Evolution of Galapagos Island Lava Lizards (Iguania: Tropiduridae: Microlophus)

Transcription

1 Molecular Phylogenetics and Evolution 32 (2004) MOLECULAR PHYLOGENETICS AND EVOLUTION Evolution of Galapagos Island Lava Lizards (Iguania: Tropiduridae: Microlophus) David Kizirian, a,b, * Adrienne Trager, c Maureen A. Donnelly, b and John W. Wright d a Department of Organismic Biology, Ecology and Evolution, University of California Los Angeles, Los Angeles, CA , USA b Department of Biological Sciences, Florida International University, Miami, FL 33199, USA c Moorpark College, 7075 Campus Road, Moorpark, CA 93021, USA d Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, CA 90007, USA Received 18 June 2003; revised 24 March 2004 Available online 2 June 2004 Abstract Nucleotide sequences of mitochondrial genes (ND1, ND2, COI, and trnas) were determined for 38 samples representing 15 taxa of tropidurid lizards from the Galapagos Islands and mainland South America. Phylogenetically informative characters (759 of 1956) were analyzed under Bayesian, maximum likelihood, and parsimony frameworks. This study supports the hypothesis that tropidurid lizards dispersed to the Galapagos on at least two separate occasions. One dispersal event involved an eastern Galapagos clade (Microlophus habelii and M. bivittatus, on Marchena and San Cristobal islands, respectively) the sister taxon of which is M. occipitalis from coastal Ecuador and Peru; the closest mainland relative of the western Galapagos clade was not unambiguously identified. The wide-ranging M. albemarlensis is revealed to be a complex of weakly divergent lineages that is paraphyletic with respect to the insular species M. duncanensis, M. grayii, and M. pacificus. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Biogeography; Galapagos; Microlophus; Phylogeography; Tropiduridae; Tropidurus 1. Introduction Despite more than a century of intensive study there are still many unanswered questions regarding the evolutionary history of organisms inhabiting the Galapagos Islands (e.g., Grehan, 2001). A fundamental issue addressed herein is the number of species of the conspicuous and abundant Lava Lizards (Microlophus; auct. Tropidurus) occurring on the islands. Based on an analysis of morphological variation, Van Denburgh and Slevin (1913) recognized seven species of Lava Lizards in the archipelago, an arrangement that has been followed by most (e.g., Lopez et al., 1992; Wright, 1983). An alternative arrangement (Lanza, 1974, 1980; Talurri et al., 1982) based on geographic distribution and behavioral data (Carpenter, 1966, 1970) recognized 17 * Corresponding author. address: dkiziria@ucla.edu (D. Kizirian). named and unnamed lineages but has not been widely accepted. Central to the disagreement about Lava Lizard diversity is the status of the populations of M. albemarlensis, which occur on four major and at least six satellite islands. In addition to addressing species diversity we also address the biogeographic history of Lava Lizards. Previous biogeographic hypotheses inferred from the electrophoretic migration rates of allozymes (Wright, 1983, 1984) and micro-complement fixation of albumins (Lopez et al., 1992) indicated that at least two independent dispersal events from the mainland to the archipelago were required to explain Lava Lizards diversity. Herein, we revisit hypotheses about the diversity and biogeography of Galapagoan Microlophus in light of nucleotide sequence data analyzed in a phylogenetic context and broader taxonomic sampling than previous studies. At the same time, we evaluate hypotheses of monophyly for the M. occipitalis and M. peruvianus groups (Dixon and Wright, 1975) and the genus Microlophus (Frost et al., 2001; Harvey and /$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi: /j.ympev

