Bull. not. Hist. Mus. Lond. (Zool.) 65( 1 ): 1-13 



Issued 24 June 1999 



Phylogenetic relationships of Toad-headed 

 lizards (Phrynocephalus, Agamidae) based on 



morphology ^ ^^ ^ 



T 

 E.N.ARNOLD 



Department of Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK 



THENA 







LIICTni 



1999 



_ 



CONTENTS 



1RARY 



Introduction 1 



Relationships of Phrynocephalus 1 



Morphological characters used to estimate phylogeny 2 



Phylogenetic analysis 7 



Discussion 9 



Acknowledgements 1 1 



References 11 



SYNOPSIS. Phrynocephalus together with its sister-group, Bufoniceps is most closely related to other advanced Palaearctic and 

 African agamids.They have been regarded as the sister-group of all these species or derived from African Agama (Moody, 1980, 

 morphological data) or as the sister of Laudakia (Joger, 1991. albumin immunology) but reassessment of morphology suggests 

 a relationship to Trapelus. Parsimony analysis of 46 morphological characters, involving 54 derived states, of 25 species of 

 Phrynocephalus indicates that successive branches arising from the main lineage of the genus are as follows: P. mystaceus; P. 

 maculatus; P. arabicus; the P. interscapularis group - (((P. clarkorum, P. ornatus) ((P. euptilopus, P. luteoguttatus) (P. 

 interscapularis, P. sogdianus))); P. scutellatus; P. golubevi; P. reticulatus; P. raddei. There is then a group of 1 1 species in which 

 relationships are generally poorly resolved, although within this P. theohaldi, P. roborowskii and P. vlangalii are clearly closely 

 related to each other and perhaps to P. forsythii, and the tuberculated species, P. helioscopus, P. persicus, P. rossikowi and P. 

 strauchi may also form a clade. There is no clear morphological evidence that the northeastern species, P. axillaris, P. versicolor, 

 P. przewalskii. and P. guttatus (which also extends far to the west) form a holophyletic group. Phrynocephalus does not appear 

 to share its general phylogeographic pattern with other Asian reptiles and this may consequently result from dispersal rather than 

 vicariance events. The phylogeny suggests the ancestor of Phrynocephalus occurred in Arabia-NW India area whence there were 

 three independent invasions of Central Asia: by the ancestors of P. mystaceus, of P. interscapularis + P. sogdianus, and of P. 

 golubevi and its sister group, the latter later extending north and eastwards into Mongolia, China and Tibet. Phrynocephalus 

 appears to have primitively occupied aeolian sand habitats but to have spread to harder substrates from which sandy habitats were 

 sometimes reinvaded. Degeneration of the outer and middle ear occurred in the early history of Phrynocephalus but was partly 

 reversed in P. axillaris and the P. theobaldi group. 



INTRODUCTION 



Toad-headed agamids, Phrynocephalus Kaup 1825, are a found in 

 the mainly Palearctic desert regions of Asia, from Eastern Turkey 

 and Russia to Mongolia, and southwards to southern Arabia and 

 Pakistan. Species in the south and centre of the range of the genus 

 are, in the main, well defined but, in the northeast, boundaries 

 between them are often less clear and numerous nominal taxa have 

 been described (see e.g. Zhao & Adler, 1993). This makes the total 

 number of species in the genus uncertain but it is likely to be in 

 excess of 30. In this paper, an estimate of phylogeny is made for 25 

 of the better defined species using morphological characters, includ- 

 ing external features and some internal ones derived from the 

 skeleton, middle ear, shoulder muscles and abdominal arteries. 



RELATIONSHIPS OF PHRYNOCEPHALUS 



Phrynocephalus is the sister group of the monotypic Bufoniceps 



© The Natural History Museum, 1999 



Arnold 1 992 which was created for Phrynocephalus laungwalaensis 

 Sharma, 1978. Moody (1980) placed Phrynocephalus (including P. 

 laungwalaensis) with what at the time was usually called Agama 

 Daudin 1 802, in his group 6 of the Agamidae. WithinAgama, as then 

 understood, this author recognised several separate genera: Agama 

 s. str., Xenagama Boulenger 1895, Pseudotrapelus Fitzinger 1843, 

 Trapelus Cuvier 1817 and Ste/Z/o Laurenti 1768. However, the name 

 Stellio is unavailable (Stejneger, 1933) and the assemblage it was 

 used to denote by Moody is paraphyletic, comprising distinct 

 Palaearctic and mainly African assemblages (Joger, 1991; Baig & 

 Bohme, 1997) of which the former is probably a clade and the 

 members of the latter more closely related to such taxa as Agama, 

 Pseudotrapelus and Trapelus (personal observations). Leviton, 

 Anderson, Adler & Minton (1992) argue for the use of Laudakia 

 Gray, 1845 for the Palaearctic forms, a course followed here. The 

 more recent suggestion (Henle, 1995), that Laudakia should be 

 confined to some members of this assemblage and the rest placed in 

 Placoderma Blyth, 1854, requires more thorough assessment of the 

 relationships of these lizards before it is adopted. The name 

 Acanthocercus Fitzinger, 1843 is available for the remainder of the 



