknowledgments 2 Character Analysis 3 Analysis of Relationships 5 Sister Groups within the Phylogeny 10 Comparison with Previous Schemes 10 Systematic Conclusions 11 RHEOBATRACHINAE, new subfamily 11 MYOBATRACHINAE Schlegel, 1850 11 Crinia Tschudi, 1838 11 Geocrinia Blake, 1973 11 Pseudophryne Fitzinger, 1843 12 Ranidella Girard, 1853 12 Paracrinia, new genus 12 Australocrinia, new genus 12 Kankanophryne, new genus 12 Phyletic Character Analysis 12 Evolutionary Zoogeography 15 Generic Distribution Patterns 15 Evolutionary Trends within the Subfamilies 23 Subfamily RHEOBATRACHINAE 23 Subfamily MYOBATRACHINAE 23 Subfamily LIMNODYNASTINAE 24 Major Trends among Subfamilies 24 Comparisons with Previous Zoogeographic Schemes 25 Myobatrachid-Leptodactylid Comparisons 26 Literature Cited 28

in

Analysis of the Intergeneric Relationships of the Australian Frog Family Myobatrachidae

W. Ronald Heyer and David S. Liem

Introduction mine directionality of states, are also deleted. The taxa used for analysis are those recognized The intergeneric relationships of the Australian by Liem (in press) with certain exceptions. Blake frog family Myobatrachidae (Leptodactylidae of (1973) and Liem (in press) used similar numerical most authors) have been analyzed recently. Lynch techniques of analysis on some of the same genera (1971) used a standard systematic approach to the with conflicting systematic results. For purposes of problem while Liem (in press) used a numerical analysis, the smallest generic groupings are used taxonomy approach. Neither author used the to test the hypotheses of Blake and Liem. In addi- phylogenetic approach as outlined by Hennig tion to the genera recognized by Liem (in press), (1966). The purpose of this paper is to apply the the following units are added: (1) Crinia geor- same methods of analysis of relationships to the giana, (2) Crinia haswelli, (3) Ranidella (as used Australian myobatrachids as have been applied by Blake, 1973), and (4) Pseudophryne occi- recently to the Neotropical leptodactylids (Heyer, dentalis. 1975a). The criteria used for determining directionality METHODS AND MATERIALS.—Character states are of character states are the same as those used categorized for a number of characters from the previously (Heyer, 1975a). In those cases where study sample. The directionality of states is the character states of this study are the same as analyzed and the information from derived states those used in the leptodactylid study (Heyer, is used to generate possible phylogenetic trees. 1975a), the arguments for directionality are the The characters used are mostly those described same and are not recounted here. This omission by Liem (in press). Liem's study included mem- is occasioned by the other study's having used the bers of the family Pelodryadidae. Characters in myobatrachids as the major outgroup for determi- Liem's study which have uniform states in the nation of state directionality. A primitive state for Myobatrachidae are not used. Some other of the leptodactylids can also be assumed to be the Liem's characters, for which not enough compar- primitive state for the myobatrachids if the able information is known in other frogs to deter- assumption is made that the myobatrachids and W. Ronald Heyer, Division of Reptiles and Amphibians, De- leptodactylids had a common ancestor. The primi- partment of Vertebrate Zoology, National Museum of Nat- tive state is always coded as zero in the character ural History, Smithsonian Institution, Washington, D.C. analysis section. 20560. David S. Liem, Wildlife Branch, Department of Natural Resources, P.O. Box 5140, Boroko, Papua, New Life history and geographic range data were Guinea. taken from the published literature (most of the SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY papers included in "Literature Cited"), Ian R. tanelle; state 1, frontoparietal fontanelle covered, Straughan's unpublished dissertation (1966), and at least in some species. personal observations (DSL). Heyer (1975a) showed that state 0 is the primi- ACKNOWLEDGMENTS.—We wish to thank Ian tive state. Straughan, Southern California Edison Company CHARACTER 5, sphenethmoid: state 0, sphen- and George R. Zug, Smithsonian Institution, for ethmoid entire; state 1, sphenethmoid divided, at their careful reading of the manuscript. least in some species. Thanks are due also to Hobart M. Smith, Uni- Trueb (1973) thought state 1 reflects reduced versity of Colorado, who originally suggested the ossification. Among the New World and collaboration that produced this manuscript. African leptodactyloids, only Euparkerella and This work was supported by a Smithsonian Tomodactylus have state 1 (Lynch, 1971). State Research Foundation Award to the first author 1 is considered the derived state (criterion II of and by ARGC grants (D70/17419 and D70/17559) Heyer, 1975a). and the University of Queensland Research Allo- CHARACTER 6, squamosal: state 0, zygomatic and cation to the Department of Zoology to the second otic rami about equal length, both moderately author during his tenure at that Department. long; state 1, zygomatic ramus shorter than otic ramus, at least in some species. Character Analysis Heyer (1975a) showed state 0 to be primitive. CHARACTER 7, vomer and vomerine teeth: state CHARACTER 1, maxillary teeth: state 0, maxillary 0, vomer and vomerine teeth present uniformly; teeth present, pedicellate; state 1, maxillary teeth state 1, intrageneric variation, all species with present, fanglike, at least in some species; state 2, vomer present, at least some species with vomerine maxillary teeth uniformly absent or absent in teeth absent; state 2, vomer present, vomerine some species. teeth absent uniformly; state 3, intrageneric varia- Trueb (1973:79) indicated that states 1 and 2 tion, at least some species with vomer and vomer- are independently derived states from state 0. The ine teeth absent; state 4, vomer and vomerine directions of change of character states are: teeth uniformly absent. Heyer (1975a) indicated state 0 to be the primi- 1 « 0 • 2 tive state. The direction of change of character CHARACTER 2, nasal contact with frontoparietal: states is: state 0, nasals not in contact with frontoparietal; state 1, nasals in contact with frontoparietal. Heyer (1975a) indicated that state 0 is primi- CHARACTER 8, intervertebral disc: state 0, inter- tive. vertebral disc fused to the posterior end of the CHARACTER 3, carotid canal: state 0, exposed centrum (procoelous); state 1, intervertebral disc shallow channel or carotid canal absent; state 1, free between centra of presacral vertebrae (amphi- exposed shallow canal present, at least in some coelous). species; state 2, carotid canal present, at least in Trueb (1973) considered amphicoely as the some species. probable ancestral frog condition, but thought it Lynch (1971) suggested that the carotid canal likely that amphicoely in modern frogs is a is formed by thickening of the frontoparietal paedomorphic trait, hence derived. The African posteriorly and/or involving dermal coossification leptodactyloid, Heleophryne, is amphicoelous, the resulting in the enclosure of the carotid artery in New World leptodactylids are uniformly pre- a bony canal. Since the vast majority of frogs have coelous. Thus, the closest outgroup members do state 0, it is considered the primitive state (crite- not clearly indicate which state is primitive. rion II of Heyer, 1975a). The direction of change Expansion of the outgroup clarifies the situation. of character states is: Bufonids, hylids, pseudids, dendrobatids, and 0 • 1 • 2 centrolenids are uniformly procoelous. State 0 is CHARACTER 4, frontoparietal fontanelle: state 0, considered the primitive state. frontoparietals separated medially, exposing fon- CHARACTER 9, cervical vertebrae: state 0, cervical NUMBER 233 and second vertebrae free; state 1, cervical and have state 0 (Lynch, 1971). State 0 is assumed the second vertebrae fused. primitive state (Criterion I of Heyer, 1975a). Trueb (1973) stated the fused condition to be CHARACTER 18, alary process of the hyoid: state derived. Lynch (1971) found state 0 uniformly in 0, alary process broad and wing-like; state 1, alary the African and Neotropical leptodactyloids. State process stalked, distally expanded into rounded 0 is considered the primitive state (criterion I of or oval plates. Heyer, 1975a). Heyer (1975a) considered state 1 to be the primi- CHARACTER 10, cervical cotyles: state 0, cervical tive state, but found it to be a poor predictor of cotyles narrowly separated; state 1, cervical cotyles phylogenetic relationships. In a consideration of widely separated. the relationships among the species of the genus Heyer (1975a) indicated state 0 to be the primi- Adenomera (Heyer, 1975b), the better assumption tive state. appeared to be that state 0 is the primitive state. CHARACTER 11, relationship of transverse process State 0 is here considered the primitive state. of last presacral vertebra to sacral vertebra: state CHARACTER 19, anterior process of the hyale: 0, last presacral vertebra noticeably shorter than state 0, anterior process present; state 1, anterior sacrum; state 1, last presacral vertebra about same process absent. width as sacrum. Heyer (1975a) indicated state 0 to be primitive. Heyer (1975a) indicated the primitive state to CHARACTER 20, depressor mandibulae muscle: be 0. state 0, slip from dorsal fascia present; state 1, CHARACTER 12, sacral diapophyses: state 0, sacral intrageneric variation, some species with state 0, diapophyses expanded; state 1, sacral diapophyses others with state 2; state 2, no dorsal fascia slip. rounded uniformly, in at least some species. The condition for Assa is not known and Heyer (1975a) indicated the primitive state to assumed to be primitive for purposes of analysis. be 0. Heyer (1975a) indicated that state 0 is primitive CHARACTER 13, coccyx: state 0, prezygapophyses relative to state 2 as described here. The direction and transverse processes present: state 1, prezy- of change of character states is: gapophyses and transverse processes absent. Trueb (1973) suggested state 0 to be primitive. CHARACTER 14, dorsal crest of ilium: state 0, no CHARACTER 21, adductor mandibularis muscle: dorsal crest; state 1, well denned crest -present. state 0, both adductor mandibulae posterior sub- Heyer (1975a) indicated the primitive state to externus and adductor mandibulae externus super- be 0. ficialis present; state 1, adductor mandibulae CHARACTER 15, carpals: state 0, first, second, posterior subexternus only present; state 2, adduc- third carpale and naviculare free; state 1, at least tor mandibulae externus superficial only present. one element of state 0 demonstrating fusion. Heyer (1975a) indicated state 0 is primitive. Geocrinia victoriana is the only member of the Directions of change of character states are: genus demonstrating state 1. Since all other species have state 0, Geocrinia is coded as state 0. Trueb (1973) and Liem (1970), following Howes and CHARACTER 22, supracoracoideus superficialis and Ridewood (1888), indicated state 0 to be the profundus muscles: state 0, muscles separate; state primitive state. 1, muscles fused, at least in some species. CHARACTER 16, prehallux: state 0, prehallux of Jones (1933) demonstrated that state 0 is the 1 to 3 segments with blunt or pointed distal seg- general state in frogs. State 0 is assumed the ments; state 1, prehallux of one shovel shaped primitive state (Criterion II of Heyer, 1975a). segment, at least in some species. CHARACTER 23, adductor longus muscle: state 0, Heyer (1975a) indicated state 0 to be primitive. muscle inserting on knee; state 1, muscle inserting CHARACTER 17, cricoid: state 0, cricoid forming on distal 1/j, to 1/3 of adductor magnus muscle or a complete ring; state 1, cricoid not forming a absent. complete ring. Heyer (1975a) indicated that state 0 is primitive. All Neotropical and African leptodactyloids CHARACTER 24, semitendinosus muscle distal ten- SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY don insertion: state 0, distal tendon insertion ven- Heyer (1975a) indicated the direction of change tral to femoral tendon of gracilis muscle; state 1, of states is: distal tendon insertion dorsal to femoral tendon 0 • 1 • 2 of gracilis muscle; state 2, distal tendon piercing CHARACTER 29, geniohyoideus medialis muscle: the gracilis muscle or femoral tendon of the state 0, muscle contiguous medially; state 1, muscle gracilis muscle. separated medially. State 0 is the common state among the frogs Heyer (1975a) indicated state 0 is the primitive considered to be related to, or primitive to, the state. myobatrachids (Dunlap, 1960). State 0 is assumed CHARACTER 30, sternohyoideus ventralis muscle the primitive state (Criterion II of Heyer, 1975a). insertion: state 0, muscle insertion near lateral The directions of changes of character states are: edge of hyoid body; state 1, muscle insertion extending to midline of hyoid body. Heyer (1975a) indicated state 0 is the primitive Taiidactylus species demonstrate both states 1 state. and 2. As these states are coded as independently CHARACTER 31, petrohyoideus anterior muscle: derived, the genus is coded as having state 0 for state 0, muscle insertion on lateral edge of hyoid purposes of analysis. plate; state 1, muscle insertion on ventral body CHARACTER 25, adductor magnus muscle: state 0, of hyoid. accessory head absent; state 1, accessory head Heyer (1975a) indicated state 0 is the primitive present. state. State 0 is found in the Ascaphidae, Discoglos- CHARACTER 32, petrohyoideus posterior muscles: sidae, Liopelmatidae, Microhylidae, Pelobatidae, state 0, three slips present; state 1, two slips pres- Pipidae, and Rhinophrynidae; state 1 is the usual ent. state for the remaining families (Dunlap, 1960). Most frogs have state 0 (Trewavas, 1933; Liem, State 0 is assumed to be the primitive state 1970); the state is assumed primitive (Criterion II (Criterion II of Heyer, 1975a, for expanded out- of Heyer, 1975a). group). CHARACTER 33, dilatator laryngis muscles: state CHARACTER 26, extensor cruris brevis muscle: 0, one pair of slips present; state 1, some species state 0, muscle as long as tibio-fibula; state 1, with one pair, others with two pairs of slips pres- muscle from y to 2/ length of tibio-fibula. 3 3 ent; state 2, two pairs of slips present. The common condition among frogs is state 0 The usual state for frogs is state 0 (Trewavas, (Dunlap, 1960). State 0 is assumed the primitive 1933); state 0 is assumed the primitive state (criter- state (Criterion II of Heyer, 1975a). ion II of Heyer, 1975a). The direction of change CHARACTER 27, abductor brevis plantaris hallucis of states is: muscle: state 0, muscle absent; state 1, muscle present, no accessory slip; state 2, muscle and 0 • l • 2 accessory slip present. CHARACTER 34, pupil shape: state 0, pupil verti- The muscle is absent in Ascaphus, Leiopelma, cal; state 1, pupil horizontal. Pipa, and Rhinophrynus (Dunlap, 1960), suggest- Heyer (1975a) indicated state 0 is the primitive ing that this is the primitive state (Criterion II of state. Heyer, 1975a). Liem (in press) stated that an CHARACTER 35, abdominal skin texture: state 0, accessory slip was essentially unique to certain skin of abdomen smooth; state 1, skin of abdomen myobatrachids. The direction of change of char- coarsely granular in some or all species. acter states is: The majority of frogs have state 0; it is assumed primitive (criterion II of Heyer, 1975a). CHARACTER 36, tongue: state 0, tongue free CHARACTER 28, Omohyoideus: state 0, muscle posteriorly; state 1, tongue adherent to mouth present in all species; state 1, muscle present in floor. some species, absent in others; state 2, muscle About 10 genera of frogs have state 1; the state absent in all species. is considered derived (criterion II of Heyer, 1975a). NUMBER 233