2 762 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) Gutberlet, 2000). Lastly, we employ the principles of phylogenetic classification (de Queiroz and Gauthier, 1990) to represent nested diversity at the species level. 2. Materials and methods 2.1. Taxon sampling and data Exemplars for this study include some of the tissue samples used by Wright (1983) and Lopez et al. (1992) as well as additional samples of mainland species not previously examined (Table 1). Genomic DNA was extracted from frozen tissues (1 25 mg) using the DNEasy Tissue Kits (Qiagen-Operon Catalog # 69504, 69506). Nucleotide sequence data were collected using polymerase chain reaction (PCR) and automated DNA sequencing technology. Approximately 2 kb of the mitochondrial genome, including NADH 2, portions of NADH 1 and CO1, and the intervening trnas were amplified using the primers in Macey et al. (1997a,b) and Jackman et al. (1999) as well as [5 0 ] taarataagtcattttggg [3 0 ], which was designed for use with some light strand primers from those studies. Some fragments were amplified using a capillary thermal-cycler (Rapidcycler) and DNA amplification kits available from Idaho Technology (Salt Lake City, Utah). PCR cocktails comprised water (25 ll), 10 X buffer (20, 30, or 40 mm MgCl 2, 5.0 ll), dntps (2 mm each; 5.0 ll), bovine serum albumin (2.5 mg/ml; 5.0 ll), light and heavy strand primers (10 mm; 2.5 ll each), Taq polymerase (ProMega; 0.3 ll), and template DNA ( ll). Reactions were subjected to an initial heating (15 s, 94 C), then cycles of denaturation (15 s, 94 C), annealing (15 s, 53 C), and extension (35 s, 71 C), and then a final extension (60 s, 71 C). Most fragments were amplified using a block thermal-cycler (Perkin Elmer 9700) and AmpliTaq Gold Table 1 Samples of Microlophus and Tropidurus used in this study Species Voucher # (Tissue #) Locality M. albemarlensis LACM (G302) Galapagos: Baltra M. albemarlensis LACM (G216) Galapagos: Bartolome M. albemarlensis LACM (G217) Galapagos: Bartolome M. albemarlensis LACM (G327) Galapagos: Daphne M. albemarlensis LACM (G321) Galapagos: Daphne M. albemarlensis LACM (G150B) Galapagos: Fernandina: Punta Espinosa M. albemarlensis LACM (G142) Galapagos: Fernandina: Punta Espinosa M. albemarlensis LACM (G170) Galapagos: Isabela: Black Cove M. albemarlensis LACM (G162) Galapagos: Isabela: Black Cove M. albemarlensis LACM (G280B) Galapagos: Isabela: Cartago Bay M. albemarlensis LACM (G341) Galapagos: Santiago: James Bay M. albemarlensis LACM (G232) Galapagos: Santiago: Sullivan Bay M. albemarlensis LACM (G231) Galapagos: Santiago: Sullivan Bay M. albemarlensis LACM (G290) Galapagos: Santa Cruz: Conway Bay M. albemarlensis LACM (G11) Galapagos: Santa Cruz: Academy Bay M. albemarlensis LACM (G17) Galapagos: Santa Cruz: Academy Bay M. bivitattus LACM (G43) Galapagos: San Cristobal: Wreck Bay M. bivitattus LACM (G49) Galapagos: San Cristobal: Wreck Bay M. delanonis LACM (G104) Galapagos: Espanola: Gardner Bay M. delanonis LACM (G92) Galapagos: Gardner: near Espanola M. delanonis LACM (G93) Galapagos: Gardner: near Espanola M. duncanensis LACM (G261) Galapagos: Duncan (Pinzon): east side M. grayii LACM (G132) Galapagos: Floreana: Black Beach M. habelii LACM (G205) Galapagos: Marchena: south side M. habelii LACM (G210) Galapagos: Marchena: south side M. koepckeorum LACM (P6-193) Peru: Lambayeque: Cerro de la Vieja: ca. 7 km S (by rd) Motupe M. occipitalis LACM (G4-78) Ecuador: Ancon at Basurera M. pacificus LACM (G192) Galapagos: Pinta: southwest side M. pacificus LACM (G193) Galapagos: Pinta: southwest side M. peruvianus LACM (P6-381) Peru: Lima: El Paraiso Peninsula: ca 6.2 km W (by rd) jct Pan Am Hwy M. peruvianus LACM (G4-91) Ecuador: Santa Elena Peninsula, Punto Carnero M. stolzmanni LACM (P6-290) Peru: Cajamarca: ca 13 km SSE (by rd) Hacienda Molino Viejo (Ochentiuno) M. stolzmanni LACM (P6-258) Peru: Cajamarca: Bellavista M. theresiae LACM (P6-389) Peru: Lima: El Paraiso Peninsula: ca 6.2 km W (by rd) jct Pan Am Hwy M. tigris LACM (P6-84) Peru: Lima: ca 3 km SE Asia Vieja T. etheridgei [at LACM] (TC-921) Bolivia: 0.5 km SW Parotania RR station T. hispidus [at LACM] (4-94) Venezuela: Isla Margarita T. hispidus [at LACM] (4-76) Venezuela: Playa Guiria

