A MORPHOLOGY-BASED PHYLOGENY OF PHYMATURUS (IGUANIA: LIOLAEMIDAE) WITH THE DESCRIPTION OF FOUR NEW SPECIES FROM ARGENTINA

Phylogenetic relationships within the liolaemid lizard genus Phymaturus were studied using parsimony analysis of morphological data. The data set includes 133 characters: 28 described in the literature as apomorphies of the three genera of Liolaemidae (Ctenoblepharys, Liolaemus, and Phymaturus), 21 published characters of allozymes and karyology, 53 characters taken from external morphology across all terminals of Phymaturus, and 31 from the skeletal anatomy. This data set includes representatives of 10 of the 12 species currently recognized in the literature plus twelve other terminals considered in this study and representing independent lineages assigned to patagonicus or palluma. Four of these terminals are described in the present study as new species, one belonging to the palluma group and the other three to the patagonicus group. We performed four analyses using different methods of coding binary polymorphic characters, and a new method for treating continuous characters. The traditional division of the genus in two groups is not supported here, with the patagonicus group resulting paraphyletic in some of the analyses. The palluma group is monophyletic and supported by many characters. A majority rule consensus tree of all runs recovers a reasonably well-resolved topology of the group. All analyses recovered a northern subclade within the palluma group, formed by species distributed in Argentina from northern of San Juan province (north to 30 degrees of latitude). In this analysis palluma from El Planchón (Chile) was found to be more closely related to this northern subclade than any other “palluma” form.


INTRODUCTION
Phymaturus comprises a group of iguanian lizards inhabiting rocky places of Patagonia and both eastern and western mid-to high-elevation slopes of the Andes.This group of viviparous and herbivorous lizards reaches its northern limit in the Puna region of among other characters that are exclusive and provide strong evidence of monophyly.
At the time of Donoso Barros (1964) and Peters and Donoso Barros (1970), the genus included only one species with two subspecies: palluma palluma (Molina) 1782, and palluma patagonicus (Koslowsky) 1898).Most taxa recognized as species today were described during the 1970s-80s: P. patagonicus payunae, P. p. somuncurensis, P. p. zapalensis, P. p. indistinctus, and P. p. nevadoi by Cei & Castro (1973); P. mallimaccii by Cei (1980); P. punae by Cei et al., (1983); and P. antofagastensis by Pereyra (1985).Despite publication of several systematic studies on the genus over the last 30 years (Cei & Lescure, 1985; Lescure & Cei, 1991;Pereyra, 1992a; among others), the phylogenetic relationships among Phymaturus species remain enigmatic.Pereyra (1992a) studied the external morphology, karyology, and allozymes of six species/populations of Phymaturus, but his study applied phenetic techniques (cluster analysis), which established phenetic distances among taxa but did not provide a genealogical reconstruction.Much of Pereyra's (1992a) information was included and reevaluated in the present study.Etheridge (1995) provided the first extensive study of genera now included in the Liolaemidae (Phymaturus, Liolaemus, and Ctenoblepharys), and proposed a cladistic taxonomy based on morphological characters.Within Phymaturus Etheridge (1995) recognized two groups of species, the palluma group with four species, and the patagonicus group formed by six species (formerly considered subspecies of patagonicus by Cei [1986]).Characters defining the palluma group were the non-imbricate superciliaries, five or more suboculars, three to four rows of lorilabials, mental narrower than rostral and usually in contact with sublabials, caudal spines well developed, and two annuli per segment.For the patagonicus group, Etheridge (1995) identified a splenial short and a fused Meckel's groove.However, Etheridge (1995) noted that these groups may not necessarily be monophyletic.Following this work, no additional studies were published that sought to recover the genealogical relationships within the genus, yet, many questions remain, like the validity and or monophyly of Etheridge's (1995) groups of species, and the phylogenetic relationships among species.Recently, Cei & Videla (2003) and Scolaro & Cei (2003) described two new species for the genus and discussed the unresolved problem of the identity of the true palluma, a lizard collected by Charles Darwin on his voyage on the Beagle and posteriorly described by Bell (1843).The latter issue remains a problem that needs to be resolved through the examination of populations on both sides of the Andes, and compared to the type material deposited at the British Museum of Natural History.
The goal of this study is to provide a first comprehensive analysis of Phymaturus, using cladistic methods.We studied the external morphology and skeletons of an extensive sample of terminals and individuals.We examined and described more than 100 informative characters, most of them used in the systematics of Phymaturus for the first time, as well as other characters taken from the literature.