CHARACTER 37, toe webbing; state 0, toes conclusions can be drawn from the phylogeny of webbed; state 1, toes fringed in some or all species; Figure 1. There are three major groupings of state 2, toes free in some or all species. myobatrachid frogs: Rheobatrachus forms a group Heyer (1975a) indicated the direction of change by itself and apparently has little if any relation- of character states to be: ship to the remaining genera; the remaining genera form two major clusters, corresponding with taxo- nomic proposals made by previous workers. The CHARACTER 38, inner metatarsal tubercle: state cluster including Myobatrachus corresponds with 0, inner metatarsal tubercle not enlarged and the subfamily Myobatrachinae; the cluster includ- shovel-shaped; state 1, inner metatarsal tubercle ing Limnodynastes corresponds with the subfamily enlarged, shovel-shaped, compressed, and keeled. Limnodynastinae (Liem, in press) (=Cycloraninae State 0 is considered primitive (Heyer, 1975a). of Lynch, 1971, and Parker, 1940, except for the CHARACTER 39, outer metatarsal tubercle: state exclusion of Cyclorana). The basal clusters defin- 0, absent; state 1, absent in some species, present ing the two groups each contain unique state in others; state 2, present, unmodified; state 3, appearances within the tree. As discussed later, present, strongly compressed. three of the muscle characters may be miscoded; Heyer (1975a) indicated state 0 to be the primi- nevertheless, each cluster still has unique state tive state. The direction of change of character appearances within the tree. These two groups are states is: convincing sister groups in terms of Hennig's (1966) terminology. There are no other meaning- ful arrangements of these two groups. CHARACTER 40, intermandibularis muscle: state There are several rearrangements possible within 0, no intermandibularis-submentalis connection; each group, however, that need detailed comment. state 1, intermandibularis-submentalis connection The cluster of Assa, Metacrinia, Pseudophryne, present. and Pseudophryne occidentalis is the strongest The generalized distribution of state 0 (Tyler, cluster in the data set, indicating that the closest 1971) suggests this state to be primitive (criterion relationship of any of these taxa is with the other II of Heyer, 1975a). three. Characters not used in the analysis provide CHARACTER 41, egg mass: state 0, not foamy; evidence to suggest a rearrangement within the state 1, foamy. cluster. First, however, the cluster of Pseudophryne Heyer (1975a) indicated state 1 to be derived. occidentalis and Pseudophryne is substantiated: all CHARACTER 42, site of egg development: state members lack tympani and columellae, have a 0, aquatic; state 1, non-aquatic, including terres- unique locomotion of crawling (except P. douglasi) trial and parental brooding. (Liem, in press), and have a uniform mating call Heyer (1975a) indicated that state 0 is primitive. (Liem, in press). Pseudophryne does form a sister The distribution of states among the taxa is group with P. occidentalis; all Pseudophryne (ex- presented in Table 1. The recoding of states as cept P. occidentalis) have the derived state of the used in the next section is presented in Table 2. depressor mandibulae muscle and have femoral glands (a derived condition, Heyer, 1975a), while P. occidentalis has narrow sacral diapophyses, a Analysis of Relationships compressed outer metatarsal tubercle, and two The combinatorial method of Sharrock and slips of the dilatator laryngis muscles (Liem, in Felsenstein (1975) was used to organize the data. press). These morphological differences indicate The combinatorial method locates all monothetic different adaptive complexes in these two groups, clusters in the data set and prints them out. Infor- differences best reflected at the generic level. Liem mation from the character state trees was included; (in press) pointed out the distinctiveness of the clusters were formed with derived states P. occidentalis but preferred not to recognize it only. The construction of the first phylogeny at the generic level. Blake (1973) did not include (Figure 1) maximized the number of shared P. occidentalis in his analysis. derived states at each clustering point. Certain Blake (1973) proposed placing Metacrinia in the SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 1.—Distribution of character states among genera as used for computer coding ("9" indicates no information)

Character Numbers and States 1 rs i 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Genus h- i Adelotus h- i 0 0 h- i 0 0 0 0 1 0 0 1 1 1 0 0 1 1 0 1

Assa 0 0 0 0 1 1 2 1 0 1 0 0 1 0 1 0 h- i 0 1 9 1 h- i C. georgiana 0 0 0 0 M 1 0 1 0 0 1 1 0 0 0 1 0 0 0 1 h- i C. haswelli 0 0 0 0 1 1 2 1 0 1 0 h- i h- i 0 0 0 0 0 0 1 Geocrinia 0 0 0 0 1 1 1 1 0 1 0 i— i 1 0 0 0 1 0 0 0 1 h- i Ranidella 0 0 0 0 h- l 1 3 1 0 1 0 1 1 0 0 0 0 0 1 1 M Glauertia 2 0 1 0 1 1 4 h- i 0 0 0 1 1 0 1 1 1 0 1 2

Heleioporus 0 0 2 0 0 0 0 0 1 0 0 1 h- i 0 1 1 0 1 0 0 1 Kyarranus 0 0 2 0 0 0 0 1 0 0 1 1 0 1 0 0 1 1 0 1

r\ 5 M n 1 T r> Lechriodus U I 1 un V Un u 1 u 1 U X J- Un U 0 1 1 0 h- i h 1 0 1 Limnodynastes 0 0 0 0 0 0 0 1 0 0 r H 0 1 0 Metacrinia 2 0 0 0 1 1 2 1 0 1 0 1 0 1 0 0 0 1

i-i 1 1 r\ n D h- i 1 D h- i r\ r\ Mixophyes u 1 z 1 un n U u u u 1 U U 1 h- i C Myobatrachus 2 1 0 0 h- C 4 1 0 0 0 1 1 0 1 0 1 0 0 1 Neobatrachus 0 0 1 0 01 0 0 0 1 0 0 1 1 0 1 1 0 1 0 0 1

Notaden 2 0 1 0 1 1 0 0 1 0 0 1 M 0 1 1 0 1 1 0 1 i- H h- l Philoria 0 1 2 0 0 0 0 0 0 0 1 1 0 1 1 0 h- l 2 1 1 h- h- i Platyplectron 0 0 2 0 0 0 0 0 1 0 0 1 1 0 r- i 1 0 0 0

Pseudophryne 2 0 0 0 1 1 3 1 0 h- i 0 0 1 0 h- i 0 1 0 1 2 1 h- l h- i h- l P. occidentalis 2 0 0 0 1 4 1 0 h- l 0 1 0 1 0 1 0 0 1 Rheobatrachus 1 0 0 0 0 1 2 0 0 0 0 0 0 0 1 0 0 0 1 0 1

Taudactylus 0 0 0 1 1 h- l 2 1 0 h- l 0 0 1 0 0 0 1 0 1 0 1 h- l h- l Uperoleia 1 0 2 1 3 1 0 h- 1 0 0 1 0 1 0 1 0 1 2 1

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Genus

Adelotus 1 1 2 1 1 2 2 0 1 0 0 0 1 0 1 0 0 2 1 1 0 Assa 1 0 1 0 0 0 2 1 1 1 1 0 1 0 0 2 0 2 0 0 1

C. georgiana 1 0 1 0 0 0 0 h- i 1 1 0 0 1 1 0 2 0 2 0 0 0 C. haswelli 1 0 1 0 0 0 2 1 1 1 0 2 i- t 0 0 1 0 2 0 0 0

Geocrina 1 0 1 0 0 0 1 1 h- i 1 0 1 1 0 0 2 0 0 0 0 1 1

Ranidella 1 0 1 0 0 0 0 1 1 h- 0 0 h- i 1 0 2 0 2 0 0 0

Glauertia 1 0 1 0 0 0 2 1 1 1 0 Is ) 1 1 0 0 0 3 0 0 0 h- ' h- l Heleioporus 0 0 1 0 0 0 0 0 0 0 0 0 0 1 h- i 0 1 1 1