3 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) DNA polymerase (Perkin Elmer) kits. PCR cocktails comprised water (17 35 ll), 10 X buffer (5.0 ll), MgCl 2 ( ll), dntps (5.0 ll), heavy and light strand primers (10 mm; 1.0 or 2.5 ll), Taq polymerase (0.3 ll), and template DNA ( ll). Reactions were subjected to a variety of thermal-cycling profiles including an initial heating (1 5 min, 95 C) then cycles of denaturation (35 s, 94 C), annealing (35 45 s, C), extension (60 s; or 150 s plus 4 s added to each subsequent cycle; 70 C, or 72 C), and a final extension (7 min, 72 C). To reduce the risk of contamination, all reagents, primer stocks, DNA extractions, and PCR cocktails were prepared in a facility that was physically isolated from areas where post-pcr procedures were performed. Kim Wipes were used to shield aerosol created when tubes were opened. In addition, pre-pcr tasks were performed earlier in the day than less sensitive procedures to reduce risk of contamination due to accumulated contaminant on skin and clothing (Wayne et al., 1999). Amplified fragments were purified using agarose gel electrophoresis and QIAquick Gel Extraction kits (Qiagen-Operon Catalog #28706). In a few cases, fragments were re-amplified using nested primer pairs. Fragments were labeled using Big Dye (2.0) florescent dye terminators and cycle-sequencing protocols for the ABI PRISM 377 Automated Sequencer. Sequencher software (version 4.1, Gene Codes) was used to edit the raw data and to produce a preliminary alignment of fragments. Ultimately, protein-coding genes were aligned assuming translation into a functional protein and transfer RNAs were aligned assuming secondary structure models for closely related taxa (Macey et al., 1997b). Nucleotide sequences were deposited at GenBank (AY ) Data analysis Of 1956 nucleotide positions, seven (characters ) in the TwC loop of trna Trp were excluded from phylogenetic analyses because they could not be unambiguously aligned, 980 were invariant, 210 were parsimony-uninformative, and 759 were parsimony-informative. A molecular clock was not assumed in any of the analyses. Trees were rooted with Iguana iguana (GenBank llg278511). Parsimony analysis was performed assuming equal weights for all transformation series and gaps were treated as a fifth character state. Heuristic tree search settings in PAUP* (version 4.0b10; Swofford, 2002) included addseq ¼ random, nreps ¼ 50,000, swap ¼ tbr, steepest ¼ yes, hold ¼ 5 or 10. Modeltest (version 3.06; Posada and Crandall, 1998) was used to identify the likelihood model that best fits the data (i.e., GTR + I + C), which was assumed in two subsequent tree searches using PAUP* (addseq ¼ random, nreps ¼ 825; start ¼ nj, nreps ¼ 1000, hold ¼ 1). Decay indices (Bremer, 1988) were calculated using TreeRot (version 2.0) with the number of heuristic searches increased to 1000 to increase accuracy of indices (Sorenson, 1999). Bootstrap statistics were calculated using PAUP* (fast-heuristic search, nreps ¼ 1002) and assumed model parameters calculated with Model test (i.e., GTR + I + C). MrBayes (version 3.0; Huelsenbeck and Ronquist, 2001) was used to estimate posterior probabilities of clades under a likelihood model. Substitution rates were allowed to be different, subject to the constraints of timereversibility (GTR; n ¼ 6). Among-site rate variation was drawn from a gamma distribution with some proportion being invariant (rates ¼ invgamma). Model parameters for structural and protein-coding partitions were treated as unlinked [unlink revmat ¼ (all) shape ¼ (all) pinvar ¼ (all) statefreq ¼ (all)]. Four (nchains ¼ 4) heated (temp ¼ 0.5) Markov chains were calculated simultaneously and sampled every 100 generations (samplefreq ¼ 100) for 2,000,000 generations (mcmc ngen ¼ 2,000,000). Stationarity was evaluated graphically (plot) and the first 2001 (of 20,001) trees were discarded (burnin ¼ 2001). Best trees were compared to those incorporating topological constraints (i.e., monophyly of Galapagoan Microlophus and monophyly of M. albemarlensis) using Kishino Hasegawa (for parsimony trees) and Shimodaira Hasegawa (for likelihood trees) tests in PAUP*. The Shimodaira Hasegawa test results are based on model parameters calculated with Modeltest (i.e., GTR + I + C), a RELL test distribution, and 10,000 bootstrap replicates. 3. Results Low guanine content (12%) suggests that nucleotide sequences represent mitochondrial genes rather than nuclear pseudogenes (Zhang and Hewitt, 1996). Parsimony analysis yielded three cladograms (Figs. 1 and 4; L ¼ 2558; CI ¼ ; CI excluding uninformative characters ¼ ; RI ¼ ; RC ¼ ) among which topological variation is due to ambiguous relationships within island groups (i.e., Santa Cruz + Baltra + Daphne, Santiago + Bartolome, and Fernandina + Isabela). Maximum likelihood analysis under the bestfitting model (GTR + I + C) yielded a single tree (Fig. 2; )ln L ¼ 13578:23698). The likelihood of the best state for the cold chain in the Bayesian analysis was ) (Fig. 3). Topologies from likelihood and parsimony analyses are largely congruent and possess comparable levels of support for most major clades. Likelihood and parsimony trees differ in the position of the M. koepckeorum, which may be the result of missing data (418 of 1956 positions were coded as missing for this taxon). Nevertheless, this study places this enigmatic taxon (Plesiomicrolophus of Frost, 1992; Microlophus of Frost et al., 2001; Harvey and Gutberlet, 2000) in the

4 764 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) Fig. 1. Strict consensus of three shortest cladograms (L ¼ 2558). Numbers on branches are decay indices. M. occipitalis group. Trees incorporating topological constraints (monophyly of Galapagoan Microlophus and monophyly of M. albemarlensis) are significantly different from best trees (Table 2). Although our taxonomic sampling was not exhaustive, the monophyly of Microlophus, Tropidurus, the M. peruvianus group, and the M. occipitalis group is supported here. Our results are congruent with those of Lopez et al. (1992), except that this study finds unambiguous support for the placement of M. peruvianus outside the smallest clade including all Galapagos island tropidurids. Phylogenetic analyses herein agree with WrightÕs (1983) phenetic clustering of island groups (i.e., Santa Cruz + Baltra + Daphne, Santiago + Bartolome) except for Isabela and Fernandina, which we find to be sister lineages and Wright (1983) identified as neighbors. This study finds M. albemarlensis to be paraphyletic with respect to M. duncanensis, M. grayii, and M. pacificus, corroborating an analysis based on allozyme variation (Wright, 1983) and a remark in Frost (1992, p. 49) that M. albemarlensis might represent grouping by plesiomorphy rather than on an understanding of the historical relationships of these island forms. 4. Discussion 4.1. Classification The existence of geographically restricted haplotype clades suggests that there is greater species diversity than implied by the classifications in Lopez et al. (1992) and