MATERIALS AND METHODS
For the phylogenetic analysis we were able to reexamine the type series of nine of the 12 currently recognized species (with the exception of palluma, verdugo, and calcogaster).Characters used in this study were obtained primarily from the skeleton, squamation, morphometry, and body patterns.The allozyme data set plus two chromosomal characters of Pereyra (1992a) were also included.Skeletal characters were visualized from cleared and stained material following Wassersug's (1976) technique, which allows differential staining of cartilage and bone.For some specimens (MVZ and SDSU dry skeletons) however, cartilaginous structures (larynx, trachea, and hyoid apparatus) were not available because they were prepared as dry skeletons.External characters were scored from formalinpreserved specimens stored in alcohol 70%.Characters for phylogenetic analysis were scored only from adults.Ontogenetic shift in external morphology between juveniles and adults was recorded and it is presented in a separated section.Scale terminology follows a previous study on the Liolaemus chiliensis group (Lobo, 2001), measurements were taken with digital calipers (± 0.01 mm) and following the definitions of Laurent (1984).Skeleton terminology follows previous studies (Lobo & Abdala, 2001, 2002).Museum numbers and localities for specimens included in this study are listed in Appendix I. We selected characters based on previous cladistic studies in Liolaemus (Lobo, 2001;Lobo & Abdala, 2001, 2002), or those described for Phymaturus (Cei, 1980(Cei, , 1986(Cei, , 1993;;Cei & Castro, 1973); Cei et al., 1983;Etheridge, 1995), and we reevaluated many of characters considered informative for the genus by Pereyra (1992a) (among meristic, allozyme, and karyological characters he studied).For a details and a list of characters see Appendix II.He found characters discriminating payunae and cf.palluma (CH) from the rest of species (morphometric characters in page 71, Table 18), head length, number of scales on the dorsum of the head (equivalent to his LCI), and head width; from his meristic characters that he found useful for discriminating payunae and mallimaccii from other species (page 68, Table 17) we used number of scales in contact to mental, scales around midbody, dorsals along the back included in a head length, and number of suboculars.Also we included his data set of allozymes (19 characters) and two from karyology (diploid chromosomal number and number of microchromosomes in females), which were available for four of our terminals (P.antofagastensis, P. mallimaccii, P. palluma from Mendoza, and P. payunae).We added diploid chromosomal number to two other terminals, P. dorsimaculatus (from Copahue, Neuquén Province) and P. excelsus (from Ojo de Agua, Rio Negro) following Morando et al., (2001).
Phylogenetic analyses were conducted with parsimony software (TNT: Tree Analysis Using New Technology, vers.1.0; Goloboff et al., (2003).We choose TNT for our analysis because this is the only program that allows the analysis of continuous characters without first converting them in discrete characters (as proposed by Thiele, 1993).Data matrices are available at www.unsa.edu.ar/acunsa.There is more than one matrix because different coding methods were applied for binary polymorphic characters.The first block in matrix is comprised of continuous characters (26 for external morphology and 14 for skeletal anatomy), the same characters are included in the following block (characters 88-106; 109-113; 135-148).During searches we deactivate these characters if we use the range method for continuous, and the opposite to apply Thiele's gap-weighting method (Thiele, 1993).For TNT numeration of characters begins with the first block (with character "0" designating the first one), so the second block (traditional format) starts with character 40.The list of characters in text 0-132 are equivalent in TNT format of our matrices to 40-172.In our analysis we considered the following TNT characters nonadditive 68-86 (allozyme characters: 28-46); 128 and 130 (throat pattern of males and females, dorsal pattern of tails: 88-90), 168 and 170 (128 and 130, "smooth pattern" and dorsal pattern black with two rows of occelli); all remaining characters were additive.Binary polymorphic characters were analyzed in four different ways: any-instance (the polymorphic species is considered already having the novelty, so it is coded "1" as in species that this state is fixed), scaled (polymorphic species have an intermediate state "1" in an ordered series between species not having that derived state "0" and those having the derived state "2"), frequency bins (frequency of the presence of the derived state is used to score each species, percentage ranges are divided by 0-9 states for TNT analysis), and missing (question mark for polymorphic species).For more details on these methods and performance of them in cladistic analysis see Wiens (2000) TNT uses Farris'optimization (Farris, 1970) to estimate distances and costs among ranges, when ranges between two terminals overlap, TNT assumes zero cost.In this study we entered ranges for continuous data to the program considering a mean ± SD as a range for any continuous character, and we standardized all continuous characters dividing the ranges by the major data and multiplying this by ten.In this way the costs of the continous characters are similar to the others.
For frequency bins we preferred not dividing by more states than 10 because the limited sample size we had of some species.
For rooting our trees we included the other two liolaemid genera, Liolaemus (three species: L. kingii, L. tenuis, and L. pseudoanomalus) and the monotypic Ctenoblephar ys.Relationships among genera in Liolaemidae were once controversial (Frost & Etheridge, 1989;Etheridge, 1995), yet according to the last revision of intergeneric relationships (Etheridge, 1995), there are many synapomorphies supporting the monophyly of each of the three genera.Recent DNA analyses support a Phymaturus-Liolaemus sister-taxon relationship with Ctenoblepharys adspersa as the basal member of the family (Schulte et al., 2003;Espinoza et al., 2004).

1) Choice of terminal taxa
We included in the present study all recognized species of Phymaturus in literature (Etheridge & Espinoza, 2000) except the recently described P. verdugo Cei & Videla, 2003 and P. calcogaster Scolaro & Cei, 2003: cf.palluma, antofagastensis, mallimaccii, punae, patagonicus, nevadoi, payunae, zapalensis, indistinctus, and somuncurensis.Also we included spurcus, a recently revalidated species (Lobo & Quinteros, in press).In addition to these species, we included twelve populations that are assigned to named species yet show morphological evidence of evolutionary isolation (likely representing independent lineages) from nominal forms.Four of these terminals are described below as new species.We follow Cei & Videla (2003) in recognizing that the issue of the provenance of P. palluma cannot be readily resolved (i.e., populations inhabitating the eastern and western slopes of the Andes exhibit morphological [this study], karyological [Pereyra, 1992a], and DNA [R.Espinoza, unpubl.data] differences), so we use cf.palluma, rather than P. palluma in this study.

I -Description of four new species of Phymaturus
Of the twelve terminals added to this analysis to the currently recognized species, four exhibit obvious discriminating characters (squamation and patterns) that justitfy their description as new species.

