Phylogenetic relationships of horned lizards (Phrynosoma) based on nuclear and mitochondrial data: Evidence for a misleading mitochondrial gene tree
Introduction
Mitochondrial DNA (mtDNA) sequences have provided valuable data for phylogeographic and interspecific phylogenetic studies for the past two decades (Avise, 1994, Avise, 2000). This reflects a number of favorable attributes of mtDNA including ease of data collection (universal primers and large copy numbers), rapid rates of evolution (in animals), and the fact that mtDNA is functionally haploid and uniparentally inherited, and thus is expected to coalesce more rapidly on average than most nuclear genes (Edwards and Beerli, 2000, Hudson and Turelli, 2003). Nevertheless, it is well-known that mtDNA is not a panacea for phylogenetics, and can be subject to a number of confounding complications including incomplete lineage sorting of retained ancestral polymorphisms, introgression, and sex-biased life histories (Avise, 1994, Avise, 2000, Harrison, 1991, Niegel and Avise, 1986, Moore, 1995). Thus, it is important to consider these possibilities when significant discordance is observed among diverse data sets such as morphology, mtDNA, and nuclear genes (Ballard and Whitlock, 2004, de Queiroz et al., 1995, Eernisse and Kluge, 1993). Several phylogenetic studies of horned lizards (genus Phrynosoma) have identified strong incongruence between mtDNA and morphological data (Hodges and Zamudio, 2004, Reeder and Montanucci, 2001), but have not provided a definitive explanation for the discordance. The focus of the present study is to resolve the previously identified incongruence, thereby allowing for the generation of a robust, multi-locus phylogenetic estimate for Phrynosoma.
Horned lizards (Phrynosoma), with their unusual morphology and life-history, are among the most distinctive and recognizable of North American animals. Not surprisingly, these lizards have been the subjects of diverse research including many taxonomic and phylogenetic studies (e.g., Heath, 1964, Hodges, 2004a, Hodges and Zamudio, 2004, Howard, 1974, Lynn, 1965, Montanucci, 1987, Montanucci, 1989, Pianka and Parker, 1975, Presch, 1969, Reeder and Montanucci, 2001, Reeve, 1952, Schmidt et al., 1989, Sherbrooke, 1987, Sherbrooke, 1997, Sherbrooke, 2001, Sherbrooke, 2003, Sherbrooke et al., 2004, Sherbrooke and Mendoza-Quijano, 2005, Sherbrooke and Middendorf, 2001, Sherbrooke and Montanucci, 1988, Zamudio, 1998, Zamudio et al., 1997, Zamudio and Parra-Olea, 2000). Most horned lizards have diets consisting primarily of ants, a dietary specialization which might be linked to an unusual complement of adaptations including a squat, dorsoventrally compressed body, cranial horns and body fringes, camouflaging coloration, and sedentary habits (Sherbrooke, 2003). The horned lizard body plan effectively eliminates the option of rapid escape from predators; rather than taking flight, horned lizards generally attempt to avoid predation by relying on crypsis (Sherbrooke, 2001). Antipredator behavior is taken to the extreme in the bizarre rock-mimicking strategy employed by P. modestum (Sherbrooke and Montanucci, 1988). Upon capture, some species orient their heads downward to make their long temporal horns appear more formidable (Sherbrooke, 1987, Sherbrooke, 2003), or more spectacularly, will squirt blood from their ocular-sinus (Middendorf and Sherbrooke, 1992, Sherbrooke and Middendorf, 2001, Sherbrooke and Middendorf, 2004). Sherbrooke and Middendorf (2001) estimated that an individual P. cornutum may lose up to 53% of its total blood volume in a single squirt.
Prior to 1997, researchers recognized 12 species of Phrynosoma distributed from Canada to Guatemala, but recent reevaluations of species limits within three polytypic groups have increased that number to 17 (P. blainvillii, P. cerroense, P. coronatum, and P. wigginsi: Montanucci, 2004; P. douglasii and P. hernandesi: Zamudio et al., 1997; P. goodei and P. platyrhinos: Mulcahy et al., in press). For reasons outlined above, the relationships among these 17 species remain controversial, with the consequence of seriously confounding detailed comparative analyses of their peculiar morphologies and unusual life-history strategies.