5 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) Fig. 2. Unrooted phylogram resulting from maximum likelihood analysis ()ln L ¼ 13578:23698) with proposed classification. Frost (1992), particularly with respect to the Microlophus albemarlensis complex. Greater inter-island diversity in Galapagos Microlophus is also suggested by variation in behavior (Carpenter, 1966, 1970), allozymes (Wright, 1983), and microsatellite DNA (Jordan et al., 2002). Intra-island variation in sprint speed (Snell et al., 1988) and endurance capacity (Miles et al., 2001) in M. albemarlensis has also been reported. Much of the divergence among M. albemarlensis, however, is weak. For example, sequence divergence among M. albemarlensis (excluding M. duncanensis, M. grayii, and M. pacificus) varies from 0 to 6% and morphological variation is nearly non-existent (Lanza, 1974, 1980; Van Denburgh and Slevin, 1913); after examination of more than 2000 specimens from 24 islands, Van Denburgh and Slevin (1913, p. 188) stated that the populations of M. albemarlensis are so similar that we have been able to find no characters which will distinguish them. The weak divergence within the M. albemarlensis complex could reflect slower rates of evolution or relatively recent isolation; a mechanism consistent with the latter is discussed below (see Section 4.2). This study corroborates previous work based on morphology (Van Denburgh and Slevin, 1913), behavior (Carpenter, 1966), and allozymes (Wright, 1983) that M. duncanensis, M. grayii, and M. pacificus represent diagnosable entities, therefore, we continue to recognize them. Because the relatively plesiomorphic lineages of

6 766 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) Fig. 3. Majority rule consensus tree resulting from Bayesian analysis. Numbers above branches are posterior probabilities of clade support and those below are bootstrap values. M. albemarlensis (i.e., those on Fernandina, Isabela, Santa Cruz, Santiago, and satellites) by themselves do not form a natural entity, we do not associate an exclusive binominal with it. Instead, we recognize the more inclusive M. albemarlensis complex, which includes the historically nested lineages M. duncanensis, M. grayii, and M. pacificus, as well as the lineages previously recognized as M. albemarlensis. It should be emphasized that the unappended binominal M. albemarlensis does not appear in our classification (Table 3) because it is paraphyletic. Should it become desirable to identify lizards of the M. albemarlensis complex from, for example, Isabela, we suggest M. albemarlensis complex from Isabela. Alternatively, there are numerous available names (e.g., Lanza, 1974, 1980; Talurri et al., 1982) that could be resurrected to recognize formally additional insular diversity within M. albemarlensis complex. The indented classification proposed here (Table 3) employs the principles of classification in de Queiroz

7 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) Table 2 Results of topological constraints tests Tree Shimodaira Hasegawa Kishino Hasegawa )ln L P Length P < < In Tree 1, M. albemarlensis was constrained to be monophyletic. In Tree 2, the monophyly of Galapagoan Microlophus was constrained, consistent with a single dispersal event to Galapagos. Tree 3 represents the best tree(s) for likelihood and parsimony analyses, respectively. Asterisks indicate significant difference at P < 0:05. Table 3 Proposed phylogenetic (indented) classification and geographic distributions of Microlophus occipitalis group species Microlophus occipitalis group M. albemarlensis complex Western Galapagos M. pacificus Pinta (Galapagos) M. duncanensis Pinzon (Galapagos) M. grayii Floreana (Galapagos) M. bivittatus San Cristobal (Galapagos) M. delanonis Espanola, Gardner (Galapagos) M. habelii