Kyarranus 1 h- l 0 1 0 2 2 0 1 0 0 0 1 0 0 2 0 0 1 1 1

Lechriodus 1 1 2 1 1 1 0 0 0 0 0 0 1 0 0 2 0 0 1 h- 1 0 h- l Limnodynastes 1 1 2 0 2 2 0 0 0 0 0 H- I 0 0 2 0 1 1 h- l 0 h- i h- l M Metacrinia 0 1 0 0 0 2 1 1 1 0 h- l 0 2 0 2 0 0 1 Mixophyes 0 1 2 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1

Myobatrachus h- l 0 0 1 0 0 0 2 0 1 1 2 1 0 0 ts ) 0 0 0 0 1 Neobatrachus 0 1 0 1 1 0 2 0 0 0 0 0 0 0 0 0 1 0 1 0 0

Notaden h- l 0 0 0 1 0 2 0 1 0 0 0 0 0 0 0 1 1 1 0 0 Philoria 0 1 0 1 0 2 2 0 1 0 0 0 1 0 0 2 0 0 1 1 1 Platyplectron 1 1 2 1 1 2 1 0 0 0 0 0 0 0 0 0 1 0 1 1 0

Pseudophryne 1 0 1 0 0 0 2 1 1 1 0 0 1 0 0 2 0 2 0 0 h- l

P. occidentalis h- 1 1 0 1 0 0 0 2 1 1 0 0 1 0 0 ts ) 0 3 0 0 1 Rheobatrachus 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 2 0 0 1 h- l Taudactylus 0 0 0 0 1 0 1 1 1 0 1 h- l 0 0 1 0 2 0 0 0

Uperoleia h- 1 1 0 1 0 0 0 0 1 1 0 0 0 h- l 0 1 0 3 0 0 0 NUMBER 233

TABLE 2.—Recoding key of character states from states as analyzed to states used in the computer program

New state number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Character nunber 1 1 2 3 3 4 5 6 7 7 7 7 8 9 Old state nurrber 1 2 1 1 2 1 1 1 1 2 3 4 1 1

New state nunber 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Character nunber 10 11 12 13 14 15 16 17 18 19 20 20 21 21 T t i Old state nunber j. X X l i 2

New state nunber 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Character nunber 22 23 24 24 25 26 27 27 28 28 29 30 31 32 Old state nunber 1 1 1 2 1 1 1 2 1 2 1 1 1 1

New state nunber 43 44 45 46 47 48 49 50 51 52 53 54 55 56 Character nunber 33 33 34 35 36 37 37 38 39 39 39 40 41 42 Old state nunber 1 2 1 1 1 1 2 1 1 2 3 1 1 1

PSEUDOPHRYNE P. OCCIDENTAL IS 25,26. ,12,17,53 PHILORIA KYARRANUS 3,21,25,26 8,29 METACRINIA \ 17,42,46 4,5,24,40, y 48,49,56 1«,27,36,37,' 38,45,55 20,24,49,56

\ . GLAUERTIA MYOBATRACHUS PLATYPLECTRON 37,38 RANIDELLA 4,21,25,28, 3,27,42,48, 14,27,29,36, 11,17,25, 29,39,46,51, «9,56 46,45 52,53 / 37,50,55 \ 15,27,29,39, 2,11,12,17,20, 4,5,21,32,34 48,51,52^ 24,37,38,43,44 HEIEIOPORUS «»5,14,21,27, 31,«5 34,48,50,55,56 TAUDACTYLUS UPEROLEIA 30,35,43,45 1,4,5,11,20,25, 26,29,31,46,53 NEOBATRACHUS NOTAOEN / 30,33 2,7,8,24, 6,15,24,27' C. GECRGIANA GEOCRINIA V 29,<°,51 39,48,51,52. 46,51,52 9,37,«3,56 \ / 4,14,21,27, 15,17,27,29,31, 34,37,38,50 39,45,48,49 9,10 (NO) 17,20,2a, 51 \ 7,8,13,22,40,41,

RHEOBATRACHUS 1,8,9,10,20,24, 27,29,42,47,51, 52,56

FIGURE 1.—Predicted phylogenetic relationships among the myobatrachid genera, ("(NO)" — cluster is nonmonothetic; underlined states — unique appearances within phytogeny; see text.) 8 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY synonymy of Pseudophryne. This action is not has 5 unique states and Geocrinia 2. However, supported by the present study. Metacrinia and many of these states are the uniform expression Pseudophryne (including occidentalis) form robust of a variable state. For example, the omohyoideus sister groups: Pseudophryne lacks a columella and is absent in haswelli, but it is also absent in some crawls (except P. douglasi); Metacrinia has the Geocrinia. Thus these kinds of states can not be derived state of the posterior petrohyoideus mus- used to define clear cut sister groups. Each taxon cle, a granular belly, and terrestrial larvae. These has one unique state: C. haswelli has an outer differences are best expressed at the generic level. metatarsal tubercle; Geocrinia has terrestrial egg The relationships of Metacrinia appear to lie clos- deposition. A further complication in attempting est with Assa, in fact. Within the characters used to deduce relationships is occasioned by the fact for analysis, all of the derived states of Assa are that C. haswelli actually shares two more derived shared with Metacrinia. Thus, within the cluster states with the cluster of Assa, Metacrinia, and of four, Assa and Metacrinia together share the Pseudophryne (including P. occidentalis) than with derived state of the posterior petrohyoideus mus- Geocrinia. The taxonomic decision of whether to cle. In addition, both taxa have the derived condi- recognize C. haswelli as a distinct genus has a tion of reduction of a phalanx in the first finger, parallel, in part, with the Ranidella complex; the a state found additionally only in Uperoleia. decision involving C. haswelli will be deferred Metacrinia and Assa do form sister groups when until the situation with Ranidella is described. more life history information is added: Metacrinia On the basis of the characters analyzed for this has narrow sacral diapophyses and a granular study, all of the derived states of Crinia georgiana belly; Assa has male brooding pouches. This dis- are found within Ranidella, suggesting that these tinctive life history pattern is best expressed at the two taxa are most closely related to each other. generic level as currently recognized. Once the individual characters are examined in In the phylogeny of Figure 1, Crinia georgiana detail, it becomes apparent that R. riparia is dis- does not form a mbnothetic cluster with Geocrinia. tinctive within Ranidella, and for purposes of A better arrangement, in which all clusters are present discussion should be separated. The best monothetic, involves pairing C. georgiana with placement of Ranidella riparia is with C. georgiana Ranidella, and Geocrinia with C. haswelli. Inter- and the remaining Ranidella Liem (in press) estingly, a combination of these four taxa does not examined (Figure 2). States 9, 10, and 11 separate form a monothetic cluster. The combination of all C. georgiana from Ranidella (restricted sense). All four with Assa, Metacrinia, and Pseudophryne these states involve reduction of the vomer and (including occidentalis) is a monothetic cluster, as vomerine teeth. Vomerine teeth are absent in is a cluster of either C. georgiana-Ranidella or Ranidella (restricted sense), the vomer is present Geocrinia~C. hasivelli with Assa, Metacrinia, and or absent. Since there is a trend within Ranidella Pseudophryne. There is no clear advantage to be (restricted sense), it is not unreasonable to carry gained from choosing between these last alterna- the trend back one step further to include tives in terms of numbers of shared states within C. georgiana within the same genus. The larval clusters, or number of convergent states. Thus, evidence (Watson and Martin, 1973) does not sup- while all of the above taxa seem to be most closely port inclusion of C. georgiana in with Ranidella related among themselves, the combination of (restricted), because C. georgiana has a stream C georgiana, C. haswelli, Geocrinia, and Ranidella adapted tadpole (as do R. riparia and R. tasma- is a poor cluster since it is not monothetic. niensis). Of the species examined in this study, Placement of C. haswelli with Geocrinia agrees R. riparia is distinctive from a cluster of C. geor- with Liem (in press), but disagrees with Blake giana and Ranidella (restricted sense) in mandi- (1973), who placed C. haswelli together with bular musculature, terrestrial egg placement, and C. georgiana. Crinia georgiana and C. hasivelli do possibly loss of the columella (Blake, 1973, so not form a monothetic cluster in the present data indicates; Liem, in press, does not). The mating set, indicating that they are not most closely call of C. georgiana is a unique quacking call, related. Geocrinia and C. haswelli appear to be a distinctive within the Ranidella complex (Main, robust sister group (Figure 2) because C. haswelli 1968). NUMBER 233

PSEUDOPHRYNE P. OCCIDENTALIS METACRINIA KYARRANUS PHILORIA AOELOTUS LIMNOOYNASTES 25,26. 12,17,53 2,17,46 8,25 3,21,25, 1,6,15,24, 21,48,45 -I- 26 34,40,47, / 2,11 ASSA C. HASWELLI GEOGRINIA 42 10,38,44, 45,56 4,5,24,40, / / 48,45,56 29,32,51 17,43 38,*5 PLATYPLECTRON HELEIOPORUS LECHRIODUS 4,5,21,25,32, 27,50,56 6,16,19,24,29, RIPARIA RANIDELLA 34,50 32,35,45,49 25,26,56 11 \ 27,35,36,37 ,21,34,48 5,10

C. GEORGIANA TAUOACTYLUS UPEROLEIA 30,35,43,45 1,4,5,11,20,25 NEOBATRACHUS NOTADEN Y 26,25,31,46,53 30,33 2,7,8,24, 25,31, 6,9,10,24,51,52 C 29,40,51 4,21,27,34, GLAUERTIA MYOBATRACHUS -\ ./- 37,38,50 4,21,25,28, 3,27,42,48, 29,39,46, 45,56 15,27,35,48

MIXOPHYES 3,4,5,6,16,19,21, 30,32,33.34,35,56

RHEOBATRACHUS 1,8,9,10,20,24,27, 29,42,47,51,52,56

FIGURE 2.—Alternate phylogenetic relationships among the myobatrachid genera. (Underlined states = unique appearances within phylogeny; dashed lines = lineages are not sister groups; see text.)

