J. Zool., Lond. (1999) 249,49±60 # 1999 The Zoological Society of London Printed in the United Kingdom A phylogeny of the European lizard genus Algyroides (Reptilia: Lacertidae) based on DNA sequences, with comments on the evolution of the group D. James Harris, E. Nicholas Arnold* and Richard H. Thomas Department of Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD, U.K. (Accepted 19 November 1998) Abstract The four species of Algyroides Bibron & Bory, 1833 form part of the relatively plesiomorphic Palaearctic clade of lacertids comprising Lacerta and its allies. An estimate of phylogeny based on DNA sequence from parts of the 12S and 16S rRNA mitochondrial genes con®rms the monophyly of the genus already suggested by several morphological features. The molecular data also indicates that relationships within the clade are: (A. nigropunctatus (A. moreoticus (A. ®tzingeri, A. marchi))); this agrees with an estimate of phylogeny based on morphology that assumes the taxon ancestral to Alygroides was relatively robust in body form, and not strongly adapted to using crevices. Initial morphological evolution within Algyroides appears to involve adaptation to crypsis in woodland habitats. The most plesiomorphic form (A. nigropunctatus) is likely to have originally climbed extensively on tree boles and branches and there may have been two subsequent independent shifts to increased use of litter and vegetation matrices with related anatomical changes (A. moreoticus, A. ®tzingeri), and one to increased use of crevices (A. marchi). Some members of Algyroides are strikingly similar in super®cial morphology to particular species of the equatorial African genus Adolfus. This resemblance results from a combination of many shared primitive features plus a few independently acquired derived ones that are likely to give performance advantage in the relatively similar structural niches that these forms occupy. This study provides evidence that: (1) the use of a combination of molecular and morphological data may sometimes allow the estimation of ancestral anatomical features when these are otherwise unknown; (2) process considerations may permit a choice to be made in cases of character evolution where tree topology means that equally parsimonious alternatives exist; such decisions about character evolution may allow ecological shifts to be similarly assessed; (3) parallel evolution in ecological analogues may involve relatively few characters. Key words: lizards, Algyroides, DNA sequence, evolution INTRODUCTION Algyroides is the sister group of Lacerta bedriagae (Harris et al., 1998a). Within Algyroides, morphological Algyroides Bibron & Bory, 1833 is a small clade of four features that might give information about relationships well-differentiated species of lacertid lizards with largely are few with considerable con¯ict in their distributions, disjunct distributions in southern Europe (Fig. 1), that and because of the uncertainties about relationships may be originally associated with woodland habitats within the Lacerta and its allies, it is not possible to (Arnold, 1973, 1987). The genus is part of a large designate the most appropriate outgroups for phyloge- Palaearctic clade of mainly primitive lacertids that netic analysis of the genus with any certainty. includes most present members of the paraphyletic To resolve relationships within Algyroides and con®rm genus Lacerta and the wall lizards Podarcis (Harris, its monophyly, a phylogenetic analysis was conducted Arnold & Thomas, 1998a); it will be referred to here as using DNA sequences derived from sections of two Lacerta and its allies{. Detailed af®nities within this mitochondrial genes. These were then combined with assemblage have not been ®rmly resolved (Arnold, morphological information and employed to assess 1989a; Mayr & Benyr, 1994; Harris, 1997), although changes in anatomy and ecology and the general mtDNA sequence provides restricted evidence that evolution of the genus and to examine parallelism *All correspondence to: E. N. Arnold, Department of Zoology, The { This assemblage does not include Lacerta jayakari and Lacerta Natural History Museum, Cromwell Road, London SW7 5BD, U.K. cyanura which are most appropriately placed in a separate genus, E-mail: [email protected] Omanosaura Lutz & Mayer 1986 (Haris et al., 1998a). 50 D. J. Harris, E. N. Arnold and R. H. Thomas C B A C+D D Fig. 1. Distribution of the species of Algyroides in southern Europe showing their largely allopatric and disjunct ranges. A, A. marchi;B,A. ®tzingeri;C,A. nigropunctatus;D,A. moreoticus (compiled from BoÈhme, 1981). between members of Algyroides and those of the central frequently use crevices as refuges, such as Lacerta oxyce- African genus Adolfus, that show considerable super- phala. Four most parsimonious trees are produced using ®cial resemblance to each other. such an ancestor, and the strict consensus tree derived from these is uninformative. If, on the other hand, alter- native states are entered for these characters, the hypo- EVIDENCE FOR CLADE STATUS OF thetical ancestor resembles a more robustly built Lacerta