Marchena (Galapagos) M. koepckeorum Coastal Peru M. occipitalis Coastal Ecuador and Peru M. stolzmanni Intra-Andean valleys of Ecuador and Peru and Gauthier (1990), despite the lack of consensus regarding how species should be treated (Cantino et al., 1999). Until such a consensus is reached, the classification used here is sufficiently informative and unambiguous regarding the diversity of Lava Lizards, it faithfully reflects the nested historical groupings discovered by phylogenetic analysis, and all recognized taxa have meaning in an evolutionary context (i.e., no paraphyletic taxa are included) Biogeography Our biogeographic inferences assume geological hypotheses summarized in Cox (1983), particularly the plate tectonics model that explains the greater portion of the history of the archipelago. In addition, we considered the phylogeny of Lava Lizards (this study), ecological data on extant tropidurids (Dixon and Wright, 1975), and data on the Humboldt Current (Wright, 1984; Wyrtki, 1967; Wyrtki et al., 1976). A phylogenetic tree superimposed on a map of the emergent portions of the Galapagos plate (Fig. 4) reveals concordance among geographic, geological, oceanographic, and phylogenetic data suggesting that we have recovered historical signal. For example, younger (western) islands harbor younger lineages, which were likely carried there by ocean currents from older (eastern) islands. Fig. 4. Map of Galapagos including adapted topology from the parsimony analysis, island names used in this study, direction of the Humboldt Current, movement of the Nazca Plate, and proposed classification. Roman numerals represent independent dispersal events from mainland. Previous studies based on allozyme variation (Wright, 1983; unpublished), immunological distances (Lopez et al., 1992), and morphology (Wright, unpublished) found M. bivittatus (Isla San Cristobal) and Microlophus habelii (Isla Marchena) to be more closely related to the mainland species M. occipitalis than to other Galapagos species, a conclusion that is unambiguously supported by this study. Given that the Galapagos Islands have never been attached to the mainland, the most parsimonious explanation of the available data is that tropidurids have dispersed to the archipelago on at least two independent occasions. Isabela is the largest island in the Galapagos and is geologically complex with five volcanic peaks and recent vulcanism (Cox, 1983). Studies of Lava Lizards (Wright, 1983) and tortoises (Beheregaray et al., 2003; Caccone et al., 1999, 2002; Ciofi et al., 2002) suggest that the biota on this island also have complex histories. For example, phylogenetic analysis of nucleotide variation allies lizards from western Isabela with those from Fernandina, rather than samples from eastern Isabela. A close relationship between western Isabela and Fernandina Lava Lizards could be explained by relatively recent (and not improbable) dispersal across the 5 km channel between the islands. In addition, although Lava Lizards from eastern Isabela are represented here by only a single sample, divergence values (1.5%) and parsimony analyses (Figs. 1 and 4) suggest that there has been independent evolution of Lava Lizards on Isabela.

