Blackwell Science, LtdOxford, UKBIJBiological Journal of the Linnean Society0024-4066The Linnean Society of London, 2003? 2003 80? 659670 Original Article STATUS AND RELATIONSHIPS OF THE EXTINCT GIANT LIZARD GALLOTIA GOLIATH J. C. RANDO Biological Journal of the Linnean Society, 2003, 80, 659–670. With 3 figures Status and relationships of the extinct giant Canary Island lizard Gallotia goliath (Reptilia: Lacertidae), assessed using ancient mtDNA from its mummified remains N. MACA-MEYER1, S. CARRANZA2, J. C. RANDO3, E. N. ARNOLD2* and V. M. CABRERA1 1 Department of Genetics, University of La Laguna, Tenerife, Canary Islands, Spain Department of Zoology, The Natural History Museum, London, SW7 5BD, UK 3 Department of Animal Biology (Zoology), University of La Laguna, Tenerife, Canary Islands, Spain 2 Received 13 November 2002; accepted for publication 22 May 2003 Ancient mitochondrial DNA sequences (378 base pairs of cytochrome b and 368 of 12S rRNA) extracted from a mummified extinct giant lizard, Gallotia goliath, from eastern Tenerife, Canary Islands, were used to assess the species status and relationship of this form within the genus. G. goliath is clearly a member of the G. simonyi group of the western Canary islands (Tenerife, La Gomera, El Hierro and La Palma) and is not closely related to the giant G. stehlini of Gran Canaria. Contrary to recent opinion, it is phylogenetically distinct, within the G. simonyi group, from the extant G. simonyi of El Hierro and also from the recently discovered live G. gomerana on La Gomera and from G. intermedia in north-western Tenerife. It may be the sister taxon of either all the other members of the G. simonyi group or of G. intermedia. The phylogenetic distinctness of G. goliath makes Tenerife unique among oceanic islands in having had one giant and two medium-sized lizard species that were probably substantially herbivorous, the others being G. intermedia and G. galloti. Gallotia shows great community differences on other islands in the Canaries, two having a single small species, one a single giant, and three a giant and a medium-sized form. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670. ADDITIONAL KEYWORDS: 12S rRNA – ancient DNA – cytochrome b – evolution – phylogeny. INTRODUCTION Lacertid lizards of the endemic genus Gallotia Boulenger, 1916, together with the geckos Tarentola (Carranza et al. 2000, 2002; Nogales et al., 1998), and the skinks Chalcides (Brown & Pestano, 1998), are the only reptile groups to have successfully colonized the Canary Islands and radiated within them. Currently, seven living species of Gallotia are recognized, based on both morphological and molecular data (González et al., 1996; López-Jurado, Mateo & Guillaume, 1997; Rando et al., 1997; Bischoff, 1998; Hernández, Nogales & Martín, 2000; Nogales et al., 2001): G. atlantica on the eastern islands of Fuerteventura and Lanzarote, the giant G. stehlini on Gran Canaria, G. galloti on ET AL. *Corresponding author. E-mail: ena@nhm.ac.uk Tenerife and La Palma, G. caesaris on La Gomera and El Hierro, and the three giant lizards of the western Canary Islands, namely, G. simonyi on El Hierro, G. intermedia on Tenerife, and G. gomerana on La Gomera (Fig. 1). G. intermedia was only discovered in 1996 while G. gomerana, originally described from fossils, was found alive in 2000 (Hernández et al., 2000; Nogales et al., 2001). Phylogenetic analyses of the Lacertidae using morphology (Arnold, 1973, 1989) and molecules (Fu, 1998, 2000; Harris, Arnold & Thomas, 1998) shows that Gallotia is monophyletic and most probably related to Psammodromus of south-west Europe and north-west Africa. Further molecular analysis including fewer outgroups but almost all extant Gallotia species also supports this view and suggests that the ancestor of Gallotia first colonized the older eastern islands of Lanzarote, Fuerteventura or Gran Canaria, moving later to the western islands of Ten- © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 659 660 J. C. RANDO ET AL. Lanzarote (15.5 Myr) Canary Islands 29o North G. atlantica atlantica G. atlantica laurae La Palma (2 Myr) [G. goliath †] [G. simonyi †] G. galloti palmae Tenerife (12-16 Myr) La Gomera (10 Myr) 28o [G. goliath †] G. gomerana G. caesaris gomerae El Hierro (0.8-1.1 Myr) [G. goliath †] G. simonyi simonyi † G. simonyi machadoi G. caesaris caesaris Fuerteventura (23 Myr) G. atlantica mahoratae G. stehlini (introduced from Gran Canaria) Gran Canaria (14-16 Myr) [G. goliath †] [G. maxima †] G. intermedia G. galloti galloti G. galloti eisentrauti G. galloti insulanagae G. stehlini G. atlantica delibesi (G. atlantica introduced from Lanzarote) 100 km 27o 18o 17o 16o 15o 14o Figure 1. Map of the Canary Islands showing the recorded distribution of species and subspecies of Gallotia and the approximate ages of the islands. † indicates extinct forms; square brackets indicate that the taxonomic status of material concerned is uncertain. erife, La Gomera, La Palma and El Hierro, following an east–west stepping-stone model of colonization (González et al., 1996; Rando et al., 1997). Molecular analysis also shows that G. simonyi, G. intermedia and G. gomerana are members of an exclusive clade, hereafter referred to as the G. simonyi group. Apart from the living species of Gallotia, extinct forms have been described from abundant subfossil material in the western islands. Some of the individuals on