Phymaturus dorsimaculatus sp. nov.
Holotype (Figure 1A and B Diagnosis: Phymaturus dorsimaculatus belongs to the palluma group (sensu Etheridge, 1995) because it has squareshaped non-imbricate superciliaries, rugose dorsal scales of the tail, usually a fragmented subocular, and the subocular-supralabials separation is two or more scale rows.It is distinguishable from all other species in the group by its unusual dorsal pattern.Dorsal pattern from occiput to the posterior region of trunk with black transverse bands interrupted medially (Figure 1A).Phymaturus verdugo is a larger species (106-120 mm both sexes; Cei & Videla, 2003) than dorsimaculatus (76.1-92.6), has a different dorsal pattern, and a patternless tail (ringed in dorsimaculatus).This new species has a different karyotype number (2n = 36) versus 2n = 26 of P. verdugo (Morando et al., 2001;Cei & Videla, 2003).Adults of cf.palluma (PA) lack melanism on head and neck and all other members of the palluma group have patternless tails.Phymaturus dorsimaculatus never have a divided rostral scale as does many punae, antofagastensis, cf.antofagastensis SC (Sierra de Calalaste and Cuesta de Randolfo), and punae LR (La Rioja).Most specimens from northern Argentina of the palluma group exhibit a homogeneous and dense "spray" pattern on their dorsum (Figure 12C) and lack black reticulation or spots.No distinctly enlarged supraoculars.Twelve non imbricate subquadrangular supercilliaries.Sixteen upper and thirteen lower ciliaries (right side).Subocular fragmented in two (left side) and three scales (right side), separated from supralabials by three to one row of lorilabials.Ten lorilabials.Preocular separated from lorilabial row by three scales.Postocular equal in length to preocular.Chinshields not enlarged (similar size of that of sublabials) forming a longitudinal row of five scales.Scales of throat round, flat, and juxtaposed.Seventy-four gulars between auditory meatus.Lateral nuchal folds well developed, with granular scales over longitudinal fold that are smaller than dorsals.Antehumeral pocket well developed.Fifty-six scales between auditory meatus and shoulder.In ventral view gular fold not well developed and posterior gular folds present with their anterior margins lacking enlarged scales on their borders.Dorsal scales round, smooth, juxtaposed.Thirty dorsal scales along midline of the trunk in a distance equivalent to head length.Scales around midbody 210.Mid-dorsal scales slightly enlarged, becoming smaller and granular on flanks and toward belly.Ventral scales larger than dorsals; 172.Between mental and precloacal area.No precloacal pores.Brachial and antebrachial scales smooth with round posterior margins.Supracarpals laminar, round, smooth.Subdigital lamellae of fingers with five keels (more conspicuous in proximal lamellae).Number of subdigital lamellae of fingers I: 9; II: 13; III: 17; IV: 19; V: 12. Claws moderately long.Supradigital lamellae convex, smooth, imbricate.Infracarpals and infratarsals with round margins and two to three obtuse keels.Supracarpals and supratarsals smooth, with round pos- terior margins.Subdigital lamellae of toes I: 8; II: 13; III: 20; IV: 23; V: 15.
Color of holotype in alcohol (Figure 1A and B): Dorsal background color brown with a distinct pattern in black extended from occiput to the posterior region of trunk forming transverse bands that are interrupted medially.Finer irregular black reticulation distributed along the vertebral region reaching the base of the tail.The tail is ringed with darker brown bands two scales wide.Forearms and hindlimbs with fine reticulation.Head not melanic.Throat gray-light brown with reticulate brown more evident below jaws.Chest black.Abdominal region variegated.
Color in life: Not available.

Etymology:
The epithet refers to the conspicuous pattern of black spots over the neck, shoulders, and dorsum of the trunk of this new species.It is a compound name (originated from two latin words dorsum that means back and macula that means spot, mark).
Distribution: Only known from the type locality, Copahue, Neuquén province, Argentina, 37°49'14"S, 21°05'12"W; 2050 m elevation.Diagnosis: Phymaturus excelsus belongs to the patagonicus group (sensu Etheridge, 1995) because it has flat imbricate superciliaries, non-rugose dorsal scales on tail, subocular usually not fragmented, and subocularsupralabials separated by one scale row.This new species differs from all other members of this group in its unique dorsal pattern, with a dorsal background in black and a pair of longitudinal series of white occelli (fig.1C).Similar patterns with a paired series of occelli are found (but in different colors, shapes, and arrengements) in payunae, zapalensis, and spectabilis sp.nov.(described below).Phymaturus payunae and P. zapalensis are sexually dimorphic in dorsal patterns, whereas in excelsus there are no pattern differences between the sexes.Dominant colors in excelsus are black and white, whereas in spectabilis are brown and light brown, and occelli in this latter species are much wider and more symmetrical.