8 768 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) A likely mechanism involved in the biogeographic history of Microlophus is transport of lizards by ocean currents (Wright, 1983, 1984; Wyrtki, 1967; Wyrtki et al., 1976). Although we have no direct evidence of Lava Lizards dispersing in this fashion, it has been suspected (e.g., Kizirian and Cole, 1999) and documented (Censky et al., 1998) for other lizard groups. The prevailing Humboldt Current flows northwesterly past the islands at approximately 7 knots, creates dry conditions in the archipelago, and may have played a role in the presumably rarer, long-distance dispersal events, particularly those from the mainland. During El Nino years, however, the flow of the current is reversed, which brings warm water from tropical latitudes and creates more mesic conditions in the Galapagos. In these wetter periods, freshwater systems in the Galapagos flood and wash vegetative mats downstream to the ocean, where they (and stowaway lizards) may be transported among islands, particularly the central ones (Wright, unpublished). The movement of vegetative mats among islands may have been responsible for maintaining gene flow in M. albemarlensis complex and explain the weak divergence within that clade. If this mechanism is currently operating to maintain the reproductive integrity of a single species on multiple islands, then M. albemarlensis (sensu Van Denburgh and Slevin, 1913), or parts thereof, may represent a non-exclusive entity (e.g., Graybeal, 1995). In any case, the classification offered here (Table 3) accommodates multiple interpretations of the available data. Acknowledgments In addition to those acknowledged previously (Wright, 1983, p. 153), we thank Kirk Fitzhugh, Dan Geiger, Mike Hardman, Michael Harvey, and Chris Thacker for reviewing drafts of this contribution; Mike Hardman, John Huelsenbeck, and Tod Reeder for assistance with data analysis; and the Ralph M. Parsons Foundation and W.M. Keck Foundation for funding. References Beheregaray, L.B., Ciofi, C., Geist, D., Gibbs, J.P., Caccone, A., Powell, J.R., Genes record a prehistoric volcano eruption in the Galapagos. Science 302, 75. Bremer, K., Limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42, Caccone, A., Gentile, G., Gibbs, J.P., Fritts, T.H., Snell, H.L., Betts, J.E., Powell, J.R., Phylogeography and history of giant Galapagos tortoises. Evolution 56, Caccone, A., Gibbs, J.P., Ketmaier, V., Suatoni, E., Powell, J.R., Origin and evolutionary relationships of giant Galapagos tortoises. Proc. Natl. Acad. Sci. USA 96, Cantino, P.D., Bryant, H.N., De Queiroz, K., Donoghue, M.J., Eriksson, T., Hillis, D.M., Lee, M.S.Y., Species names in phylogenetic nomenclature. Syst. Biol. 48, Carpenter, C.C., Comparative behavior of the Galapagos Lava Lizards (Tropidurus). In: Bowman, R.I. (Ed.), The Galapagos. Proceedings of the Symposia of the Galapagos International Scientific Project. University of California Press, pp Carpenter, C.C., Miscellaneous notes on Galapagos Lava Lizards (Tropidurus Iguanidae). Herpetologica 26, Censky, E.J., Hodge, J.K., Dudley, J., Over-water dispersal of lizards due to hurricanes. Nature 395, 556. Ciofi, C., Milinkovitch, C., Gibbs, J.P., Caccone, A., Powell, J.R., Microsatellite analysis of genetic divergence among giant Galapagos tortoises. Mol. Ecol. 11, Cox, A., Ages of the Galapagos Islands. In: Bowman, R.I. et al. (Eds.), Patterns of Evolution in Galapagos Organisms. AAAS Symposium Volume, San Francisco, USA, pp de Queiroz, K., Gauthier, J., Phylogeny as a central principle in taxonomy: phylogenetic definitions of taxon names. Syst. Zool. 39, Dixon, J.R., Wright, J.W., A review of the lizards of the iguanid genus Tropidurus in Peru. Cont. Sci. Nat. Hist. Mus. L.A. Co. 271, Frost, D.R., Phylogenetic analysis and taxonomy of the Tropidurus group of lizards (Iguania: Tropiduridae). Am. Mus. Nov. 3033, Frost, D.R., Rodrigues, M.T., Grant, T., Titus, T., Phylogenetics of the lizard genus Tropidurus (Squamata: Tropiduridae: Tropidurinae): direct optimization, descriptive efficiency, and sensitivity analysis of congruence between molecular data and morphology. Mol. Phylogenet. Evol. 21, Graybeal, A., Naming species. Syst. Biol. 44, Grehan, J., Biogeography and evolution of the Galapagos: integration of the biological and geological