which these descriptions were based were much bigger than the largest members of surviving populations but such huge animals disappeared soon after the arrival of people about 2000 years ago (Mateo & LópezJurado, 1992). Some of these giant lizards have been named G. goliath (Mertens, 1942) and G. maxima (Bravo, 1953), both taxa based on material from Tenerife. The name G. goliath has also been applied to the remains of very large lizards from La Palma (Mertens, 1942), La Gomera (G. g. bravoana (Hutterer, 1985)) and El Hierro (Izquierdo, Medina & Hernández, 1989), and G. simonyi has been used for rather smaller remains. A recent investigation of subfossil remains of large lizards from the western Canary Islands that have been assigned to G. goliath, G. maxima and G. simonyi concluded that they are all conspecific with living G. simonyi of El Hierro (Barahona et al., 2000). Some characters previously used to distinguish the extinct taxa fall within the range of intraspecific variation found among living G. simonyi, while others probably result from allometric and other ontogenetic changes associated with increased size (Barahona et al., 2000). The relatively small size of living members of the G. simonyi group may have resulted from predation pressure and degradation of the ecosystem in which they live by people and their associated domestic animals (Pregill, 1986; Nogales et al., 1988; Nogales & Medina, 1996; García-Márquez, López-Jurado & Mateo, 1997). In spite of current opinion, there are clear indications that members of the G. simonyi group do in fact show some differentiation. Apart from differences in mitochondrial DNA (mtDNA) sequences between living G. intermedia, G. gomerana and G. simonyi (Hernández et al., 2001), these forms also differ in external features, such as colouring and aspects of scaling (Hernández et al., 2000; Nogales et al., 2001; Arnold, 2002). Mummified remains of G. goliath also show distinctive features of head scaling (Castillo, Rando & Zamora, 1994) and subfossil populations of the G. simonyi group from different islands exhibit differences in tooth number, especially in large animals (Barahona et al., 2000). Given this, it would not be unexpected for other, completely extinct populations of giant lizards from the western Canary Islands to show differences in their mtDNA. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 STATUS AND RELATIONSHIPS OF THE EXTINCT GIANT LIZARD GALLOTIA GOLIATH Molecular techniques have already been used to assess the taxonomic status of some giant extinct lizards in the Canary Islands (Carranza et al., 1999). Gallotia simonyi simonyi, known only from El Roque Chico de Salmor, off El Hierro island where it has been extinct since the 1940s, was compared with G. s. machadoi, once widespread on El Hierro but now surviving only on an almost inaccessible cliff, La Fuga de Gorreta (Machado, 1985; Pérez-Mellado et al., 1997). The study showed that both subspecies were identical in the 1725 base pairs (bp) of mtDNA compared, suggesting that they were part of the same basic population until quite recently. In this study we used techniques appropriate to ancient DNA to obtain partial sequences of two mitochondrial genes from mummified remains of G. goliath from eastern Tenerife to clarify its taxonomic status and assess its relationships to other Gallotia species. 661 MATERIAL AND METHODS MATERIAL The 25 Gallotia specimens used in the phylogenetic analysis are listed in Table 1. They comprise representatives of nearly all extant species and subspecies of Gallotia including G. goliath. Lacerta dugesii and L. lepida were used as outgroups. No representatives of the genus Psammodromus (the closest outgroup of Gallotia) could be sequenced for the 5¢ upstream region of the cytocrome b (cyt b) gene (between primers GLUD-5¢-and CB1-5; Palumbi, 1996) and therefore none were included in the analysis. EXTANT MATERIAL DNA extraction followed standard proteinase K protocols described elsewhere (Carranza et al., 1999). Segments of two mtDNA genes (cyt b and 12S rRNA) Table 1. Samples used in this study Species Locality GenBank accession number 12S rRNA/cytochrome b Outgroups Lacerta dugesii Lacerta lepida Madeira Spain Z48041/Z48037 AY151979/AY151899 Canary Islands Gallotia atlantica mahoratae-1 Gallotia atlantica mahoratae-2 Gallotia atlantica laurae Gallotia atlantica atlantica Gallotia stehlini-1 Gallotia stehlini-2 Gallotia stehlini-3 Gallotia stehlini-4 Gallotia goliath Gallotia intermedia-1 Gallotia intermedia-2 Gallotia simonyi machadoi Gallotia gomerana Gallotia caesaris gomerae-1 Gallotia caesaris gomerae-2 Gallotia caesaris caesaris-1 Gallotia caesaris caesaris-2 Gallotia caesaris caesaris-3 Gallotia galloti palmae-1 Gallotia galloti palmae-2 Gallotia galloti palmae-3 Gallotia galloti eisentrauti-1 Gallotia galloti eisentrauti-2 Gallotia galloti galloti-1 Gallotia galloti galloti-2 Tindaya (Fuerteventura) Island of Lobos (Fuerteventura) Malpaís de la Corona (Lanzarote) La Santa (Lanzarote) Juncalillo (Gran Canaria) Maspalomas (Gran Canaria) Tauro (Gran Canaria) Galdar (Gran Canaria) Barranco de las Moraditas (Tenerife) Los Gigantes (Tenerife) Los Gigantes (Tenerife) Risco de Tibataje (El Hierro) Valle Gran Rey (La Gomera) Las Rosas (La Gomera) Playa de Santiago (La Gomera) El Julan (El Hierro) Los Llanillos (El Hierro) Tamaduste (El Hierro) Fuencaliente (La Palma) Las Caletas (La Palma) Puerto Espindola (La Palma) El Sauzal (NE Tenerife) San Vicente (NE Tenerife) El Palmar (Tenerife) Garachico (Tenerife) AY154905/AY154896 AF439945/AF439950 AY154906/AY154897 AY154907/AY154899 Z48039/Z48036 AF439944/AF439949 AY151917/AY151838 AY154908/AY154899 AF306568/AF306569 AY151923/AY151844 AY154913/AY154904 AY151924/AF101219 AJ272395/AJ272396 AY154910/AY154901 AY154911/AY154902 AF439943/AF439948 AY151922/AY151843 AY154912/AY154903 AF439941/AF439946 AY151920/AY151841 AY154909/AY154900 AF439942/AF439947 AY151918/AY151839 Z48038/Z48034 AY151919/AY151840 Accession numbers of new sequences obtained for this study are in italics. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 662 J. C. RANDO ET AL. were amplified via PCR using the following primers: GLUD-5¢ (Martin, Naylor & Palumbi, 1992; Palumbi, 1996) and cytochrome b2 (Kocher et al., 1989) for the cyt b gene and 12Sa and 12Sb for the 12S rRNA gene (Kocher et al., 1989). Thermocycling consisted of an initial step of 90 s at 94∞C followed by 35 cycles of 30 s at 94∞C, 45 s at 45∞C and 1 min at 72∞C. Amplified PCR bands were purified using a silica-based method (Boyle & Lew, 1995). The PCR products were sequenced using an ABI 377 automated sequencer following the manufacturer’s protocols. MUMMIFIED SPECIMEN OF GALLOTIA GOLIATH The material of G. goliath consisted of fragmented mummified remains found in Barranco de las Moraditas in Tenerife, which have already been described in detail (Castillo et al., 1994). This material was found among the otherwise inorganic fill of a small cavity in basaltic rock of the volcanic Series III dating from the Quaternary (Ancochea et al., 1990). Ancient DNA was extracted from four vertebrae and the distal portion of a partial left forelimb comprising toe bones, claws and skin. To avoid contamination with DNA from contemporary Gallotia material, extractions were performed by only one person (N. M.-M.) who was never involved in handling samples from extant taxa. In addition, this work was carried out in a laboratory dedicated exclusively to the manipulation of ancient DNA that was physically separated from the main laboratory where other DNA work was carried out. To prevent human contamination from previous handling, approximately 0.2 g of the bone/skin sample was dipped in a 10% ClH solution, washed several times in sterile distilled water and dried under a 254-nm UV lamp. The material was then crushed in a sterilized mortar with liquid N2 and transferred to 1 mL guanidinium thiocyanate (GuSCN) extraction solution following the GuSCn-silica method (Boom et al., 1990). To eliminate any remaining contamination with alien DNA, all aliquoted solutions were first treated with UV light (Sarkar & Sommer, 1990) and then with 8methoxypsoralen (Jinno, Yoshiura & Niikawa, 1990). Three independent extraction controls, in which the sample was substituted by the same weight of sterilized filter paper, were processed in parallel with each extraction. The ancient DNA and controls were released from the silica pellet in which they were enclosed by dissolving these in 50 mL TE buffer (10 mM TrisCl, 1 mM EDTA, pH = 8). Except when being processed, samples were kept at -70∞C. The ancient DNA was PCR amplified following the same thermocycling conditions as for the extant specimens (see above). As might be expected, the DNA extracted from the mummified specimen was very degraded and so five overlapping subfragments for the 12S rRNA and four for the cyt b gene had to be amplified and sequenced using the set of primers listed in Table 2. To avoid false positives during PCR, strict conditions proposed by Kwok & Higuchi (1989) were followed whenever possible, and all the aliquoted PCR mixes were treated with UV light (Sarkar & Sommer, 1990) before the addition of 5 mL of ancient DNA template. Three negative PCR controls were included with each set of Table 2. Primers used in this study Primer Gene Position* Sequence (5¢-3¢) L14724 H14929 L14841 L14921 H15035 L15031 H15149 L1064 H1179 L1149 H1256 L1235 H1357 L1343 H1467 L1342 H1565 Cyt b Cyt b Cyt b Cyt b Cyt b Cyt b Cyt b 12S 12S 12S 12S 12S 12S 12S 12S 12S 12S 14704–14724 14954–14929 14815–14841 14896–14921 15060–15035 15006–15031 15174–15149 1043–1064 1197–1179 1128–1149 1275–1256 1217–1235 1378–1357 1323–1343 1485–1467 1412–1432 1584–1565 TGACTTGAAGAACCACCGTTG GGATGTGGGCGATGGATGAGAATGCG AAAAAGCTTCCATCCAACATCTCAGCATGATGAAA CCTAGCCATGCACTACTCACCAGAC CCGTAGTACAGGCCACGTCCGATGTG CCTCACTATTTTTTATCTGCATCTAC AAACTGCAGCCCCTCAGAATGATATTTGTCCTCA TTGACCACACGAAAGCTTAGAA TAGGTCGAATGTGGGACACC CTTTCCGCCAGAGAACTACAAG CGACGGCGGTATATAGGCTG CGATACTCCCCGCTCTACC AAAAATGTAGCCAATCTCTGCC CTAACACGTCAGGTCAAGGTG GTGTGTACGCGTCCCAGAG CAGCATGAAGGCGAATTTAGT TTCCGGTACGCTTACCATGT *Positions refer to the complete human mitochondrial genome sequence (Anderson et al., 1981). © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 STATUS AND RELATIONSHIPS OF THE EXTINCT GIANT LIZARD GALLOTIA GOLIATH amplifications. The sequencing protocol for the PCR products was the same as for the extant samples. PHYLOGENETIC ANALYSIS DNA sequences were visualized using an alignment editor (GDE, Smith et al., 1994). No gaps had to be postulated to align the cyt b sequences, which were translated into amino acids using the vertebrate mitochondrial code. No stop codons were observed in the cyt b sequences suggesting that they are probably functional. The 12S rRNA sequences were aligned with reference