Phymaturus excelsus
Description of holotype: Male.SVL 82.5 mm.Head length 16.0 mm.Head width 14.2 mm.Head height (at parietal) 9.0 mm.Axilla-groin 41.9 mm (50.8% of SVL).Tail length (complete, not regenerated) 82.8 mm (1.00 times SVL).Trunk width: 35.6 mm (43.1% of SVL).Twenty one smooth dorsal head scales.Five, five, and four scale organs in each postrostral.Nasal scale not in contact with rostral (separated by one scale), bordered by eight scales.Canthal separated from nasal by two scales.Loreal region flat.Nine enlarged supralabial scales with the seventh upturned posteriorly, not contacting subocular (separated by one lorilabial).Seven enlarged infralabials.Auditory meatus oval; auricular scale absent, three to four projecting scales on anterior margin of auditory meatus (both sides).Eleven convex, juxtaposed, smooth temporals.Rostral undivided.Mental subpentagonal, in contact with four scales.Interparietal bordered by seven scales, parietals smaller.Frontal region without an azygous scale.Supraorbital semicircles incomplete posteriorly on both sides.No distinctly enlarged supraoculars.Nine distinctly imbricate superciliaries.Thirteen upper and twelve lower ciliaries (right side).Subocular elongate, longer than eye diameter, separated from supralabials by a single row of lorilabials.Eleven lorilabials; tenth through eleventh contacting subocular.Preocular separated from lorilabial row by one scale.Chinshields forming a longitudinal row of seven or eight enlarged scales.Scales of throat round, flat, and juxtaposed.Seventy gulars between auditory meatus.Lateral nuchal folds well developed with granular scales over longitudinal fold that are smaller than dorsals.Antehumeral pocket well developed.Sixty-eight scales between auditory meatus and shoulder.In ventral view, gular fold absent, and posterior gular folds present with their anterior margins bordered by enlarged scales.Dorsal scales round, smooth, juxtaposed.Forty-three dorsal scales along midline of the trunk in a distance equivalent to head length.Scales around midbody 202.Middorsal scales same size of those on flanks.Ventral scales larger than dorsals.Ventral scales between mental and precloacal pores 182.Eleven precloacal pores forming an interrupted row.Brachial and antebrachial scales smooth with round posterior margins.Supracarpals round and smooth.Subdigital lamellae of fingers with five to three keels, in number I: 11; II: 17; III: 24; IV: 25; V: 17. Claws moderately long.Supradigital lamellae convex and imbricate.Infracarpals and infratarsals trifid with round margins.Supracarpals and supratarsals smooth with round posterior margins.Subdigital lamellae of toes with three to five keels: I: 12; II: 19; III: 24; IV: 29; V: 19.
Color of holotype in alcohol (Figure 1C and D): Dorsal background black on trunk, shoulder, neck, and head.A paired series of eight lateral white or cream occelli are conspicuous from the occiput to the thighs.Between these two rows of occelli are irregularly located small light-cream spots.Dorsum of limbs and tail variegated.Ventral coloration ligth gray to white, inmaculate exhibiting small spots on the throat.
Color in life: see Figure 3. Dorsal background of trunk, head and limbs black, with light brown ocelli over the dorsum.A reticulate pattern in black and light brown over the head, limbs and dorsal surface of tail.
Etymology: Phymaturus excelsus is Latin for "distinguished," which describes the peculiar and distinct pattern exhibited by these lizards.
Distribution (Figure 5): Phymaturus excelsus is known only from its type locality, where it lives syntopically with Phymaturus spurcus a species described by Barbour (1921) from Estancia Huanuluan, 40 km to the north (straight line).
Diagnosis: Phymaturus spectabilis belongs to the patagonicus group (sensu Etheridge, 1995) because it has flat imbricate superciliaries, non-rugose dorsal scales of the tail, subocular unique usually not fragmented, and subocular-supralabials separation given for one row of scales.It is distinguishible of all other species of the genus by its unusual dorsal pattern (Figures 2 and  4).Phenetically the species more close to P. spectabilis is P. excelsus.The last one exhibit a general dorsal pattern black with smaller and more occelli than spectabilis (7-8 among shoulder and the level of thighs versus 5-6) and several markings irregularly distributed on its vertebral field between the series of dorsal occelli.
Color of holotype in alcohol (Figure 2A and B): Dorsal background predominantly brown and black, six white and brown occelli delimited in black and exhibiting one to two small black spots in the middle of each occellation.Dorsal pattern of tail variegated.Beyond shoulders occelli are fused reaching the parietal region of head.Central region of head black flanked by a pair of brown bands that reach the nasal region anteriorly.Ventral surfaces light gray, lateral margins of abdomen and chest with small black diffuse spots, throat with few very small spots, more conspicuous under jaws.This light gray coloration becomes light brown gradually along the flanks.Ventral surface of tail is light brown variegated with dark brown.
Color in life: See Figure 4. Similar pattern to that described for excelsus, with the light brown fields widely extended (ocelli over dorsum and pattern of head, limbs and dorsal surface tailand).
Etymology: The epithet spectabilis is Latin and means "notable, showy" in reference to the distinct pattern of dorsal occellations in this new species.
Distribution (Figure 5): Only known from the type locality, 28 km south of Ingeniero Jacobacci, Rio Negro, Argentina.

Phymaturus tenebrosus sp. nov.
Holotype: MCN 1271.20  Diagnosis: Phymaturus tenebrosus belongs to the patagonicus group (sensu Etheridge, 1995) because it has flat imbricate superciliaries, non-rugose dorsal scales on tail, subocular usually not fragmented, and subocularsupralabials separated by one scale row.The new species is distinguishible from all other species of the genus by its dorsal pattern: black in most specimens, some with very fine, sparse white spots (e.g., holotype).This melanissm is dark brown in some specimens.The morphologically most similar species to P. tenebrosus are P. zapalensis and the recently described P. calcogaster, from which this new species can be readily distinguished.Phymaturus zapalensis has a more dense pattern of dorsal spots and very often a pattern of reticulation/ occellation and obvious sexual dimorphism (females with variegated pattern) and P. calcogaster, which was described solely from the male holotype (Scolaro & Cei, 2003), which has a peculiar pattern not seen in P. tenebrosus: dorsum with larger and homogeneous white spots arranged in trasverse rows, a subocular fragmented into five scales (P.tenebrosus usually lacks subocular fragmentation), and the throat of P. calcogaster is variegated.Variation: Based on 16 adult specimens (12 females and 4 males from both known localities).SVL 85.0-107.5 mm (x = 94.8;SD = 6.4).Head length 0.15-0.19%(x = 0.17; SD = 0.01) of SVL.Tail length 1.00-1.41(x = 1.21;SD = 0.18) times SVL.Scales around midbody 171-236 (x = 199.2;SD = 20.2).Dorsal head scales 16-25 (x = 20.2;SD = 2.18).Ventrals 149-185 (x = 171.4;SD = 14.15).Precloacal pores in males 7-9 (x = 8.5; SD = 1.0).Scales surrounding interparietal 5-8 (x = 6.9;SD = 0.9).Scales of neck along longitudinal fold from posterior border of auditory meatus to shoulder 58-88 (x = 70.9;SD = 8.3).Gulars 65-100 (x = 77.8;SD = 9.3).Scales between rostral and frontal 6-11 (x = 8.7; SD = 1.3).Dorsal and flank pattern in most specimens black, in some specimens with small and scacerly distributed white spots.Some specimens brown morph with conspicuous black spots on flanks (similar to that exhibited by specimens of P. zapalensis).Not all specimens have strong ventral coloration (orange in brown individuals, yet mustard or dark gray in black specimens), suggesting that this coloration may be related to season or physiological conditions.