ALGYROIDES AND INTERRELATIONSHIPS without marked crevice-using specializations, such as WITHIN THE GENUS DERIVED FROM Lacerta laevis. In this case, a single most parsimonious MORPHOLOGY tree is produced (Fig. 3). This latter version of events is more parsimonious overall, given the lack of knowledge The morphological characters mentioned in the text are about detailed relationships within the clade made up of discussed elsewhere (Arnold, 1973, 1989a, b) Lacerta and its allies, as most 0 states are more likely to A number of derived morphological features which, in be primitive in the ancestor of this assemblage. combination, de®ne Algyroides as a clade are likely to have been present in the immediate common ancestor of the genus (Table 1, characters 1±6). Characters varying EVIDENCE FOR RELATIONSHIPS FROM interspeci®cally within Algyroides are also listed in Table ALBUMIN IMMUNOLOGY 1, and their distributions shown in Table 2. To establish a hypothesis of relationships, these data were analysed Two possible phylogenies for the species of Algyroides by maximum parsimony, with exhaustive searches, using have been suggested on the basis of albumin evolution the program PAUP 3.1.1 (Swofford, 1993); the single assessed by immunological means (Mayer & Lutz, 1990). multistate character (number 7) was treated as ordered. One is that the relationships are: A. ®tzingeri (A. marchi Several of the morphological features that vary within (A. nigropunctatus, A. moreoticus)). The alternative is Algyroides also vary in the rest of the clade made up of similar except that A. marchi is excluded and appears to Lacerta and its allies, including characters 8, 9, 10, 11, 20 be most closely related to a putative clade comprising and 25. As the closest outgroups of Algyroides within this Podarcis and Lacerta graeca. The latter arrangement assemblage are not known for certain, this variation would make Algyroides polyphyletic, something that means that the immediate ancestor of Algyroides could con¯icts strongly with the morphological evidence have had a range of different character states. Which ones that the genus is a clade. are assumed affects the supposed pattern of relation- ships produced within Algyroides. At one extreme, the tree can be rooted using a hypothetical ancestor in which EVIDENCE FOR RELATIONSHIPS FROM characters 8, 9, 10 and 11 are scored as 1, and character MITOCHONDRIAL DNA SEQUENCES 20 as 0. In this case the ancestor would resemble some of the delicately built and ¯attened Lacerta species in the Given the problems of choosing appropriate outgroups Palaearctic group that live on rocky surfaces and for Algyroides when assessing the morphological Molecular phylogeny of Algyroides lizards 51 Table 1. Morphological features that de®ne Algyroides and others varying within it. Most of these features are discussed elsewhere (Arnold, 1973, 1989a) and their distribution is shown in Table 2 Features that in combination de®ne Algyroides 1. Small lips on the lobe sulci of the hemipenis (Arnold, 1973) 2. Distinctive dorsal scale micro-ornamentation in which pustullate swellings project among the upturned posterior edges of the strap-shaped cell surfaces that constitute the oberhautchen (Fig. 2). Such pustules occur elsewhere amongst lacertids only in Adolfus africanus 3. Dorsal scales strongly enlarged and markedly keeled 4. 6±8 enlarged dorsal scales in a transverse row between the hind legs 5. Lower number of presacral vertebrae in males than is usual in Lacerta and its allies. The commonest numbers are 25 in A. nigropunctatus, A. moreoticus and A. marchi, and 26 in A. ®tzingeri, overall evidence suggesting the latter ®gure is a reversal 6. Sombre dorsal colouring: usually brownish, often with a darker markings Features varying within Algyroides 7. Length of adults from snout to vent: > 60 mm (0), < 55 mm (1), < 45 mm (2) 8. Skull: robust (0), lightly built (1) 9. Supraocular osteoderms: intact in adults (0), fenestrated (1) 10. Pterygoid teeth: often present (0), absent (1) 11. Commonest number of presacral vertebrae in females: usually 26 (0), usually 27 (1) 12. On average number of long post-sternal ribs outnumber those in the more posterior series of short ribs by about three in both sexes: no (0), yes (1) 13. Shape of sternal fontanelle: oval (0), sometimes tending to heart-shape (1) 14. Number of postnasal scales: nearly always two superposed scales (0), often a single scale (1) 15. Anterior lateral edge of the parietal scale extending towards edge of the parietal table of skull: no (0), yes (1) 16. Size of lateral scales: distinctly smaller than enlarged dorsals (0), about the same size as the enlarged dorsals (1) 17. Dorsal and lateral scales lanceolate and strongly overlapping posteriorly: no (0), yes (1) 18. Keeling on dorsal scales: relatively weak (0), strong (1) 19. Micro-ornamentation on dorsal scales: raised posterior edges of epidermal cells constituting oberhautchen more or less smooth (0), denticulated (1) 20. Posterior overlap of ventral scales: small (0), more extensive (1) 21. Dark vertebral streak or series of spots: absent (0), sometimes present (1) 22.