In attempting to determine the relationships Geocrinia, (2) Crinia haswelli, (3) Crinia georgiana, among Crinia georgiana, C. haswelli, Ranidella (4) Ranidella, and (5) R. riparia and R. tasma- riparia, Geocrinia, and Ranidella, certain charac- niensis (latter was not included in Liem's, in press, ters are keys to understanding the relationships: study sample). This arrangement is consistent with vomer and vomerine teeth, omohyoideus muscle, the available larval evidence (Watson and Martin, outer metatarsal tubercle, belly texture, and egg 1973) and has the advantage that each unit is placement. There is no way to group the myoba- homogeneous with respect to life history mode. trachine taxa so that two of the derived states of Separation of tasmaniensis from Crinia, sensu lato, these characters define the same assemblage. has been previously suggested (Littlejohn, 1970). Rather, a grouping which results in a cluster hav- Of the remaining myobatrachine taxa, Taudac- ing all the taxa with the same derived state of one tylus and Uperoleia are most closely related, and character leads to convergence of states in the when joined with the taxa mentioned above, form other characters. The most reasonable assumption a monothetic cluster with a unique state appear- is that the common ancestor of the group had the ance (Figure 2). Glauertia and Myobatrachus are genetic potential to produce the states in question, closely related and can also form a monothetic but, in part, the results were a mosaic pattern. cluster at the same level as Taudactylus and Based on this assumption, the most reasonable Uperoleia of Figure 2, but the cluster does not taxonomic criteria to describe the situation is the have a unique state appearance. recognition of the following as distinct genera, The proposed relationships among the limno- indicating separate evolutionary lineages: (1) dynastines of Figure 1 contain one non-monothetic 10 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY cluster. If clusters are chosen to maximize the considered as separate generic units in order to number of appearances of unique states within present more information. the total phylogeny, one major difference in place- ment occurs: the location of Mixophyes. With SISTER GROUPS WITHIN THE PHYLOGENY Mixophyes removed basally, the remaining limno- dynastines have fused cervical and second verte- Within the phylogeny of Figure 2, those dicho- brae. Mixophyes is unique in that the nervus tomies not representing sister groups (Hennig, mandibularis lies between the levator mandibulae 1966) are indicated by dashed lines. In two cases, externus and subexternus muscles, it is the only additional information is needed to recognize myobatrachid genus with axillary amplexus (DSL), sister groups. In the case of Mixophyes and the the genus is unique in egg placement (Liem, in remaining limnodynastines, the life history infor- press) and tadpole morphology (Watson and Mar- mation has been presented previously. In the case tin, 1973), further supporting the basal dichotomy of Rheobatrachus compared with the rest of the as represented in the phylogeny of Figure 2. genera, Rheobatrachus has a unique sacro-coccygeal Liem (in press) thought Heleioporus and Neoba- articulation (Liem, 1973) and a unique develop- trachus to be closely related. This hypothesis is mental pattern of gastric larval incubation (Cor- not substantiated by this study; in fact, the two ben, Ingram, Tyler, 1974). As argued elsewhere taxa do not share a monothetic cluster. Assuming (Heyer, 1975a), the weakest areas in the phylogenies that Neobatrachus has closest relationships with occur where sister groups are not present. It seems Notaden, then the remaining limnodynastines are reasonable to assume that higher taxa recognized characterized by eggs laid in foam nests. in an analysis of this sort should be sister groups. Liem (in press) indicated a close relationship between Adelotus and Lechriodus. Apparently this COMPARISON WITH PREVIOUS SCHEMES was occasioned by shared primitive states, since analysis of the derived states indicates the closest Lynch (1971) summarized the classificatory relationships of Adelotus and Lechriodus lie else- schemes of the myobatrachids, with the exception where rather than with each other (Figure 2). of Rheobatrachus, which was unknown to science Liem (in press), following the lead of Littlejohn at that time. Parker (1940) was the first to recog- (1963), proposed placing Kyarranus in the syn- nize the distinctiveness of the myobatrachines and onymy of Philoria. These two taxa share the most limnodynastines. Lynch's (1971) study, as well as derived states among the limnodynastines. Further, the present study, supports Parker's conclusions. the two taxa uniquely share the derived life history Lynch (1971) further recognized two tribes in the pattern of nonaquatic, nonfeeding larvae among limnodynastines. The present study does not sup- the limnodynastines. Within this study, Kyarranus port that conclusion, since Lynch included and Philoria form sister groups, Kyarranus having Heleioporus, Mixophyes, Neobatrachus, and Nota- derived states of the squamosal and the supracora- den in a single tribe. The results of this study coideus muscle, Philoria having derived states of point to the distinctiveness of Mixophyes within the nasal and frontoparietal relationship, pre- the limnodynastines. hallux, and depressor mandibulae muscle. Philoria Liem (in press) included Rheobatrachus in the has undergone a separate evolutionary history subfamily Myobatrachinae. In contrast to the pres- from Kyarranus although both had a common ent study, Liem (in press) also drew some different ancestor. It seems rather a point of personal pref- systematic conclusions at the generic level, as did erence whether the emphasis is on the common Blake (1973) in his study of the Myobatrachinae. ancestry, resulting in recognition of a single genus, Differences between this study and Liem's and or on the evolutionary divergence, suggesting two Blake's studies are due to the different analytic genera. In fact, we disagree on this one point: the procedures used. Both Liem (in press) and Blake first author prefers to recognize two genera, the (1973) used a phenetic approach in which primi- tive and derived states were used to produce phylo- second author prefers to recognize one. For pur- genetic hypotheses. In this study, only derived state poses of this study, Kyarranus and Philoria are information was used to produce phylogenetic NUMBER 233 11 hypotheses. We think that clusters based on shared The genera of each of the subfamilies as rec- primitive states are weak, since there is no infor- ognized in this study are: mation with respect to change from an ancestral NEW SUBFAMILY New genus for Ranidella condition. All taxa can be systematically thought Rheobatrachus riparia and tasmaniensis of as collections of primitive and derived character MYOBATRACHINAE LlMNODYNASTINAE states. The primitive states indicate no change in Assa Adelotus that character from the ancestral condition. The Crinia Heleioporus derived states do indicate a change from an ances- Geocrinia Kyarranus tral condition. If groups of taxa share the same Glauertia Lechriodus Metacrinia Limnodynastes derived states, they may form a common evolu- Myobatrachus Mixophyes tionary lineage. From a philosophical point of Pseudophryne Neobatrachus view, we think the approach used in this study is Taudactylus Notaden superior for deducing relationships than is a phene- Uperoleia Philoria tic analysis, which indiscriminately uses primitive New genus for Pseudo- Platyplectron phryne occidentalis and derived character state information. The New genus for Crinia has- approximation of reality in the kind of analysis welli presented herein depends on the correctness of determination of primitive and derived states. RHEOBATRACHINAE, new subfamily TYPE-GENUS—Rheobatrachus Liem, 1973. Systematic Conclusions DIAGNOSIS—Maxillary teeth fanglike; fontanelle not roofed over by frontoparietals; columella The relationships, as presented in the phylo- present; cervical cotyles narrowly separated; sacral genies (Figures 1, 2), are best represented by rec- diapophyses broad; sacro-coccygeal articulation ognizing three subfamilies: one for Rheobatrachus, monocondylar; coccygeal prezygapophyses and one for the myobatrachines, and a third for the transverse processes present; cricoid a complete limnodynastines. The limits of this study preclude ring; alary process of hyoid broad and winglike; an evaluation of the familial status of these three pupil vertical; tongue adherent to mouth floor; units. It may be that all three units are actually outer metatarsal tubercle present; female gastric familial units. Some authors (Liem, in press; Wat- brooding of young. son and Martin, 1973) suggest that the myobatra- CONTENT.—Monotypic, Rheobatrachus silus. chines and limnodynastines are familial units; this study demonstrates that Rheobatrachus is very MYOBATRACHINAE Schlegel, 1850 distinctive from the myobatrachines and limno- dynastines. Recognizing each as a subfamily has Crinia Tschudi, 1838 the advantage of pointing out the major differ- TYPE-SPECIES.—Crinia georgiana Tschudi, 1838. ences within the total study group, while not add- DIAGNOSIS.—A myobatrachine genus; cervical ing taxonomic confusion to any future situation cotyles widely separated; vomer and vomerine (a suprageneric taxon is available for a family teeth present; narrow sacral diapophyses; depressor name). mandibulae muscle with slip from dorsal fascia; The only possible formal classification below the omohyoideus muscle present; tympanum present; subfamily level that could be proposed confidently belly smooth; toes without fringe or web; outer would be to recognize a tribe for Mixophyes. There metatarsal tubercle present; eggs placed in water; seems, however, to be no advantage to emphasizing stream larvae. the distinctiveness of Mixophyes in a formal action. CONTENT.—Monotypic, Crinia georgiana. The new taxonomic categories or combinations arising from this study are presented formally. Geocrinia Blake, 1973 Diagnostic definitions only are provided since the full character complements of the taxa are pre- TYPE-SPECIES.—Pterophrynus laevis Gunther, sented elsewhere within the paper (Table 1). 1864. 12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY DIAGNOSIS.—A myobatrachine genus; cervical teeth absent; narrow sacral diapophyses; depressor cotyles widely separated; vomer present, vomerine mandibulae muscle with slip from dorsal fascia; teeth present or absent; narrow sacral diapophyses; omohyoideus muscle absent; tympanum present; depressor mandibulae muscle with slip from dorsal belly smooth; toes fringed; outer metatarsal tuber- fascia; omohyoideus muscle present or absent; tym- cle present; eggs placed in water; pond larvae. panum present; belly smooth; toes lacking fringe CONTENT.—Monotypic, Paracrinia haswelli. or web; outer metatarsal tubercle absent; eggs laid ETYMOLOGY.—From the Greek para ( = beside), in concealed sites on land; pond larvae or terres- indicating the fact that the relationships of this trial, non-feeding larvae. genus fall within the Crinia complex. The genus is CONTENT.—Geocrinia laevis, leai, lutea, rosea, feminine in gender. victoriana. Australocrinia, new genus Pseudophryne Fitzinger, 1843 TYPE-SPECIES.—Pterophrynus tasmaniensis Giin- TYPE-SPECIES.—Phryniscus australis (in part) ther, 1864. Dume'ril and Bibron, 1841 (—Pseudophryne semi- DIAGNOSIS.—A myobatrachine genus; cervical mar morata Lucas, 1892). cotyles widely separated: vomer present, vomerine DIAGNOSIS.—A myobatrachine genus; cervical teeth present or absent; narrow sacral diapophyses; cotyles widely separated; vomer present or absent, depressor mandibulae muscle lacking slip from vomerine teeth absent; broad sacral diapophyses; dorsal fascia; omohyoideus muscle present; tym- depressor mandibulae muscle lacking a slip from panum present; belly granular; toes fringed or dorsal fascia; omohyoideus muscle absent; col- free; outer metatarsal tubercle present; eggs placed umella absent; belly smooth; toes free, no fringe on land or water; stream or pond larvae. or web; outer metatarsal tubercle present; eggs CONTENT.—Australocrinia riparia, tasmaniensis. placed on land (except P. douglasi, which places ETYMOLOGY.—From the Latin australis ( = south- eggs in water); pond larvae. ern), indicating the southern geographic distribu- CONTENT.—Pseudophryne australis, bibroni, cori- tion of this member of the Crinia complex. The acea, corroboree, dendyi, douglasi, guentheri, ma- genus is feminine in gender. jor, semimarmorata. Kankanophryne, new genus Ranidella Girard, 1853 TYPE-SPECIES.—Pseudophryne occidentalis Parker, TYPE-SPECIES.—Ranidella signifera Girard, 1853. 1940. DIAGNOSIS.—A myobatrachine genus; cervical DIAGNOSIS.—A myobatrachine genus; cervical cotyles widely separated; vomer present or absent, cotyles widely separated; vomer and vomerine teeth vomerine teeth absent; narrow sacral diapophyses; absent; narrow sacral diapophyses; depressor man- depressor mandibulae muscle with slip from dorsal dibulae muscle with slip from dorsal fascia; fascia; omohyoideus muscle present; tympanum omohyoideus muscle absent; columella absent; present; belly granular; toes fringed or free; outer belly smooth; toes free of fringe or web; outer metatarsal tubercle present; eggs placed in water; metatarsal tubercle present; eggs placed on land; pond larvae. pond larvae. CONTENT.—Ranidella glauerti, insignifera, parin- CONTENT.—Monotypic, Kankanophryne occident- signifera, pseudinsignifera, remota, signifera, sloa- alis. nei, subinsignifera, tinnula. ETYMOLOGY.—From the Greek kankanos (= dry), in reference to the arid distribution of this genus. Paracrinia, new genus The genus is feminine in gender. TYPE-SPECIES.—Crinia haswelli Fletcher, 1894. DIAGNOSIS.—A myobatrachine genus; cervical Phyletic Character Analysis cotyles widely separated; vomer present, vomerine The analysis of characters into primitive and NUMBER 23S 13 derived states seems overall to be intuitively sound, the states of a given character differ with respect since the recognition of the myobatrachines and to clustering behavior. State 1 of the vomer- limnodynastines as distinct units is in agreement vomerine teeth complex (character 7) has good with every other systematist that has worked with clustering behavior, states 2 and 3 are average, the myobatrachids in any detail. state 4 is poor. State 1 of the adductor mandibulae As discussed in detail elsewhere (Heyer, 1975a), muscle (character 21) has good clustering behavior, an analysis of the phylogenetic usefulness of each state 2 is poor. State 1 of the semitendinosus is character state can be made after the fact. This good (character 24), state 2 is poor. State 1 of the is done by using the information from the better abductor brevis plantaris hallucis muscle (character phylogenetic tree to determine the number of 27) is average, state 2 is best. State 1 of the dila- appearances of each state (as printed) on the tree tator laryngis muscle (character 33) is average, state divided by the total number of taxa that actually 2 is poor. State 1 of toe webbing (character 37) is have that state. The numeric values for each state good, state 2 is average. States 1 and 2 of the outer are then ranked and grouped into categories. A metatarsal tubercle are average, state 3 is poor. value of 1 indicates no phylogenetic information; Examination of the characters indicates a dis- the smaller the number, the better the state links crepancy with respect to the hyoid complex. As the taxa as found in the preferred phylogeny. noted in the character analysis, the broad, wing- The phylogeny of Figure 2 was used for analysis, like state of the alary process is thought to be except that riparia and Ranidella were considered the primitive state, in contrast to previous coding as a single unit to be consistent with the original (Heyer, 1975a). The associated muscles, correspond- data. For multistate characters, averages are used. ing with broad winglike alary process, however, Best characters have values of 0.04 to 0.20, good were coded as derived states, as in the previous characters have values of 0.25 to 0.42, average char- study (Heyer, 1975a). The distribution of states acters have values of 0.50 to 0.60, poor characters indicates that in order to be consistent, the primi- have values of 0.67 to 1.00. tive and derived states should be reversed for the The best characters with a unique appearance following characters: geniohyoideus medialis mus- in the phylogenetic tree are: intervertebral disc, cle (character 29), sternohyoideus muscle (character cervical vertebral fusion, cervical cotyles, coccyx, 30), and anterior petrohyoideus muscle (character cricoid, alary process of the hyoid, anterior petro- 31). hyoideus muscle, intermandibularis muscle, egg The purpose of this entire analysis is to produce mass type. The best characters with two or more a likely phylogenetic hypothesis for the myoba- appearances in the tree are: sphenethmoid and trachids. The analysis thus far concludes that there geniohyoideus muscle. are several changes that could be made in the Good characters are: squamosal, sacral dia- original data set that would improve the phylo- pophyses, carpals, adductor longus muscle, adduc- genetic hypothesis. The changes are: (1) deletion tor magnus muscle, omohyoideus, sternohyoideus, of Rheobatrachus, since its evolutionary history pupil shape, and toe webbing. appears to be distinct from the evolutionary history Average characters are vomer and vomerine of the other two subfamilies; (2) addition of Ans- teeth, anterior process of the hyale, supra coracoi- tralocrinia; (3) reversal of direction of character deus muscle, semitendinosus muscle. states for character numbers 29, 30, and 31; (4) de- Poor characters are: maxillary teeth, nasal con- letion of characters in the poor category (as ana- tact with frontoparietial, carotid canal, frontoparie- lyzed above). The purpose of making the changes tal fontanelle, transverse process width, dorsal crest in the data set is not to draw new taxonomic con- of ilium, prehallux, depressor mandibulae muscle, clusions. Taxonomic conclusions were drawn on adductor mandibulae muscle, extensor cruralis the basis of the entire original data set. Rather, brevis muscle, petrohyoideus posterior muscles, the purpose of making the changes is to determine dilatator laryngis muscle, belly texture, tongue, whether the changes allow new insights into the inner metatarsal tubercle, outer metatarsal tuber- relationships among the taxa. cle, site of egg development. The best arrangement of the taxa in terms of There are seven multistate characters in which maximization of character states at each cluster, 14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY numbers of monothetic clusters, and unique state With the data in the revised set, Notaden forms appearances is presented in Figure 3. All clusters a sister group with all other limnodynastines are monothetic. whereas Mixophyes does not. As presented earlier, Within the myobatrachines, sister groups are data not originally coded will allow Mixophyes to represented at the basal branchings involving form a sister group with all other limnodynastines. Taudactylus, Glauertia, and Myobatrachus. None We prefer the arrangement depicted in Figure 3. of the branchpoints from Geocrinia onward are The cluster involving Adelotus, Limnodynastes, and sister groups. This pattern contrasts with the pat- Platyplectron in Figure 3 represents the taxonomic tern seen in Figure 2, where the distal branchings thinking of most previous workers, as Platyplectron all represented sister groups, but the basal branch- has usually been included in Limnodynastes. Ade- ings did not represent sister groups. As discussed lotus does form a sister group with Limnodynastes previously, it appears that evolution in the Crinia- and Platyplectron. As in the phylogeny of Figure complex involved a mosaic pattern of characters. 2, the relationships proposed in Figure 3 involve This hypothesis is consistent with a phylogeny in the unique development of the foam nest. Helei- which no sister groups are present. A mosaic pat- oporus does not have expanded fingers in the fe- tern of evolution in the Crin ja-complex is best de- male as do all the other genera with foam nests. picted in the phylogeny of Figure 3. The female finger flanges are used to beat up the Among the limnodynastines, there are two alter- foam. The method of foam nest construction in nate sequences at the basal branching. The Heleioporus is unknown (Martin, 1972). The lack position of Notaden and Mixophyes can be inter- of finger flanges in Heleioporus suggests that changed. In the interchanged arrangement, char- Heleioporus represents a divergent evolutionary acter state 14, involving fusion of the cervical lineage within the foam nest builders. This di- vertebrae, is not a uniquely derived state, but the vergent pattern is best expressed in Figure 3. state of the adductor magnus muscle (state 33) is. The revised data set does allow construction of