Reeve (1952) presented the first phylogenetic hypotheses for Phrynosoma based on non-cladistic treatments of morphological and osteological characteristics. This was followed by the implicitly cladistic osteological study of Presch (1969), which generated a phylogenetic hypothesis for horned lizards very much at odds with that of Reeve (1952). Montanucci (1987) conducted the first explicitly cladistic analysis of the group, basing his analysis on osteology and scalation. The first molecular study was undertaken by Reeder and Montanucci (2001), in which relationships were inferred on the basis of 671 bp of 12S and 16S rRNA (mtDNA) sequence data. The Reeder and Montanucci (2001) study also included a reanalysis of Montanucci’s original morphological data (32 informative characters). Their preferred phylogeny, based on a combination of morphological and genetic data, was fully resolved, but with only three strongly supported clades: (1) P. taurus and P. braconnieri, (2) P. ditmarsi, P. hernandesi, and P. orbiculare, and (3) P. platyrhinos and P. mcallii. Reeder and Montanucci (2001) detected strong incongruence between their mtDNA and morphological data sets based on consensus tree methods and Wilcoxon signed-ranks tests. Hodges and Zamudio (2004) expanded the Reeder and Montanucci (2001) study by adding 1797 additional mtDNA characters (Cyt b and ND4) and including genetic data for P. braconnieri and P. douglasii. Their phylogenetic estimate based on combined mtDNA and morphological data supported five clades with strong support: (((P. ditmarsi + P. hernandesi) + P. douglasii) + P. orbiculare); (P. platyrhinos + P. mcallii); (P. taurus + P. braconnieri). However, they also identified strong incongruence between their morphological and mtDNA topologies using partition homogeneity tests and parametric bootstrapping. The only clade retained in a strict consensus of their preferred mtDNA and morphology trees was (P. taurus + P. braconnieri).
In this paper, we take a multi-locus approach and analyze data from three nuclear genes, as well as six mitochondrial genes and their flanking tRNAs. Our nuclear genes (>2.2 kb) include recombination activating gene-1 (RAG-1), brain-derived neurotrophic factor (BDNF), and glyceraldehyde-3-phosphate dehydrogenase (GAPD). We expand the existing mtDNA data set to include over 5.1 kb of sequence data including 12S, 16S, ND1, ND2, ND4, Cyt b, and associated tRNA genes. Because one of our primary objectives is to identify and resolve the factors generating strong incongruence among competing Phrynosoma topologies, we compare the trees inferred from our mtDNA and nuclear data using local and global tests of congruence to identify specific problematic relationships. Processes potentially responsible for incongruence could include character convergence (applicable to morphology and DNA), incomplete lineage sorting of ancestral polymorphisms (applicable to genetic data), and reticulation (introgressive hybridization, recombination, and horizontal gene transfer). Interspecific hybridization is documented in Phrynosoma (Baur, 1984, Montanucci, 2004, Mulcahy et al., in press), and therefore introgression via cytoplasmic capture could provide an explanation for discordance between data sets (as suggested previously by Reeder and Montanucci, 2001).
We use our preferred phylogeny of Phrynosoma to reconstruct the evolutionary history of several interesting characters, reinterpret the biogeographic history of the group, and provide a phylogenetic taxonomy. Anti-predator blood-squirting is considered to be the ancestral state in Phrynosoma, but the number of subsequent losses of this character remains unclear (Hodges, 2004b, Sherbrooke et al., 2004, Sherbrooke and Mendoza-Quijano, 2005, Sherbrooke and Middendorf, 2001). Variation in crainial horn morphology is striking within Phrynosoma, with some species exhibiting an impressive armament of horns (e.g., P. solare and P. mcallii), while other species almost lack horns entirely (e.g., P. ditmarsi and P. douglasii). We reevaluate the traditional division of Phrynosoma into southern and northern radiations (Montanucci, 1987) and present a revised biogeographic scenario. Finally, we provide a phylogenetic taxonomy to emphasize the strongly supported clades that should be of particular interest for future comparative studies within Phrynosoma.