Description of holotype
Color of holotype in alcohol (Figure 2C and D): Dorsal background with black spots and small white markings that are smaller and more densely distributed middorsally.This pattern extends over the head and dorsal surfaces of the limbs.The tail is patternless.General coloration of ventral surfaces (throat, limbs, and tail) gray, with slight yellow color in the posterior region of abdomen continuous over the cloaca and thighs.
Color in life: See Figure 4. Dorsum black scattered with small white spots.Margins of ventral region gray and central areas of chest, abdomen, ventral surfaces of thighs and cloaca mustard.Females can exhibit their bellies light gray to orange.

Etymology:
The epithet tenebrosus Latin word that means "dark, gloomy" in reference to the dark dorsal coloration of this new species.
Distribution (Figure 5): This new species is known to be found between Bariloche and Pilcaniyeu northward to 20 km south of Cerro Alto (National Road Nº 40), Rio Negro, Argentina.
Phymaturus cf.antofagastensis (SC), from Cuesta de Randolfo and Sierra de Calalaste: according to Pereyra (1991) the type locality of Phymaturus antofagastensis is Los Nacimientos near Paso San Francisco.Both terminals (Paso San Francisco and Randolfo-Calalaste) have enlarged scales in the central area of chest, as described in the original description (Pereyra, 1985).Specimens from Cuesta de Randolfo and Sierra de Calalaste exhibit an undivided rostral and a homogeneous dense "spray" dorsal pattern of coloration (Figure 12C), whereas most specimens from Paso San Francisco have a divided rostral scale and an "aggregate dorsal" pattern of coloration (Figure 12D).
Phymaturus cf.palluma (CH): (Chillán, Chile): lizards from this population, as in P. cf. dorsimaculatus (Copahue, Neuquén province), have ringed tails but they lack the reticulate pattern typical of other Phymaturus.Females from this locality have immaculate throats, whereas all other P. cf. palluma have spotted or black throats.Specimens from this population also have lorilabials occasionally in contact with the subocular, which never occurs in other P. cf. palluma, and members of this population have fewer scales in contact with the nasal than in all other populations.
Phymaturus cf.palluma (EP): (El Planchón and San Pedro, Chile): females can exhibit precloacal pores (not found in other P. cf. palluma), temporal scales are protruded conical shaped not as in dorsimaculatus and P. cf. palluma (CH), anterior gular fold present (absent in P. cf. palluma of Payunia), the lorilabials never contact the subocular (as in P. cf. palluma of Chillán), there are no enlarged scales on the margins of posterior gular fold as in P. cf. palluma (ME).Specimens from El Planchón are quite similar of those from Maule (MVZ 232506-07).
Phymaturus cf.palluma (ME): differs from P. cf. palluma (PA) because adult specimens of the former have partially or completely black heads and neck foldings, sexual dimorphism in dorsal pattern, but never a divided rostral (which is common in P. cf. palluma PA).This population differs also in similar way from P. cf. palluma (LB) because this last population lacks sexual dimorphism, and temporal scales in adult specimens are strongly projected (spinose).
Phymaturus cf.palluma (LB) (Laguna Blanca): is smaller than P. verdugo: 86-110 mm, whereas in P. verdugo: 106-120 mm SVL.From P. dorsimaculatus sp.nov.because do not exhibit the same pattern of black transversal bars over theirs neck and shoulders and ringed pattern on tails (this last character discriminate it also from P. cf. palluma CH).Does not exhibit sexual dimorphism in pattern like P. cf. palluma (ME) (with both states agregate and dense "spray" Figure 12).Is allopatrically distributed very far from the northern group of species (P.punae, P. cf. punae LR, P. antofagastensis, and P. cf. antofagastensis SC) and no individuals have a divided rostral.
Phymaturus cf.palluma (PA) (Payunia): these lizards are morphologically similar to those here called "Mendoza" but the adult specimens lack the distinct black head and neck present in other forms of "palluma".This character discriminates this form from other members of the palluma group.It shares with northern species of the group the a divided rostral in many specimens.
Phymaturus cf.patagonicus (EC) (San Antonio del Cuy, Rio Negro): Phymaturus with a gray background coloration with white spots (like in P. patagonicus from Dolavon, Chubut province), spots are smaller and usually these lizards have a dark gray to black dorsolateral band.This dark band is also present in individuals of P. tenebrosus and most females of P. zapalensis (Figure 12A), but P. patagonicus (EC) lacks black and brown morphs and has a smaller body size.