PSEUOOPHRYNE OCCIDENTAL IS \ 12,17,53 ASSA METACRINIA HASUELLI RANIOELLA RIPARIA UPEROLEIA \ \ 17 17,37,38, 11,17,45,49 17,45,49 11,20,24, -\—-\—I- 45 \ / ^ ^53 20,24,37,38,45,49^^=_^_^ \ / GEORGIAN]! 9,10 LIMNOOYNASTES PLATYPtECTRON 17,45,49 / 38,45,48,49,51 / GEOCRINIA 51,52 40" ADELOTUS 9,17,37, / KYARRANUS PHILORIA \ 24,38,45.51,52 8,29 / \ / \ / 27,36,37 LECHRIOOUS MYOBATRACHUS 15,29 24,27,36,37, w 24,n40",45,48,49 9,10,11,12,17 38,45,48,49 20,24,37,39. -3,-.45,4, / 29,32 N GLAUERTIA 27,48 9,10,11,12,17,20,24,28, 29,37,38,45,51,52,53 \/ 31 TAUOACTYLUS \ 9.10,15.24,27.30 \ 35,45,48,51,52 7,8,13,22

30,33 NOTAOEN 7,8,24,27,29,37,38,51 \ il MIXOPHYES \ 30,32,33,35,"40"

17,20,20, "39","4i",51

FIGURE 3.—Preferred phylogenetic relationships among the myobatrachid genera. (Underlined states - unique appearances within phylogeny; dashed lines = lineages are not sister groups; quotation marks - direction of change of state has been reversed; see text.) NUMBER 235 15 a phylogenetic tree that differs from the trees pro- 4. Open woodland and shrubland distributions, duced by the original data set. We think the neither wet forest nor primarily desert, < 500 mm phylogeny based on the revised data set is a more isohyet. Glauertia, Kankanophryne, Neobatrachus, plausible phylogenetic hypothesis, representing the and Notaden share this pattern. Of these genera, best fit of the data with the probable phylogeny of Notaden has the most arid distribution pattern, the group. being found in shrubland, spinifex plains, and sand dunes in the rainfall distribution region < 360 mm isohyet. For zoogeographic purposes, the arid Evolutionary Zoogeography distribution pattern of Notaden is not recognized as a distinct pattern because there is only one GENERIC DISTRIBUTION PATTERNS genus demonstrating the pattern. The known geographic distributions of the The historical record of vegetation types in the genera, summarized in map form, are essential data Australian region allows some generalization on for the determination of distribution patterns. evolutionary trends and sequences within the Myo- This information for all genera is not available in batrachidae. Details of past vegetation type dis- a single source. Since the discussion that follows is tribution and paleoclimates in the Australian based upon the basic distribution data, those data region are not well known. The following sum- are presented here (Figures 5-28). In grouping the mary, which appears to be sufficient to explain individual distribution maps into similar clusters major evolutionary and zoogeographic trends in and attempting to associate those patterns with myobatrachids, seems reasonably substantiated by features of the Australian climate, landscape, and/ the available data. The antarctic forest type (Notho- or vegetation, certain facts become apparent. No (Continued on p. 22) genus of myobatrachid frog is primarily adapted to the desert habitat (compare Figure 20, Notaden, the genus with the most extensive desert distribu- tion, with Figure 4). In associating the distribution patterns with other features, the distributions can be understood best in terms of the vegetation types found in the region. The geographic distributions of the genera (Figures 5-28) show four major pat- terns associated with broad vegetational categories: 1. Distributions primarily restricted to the ant- arctic vegetation unit (temperate rainforest and subtropical rainforest), > 1520 mm rainfall isohyet. Assa, Kyarranus, Lechriodus, Rheobatrachus, and Taudactylus share this pattern. 2. Distributions limited to wet forest vegetation types (wet sclerophyll forest), ± 1000 mm isohyet. The genera are either found only in the eastern wet forest (Adelotus, Mixophyes, Paracrinia, Phi- loria), the southwestern wet forests (Crinia, Meta- crinia), or both eastern and southwestern wet forests (Geocrinia). 3. Distributions in both wet forest and sub- humid woodland, about 500 mm to 1000 mm iso- hyts. The distributions are either primarily eastern (Australocrinia, Limnodynastes), western (Myo- FIGURE 4.—Broad vegetation zones critical to analysis of batrachus), or eastern and western (Heleioporus, myobatrachid distribution patterns. (Solid areas = antarctic Platyplectron, Pseudophryne, Ranidella, Upero- vegetation unit; hatched areas = wet forests; stippled areas leia). = sandy deserts.) 16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 5.—Geographic distribution of the myobatrachid FIGURE 6.—Geographic distribution of the myobatrachid genus Adelotus. genus Assa.

FIGURE 7.—Geographic distribution of the myobatrachid FIGURE 8.—Geographic distribution of the myobatrachid genus Australocrina. genus Crinia. NUMBER 233 17

FIGURE 9.—Geographic distribution of the myobatrachid FIGURE 10.—Geographic distribution of the myobatrachid genus Geocrinia. genus Glauertia.

FIGURE 11.—Geographic distribution of the myobatrachid FIGURE 12.—Geographic distribution of the myobatrachid genus Heleioporus. genus Kankanophryne. 18 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 13.—Geographic distribution of the myobatrachid FIGURE 14.—Geographic distribution of the myobatrachid genus Kyarranus. genus Lechriodus.

FIGURE 15.—Geographic distribution of the myobatrachid FIGURE 16.—Geographic distribution of the myobatrachid genus Limnodynastes. genus Metacrinia. NUMBER 233 19

FIGURE 17.—Geographic distribution of the myobatrachid FIGURE 18.—Geographic distribution of the myobatrachid genus Mixophyes. genus Myobatrachus.

V FIGURE 19.—Geographic distribution of the myobatrachid FIGURE 20.—Geographic distribution of the myobatrachid genus Neobatrachus. genus Notaden. 20 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 21.—Geographic distribution of the myobatrachid FIGURE 22.—Geographic distribution of the myobatrachid genus Paracrinia. genus Philoria.