In addition, as a by-product of our outgroup sampling and rooting strategy we are able to infer the phylogenetic relationships among the remaining phrynosomatid sand lizard genera (Cophosaurus, Callisaurus, Holbrookia, and Uma). The relationships among these lineages are unclear, and previous studies based on morphology (de Queiroz, 1989), allozymes (de Queiroz, 1992), and mtDNA (Wilgenbusch and de Queiroz, 2000) have produced conflicting results. The two points of contention among competing hypotheses are whether Uma or Holbrookia was the first lineage to diverge and if the two “earless” genera with concealed tympanic membranes (Cophosaurus and Holbrookia) form a clade. Although Wilgenbusch and de Queiroz (2000) found strong support for an “earless” clade based on Cyt b and 12S mtDNA data, they could not infer the placement of Uma with strong statistical support.
Section snippets
Taxon sampling
We included exemplars of all currently recognized species of Phrynosoma in our analysis, including new species from recent systematic revisions. Our taxon sampling scheme enables us to infer the phylogenetic relationships among all recognized species within Phrynosoma, but it does not enable us to test the monophyly of any species. We included representatives of the four evolutionary species (P. blainvillii, P. cerroense, P. coronatum, and P. wigginsi) within what was previously considered P.
Mitochondrial DNA
The 12S rRNA data included 778 aligned nucleotide positions, 84 of which were excluded due to ambiguous alignment, leaving a final data set of 694 characters. When combined with the 16S rRNA and tRNA data (leucine, isoleucine, glutamine, tryptophan, and alanine), the combined structural RNA data set contained 1416 aligned positions and 133 excluded sites. The ND1 protein-coding gene was 969 bp for most taxa, except for Uta stansburiana and Phrynosoma mcallii, which both shared an amino acid
Introgression and the phylogenetic relationships of horned lizards
Previous phylogenetic analyses of Phrynosoma incorporating at least two data sets have detected strong incongruence (Hodges and Zamudio, 2004, Reeder and Montanucci, 2001). We attempted to resolve this incongruence by generating a multi-locus phylogenetic estimate for Phrynosoma based on three nuclear genes, but in doing so discovered additional sources of conflict. Although the topologies inferred from the mtDNA and nuclear data partitions were significantly different, we were able to identify
Acknowledgments
We are grateful to Wendy Hodges, Brad Hollingsworth, Dan Mulcahy, Tod Reeder, and Erica Rosenblum for providing tissue samples, and Matt Brandley and James Schulte for providing unpublished primer sequences. We thank the following lab groups and individuals for providing useful comments on earlier versions of this manuscript: the McGuire and Wake Labs at UC Berkeley, the Quasi-Independent Phylogenetics Discussion Group at KU, C. J. Cole, A. Larson, R. Montanucci, D. Mulcahy, T. Papenfuss, W.
References (100)
- et al.
Horned lizard (Phrynosoma) phylogeny inferred from mitochondrial genes and morphological characters: understanding conflicts using multiple approaches
Mol. Phylogent. Evol.
(2004) Phylogenetic relationships among phrynosomatid lizards as inferred from mitochondrial ribosomal DNA sequences: substitutional bias and information content of transitions relative to transversions
Mol. Phylogenet. Evol.
(1995)- et al.
Stone-mimicry in the round-tailed horned lizard Phrynosoma modestum (Sauria: Iguanidae)
J. Arid Environ.
(1988) - et al.
Molecular homology assessment and phylogeny in the lizard family Opluridae (Squamata: Iguania)
Mol. Phylogenet. Evol.
(1996) Mother species-father species: unidirectional hybridization in animals with female choice
Anim. Behav.
(1999)- et al.
Comment on “How the horned lizard got its horns”
Science
(2004) - et al.
Mitochondrial DNA sequence divergence and phylogenetic relationships among eight chromosome races of the Sceloporus grammicus complex (phrynosomatidae) in central Mexico
Syst. Biol.
(1994) Molecular Markers, Natural History, and Evolution
(1994)Phylogeography. The History and Formation of Species
(2000)Description of new genera and species of North American lizards in the museum of the Smithsonian Institution
Proc. Acad. Nat. Sci. Philadelphia
(1858)