2) Phylogenetic analysis of Phymaturus comparison of four matrices
The analysis applying the ANY INSTANCE coding method for binary polymorphic characters brought one most parsimonious tree of 522.48 steps (CI = 0.57; RI = 0.66) (Figure 6).This topology recovered P. patagonicus as the most basal species of the genus, the patagonicus group is paraphyletic, P. payunae is the sister taxon of a subclade formed by P. somuncurensis, P. spectabilis, and P. spurcus, this group is related to another formed by P. indistinctus as the most basal species, P. cf. patagonicus (EC), and P. nevadoi as sister taxon to the pair formed by P. zapalensis and P. tenebrosus.Phymaturus excelsus is related to the palluma group for which P. dorsimaculatus is the most basal species.The topology of the palluma group in this tree is unbalanced, with species subsequently related as follows: P. cf. palluma (CH), P. cf. palluma (LB), P. cf. palluma (ME), P. cf. palluma (EP), P. cf. palluma (PA), P. punae, P. punae (LR), P. cf. antofagastensis (SC), P. antofagastensis, and P. mallimaccii.Jacknifing recovered four nodes with good support in the consensus tree: 52% for P. tenebrosus-P.zapalensis, 90% for P. spectabilis-P.spurcus, 94% for the palluma group, and 61% for the palluma group except P. dorsimaculatus and P. cf. palluma (CH).
The analysis applying the MISSING coding method for binary polymorphic characters recovered two equally parsimonious trees of 489.59 steps (CI = 0.60; RI = 0.68).In Figure 7 we show one of these trees.In this analysis the patagonicus group is not monophyletic, common to these trees are the pair formed by tenebrosus and zapalensis, and the subclade including P. somuncurensis sister taxon of P. spurcus-P.spectabilis.Phymaturus dorsimaculatus and P. cf. palluma (CH) are the basal subgroup in the palluma group, P. cf. palluma (LB) is the most basal species of a big clade formed by all remaining species.Phymaturus cf.palluma (PA) and P. cf. palluma (ME) are related to a group including cf. palluma (EP) and all northern species; P. cf. punae (LR) is basal to a couple of sister taxa, one including P. punae and P. cf. antofagastensis (SC) and the other mallimaccii and antofagastensis.For this analysis Jacknifing reports eight nodes supported over 50% of frequency.The palluma group at 99%, P. cf. palluma (EP) plus all northern species: P. punae, P. punae (LR), P. mallimaccii, P. antofagastensis, and P. cf. antofagastensis at 85%, all northern species at 83%, P. cf. palluma (LB)-P.cf.palluma (PA) at 81% (not found in both trees); the subclade formed by P. punae and P. antofagastensis (SC) at 79%, P. spurcus-P.spectabilis at 68%, all northern species without P. punae (LR) at 53% and a node not found in the two shortest trees formed by P. antofagastensis, P. punae, and P. cf. antofagastensis (SC) at 74%.Supported group common in all analysis (4) was the palluma group (94, and three 99%), P. cf. palluma (PA)-P.cf.palluma (LB) in three analysis (53, 81, 86%) (not supported in the any instance analysis).In three runs also is recovered the node P. spurcus-P.spectabilis (not in the frequency bins analysis; 52, 64, 68%).Phymaturus punae is related to antofagastensis and P. cf. antofagastensis (SC) in three jacknife consensus trees (missing, scaled and frequency) (74, 54 and 52%).The northern subclade formed by punae, punae (LR), P. mallimaccii, P. antofagastensis, and P. cf. antofagastensis (SC) have support in the missing and scaled analysis (80-83%).Phymaturus dorsimaculatus is the most basal species in two of the jacknife consensus trees (frequency and scaled analysis) (53-57%).