FIGURE 23.—Geographic distribution of the myobatrachid FIGURE 24.—Geographic distribution of trie myobatrachid genus Platyplectron. genus Pseudophryne. NUMBER 233 21

FICURE 25.—Geographic distribution of the myobatrachid FICURE 26.—Geographic distribution of the myobatrachid genus Ranidella. genus Rheobatrachvu.

V FIGURE 27.—Geographic distribution of the myobatrachid FIGURE 28.—Geographic distribution of the myobatrachid genus Taudactylus. genus Uperolcia. 22 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY fagus, podocarps) has had a long history in the that could also live in more open formations. Adap- Australian region and during the early Tertiary, tive trends in this group of genera include stream at least, had a broader distribution than present. adaptations, burrowing, and possibly direct de- Cookson (1953), Cookson and Pike, (1953, 1954), velopment. and Darlington (1965) have shown that fossil A fourth stage in the evolutionary history of the remains of the antarctic vegetation have been myobatrachids involved adaptations leading to a found at several localities across southern Australia, primary radiation in open, rather arid vegetation suggesting a former continuous forest belt. It is types. The genera with this pattern either show a probable that the extent of this forest unit has distribution associated with water (oasis effect) been affected by world wide climatic changes, within a more arid region (Glauertia, Kankano- especially glaciation cycles. Arid environments phryne), or burrowers that can utilize temporary have long been a feature of the Australian con- bodies of water (Neobatrachus, Notaden). tinent. The Great Dividing Range was formed at To summarize, the major adaptations associated the end of the Mesozoic, and arid environments with the four generic distribution patterns are as have been present in Australia throughout the follows: Tertiary to the present. One of the most striking features in the evolu- Pattern 1 (Temperate/subtropical rainforest) Gastric brooding: Rheobatrachus tion of myobatrachid frogs has been the lack of Inguinal pouch brooding: Assa radiation in the strict desert environment. The Terrestrial egg desert habitat has been present in Australia for Nonfeeding larvae: Kyarranus a long enough time to allow a radiation. The Steam adapted larvae: Taudactylus Australian deserts are apparently too harsh an Montane stream adaptations: Rheobatrachus, Taudactylus environment for frogs; some particular species can Pattern 2 (Wet sclerophyll forest) Terrestrial egg tolerate desert conditions, but none are primarily Stream adapted larvae: Crinia, Mixophyes adapted for desert life. A similar situation occurs Nonfeeding larvae: Philoria with the Namib desert anuran fauna in South Trend toward direct development: Geocrinia, part (leai, Africa (Channing, 1974). laevis) Direct development: Geocrinia, part (lutea, rosea), Meta- The antarctic forest vegetation appears to be crinia the vegetation type associated with the ancestral Pattern 3 (Wet forest and subhumid woodland) myobatrachid stock. This assumption is supported Rapid egg development: Platyplectron, part (ornatus, by the historical presence of this vegetation type spenceri) and by the distributions of several genera limited Burrowing adaptations: Heleioporus, Platyplectron strictly to this vegetation type. The genera pres- Terrestrial egg Steam adapted larvae: Australocrinia ently restricted to the antarctic vegetation type Advanced larval development: Pseudophryne (except represent the descendants of the original stock douglasi) which was associated with the temperate antarctic Foam nest in burrow: Heleioporus forests. If a generalized life history pattern is as- Possible direct development: Myobatrachus sumed for the ancestral stock, then the following Pattern 4 (Open woodland and shrubland) Short larval life: Kankanophryne, Neobatrachus, Notaden major evolutionary trends have occurred within Burrowing adaptations: Neobatrachus, Notaden the beech and podocarp forests: trends toward Opportunistic breeding: Neobatrachus (part), Notaden nonaquatic larvae; either direct development, in- guinal pouch brooding, or gastric brooding; and a All four stages occurred roughly contempora- stream adapted life history. neously, early in the Tertiary at the latest. If the A second stage in the historical zoogeography four stages occurred in sequence, there should be of the myobatrachids was the association with tem- an overall reduction in the average number of perate wet forest (including Eucalyptus) types. The primitive states per genus from the first to the same adaptive trends are seen in this assemblage fourth stage. This is not the case. The first three of genera as in the previous assemblage: trends to- groups all have an average of 55% primitive states ward direct development and stream adaptations. per genus, the fourth group has an average of 54%. A third stage was the evolution of forest forms A rough indication of recency of evolutionary NUMBER 233 activity can be gained by examining the average EVOLUTIONARY TRENDS WITHIN THE SUBFAMILIES number of species per genus. The assumption is that lower number of species per genus represents When the historical zoogeography of the myo- an older radiation, a higher number of species batrachids is viewed in terms of evolutionary re- per genus represents more recent speciation. The lationships, additional trends can be postulated. first distribution pattern group has an average SUBFAMILY RHEOBATRACHINAE.—The sole mem- number of species per genus of 2.6, the second ber of this subfamily, Rheobatrachus silus, repre- group an average of 2.0, the third group an average sents an early stream adaptational complex from of 5.0, and the fourth group an average of 3.25. If the ancestral myobatrachid stock. Rheobatrachus the difference between the first and second group has remained in the temperate antarctic forest is real, it could be interpreted in the following vegetation and represents a combination of a large way. The general wet forest associated group has number of primitive states (67%) together with had a more continuous habitat available since the some of the most bizarre derived states known Tertiary, while the antarctic forest associated group involving life history (Corben, Ingram, and Tyler, has had a more fragmentary distribution, with pop- 1974). ulations of frogs becoming isolated in forest frag- SUBFAMILY MYOBATRACHINAE.—Members of the ments sometime in the Tertiary. The isolation of Myobatrachinae demonstrate all four geographic these latter populations has led to relatively more patterns; however, there are definite emphases speciation in this group than in the more general within the patterns, most of the genera being as- wet forest associated group. This hypothesis is sup- sociated with wet forests. Taudactylus represents, ported by the fact that most of the antarctic forest within the antarctic forest type, an early specializa- associated genera that are not monotypic typically tion to stream adaptations, with respect to the have allopatrically distributed species. The fourth larvae of all known species and all adults (Liem group, primarily arid adapted, has an intermediate and Hosmer, 1973). The Crmia-complex, repre- amount of relatively recent speciation, while the sented by Assa, Australocrinia, Crinia, Geocrinia, forest-open formation group has the greatest Paracrinia, Ranidella, all appear to represent more amount of recent speciation. Apparently, in terms or less independent specializations from the an- of fluctuating environments there has historically cestral wet forest myobatrachine stock. Each generic been greater fragmentation and coalescing of open unit likely represents a previous isolation in a wet formation habitats. This has allowed speciation to forested area with concomitant adaptational re- continue throughout the Tertiary to the present. sponses to that given environment. The result of The phenomenon of more recent speciation in several independent adaptational responses in a open eucalypt habitat compared to rainforest habi- generalized myobatrachine is a mosaic pattern of tat has also been observed in Australian birds derived states. The generalized myobatrachine (Schodde and Calaby, 1972). morphology has been retained, so the similarity In summary, it appears that the major expansion among these taxa is based on a suite of primitive, from the ancestral temperate antarctic forest habi- rather than derived, states. Ranidella had the tat into other habitat types occurred more or less genetic plasticity to adapt to environments border- synchronously; that the radiations within the ma- ing the wet forest habitat. The genera demon- jor habitats are old, stemming from the early strating the arid adaptation geographic pattern, Tertiary; and that the most recent evolutionary Glauertia and Kankanophryne, are exploiting local activity is associated with open formation habitats, mesic habitats within an arid landscape. The di- intermediate between forest and desert environ- vergence of Glauertia into this specialized habitat ments. type probably occurred earlier than for Kankan- Another overall trend is that direct development ophryne. Pseudophryne, Ranidella, and Uperoleia is approached or attained only in wet forest genera, have adapted in part to open formation vegetation as pointed out previously by Main, Littlejohn, and types. The large terrestrial egg of Pseudophryne Lee (1959). The only possible exception is Myo- with initial development out of water has adaptive batrachus, which has huge ovarian eggs, suggesting value in areas of unreliable rainfall. There are no direct development. particular obvious adaptations in the life history 24 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY patterns of Ranidella and Uperoleia to predict a Lechriodus seems to be associated with the podo- habitat shift to open formation vegetation. Myo- carp vegetation unit of the temperate antarctic batrachus is a very specialized form, a subterranean vegetation and has undergone a minor radiation termite eater. It is the only burrowing myoba- associated with, in part, the podocarp forests of trachid that can burrow head first and is restricted New Guinea. There is no evidence, however, to to sandy areas in southwest Australia. Within the suggest that the genus arose in New Guinea; in Myobatrachine, it appears that Taudactylus, fact, the Australian species is the most primitive in Glauertia, and Myobatrachus represent the earliest the genus (Zweifel, 1972). specializations and divergences from the common myobatrachine ancestral stock. The remainder of the myobatrachine genera appear to have had a MAJOR TRENDS AMONG SUBFAMILIES common ancestor that was widespread throughout the late Mesozoic-early Tertiary wet temperate Even though the adaptive trends within the forests, and that fragmentation of this common Myobatrachinae and Limnodynastinae are similar stock associated with fragmentation of the temper- in that they include adaptational complexes asso- ate forests, with evolution proceeding indepen- ciated with open formations and arid environ- dently in each unit, gave rise to the mosiac com- ments, direct development and stream habitats, plex of genera present today. the adaptations associated with these shifts are SUBFAMILY LIMNODYNASTINAE.—Mixophyes ap- different. The myobatrachines have no truly arid pears to be an early specialization of the limnody- adapted genera; the limnodynastines have two. nastine ancestral stock to the stream environment. Direct development involves a foam nest in the Neobatrachus and Notaden represent early speciali- limnodynastines; a foam nest is not involved in zations to the arid environment involving fossorial- direct development in the myobatrachines. The ity in the adult stage. Notaden and Neobatrachus foam nest is the key preadaptation to open forma- are the most arid adapted myobatrachid genera tion utilization by some of the limnodynastines; and, interestingly, have almost completely allo- the foam nest is not a feature of the open forma- patric distributions. The evolutionary history of tion myobatrachines. the remaining limnodynastines is associated with Within all three subfamilies, the ancestral pat- the foam nest. Four of the genera having a foam tern appears to be an association with temperate nest are wet forest genera, the other three are antarctic vegetation and a generalized life history forest and open formation forms. Either the foam pattern involving a free living larva. The most nest was initially evolved in response to selective parsimonious explanation of the early evolution of pressures associated with wet forest environments the myobatrachids is that the present myoba- or open formation environments. The evidence trachids represent three very early divergences in points to the original adaptive significance of the the myobatrachid ancestral stock. One lineage foam nest as having been associated with wet became adapted as adults to the stream habitats forests. None of the genera that are primarily arid with an extremely specialized life history pattern adapted have a foam nest, while several of the of gastric brooding. The degree of specialization genera with foam nests are found only in wet forest apparently led to an evolutionary dead end repre- habitats. The wet forest foam-nesting genera have sented in Rheobatrachus silus. The early diver- an average of 2.5 species per genus, contrasted gence of the myobatrachine and limnodynastine with a 5.7 average for the open formation foam- stocks involved morphological features which at nesting genera, suggesting the wet forest radiation present are difficult to associate with adaptive is older than the open formation radiation. The trends. Both groups were associated with the tem- foam nest, then, served as a preadaption for ex- perate antarctic vegetation, and they likely had a ploitation of a more seasonal environment associ- generalized life history pattern. The basic morpho- ated with open formation vegetations. Lechriodus logical differences between the two stocks allowed is noteworthy among the myobatrachids in that their evolution to act in independent ways in rela- it is the only genus in which more species are tion to the same environmental changes of increas- found outside of Australia than within Australia. ing aridity in the late Mesozoic to present. NUMBER 233 25