to the published secondary structure of this gene (Hickson et al., 1996) and gaps were inserted to resolve length differences between sequences. Two hypervariable regions of approximately 14 bp (between stems 36 and 38 of Hickson et al. (1996) and 8 bp (between stems 42 and 42¢ of Hickson et al. (1996)) could not be aligned unambiguously and they were therefore excluded from further analyses. A saturation analysis was carried out in which the observed proportions of transitions (ts) and transversions (tv) were plotted against the uncorrected genetic distances, the 12S rRNA and the highly variable 3rd codon positions of the cyt b gene being analysed independently. The results showed no evidence of saturation in the 12S rRNA ts and tv, or in the cyt b 3rd codon tv, but the cyt b 3rd codon ts show a small degree of saturation for all comparisons between Gallotia species and the outgroups (results not shown). Because of this, cyt b 3rd codon ts were given a weight of 0 in one of the parsimony analyses. The two gene fragments were tested for incongruence using the incongruence length difference test (ILD) (Mickevich & Farris, 1981; Farris et al., 1994); 10 000 heuristic replicates were used, and the invariable characters were removed before starting the analysis (Cunningham, 1997). The result of the ILD test (ILD P > 0.172) indicated both genes to be congruent and therefore they could be combined in a total evidence analysis. Three different methods of phylogenetic analysis were employed and the results compared. These were maximum likelihood (ML), maximum parsimony (MP) and neighbour-joining (NJ). Modeltest version 3.06 (Posada & Crandall, 1998) was used to select the most appropriate model of sequence evolution for the NJ and ML analyses. This was the general time reversible model (GTR) taking into account the number of invariable sites (I) and the shape of the gamma distribution (G) (i.e. GTR + I + G). A simpler analysis was also carried out using the Kimura’s two-parameter (K2P) model (Kimura, 1980), which takes into account the differences between the number of ts and tv (ts/tv ratio) but assumes equal frequencies for the four types of nucleotide. This additional analysis was undertaken because, although it has been demonstrated that ML 663 performs better with the most correct model of sequence evolution rather than with a simpler, more incorrect model (Yang, 1996; Rosenberg & Kumar, 2001), it has also been found that with a complex model ML does not work well when the true evolutionary pattern is simple (Yang, 1996). The MP and ML analyses were heuristic searches involving tree bisection and reconnection (TBR) branch swapping with 1000 and 10 random stepwise additions of taxa, respectively. In the MP analyses gaps were considered as a fifth state. The weight given to tv was varied relative to that of ts (tv was allocated the same, two times, four times, six times and ten times the weight of ts). The weight of the gaps was always equal to the maximum weight assigned to either ts or tv. The reason for using these weights was because all fell around the estimated ts/tv ratio for our dataset (ts/tv = 5.6). As noted, the cyt b 3rd codon ts were given a weight of 0 in some analyses. In total, 11 different kinds of phylogenetic analyses were used: (1) MP (ts = tv); (2) MP (ts = 1, tv = 2); (3) MP (ts = 1, tv = 4); (4) MP (ts = 1, tv = 6); (5) MP (ts = 1, tv = 10); (6) MP (cyt b 3rd codon ts = 0); (7) NJ (GTR + I + G); (8) NJ (K2P); (9) ML (GTR + I + G); (10) ML (K2P); (11) ML (enforcing molecular clock) (GTR + I + G). All analyses were performed in PAUP* version 4.0b10 (Swofford, 1998) except where stated. Robustness of the trees was assessed by bootstrap analysis (Felsenstein, 1985) and involved 500 pseudoreplications for the ML (GTR + I + G) tree and 1000 for all the other analyses. Where appropriate, topological constraints were generated using MacClade version 4.0 (Maddison & Maddison, 1992) and they were compared with our optimal topologies using the Kishino-Hasegawa (1989) and Shimodaira-Hasegawa (1999) tests employing RELL bootstrap with 1000 bootstrap replicates. In order to assess the age of speciation events, molecular clock assumptions were incorporated. The likelihood ratio test (Huelsenbeck & Crandall, 1997) was used for testing the statistical significance of the difference between the log likelihood of the trees calculated with and without the clock assumptions. RESULTS PHYLOGENETIC RELATIONSHIPS OF GALLOTIA GOLIATH AND ITS INDEPENDENT STATUS The sequences obtained from the mummified remains of G. goliath differed from all other species and subspecies of Gallotia for which these gene fragments have been studied, increasing our confidence that the sequences belonged to G. goliath itself. After excluding all the regions that could not be unambiguously aligned, the dataset contained 746 characters (378 bp © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 664 J. C. RANDO ET AL. of cyt b and 368 bp of 12S rRNA) of which 221 were variable and 180 parsimony-informative. The strict consensus tree of all 26 trees obtained from the 11 different analyses employed is shown in Figure 2. Additional information about the ML and MP analyses is given in Table 3. The consensus tree clearly showed that G. goliath belongs to the G. simonyi group, together with G. simonyi, G. gomerana and G. intermedia. Within this grouping, G. s. machadoi and