Consensus tree and common apomorphies
In this study there are competing hypothesis that are shown in the four original topologies (Figures 6  and 7).From these four original analysis we built a majority rule consensus tree shown in Figure 8, because a majority rule tree can choose among hypothesis incongruent among the original runs observations on those contradictory relationships are given below.
Phymaturus somuncurensis is sister taxon of spurcus -spectabilis (Node 5) in the any instance and missing analysis, somuncurensis is related to the pair payunaezapalensis in the frequency bins run while in the scaled analysis is basal to a group formed by other four species.Phymaturus cf.patagonicus (EC) is sister taxon of the group formed by nevadoi and the pair formed by zapalensis and tenebrosus (Node 6) in any instance and scaled runs, cf.patagonicus (EC) is sister taxon of patagonicus in the frequency bins analysis.Phymaturus dorsimaculatus is sister taxon of cf.palluma (CH) (Node 3) in scaled and missing runs, in the frequency bins run dorsimaculatus is the basal species of the palluma group not related to cf. palluma (CH).Northern species (Node 12) are present in any instance and missing runs.Phymaturus mallimaccii is related to antofagastensis in the any instance and missing runs (Node 13); mallimaccii is related to antofagastensis and cf.antofagastensis (SC) in the scaled analysis and to cf. palluma (EP) in the frequency bins run.
The palluma group (Node 1) is supported by the following common apomorphies (all runs) characters 5 (number of ventral scales), character 7 (number of gulars), character 11 (number of upper ciliars), character 13 (supralabial upturned), character 18 (preocular scale separation from lorilabial row) and character 26 (superciliary scales shape subcuadrangular not imbricated).The patagonicus group (Node 2) is not monophyletic in all runs and had no support as the palluma group, in those topologies where it is found monophyletic common apomorphies involves changes in characters 4 (number of dorsal scales in a head-length), 21 (males trunk length/snout vent length ratio), 28 (number of pterigoid teeth) 119 (lorilabials-subocular scale separation) and 149 (number of scleral ossicles).Evidence for the sister taxa relationship between dorsimaculatus and cf.palluma (CH) (Node 3) are changes on characters 10 (number of superciliaries) and 14 (subocular fragmentation, Figure 9).Character changes supporting Node 4 are related to characters 8 (number of scales in contact to interparietal), 13 (number of supralabial upturned), 16 (scales in contact to mental, Figure 9) and 130 (dorsal pattern of tails, Figure 10).Node 5 includes somuncurensis, spectabilis and spurcus, FIGURE 6.Two hypotheses of phylogenetic relationships among species of Phymaturus using two different methods for coding binary polymorphisms (any-instance and frequency-bins analysis).Values at some nodes refer to Jacknife support >50% from 1000 replications.FIGURE 7. Two additional methods for coding binary polymorphic characters (scaled and missing analyses).An additional topology was recovered differing only in the relationships within the patagonicus group (see text for details).Values at some nodes refer to Jacknife support >50% from 1000 replications.and changes are on characters 7 (number of gulars), 109 (male/female SVL ratio) and 151 (lacrimal foramina opening).Node 6 (within patagonicus group) is supported by changes on characters 6 (number of scales on the lateral wall of neck, between earing and shoulder) and 23 (males tibia length (SVL ratio).Node 7 (within the palluma group) is supported by the following changes, characters 9 (number of infralabials), 17 (number of scale organs on posrostrals), 18 (number of scales separating preocular from lorilabial row), 19 (maximun SVL found), 109 (male/female SVL ratio) and 128 (throat of males pattern).Characters supporting the relationship between spurcus and spectabilis (Node 8) are 14 (subocular fragmentation, Figure 9), 19 (maximun SVL found) and 110 (males trunk length/ SVL ratio).Node 9 exhibit changes on characters 10 (number of superciliaries), 13 (number of supralabial upturned), 15 (scales in contact to nasal) and 18 (number of scales separating preocular from the lorilabial row).Phymaturus cf.palluma (EP) is related to the northern subclade in Node 10: characters 0 (number of dorsal head scales), 7 (number of gulars), 8 (number of scales in contact with interparietal), 17 (number of scale organs on posrostrals) and 18 (number of scales separating preocular from the lorilabial row).Phymaturus tenebrosus is sister taxon of zapalensis at Node 11, character 21 (males trunk length/SVL ratio).The northern Argentina subclade is supported by characters 6 (number of scales on the lateral wall of neck, between earing and shoulder) and 12 (number of scales between frontal and rostral).Phymaturus antofagastensis and mallimaccii are sister taxa at Node 13, characters 19 (maximun SVL found) and 110 (males trunk length/ SVL ratio).

Search of additional topologies
Analyzing with the range method of TNT for continuous characters we had an unique shortest tree for three analysis (any instance, frequency bins.and scaled analysis) while running with the missing coding method for binary polymorphic characters we had two topologies.Continuous characters ranges were entered with three decimals, so length measures of trees are given with three decimals, this fact make less expectable reaching more than one topology for the same run.For this reason we wanted to know how many other topologies are close to the shortest one for each analysis, and look if exist a tree or/and trees repeated among these four different analysis.We made a search looking for suboptimal trees up to one step longer having a total of 383 topologies (any instance-analysis: 57 trees, scaled-analysis: 199, frequency bins-analysis: 29 and missing-analysis: 98).In this analysis we got three topologies repeated in two of the four runs (tree 13 of any instance-analysis = tree 369 of the missinganalysis; tree 50 of any instance-analysis = tree 84 of the scaled-analysis and tree 177 of the scaled-analysis = tree 296 of the missing-analysis) and two trees were recovered in three of the four analysis: tree 18 of any instance-analysis = tree 82 of the scaled-analysis = tree 295 of the missing-analysis; and tree 54 of the any instance-analysis = tree 176 of the scaled-analysis = tree 294 of the missing-analysis.Trees 18 and 54 repeated in three of the four analysis differs between them in the arrengement of the patagonicus group.The palluma group in these trees show the same topology obtained in the missing-analysis, for the patagonicus group recover the following topologies: ((((((((tenebrosus zapalensis) nevadoi) ((spurcus spectabilis) somuncurensis)) payunae) cf.patagonicus EC) patagonicus) indistinctus) excelsus); and the other differing only in the position of payunae, here as follows: (((tenebrosus zapalensis) nevadoi) (((spectabilis spurcus) somuncurensis) payunae)).Terminal arrengements of the patagonicus group are the same recovered in the majority rule consensus tree and in other topologies see Figures 6 to 8.