evidence. Biol. J. Linn. Soc. 74, Harvey, M.B., Gutberlet Jr., R.L., A phylogenetic analysis of the tropidurine lizards (Squamata: Tropiduridae), including new characters of squamation and epidermal microstructure. Zool. J. Linn. Soc. 128, Huelsenbeck, J.P., Ronquist, F., MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17, Jackman, T.R., Larson, A., de Queiroz, K., Losos, J.B., Phylogenetic relationships and tempo of early diversification in Anolis lizards. Syst. Biol. 48, Jordan, M.A., Hammond, R.L., Snell, H.L., Snell, H.M., Jordan, W.C., Isolation and characterization of microsatellite loci from Galapagos lava lizards (Microlophus spp.). Mol. Ecol. 2, Kizirian, D.A., Cole, C.J., Origin of the unisexual lizard, Gymnophthalmus underwoodi (Gymnophthalmidae), inferred from mitochondrial DNA sequence data. Mol. Phylogenet. Evol. 11, Lanza, B., In: Le Isole Galpagos. Con la Spedizione Mares G.R.S.T.S. allõ Archipelago di Colombo. Universo, Firenze, vol. 54, pp , Lanza, B., Notes on the natural history of some minor Galapagoan islands. Museo Zoologico DellÕUniversita di Firenze. Lopez, T.J., Hauselman, E.D., Maxson, L.R., Wright, J.W., Preliminary analysis of phylogenetic relationships among Galapagos Island lizards of the genus Tropidurus. Amphibia-Reptilia 13, Macey, J.R., Larson, A., Ananjeva, N.B., Fang, Z., Papenfuss, T.J., 1997a. Two novel gene orders and the role of light-strand replication in rearrangement of the vertebrate genome. Mol. Biol. Evol. 14, Macey, J.R., Larson, A., Ananjeva, N.B., Papenfuss, T.J., 1997b. Evolutionary shifts in three major structural features of the

9 D. Kizirian et al. / Molecular Phylogenetics and Evolution 32 (2004) mitochondrial genome among Iguanian lizards. J. Mol. Evol. 44, Miles, D.B., Snell, H.L., Snell, H.M., Intrapopulational variation in endurance of Galapagos lava lizards (Microlophus albemarlensis): evidence for an interaction between natural and sexual selection. Evol. Ecol. Res. 3, Posada, D., Crandall, K.A., MODELTEST: testing the model of DNA substitution. Bioinformatics 14, Snell, H.L., Jennings, R.D., Snell, H.M., Harcourt, S., Intrapopulational variation in predator-avoidance performance of Galapagos lava lizards: the interaction of sexual and natural selection. Evol. Ecol. 2, Sorenson, M.D., TreeRot (version 2). Boston University, Boston, MA. Swofford, D.L., PAUP*. Phylogenetic Analysis Using Parsimony (* and other methods). Version 4.0b10. Sinauer Associates, Sunderland, MA, USA. Talurri, M.V., Dallai, R., Lanza, B., The karyotype of some Tropidurus (Reptilia Iguanidae) from the Galapagos Islands. In: Galapagos: Studi e ricerche, Spedizone L. Mares-G.R.S.T.S. Museo Zoologico DellÕUniversita di Firenze, pp Van Denburgh, J., Slevin, J.R., Expedition of the California Academy of Sciences to the Galapagos Islands, , IX, The Galapagoan lizards of the Genus Tropidurus; with notes on the Iguanas of the Genera Conolophus and Amblyrhynchus. Proc. Calif. Acad. Sci., IV, 2, Wayne, R.K., Leonard, J.A., Cooper, A., Full of sound and fury: The recent history of ancient DNA. Annu. Rev. Ecol. Syst. 30, Wright, J.W., The evolution and biogeography of the lizards of the Galapagos Archipelago: evolutionary genetics of Phyllodactylus and Tropidurus populations. In: Bowman, R.I. et al. (Eds.), Patterns of Evolution in Galapagos Organisms. AAAS Symposium Volume, San Francisco, USA, pp Wright, J.W., The origin and evolution of lizards of the Galapagos Islands. Terra 1984, Wyrtki, K., Circulation and water masses in the eastern Equatorial Pacific Ocean. Int. J. Oceanol. Limnol. 1, Wyrtki, K., Stroup, E., Patzert, W., Williams, R., Quinn, W., Predicting and observing El Nino. Science 191 (4225), Zhang, D. X., Hewitt, G.M., Nuclear integrations: challenges for mitochondrial DNA markers. Trends Ecol. Evol. 11,