Cladogenesis and Vicariance Patterns in the Toad-Headed Lizard Phrynocephalus versicolor Species Complex

Using mtDNA sequence data and a phylogenetic approach, we examined the divergence and evolutionary history of the Toad-Headed Lizards of the Phrynocephalus versicolor species complex, which are currently assigned to three species, P. versicolor, Phrynocephalus frontalis, and Phrynocephalus przewalskii. Both parsimony and likelihood-based analyses resulted in three clades, which do not conform to the current species designation. The three species did not form mutually exclusive monophyletic groups. Instead, the populations from Tengger Desert, Ordos Highland, and further east, which include P. frontalis, P. przewalskii, and several populations of P. versicolor, formed a clade and we recommend recognizing them as a single species under the name P. przewalskii. The populations from the northwestern part of the complex's distribution also formed a monophyletic group, for which we recommend retaining the name P. versicolor. Finally, the southwestern populations share a unique evolutionary history and should be recognized as a new species, Phrynocephalus sp. The resulting phylogeny suggests that the species diversity and distribution patterns of P. przewalskii, P. versicolor, and Phrynocephalus sp. can be better explained by vicariance, rather than the traditional west-to-east dispersal hypothesis.

Fig. 1.

Map of northern China and the sampling sites. 1 = Abag; 2 = Sonid Zouqi; 3 = Erenhot; 4 = Sonid Youqi; 5 = Bailingmiao; 6 = Dongsheng; 7 = Yulin; 8 = Wuda; 9 = Bayan Hot; 10 = Zhongwei; 11 = Wuwei; 12 = Jinchang; 13 = Jiayuguan; 14 = West of Jiayuguan; 15 = Qiaowan; 16 = Xingxingxa; 17 = Luyuan; 18 = Ejin Qi; 19 = Heyingshan; 20 = Shibanjing; 21 = Hongliu Spring; 22 = Hami; 23 = Kuytun; 24 = Bole; 25 = Burgin. Solid cycles are Phrynocephalus przewalskii; Solid triangles are Phrynocephalus sp.; solid squares are Phrynocephalus versicolor; solid pentagons are Phrynocephalus guttatus; open cycles are reference cities; cross is Djungar Gate

Fig. 2.

The strict consensus tree from the maximum parsimony analysis. The taxa are the haplotypes, which are designated by locations and letters. Numbers above the branches are bootstrap proportions greater than 90 and numbers below the branches are decay indices. Numbers in front of each haplotype are location numbers in Figure 1. Populations <6> and <7> are Phrynocephalus frontalis sensu lato, which are labeled by f.; populations <10>, <11>, and <12> are Phrynocephalus przewalskii sensu lato, which are labeled by p.; the remaining populations are Phrynocephalus versicolor sensu lato, which are labeled by v.; Ejin Qi A = Hongliu Spring B; Hongliu Spring A = Heyingshan B; Hami A = Hami B. Xingxingxa A = Xingxingxa B. The vertical lines show the major clades

Fig. 3.

The single maximum likelihood tree. Taxon annotation is the same as that in Figure 2. The 50% majority consensus tree from the Bayesian analysis shares the same topology. The branch length is determined by likelihood calculations. The numbers by the nodes are the Bayesian posterior probabilities greater than 90%. The vertical lines show the major clades and species designation

Contributor Notes

(YW) Chengdu Institute of Biology, Chinese Academy of Sciences, Chengdu, Sichuan 610041, China; and (JF) Department of Zoology, University of Guelph, Guelph, Ontario N1G 2W1, Canada. (YW) arcib@cib.ac.cn; and (JF) jfu@uoguelph.ca Send reprint requests to JF.

Accepted: 17 Nov 2003