COMPARISONS WITH PREVIOUS ZOOGEOGRAPHIC The Spencerian scheme appears to account much SCHEMES better for reptile distributions than for myoba- trachid distributions (e.g. Keast, 1959; Kluge, 1967). Previous workers in discussing the zoogeography We think that the Spencerian subregion scheme of the Australian anuran fauna have proposed associates best with those faunal units having a three major ideas. The first is that a segment of tropical origin. For those faunal units having a the anuran fauna is old, stemming at least from temperate origin, such as the myobatrachids, the the early Tertiary. The second is that the anuran scheme proposed herein apparently is a more fauna can be explained best by using a modified realistic zoogeographic hypothesis. Spencerian (1896) scheme of four faunal subregions Lee (1967:428-429) summarized the speciation in Australia. The third is the proposal of a mode scheme proposed for certain genera: of speciation in some genera involving evolution of species in southeast Australia with dispersal to southwest Australia. Faced with the incongruous situation of a multiplicity of species within Heleioporus and two other genera of frogs This study confirms the importance of recog- (Crinia and Neobatrachus) in the topographically featureless nizing the antiquity of the myobatrachid frog south-western pocket. Main, Lee, and Littlejohn (1958) fauna. The myobatrachids were certainly present suggested the following explanation for their origin: in the Australian region by late Mesozoic and (1) Within each genus two elements or species groups were much of the radiation of the group may well have recognized and identified as bassian and eyrean elements (after the terminology of Serventy and Whittell (1951). occurred in early Tertiary. The breeding biology of the bassian frogs was adapted The faunal subregions used by most workers to to uniform rainfall or continuously moist situations, and explain the geographic distribution are as follows that of the eyrean species to seasonal rainfall and sea- (from Littlejohn, 1967:152): sonally arid conditions. Designation of the eastern species based on these adaptations presented no problems, but (1) Bassian—coastal, cool temperate, relatively uniform rain- with the specialization of the western fauna to the pro- fall; east and south-east of the Eastern Highlands and south nounced seasonal conditions of the area, use had to be of the Clarence River; (2) Torresian—northern and eastern made of similar morphology to provide clues of relation- tropical area north of the Clarence River with a summer ship. Judged by their morphological dissimilarity, this maximum rainfall; (3) Eyrean—all the drier inland portions division into two elements was a pre-Pleistocene event. of the continent, including the north-west coast; (4) South- (2) South-eastern rather than south-western Australia was west—the humid south-western corner of Western Australia favoured as the centre of origin of the genera. The south- which receives a reliable winter maximum rainfall. east is more diverse topographically and ecologically, and despite the greater number of species in these genera Moore (1961) has been the only recent worker in south-west, there are more species and genera of frogs in the south-east. that did not use the above scheme. Moore's scheme, (3) The multiplicity of species within each genus in the in part, recognized two additional regions and he south-west resulted from a series of invasions of the emphasized that only one of his regions was truly eastern representatives of each element into the south- distinctive: that corresponding with the South-west western pocket where some persisted. The invasions were pattern described above. separated by periods of isolation of sufficient duration for the populations to retain separate identity when brought None of the myobatrachids have a strictly Tor- together again. resian distribution. Several of the genera have (4) The invasions were considered to be coincident with the distributions combining Bassian and Torresian pluvial or glacial phases, and isolation with the arid in- distributions. Most of the genera with distributions terglacial phases of the Pleistocene. The Pleistocene in the Eyrean region also are found in the Bassian, events of the northern hemisphere were probably paral- leled in the southern hemisphere, but only with minor the Torresian, and/or the South-west regions. continental glaciation. Browne (1957) and Browne and Neither the modified Spencerian nor the Moore Vallance (1963) have been able to recognize three phases (1961) schemes adequately describe the distribution of glaciation in southeastern Australia, and this cor- patterns of the myobatrachids. We think the key responds to the number of western eyrean representatives to understanding the zoogeography of the myoba- in each genus, but not to the number of bassian species, of which there are either one or two. Main, Lee, and trachids is the recognition of the temperate antarc- Littlejohn (1958) suggested that early bassian invaders tic vegetation associated myobatrachid fauna as a may have succumbed to the marked arid conditions of distinctive unit. later interglacial phases. 26 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

(5) Finally, the disproportionate number of species in east- ful comparison is to compare that portion of the ern and western Australia was attributed to the failure leptodactylid radiation that corresponds with the of the western representatives to reinvade the east after myobatrachid radiation. The myobatrachid history specialization during isolation. includes original association with temperate forests The speciation scheme includes recognizing and evolution into open formations and arid en- species pairs in southeast and southwest Australia. vironments. A comparable situation is found in The kinds of information used to determine rela- South America in the case of the anuran fauna of tionships have been morphological similarity, mat- Chile and adjacent Argentina, including the Gran ing call similarity, and hybridization results. The Chaco, but excluding Misiones and east coastal authors who support this mode of speciation do Argentina. In this Neotropical region, the original not distinguish primitive and derived states; thus association of the leptodactylids was with the tem- the morphological and mating call similarities perate forests and radiations have occurred into the could be based on primitive states and contain no open formations and arid environments. information regarding relationships. Hybridization Although the fauna of southern South America data as an index to relationships among anurans is reasonably well understood, the leptodactylids of must be used with great caution. There are many Chile and Argentina are not as well understood examples of viable intergeneric crosses. Postmating systematically as are the myobatrachids of Australia. isolating mechanisms involving hybrid inviability There are 15 genera and 50 species of lepto- may only be selected for in closely related forms dactylids found in the area of comparison in Chile that have had historical overlap in geographic and Argentina. This larger number of taxa indi- ranges. We present these arguments not to invali- cates that the myobatrachid radiation has been date the conclusions of the authors who have pro- larger than the leptodactylid radiation in sim- posed this mode of speciation, but to indicate that ilar habitats. Interestingly, both the southern there is room for another hypothesis. leptodactylids and myobatrachids have the same average number of species per genus, 3 1/3, suggest- The hypothesis we favor is that for such genera ing that the radiations are of equivalent age. as Heleioporiis, during glacial periods, when the mesic forests would have had an expanded distri- A feature of the myobatrachid radiation is that bution, local arid habitats may have persisted due there are relatively few species per genus, Ranidella to such factors as drainage peculiarities of the soil. and Pseudophryne being high with 8 and 9 known These isolated arid habitats would serve as refugia species. This compares poorly with the 300 plus for the arid adapted fauna. Due to the isolated Neotropical species of Eleutherodactylus or even nature of these arid refugia, evolution could oper- the 20-40 species in such Neotropical genera as ate independently on the faunas of each of them Leptodactylus and Physalaemus. When one exam- allowing allopatric speciation to occur. The origin ines the leptodactylid fauna of Chile and Argen- of the species would be within the same general tina, the situation is exactly that found in the region where they are found at present, not in myobatrachids. Leptodactylus, Pleurodema, and southeast Australia exclusively. Telmatobius are all high with a total of 8 known species. The more extensive radiation of the myoba- MYOHATRACHID-LEPTODACTYLID COMPARISONS trachids than the southern leptodactylids appears to be valid. The only other group of frogs with The leptodactylid radiation in the Neotropics extensive distributions in Australia is the tree frog resulted in at least 47 genera and 500 species. The family Pelodryadidae. The tree frog family Hylidae vastness of this radiation appears to overwhelm is well represented in southern South America. In the myobatrachid radiation which resulted in 24 addition, there are two bufonid genera, two genera and about 80 species. The comparison of microhylid genera, and one pseudid genus with the total leptodactylid and myobatrachid radiations extensive distributions in Chile and Argentina. is in one sense unrealistic, since the myobatrachids This last set of genera appears to have ecological have not radiated into tropical environments in equivalents among the myobatrachids. The most the way that leptodactylids have. A more meaning- striking example is the fossorial termite eating NUMBER 2S3 27 myobatrachid Myobatrachus and the fossorial ter- habitats. The known methods of foam nest forma- mite eating microhylid Dermatonotus mulleri. tion are very different, as pointed out by Martin There is, incidently, only one such species in each (1970). In the myobatrachids, the fingers of the region. If this last set of genera is added to the female are modified to whip the jelly into foam, leptodactylids, there is a total of 20 genera and while in the leptodactylids the hind legs of the 64 species. This figure is rather close to the male whip the jelly into foam. myobatrachid figure, taking into account the fact Recently, a complex association of specialized that the southern South American anuran fauna larval microhabitats and morphology has been is not as well known as the Australian anuran described for Leptodactylus pentadactylus (Heyer, fauna and the total area of comparison is smaller McDiarmid, and Weigmann, 1975). Briefly, L. for southern South America. It appears as though pentadactylus, a foam nest breeder, utilizes small the total anuran frog faunas of these two regions aquarium sized puddles as larvae. These puddles, are equivalent and that the myobatrachid radiation occurring in wet environments, are replenished by has been more extensive than the leptodactylid. almost daily rainfall, so they do not dry up. The Convergence of adaptive complexes between the larvae are facultatively carnivorous, a rare condi- myobatrachids and southern leptodactylids is strik- tion among tadpoles. It was postulated that the ing. In both groups there are riparian forms, carnivorous tadpole mode is possible in these par- fossorial forms, and trends toward and including ticular microhabitats because there are no fish in direct development. Certain of these are worth these small puddles (see Heyer, McDiarmid, and more detailed comment. Weigmann, 1975, for a fuller explanation). Parental brooding of the larvae is found in both Lechriodus fletcheri, a wet forest associated, foam the myobatrachids and southern leptodactylids. nest breeding myobatrachid, also utilizes small The evolution of this pattern, however, has clearly aquarium sized puddles and has facultatively car- occurred three different times. The myobatrachid nivorous larvae. Assa has male inguinal brooding pouches, the The myobatrachids and leptodactylids most myobatrachid Rheobatrachus has a female gastric likely had a common ancestor that was widespread brooding pouch, and the leptodactyloid throughout the Antarcto-Tertiary Geoflora of Rhinoderma has a male vocal sac brooding pouch. Gondwanaland. Both the myobatrachids and The evolution of the foam nest has had a simi- leptodactylids have members that are still asso- lar, but independent, evolutionary history in both ciated with this vegetation. Arid centers developed the myobatrachids and leptodactylids. In both independently in Australia and South America. groups, the original adaptive significance of the The convergence of life histories and morphologies foam nest was in response to selective pressures of the myobatrachids and leptodactylids with deriv- associated with wet forest environments. The pres- ative temperate forests, open formation vegetations, ence of the foam nest then served as a preadapta- and arid environments is nothing short of spec- tion allowing radiation into open formation tacular. Literature Cited