G. gomerana formed a monophyletic group with a bootstrap support of 100 in all the independent analyses with the exception of MP (cyt b 3rd ts = 0), which had a bootstrap value of only 69. This underlines the role of the highly variable 3rd codon ts of the cyt b gene in resolving recent cladogenetic events. Other relationships within the G. simonyi group were not consistently resolved. In the NJ trees and the two ML trees, in which the molecular clock was not enforced, G. goliath was sister to all the other lizards of the G. simonyi group but with low bootstrap support. All MP analyses except that in which the cyt b 3rd codon ts were given a null value (MP (cyt b 3rd ts = 0)), produced an alternative topology of exactly the same number of steps, in which G. goliath was sister to G. intermedia (Fig. 2B). The ML tree enforcing the molecular clock and the tree from the MP (cyt b 3rd ts = 0) analysis also supported the hypothesis that G. goliath is sister to G. intermedia. To investi- gate these relationships further, the ML (GTR + I + G) topology in which the molecular clock was not enforced (G. goliath sister to all other lizards of the G. simonyi group) was compared with a tree constrained so that G. goliath was sister to G. intermedia. The results are presented in Table 4; the difference between the two trees was not significant. Moreover, both trees presented the same number of steps (379). Genetic distances between all the members of the G. simonyi group indicated G. goliath is genetically well differentiated from all the rest, the average divergence ranging between 2.71% and 4.20% (Table 5). This range is higher than the genetic distances between G. gomerana and G. simonyi (0.94%) and at the same general level as the genetic distances between these and G. intermedia (3.46%– 3.89%). TIMING OF EVENTS IN THE HISTORY OF THE GALLOTIA SIMONYI GROUP Inferring the ages of molecular divergence events and island colonization often requires the use of a molecular clock calibration. Several sources of error may affect this calibration and should be taken into account when drawing any conclusions from the results. Factors that can affect clock calibrations on islands include increased stochastic variation at low Table 3. Additional data for the maximum parsimony (MP) and maximum likelihood (ML) trees Type of analysis Trees (N) Steps (N) CI RI –log likelihood MP MP MP MP MP MP ML ML ML 4 4 4 4 4 2 1 1 1 379 477 673 869 1261 220 – – – 0.628 0.652 0.679 0.694 0.710 0.668 – – – 0.844 0.850 0.857 0.860 0.865 0.849 – – – – – – – – – 2965.89073 3207.86958 2984.43213 (ts = tv) (ts = 1; tv = 2) (ts = 1; tv = 4) (ts = 1; tv = 6) (ts = 1; tv = 10) (cyt b 3rd ts = 0) (GTR + I + G) (K2P) (GTR + I + G) enforcing molecular clock MP values are after excluding uninformative sites. CI, consistency index; GTR + I + G, general time reversible model taking into account the number of invariable sites and the shape of the gamma distribution; K2P, Kimura’s two-parameter model (Kimura, 1980); RI, retention index; ts, transitions; tv, transversions. 䉴 Figure 2. (A) Strict consensus of the 26 trees obtained from the combined analysis of the cyt b and 12S rRNA genes using 11 different approaches (see Material and Methods and Table 3 for details). Where bootstrap percentages vary, numbers by nodes indicate support for the individual methods and parameters used, given in the order indicated in the box below the tree. When differences between bootstrap figures are <10% only the average value is shown. (B) Support for alternative phylogenetic hypotheses at unresolved nodes in the strict consensus tree (A) for each of the 11 methods used. Dark boxes indicate all trees of a particular analysis support a hypothesis; grey boxes indicate support by some of the equally parsimonious trees but not all. Figures indicate bootstrap values and a question mark that no bootstrap assessment was made. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 STATUS AND RELATIONSHIPS OF THE EXTINCT GIANT LIZARD GALLOTIA GOLIATH simonyi © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 665 666 J. C. RANDO ET AL. Table 4. Statistical support for alternative hypotheses on Gallotia phylogeny Tree –log likelihood D-log likelihood Unconstrained ML (GTR + I + G) tree (Fig. 2). G. goliath sister to all other lizards of the G. simonyi group. Constrained so G. goliath is sister to G. intermedia. 2965.89073 (best) 2968.67547 2.78474 SH P KH P 0.208 0.4364 P > 0.05 suggests that the constrained and unconstrained solutions are not significantly different. GTR + I + G, general time reversible model taking into account the number of invariable sites and the shape of the gamma distribution; KH, Kishino-Hasegawa (1989) test; SH, Shimodaira-Hasegawa (1999) test; ML, maximum likelihood. Table 5. Genetic divergence of cyt b and 12S rRNA genes between all members of the Gallotia simonyi group calculated using the Kimura 2-parameter correction Comparison Maximum genetic variability (%) Minimum genetic variability (%) Average (%) G. goliath vs. G. intermedia G. goliath vs. G. simonyi G. goliath vs. G. gomerana G. intermedia vs. G. simonyi G. intermedia vs. G. gomerana G. gomerana vs. G. simonyi 2.75 3.70 4.20 3.46 3.89 0.94 2.47 3.70 4.20 3.17 3.60 0.94 2.61 3.70 4.20 3.31 3.75 0.94 levels of sequence divergence, possible extinct or unsampled lineages, and the assumption that islands are colonized immediately after their appearance (Emerson, Oromi & Hewitt, 2000a,b; Emerson, 2002 and references therein). The –log likelihood value of the ML tree (GTR + I + G) (2965.89073) was compared with that of the same tree constructed under molecular clock assumptions (2984.43213). The results showed that no significant difference between the likelihoods of the two trees (likelihood ratio test statistic (-2logD) = 37.0828, which approximates to X225 distribution under the null hypothesis; P > 0.05). So the sequences could be used for estimating dates. The clock was calibrated using the ML (GTR + I + G) genetic distances and two approximate ages for the island of El Hierro: 0.8 Myr (Abdel-Monem, Watkins & Gast, 1972; Fuster et al., 1993) and 1.1 Myr (Guillou et al., 1996)). El Hierro is quite close to the much older island of La Gomera and presumably received its lizard colonists from there when they rafted on the prevailing ocean currents which run in an appropriate direction. This would apply both to G. caesaris caesaris and G. simonyi, the colonization events being indicated on the phylogeny by the nodes marking the separation of these taxa from their sister lineages on La Gomera, G. caesaris gomerae and G. gomerana, respectively. The degree of genetic differentiation between G. c. caesaris and G. c. gomerae was greater than that between G. gomerana and G. simonyi, sug- gesting that G. c. caesaris colonized El Hierro first. The maximum age of this event would be the age of the island itself. Using this as a basis for calibration, the evolutionary rate of Gallotia for the combination of gene fragments used in this investigation would range between 2% and 1.48% per Myr, depending on which estimate of the age of El Hierro was used (0.8 Myr or 1.1 Myr, respectively). Dates derived from this calibration are shown in Figure 3. They suggest that the G. simonyi group separated from the G. galloti– G. caesaris clade about 5–7 Mya and that its currently recognized lineages began to diverge 2.1–2.8 Mya. Whether G. goliath is sister to the rest of the G. simonyi group or to G. intermedia, the age of its exclusive lineage is much greater than those of G. simonyi and G. gomerana. DISCUSSION Gallotia goliath is clearly a member of the G. simonyi group of very large lacertid lizards in the western Canary islands (Tenerife, La Gomera, El Hierro and La Palma) and is not closely related to the quite similar giant G. stehlini of Gran Canaria. Gallotia goliath is one of several distinct forms in the G. simonyi group and is separable on the basis of morphology and mtDNA sequence from the extant G. simonyi of El Hierro, from the recently discovered living G. gomerana on La Gomera, and from G. intermedia in north-western Tenerife. Within the G. simonyi group, © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 STATUS AND RELATIONSHIPS OF THE EXTINCT GIANT LIZARD GALLOTIA GOLIATH 667 Figure 3. Maximum likelihood (GTR + I + G) tree enforcing the molecular clock. Numbers at nodes indicate millions of years since the cladogenetic event represented. Upper numbers assume an age for El Hierro of 0.8 Myr, lower numbers one of 1.1 Myr. The calibration point is highlighted with a grey box. G. goliath may be sister either to all the other members, or to G. intermedia. The distinctness of G. goliath makes Tenerife unique among oceanic islands in having had one giant and two medium-sized lizard species that were probably substantially herbivorous, the others being G. intermedia and G. galloti. Gallotia shows great community differences on other islands in the Canaries, two having a single small species, one a single giant, and three a giant and a medium-sized form. This irregularity is similar to that found in Phelsuma geckos on the Mascarene islands in the south-west Indian Ocean (Arnold, 2000) but contrasts with the uniform patterns of radiation seen in some Anolis (Iguanidae) clades in the Greater Antilles (Losos et al., 1998). ACKNOWLEDGEMENTS We are very grateful to the ‘Viceconsejería de Medio Ambiente del Gobierno de Canarias’ for their help. This work was partially supported by grants form the Natural Environment Research Council (GR9/03327) and the Royal Society (European Science Exchange © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 668 J. C. RANDO ET AL. Programme, 1999–2001). S. Carranza was supported by a Marie Curie Individual Fellowship under the European Union Human Potential Programme (MCFI-1999-00523). REFERENCES Abdel-Monem A, Watkins NO, Gast PW. 1972. Potassiumargon ages, volcanic stratigraphy and geomagnetic polarity history of the Canary Islands: Tenerife, La Palma and Hierro. American Journal of Science 272: 803–825. Ancochea E, Fuster JM, Ibarrola E, Cendrero A, Coello J, Hernan F, Cantagrel JM, Jamond C. 1990. Volcanic evolution of the island of Tenerife (Canary Islands) in the light of new K-Ar data. Journal of Volcanology and Geothermal Research 44: 231–249. Anderson S, Bankier AT, Barrell BG, de Bruijn MHL, Coulson AR, Drouin J, Eperon IC, Nierlich DP, Roe BA, Sanger F, Schreier PH, Smith AJH, Staden R, Young IG. 1981. Sequence and organisation of the human mitochondrial genome. Nature 290: 457–465. Arnold EN. 1973. Relationships of the Palearctic lizards assigned to the genera Lacerta, Algyroides and Psammodromus (Reptilia: Lacertidae). Bulletin of the British Museum (Natural History), Zoology 25: 289–366. Arnold EN. 1989. Towards a phylogeny and biogeography of the Lacertidae: relationships within an Old-World family of lizards derived from morphology. Bulletin of the British Museum (Natural History), Zoology 55: 209–257. Arnold EN. 2000. Using fossils and phylogenies to understand evolution of reptile communities on islands. Bönner Zoologische Monographier 46: 309–323. Barahona F, Evans SE, Mateo JA, García-Márquez M, López-Jurado LF. 2000. Endemism, gigantism and extinction in island lizards: the genus Gallotia on the Canary Islands. Journal of Zoology 250: 373–388. Bischoff W. 1998. Bemerkungen zu den ‘fossilen’ Rieseidenchsen der Kanarischen Inslen. In: Handbuch der Reptilien und Amphibien Europas. Weisbaden: AULA-Verlag, 387–407. Boom R, Sol CJA, Salimans MMM, Jansen CL, Wertheimvan Dillen PME, Van der Noordaa J. 1990. Rapid and simple method for purification for nucleic acids. Journal of Clinical Microbiology 28: 495–503. Boyle JS, Lew AM. 1995. An inexpensive alternative to glassmilk for DNA purification. Trends in Genetics 11: 8. Bravo T. 1953. Lacerta maxima n. sp. de la fauna continental extinguida en el Pleistoceno de las Islas Canarias. Estudios Geológicos Instito de Investigaciones geológicas Lucas Mallado 9: 7–34. Brown RP, Pestano J. 1998. Phylogeography of skinks (Chalcides) in the Canary Islands inferred from mitochondrial DNA sequences. Molecular Ecology 7: 1183–1191. Carranza S, Arnold EN, Mateo JA, Geniez P. 2002. Relationships and evolution of the North African geckos, Geckonia and Tarentola (Reptilia: Gekkonidae), based on mitochondrial and nuclear DNA sequences. Molecular Phylogenetics and Evolution 23: 244–256. Carranza S, Arnold EN, Mateo JA, López-Jurado LF. 2000. Long-distance colonization and radiation in gekkonid lizards, Tarentola (Reptilia: Gekkonidae), revealed by mitochondrial DNA sequences. Proceedings of the Royal Society of London B 267: 637–649. Carranza S, Arnold EN, Thomas RH, Mateo JA, LópezJurado LF. 1999. Status of the extinct giant lacertid lizard Gallotia simonyi simonyi (Reptilia: Lacertidae) assessed using mtDNA sequences from museum specimens. Herpetological Journal 9: 83–86. Castillo C, Rando JC, Zamora JF. 1994. Discovery of mummified extinct giant lizards (Gallotia goliath, Lacertidae) in Tenerife, Canary Islands. Bönner Zoologische Beiträge 45: 129–136. Cunningham CW. 1997. Is congruence between data partitions a reliable predictor of phylogenetic accuracy? Empirically testing an iterative procedure for choosing among phylogenetic methods. Systematic Biology 46: 464–478. Emerson BC. 2002. Evolution on oceanic islands: molecular phylogenetic approaches to understanding pattern and process. Molecular Ecology 11: 951–966. Emerson BC, Oromi P, Hewitt G. 2000a. Colonization and diversification of the species Brachyderes rugatus (Coleoptera) on the Canary Islands: evidence from mitochondrial DNA COII gene sequences. Evolution 54: 911–923. Emerson BC, Oromi P, Hewitt G. 2000b. Tracking colonization and diversification on insect lineages on islands: mitochondrial DNA phylogeography of Tarphius canariensis (Coleoptera: Colydiidae) on the Canary Islands. Proceedings of the Royal Society of London Series B 267: 2199–2205. Farris JS, Kallersjo M, Kluge AG, Bult C. 1994. Testing significance of incongruence. Cladistics 10: 315–319. Felsenstein J. 1985. Confidence-limits on phylogenies – an approach using the bootstrap. Evolution 39: 783–791. Fu J. 1998. Toward the phylogeny of the family Lacertidae: implications from mitochondrial DNA 12S and 16S gene sequences (Reptilia: Squamata). Molecular Phylogenetics and Evolution 9: 118–130. Fu J. 2000. Towards the phylogeny of the family Lacertidae – why 4708 base pairs of mtDNA sequences cannot draw the picture. Biological Journal of the Linnean Society 71: 203– 217. Fuster JM, Hernan F, Cendrero A, Coello J, Cantagrel JM, Ancochea E, Ibarrola E. 1993. Geocronología de la isla de El Hierro (Islas Canarias). Bolétin de la Real Sociedad Española de Historia Natural, Sección Geológica. 88: 85–97. García-Márquez M, López-Jurado LF, Mateo JA. 1997. Predación de Gallotia simonyi por gatos cimarrones. Boletín de la Asociación Herpetológica Española 8: 20–23. González P, Pinto F, Nogales M, Jiménez AJ, Hernández M, Cabrera VM. 1996. Phylogenetic relationships of the Canary Islands endemic lizard genus Gallotia (Sauria: Lacertidae), inferred from mitochondrial DNA sequences. Molecular Phylogenetics and Evolution 6: 63–71. Guillou H, Carracedo JC, Torrado FP, Badiola ER. 1996. K-Ar ages and magnetic stratigraphy of a hotspot- © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 STATUS AND RELATIONSHIPS OF THE EXTINCT GIANT LIZARD GALLOTIA GOLIATH induced, fast grown oceanic island: El Hierro, Canary Islands. Journal of Volcanology and Geothermal Research 73: 141–155. Harris DJ, Arnold EN, Thomas RH. 1998. Relationships of lacertid lizards (Reptilia: Lacertidae) estimated from mitochondrial DNA sequences and morphology. Proceedings of the Royal Society of London B 265: 1939–1948. Hernández M, Maca-Meyer N, Rando JC, Valido A, Nogales M. 2001. Addition of a new living giant lizard from La Gomera Island to the phylogeny of the endemic genus Gallotia (Canarian archipelago). Herpetological Journal 11: 171–173. Hernández E, Nogales M, Martín A. 2000. Discovery of a new lizard in the Canary Islands, with a multivariate analysis of Gallotia (Reptilia: Lacertidae). Herpetologica 56: 63– 76. Hickson RE, Simon C, Cooper A, Spicer GS, Sullivan J, Penny D. 