Blake. A. J. D. Hennig, W. 1973. Taxonomy and Relationships of Myobatrachine 1966. Phylogenetic Systematics. Translated by D. D. Davis Frogs (Leptodactylidae): A Numerical Approach. and R. Zangerl. 263 pages, 69 figures. Urbana: Australian Journal of Zoology, 21:119-149. University of Illinois Press. dimming. A. £. Heyer, W. R. 1974. Low-Rainfall-Tolerant South African Anurans, with 1975a. A Preliminary Analysis of the Intergeneric Re- Particular Reference to Those of the Namib. 70 lationships of the Frog Family Leptodactylidae. pages. M.S. thesis, University of Natal. Smithsonian Contributions to Zoology, 199:1-55. Cookson, I. C. 1975b. Adenomera lutzi (Amphibia, Leptodactylidae), a 1953. The Identification of the Sporomorph Phyllocladi- New Species of Frog from Guyana. Proceedings of dites with Dacrydium and Its Distribution in South- the Biological Society of Washington, 88:315-318. ern Tertiary Deposits. Australian Journal of Botany, Heyer, W. R., R. W. McDiarmid, and D. L. Weigmann 1:64-70. 1975. Tadpoles, Predation and Pond Habitats in the Cookson, I. C, and K. M. Pike Tropics. Biotropica, 7:100-111. 1953. The Tertiary Occurrence and Distribution of Podo- Howes, G. B., and W. Ridewood carpus (Section Dacrycarpus) in Australia and Tas- 1888. On the Carpus and Tarsus of the Anura. Proceed- mania, Australian Journal of Botany, 1:71-82. ings of the Zoological Society of London, 1888:141- 1954. The Fossil Occurrence of Phyllocladus and Two 182. Other Podocarpaceous Types in Australia. Aus- Jones, E. I. tralian Journal of Botany, 2:60-68. 1933. Observations on the Pectoral Musculature of Am- Corben, C. J., G. J. Ingram, and M. J. Tyler phibia Salientia. Annals and Magazine of Natural 1974. Gastric Brooding: Unique Form of Parental Care History, series 10, 12:403-420. in an Australian Frog. Science, 186:946-947. Keast, A. Darlington, P. J., Jr. 1959. The Reptiles of Australia. Monographiae Biologi- 1965. Biogeography of the Southern End of the World. cae, 8 (Biogeography and Ecology in Australia): x + 236 pages, 38 figures. Cambridge: Harvard 115-135. University Press. Kluge, A. G. Dumeril, A. M. C, and G. Bibron 1967. Systematics, Phylogeny, and Zoogeography of the 1841. Erpe'tologie generate, ou historie naturelle complete Lizard Genus Diplodactylus Gray (Gekkonidae). des reptiles. Volume 8, 784 pages. Paris. Australian Journal of Zoology, 15:1007-1108. Dunlap. D. G. 1960. The Comparative Myology of the Pelvic Append- Lee, A. K. age in the Salientia. Journal of Morphology, 1967. Studies in Australian Amphibia, II: Taxonomy, 106:1-76. Ecology, and Evolution of the Genus Heleioporus Fitzinger, L. Gray (Anura: Leptodactylidae). Australian Journal 1843. Systema Reptilium. vi + 106 pages. Vienna: of Zoology, 15:367-439. Braimiullcr und Seidel. Liem, D. S. Fletcher, J. J. 1970. The Morphology, Systematics, and Evolution of the 1894. Description of a New Cystignathoid Frog from Old World Treefrogs (Rhacophoridae and Hyper- New South Wales. Proceedings of the Linnean oliidae). Fieldiana: Zoology, 57:1-145. Society of New South Wales, 8:522-523. 1973. A New Genus of Frog of the Family Leptodacty- Girard, C. lidae from S.E. Queensland, Australia. Memoirs of 1853. Descriptions of New Species of Reptiles Collected the Queensland Museum, 16:459-470. by U.S. Exploring Expedition under the Command In press. The Morphology and Systematics of Australo- of Captain Wilkes, U.S.N., (Second Part) Including Papuan Leptodactyloid Frogs (Anura: Myobatra- the Species of Batrachians Exotic to North America. chidae). Records of the Australian Museum. Proceedings of the Academy of Xatural Sciences of Liem, D. S., and W. Hosmer Philadelphia, 6:420-424. 1973. Frogs of the Genus Taudactylus with Descriptions Giinther, A. of Two New Species (Anura: Leptodactylidae). 1864. Third Contribution to Our Knowledge of Batrach- Memoirs of the Queensland Museum, 16:435-457. ians from Australia. Annals and Magazine of Littlejohn, M. J. Natural History, series 3, 14:311-315. 1963. The Breeding Biology of the Baw Baw Frog

28 NUMBER 233 29

Philoria frosti Spencer. Proceedings of the Linnean and Mammal Faunas in Relation to Torres Strait. Society of New South Wales, 88:273-276. Pages 257-300 in D. Walker, editor, Bridge and 1967. Patterns of Zoogeography and Speciation in South- Barrier: The Natural and Cultural History of eastern Australian Amphibia. Pages 150-174 in Torres Strait, xxii + 437 pages. Canberra: Aus- A. H. Weatherly, editor, Australian Inland Waters tralian National University. and Their Fauna: Eleven Studies. Canberra: Aus- Sharrock, C, and J. Felscnstein tralian National University Press. 1975. Finding All Monothetic Subsets of a Taxonomic 1970. Crinia tasmaniensis (Anura: Leptodactylidae): Geo- Group. Systematic Zoology, 24(3):373-377. graphic Distribution, Mating Call Structure and Spencer, B., editor Relationships. Proceedings of the Linnean Society 1896. Report on the Work of the Horn Scientific Expe- of New South Wales, 94:119-127. dition to Central Australia. 2 volumes. Melbourne: Lucas, A. H. S. Melville, Mullen and Slade. 1892. Notes on the Distribution of Victorian Batrachians, Straughan, Ian R. with Descriptions of Two New Species. Proceedings 1966. An Analysis of Species Recognition and Species Iso- of the Royal Society of Victoria, 4:59-64. lation in Certain Queensland Frogs. Ph.D. thesis. Lynch, J. D. University of Queensland. 1971. Evolutionary Relationships, Osteology, and Zoo- Trcwavas, E. geography of Leptodactyloid Frogs. University of 1933. The Hyoid and Larynx of the Anura. Philosophical Kansas Museum of Natural History Miscellaneous Transactions of the Royal Society of London, series Publications, 53:1-238. B, 222:401-527. Main, A. R. Trueb, L. 1968. Ecology, Systematics and Evolution of Australian 1973. Bones, Frogs, and Evolution. Pages 65-132 in J. L. Frogs. Advances in Ecological Research, 5:37-86. Vial, editor, Evolutionary Biology of the Anurans, Main, A. R., M. J. Littlejohn, and A. K. Lee Contemporary Research on Major Problems, xii + 1959. Ecology of Australian Frogs. Monographiae Biologi- 470 pages. Columbia: University of Missouri Press. cae, 8 (Biogcography and Ecology in Australia): Tschudi, J. J. 396-411. 1838. Classification der Batrachier, mit Berucksichtigung Martin, A. A. der fossilen Thiere dieser Abtheilung der Reptilien. 1970. Parallel Evolution in the Adaptive Ecology oi Memoires de la Societe des Sciences Naturelles de Leptodactylid Frogs of South America and Aus- Neuchatel, 2:1-102. tralia. Evolution, 24:643-644. Tyler, M. J. 1972. Studies in Australian Amphibia, III: The Limno- 1971. Superficial Mandibular Musculature, Vocal Sacs dynastes dorsalis Complex (Anura: Leptodactylidae). and the Phylogeny of Australo-Papuan Leptodacty- Australian Journal of Zoology, 20:165-211. lid Frogs. Records of the South Australian Museum, Moore, J. A. 16:1-20. 1961. The Frogs of Eastern New South Wales. Bulletin Watson, G. F., and A. A. Martin of the American Museum of Natural History, 1973. Life History, Larval Morphology and Relationships 121:149-386. of Australian Leptodactylid Frogs. Transactions of Parker, H. W. the Royal Society of South Australia, 97:33-45. 1940. The Australasian Frogs of the Family Leptodacty- Zweifel, R. G. lidae. Novitates Zoologicae, 42:1-106. 1972. A Review of the Frog Genus Lechriodus (Lep- Schodde, R., and J. H. Calaby todactylidae) of New Guinea and Australia. Ameri- 1972. The Biogeography of the Australo-Papuan Bird can Museum Novitates, 2507:1-41.

•it U.S. GOVERNMENT PRINTING OFFICE: 1*7* Z1l-SB*/l«

Publication in Smithsonian Contributions to Zoology Manuscripts for serial publications are accepted by the Smithsonian Institution Press, sub- ject to substantive review, only through departments of the various Smithsonian museums. Non- Smithsonian authors should address inquiries to the appropriate department. If submission is invited, the following format requirements of the Press will govern the preparation of copy. Copy must be typewritten, double-spaced, on one side of standard white bond paper, with W'i" top and left margins, submitted in ribbon copy with a carbon or duplicate, and accompa- nied by the original artwork. Duplicate copies of all material, including illustrations, should be retained by the author. There may be several paragraphs to a page, but each page should begin with a new paragraph. Number consecutively all pages, including title page, abstract, text, litera- ture cited, legends, and tables. The minimum length is 30 pages, including typescript and illus- trations. The title should be complete and clear for easy indexing by abstracting services. Taxonomic titles will carry a final line indicating the higher categories to which the taxon is referable: "(Hymenoptera: Sphecidae)." Include an abstract as an introductory part of the text. Identify the author on the first page of text with an unnumbered footnote that includes his professional mailing address. A table of contents is optional. An index, if required, may be supplied by the author when he returns page proof. Two headings are used: (1) text heads (boldface in print) for major sections and chap- ters and (2) paragraph sideheads (caps and small caps in print) for subdivisions. Further headings may be worked out with the editor. In taxonomic keys, number only the first item of each couplet; if there is only one couplet, omit the number. For easy reference, number also the taxa and their corresponding headings throughout the text; do not incorporate page references in the key. In synonymy, use the short form (taxon, author, date:page) with a full reference at the end of the paper under "Literature Cited." Begin each taxon at the left margin with subse- quent lines indented about three spaces. Within an entry, use a period-dash (.—) to separate each reference. Enclose with square brackets any annotation in, or at the end of, the entry. For references within the text, use the author-date system: "(Jones, 1910)" and "Jones (1910)." If the reference is expanded, abbreviate the data: "Jones (1910:122, pi. 20: fig. 1)." Simple tabulations in the text (e.g., columns of data) may carry headings or not, but they should not contain rules. Formal tables must be submitted as pages separate from the text, and each table, no matter how large, should be pasted up as a single sheet of copy. Use the metric system instead of, or in addition to, the English system. Illustrations (line drawings, maps, photographs, shaded drawings) can be intermixed throughout the printed text. They will be termed Figures and should be numbered consecu- tively; however, if a group of figures is treated as a single figure, the components should be indicated by lowercase italic letters on the illustration, in the legend, and in text references: "Figure 9b." If illustrations (usually tone photographs) are printed separately from the text as full pages on a different stock of paper, they will be termed Plates, and individual components should be lettered (Plate 96) but may be numbered (Plate 9: figure 2). Never combine the numbering system of text illustrations with that of plate illustrations. Submit all legends on pages separate from the text and not attached to the artwork. An instruction booklet for the preparation of illustrations is available from the Press on request. In the bibliography (usually called "Literature Cited"), spell out book, journal, and article titles, using initial caps with all words except minor terms such as "and, of, the." For capitalization of titles in foreign languages, follow the national practice of each language. Underscore (for italics) book and journal titles. Use the colon-parentheses system for volume, number, and page citations: "10(2) :5-9." Spell out such words as "figures," "plates," "pages." For free copies of his own paper, a Smithsonian author should indicate his requirements on "Form 36" (submitted to the Press with the manuscript). A non-Smithsonian author will receive 50 free copies; order forms for quantities above this amount with instructions for pay- ment will be supplied when page proof is forwarded. K 1

•

"iHr ^ ^MBI