1996. Conserved sequence motifs, alignment and secondary structure for the third domain of animal 12S rRNA. Molecular Biology and Evolution 13: 150–169. Huelsenbeck JP, Crandall KA. 1997. Phylogeny estimation and hypothesis testing using maximum likelihood. Annual Review of Ecology and Systematics 28: 437–466. Hutterer R. 1985. Neue Funde von Rieseneidechsen (Lacertidae) auf der Insel Gomera. Bönner Zoologische Beiträge 36: 365–394. Izquierdo I, Medina AL, Hernández JJ. 1989. Bones of giant lacertids from a new site on El Hierro (Canary Islands). Amphibia-Reptilia 10: 63–69. Jinno Y, Yoshiura K, Niikawa N. 1990. Use of psoralen as extinguisher of base substitiuons through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: 111–120. Kimura M. 1980. A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: 111–120. Kishino H, Hasegawa M. 1989. Evaluation of the maximum likelihood estimate of the evolutionary tree topologies from DNA sequence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29: 170–179. Kocher TD, Thomas WK, Meyer A, Edwards SV, Paabo S, Villablanca FX, Wilson AC. 1989. Dynamics of mitochondrial DNA evolution in animals: amplification and sequencing with conserved primers. Proceedings of the National Academy of Sciences, USA 86: 6196–6200. Kwok S, Higuchi R. 1989. Avoiding false positives with PCR. Nature 339: 237–238. López-Jurado LF, Mateo JA, Guillaume CP. 1997. El complejo Gallotia galloti (Oudart, 1839) (Sauria: Lacertidae) de las Islas Canarias: nuevos datos para la interpretación del proceso evolutivo del grupo. Revista Española de Herpetologia 11: 35–46. Losos JB, Jackman TR, Larson A, De Queiroz K, Rodriguez Schettino L. 1998. Contingency and determinism in replicated adaptive radiations of island lizards. Science 279: 2115–2118. 669 Machado A. 1985. Hypothesis on the reasons for the decline of the large lizards in the Canary Islands. Bönner Zoologische Beiträge 36: 563–575. Maddison WP, Maddison DR. 1992. Macclade, Version 3.06. Sunderland, MA: Sinauer Associates. Martin AP, Naylor GJP, Palumbi RS. 1992. Rates of mitochondrial DNA evolution in sharks are slow compared to mammals. Nature 357: 153–155. Mateo JA, López-Jurado LF. 1992. Study of dentition in lizards from Gran Canaria Island (Canary Islands) and its ecological and evolutionary significance. Biological Journal of the Linnean Society 46: 39–48. Mertens R. 1942. Lacerta goliath n. sp., eine ausgestorbene Rieseneidechse von den Kanaren. Senckenbergiana 25: 330– 339. Mickevich MF, Farris JS. 1981. The implications of congruence in Menidia. Systematic Zoology 30: 351–370. Nogales M, López M, Jiménez AJ, Larruga JM, Hernández M, González P. 1998. Evolution and biogeography of the genus Tarentola (Sauria: Gekkonidae) in the Canary Islands, inferred from mitochondrial DNA sequences. Journal of Evolutionary Biology 11: 481–494. Nogales M, Martín A, Delgado G, Emmerson K. 1988. Food spectrum of the feral cat (Felis catus L., 1758) in the juniper woodland on El Hierro (Canary Islands). Bönner Zoologische Beiträge 39: 1–6. Nogales M, Medina FA. 1996. A review of the diet of feral domestic cats (Felis silvestris, F. catus) on the Canary Islands, with new data from the laurel forest of La Gomera. International Journal of Mammalian Biology 61: 1–6. Nogales M, Rando JC, Valido A, Martín A. 2001. Discovery of a living giant lizard, genus Gallotia (Reptilia: Lacertidae), from La Gomera, Canary Islands. Herpetologica 57: 169–179. Palumbi SR. 1996. The polymerase chain reaction. In: Hillis DM, Moritz C, Mable BK, eds. Molecular systematics 2nd edn. Sunderland, MA: Siinauer Associates 205–247. Pérez-Mellado V, Arano B, Astudillo G, Cejudo D, GarcíaMárquez M, Llorente G, Márquez R, Mateo JA, Orrit N, Romero-Bevia M, López-Jurado LF. 1997. Recovery plan for the giant lizard of El Hierro (Canary islands), Gallotia simonyi: project outline and preliminary results. In: Bohme W, Bischoff W, Ziegler T, eds. Herpetologica Bonnensis. Bonn, Germany: Societas Europaea Herpetologica. Posada D, Crandall KA. 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817–818. Pregill G. 1986. Body size of insular lizards – a pattern of Holocene dwarfism. Evolution 40: 997–1008. Rando JC, Hernández E, López M, González AM. 1997. Phylogenetic relationships of the Canary Islands endemic lizard genus Gallotia inferred from mitochondrial DNA sequences: Incorporation of a new subspecies. Molecular Phylogenetics and Evolution 8: 114–116. Rosenberg MS, Kumar S. 2001. Traditional phylogenetic reconstruction methods reconstruct shallow and deep evolutionary relationships equally well. Molecular Biology and Evolution 18: 1823–1827. Sarkar G, Sommer SS. 1990. Shedding light on PCR contamination. Nature 343: 27. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670 670 J. C. RANDO ET AL. Shimodaira H, Hasegawa M. 1999. Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Molecular Biology and Evolution 16: 114–116. Smith SW, Overbeek R, Woese CR, Gilbert W, Gillevet PM. 1994. The Genetic Data Environment: an expandable GUI for multiple sequence-analysis. Computational and Applied Biosciences 10: 671–675. Swofford DL. 1998. PAUP*: phylogenetic analysis using parsimony (and other methods). Version 4.0. Sunderland, MA: Sinauer Associates. Yang ZH. 1996. Phylogenetic analysis using parsimony and likelihood methods. Journal of Molecular Evolution 42: 294– 307. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80, 659–670