3) Observations on the ontogenetic shift of morphological characters
A sample of 15 juvenile Phymaturus cf.palluma (PA), presumably in their first range from 65.6-83.0mm SVL allowed us to make observations on their dimorphic differentiation.Dissection of these specimens were disected to determine sex for all but two individuals.SVL X = 75.2(SD = 4.8; range = 69.3-80.5)for males, and X = 75.0(SD = 4.8; range = 67.3-83.1)for females.All specimens have the same dorsal pattern as adults, three of five males have their abdomi-  nal region variegated as do two of eight females.All male juveniles lack precloacal pores, but four of them (4/5) have a row of slightly modified scales suggesting the development of precloacal pores.All male juveniles have 2-3 enlarged scales on the base of the tail adjacent to the posterior margin of the cloacal apperture as can be seen in most adult males of the palluma group.This character is more obvious than the presence or absence of precloacal pores and be used to sex very early stages of neonate and new-born specimens without the need for dissections.Two individuals of undetermined sex were among the smallest examined: 66.1 and 65.6 mm SVL (SDSU 1955 and1961, respectively), both lack enlarged scales on the base of the tail and no traces of precloacal pores.
Four embryos taken from two gravid dissected females of P. dorsimaculatus exhibit the following differences in comparison to adults of their species: more, smallerscales on the dorsum and flanks, counting dorsal scales in a head length (moving with the caliper a head length on the midle of the back) in these embryos are 46-60 (36-44 in adults), fewer scale organs on the postrostrals 2-3 in two embryos but not fomed in the other two (X = 2.20; SD = 0.45) than in adults (X = 3.4; SD = 1.27); temporals are flat and forming a pavement in embryos, whereas in adults these become more prominent and, in some specimens, conical; dorsal scales of the tail are smooth, whereas in adults these are rugose (as for members of the palluma group).All other characters of squamation and those referred to dorsal, ventral, and tail pattern are the same as the adults.Almost the same differences were observed between juvenile and adult P. cf. palluma (CH): more, smaller scales in the dorsum of body (39-42 in juvenile versus 32 in both adult females), temporals flat in juveniles vs. conical to spiny in adults, and dorsal scales of the tail are smooth in juveniles, whereas these are conspicuously rugose in adults.There was no distinction in the number of scale organs on the postrostrals as was detected between embryos and adults of P. dorsimaculatus.Later conditions of characters (exhibited by adults) described here having change from embryo-juvenileadult specimens like the number of dorsal body scales (in a head length), shape of the temporals, and dorsal tail rugosity are present only in species belonging to the palluma group.In species of the "patagonicus" group, those earlier stages (greater number of dorsal scales, temporals flat, tail scales smooth) for all Phymaturus species are present also in adults, and because this condition is the same in adults of Ctenoblepharys and Liolaemus, we assume that terminal additions on the ontogeny of those characters happened in the common ancestor of the palluma group.

4) Biogeography of Phymaturus
Taking into account endemism areas described for this southern region of South America ( and Puna (PUNA) for the last two species.Curiously this subclade of northern populations of Phymaturus is distributed exclusively in the transitional Puna subdistrict delimited by Martínez Carretero (1995) that occupies western region of Catamarca, La Rioja and northwestern San Juan.Spliting this group in two areas in any further biogeographic analysis (like Fitch optimization, DIVA, etc.) can carry to make extra assumptions of historical events (vicariance, extinctions or/and dispersals).Delimitation of areas among members of the northern subclade of the palluma group between Prepuna and Puna it is not enough explicit and assigning one or other area to these taxa becomes problematic.In cladistic biogeography comparing original cladograms of taxa with general area cladograms is a way to asses how our historical hypothesis are congruent with that history of the geography they occupies as it was recovered from the phylogenies of other taxa (plants or animals).Our four hypothesis for Phymaturus are incongreunt among them in many nodes of the tree, but some part of the relationships recovered are congruent to the general area  these two pairs of areas is not seen here in Phymaturus cladograms, in the four analysis species distributed in that area are close related to the norther subclade (Transitional Puna).Area relationships suggested by the frequency bins analysis is not congruent with the other three analysis of Phymaturus and with general area cladograms known.In this hypothesis (Figure 6) a repeated vicariant hypothesis is suggested within both groups of species: there exist a pair of sister taxa linking the Payunia plateau (Mendoza) with Laguna Blanca (Neuquén), cf.palluma (PA) with cf.palluma (LB) in the palluma group and other couple of species in the patagonicus group, payunae and zapalensis. In

Taxonomic remarks
The taxonomic use of palluma or flagellifer for designating the oldest known species of the genus should be resolved soon by the International Commision of Zoological Nomenclature (ICZN).There are two com- peting positions: Cei & Lescure (1985) and Lescure & Cei (1991) proposed the use of flagellifer instead of palluma Molina, 1782, owing to confusion involving the type specimen, which was an individual of the Chilean teiid Callopistes maculatus.Recently, Etheridge & Savage (2003) proposed conservating the usage of the name palluma by designation of a neotype for Lacerta palluma Molina, 1782, because its wide use in the literature for the last 150 years.Both requests are reasonable and will be considered by the ICZN in 2005.
The name Phymaturus adrianae, proposed by Pereyra (1992b) for a new species in an meeting abstract is a nomen nudum (ICZN, Art. 9, 1999).However, Cei & Videla (2003) used the name and provided a photograph and some characteristics to diagnose the species, and thus may be considered the authors of the name Phymaturus adrianae.But because there was no type material indicated in this publication, we do not consider this name valid.Considering that at this time we do not know certainly the identity of the true palluma (if it is assignable to populations of Mendoza in Argentina or to populations of Chile, as noted by Cei & Videla, 2003) exist the risk that adrianae (populations of Argentine side of the Andes) could be a synonym of palluma.The aim of this study was extracting phylogenetic information from morphological characters and the description of new taxa, the resolution of those taxonomic problems requires others kind of information.
Recently Scolaro & Cei (2003) described a new species of Phymaturus based on a single specimen collected in the area of Esquel (Chubut, Argentina).Characters provided in this description make it difficult to assign this new taxon to either species group, but some comments are warranted.According to their description, the specimen has imbricate superciliaries (pg.109) and a spiny tail, but not as spiny as that seen among members of the palluma group (pg.111).These characters suggest this new species is allied outside of the palluma group, but the authors also describe the new taxon as possessing four suboculars separated from the supralabials by two rows of lorilabials -a character that is shared by all members of the palluma group.However, fragmentation of the subocular was also described for P. spurcus (Lobo & Quinteros, in press).The dorsal pattern of this lizard resembles specimens of P. tenebrosus and some P. zapalensis, as well as other species of the "patagonicus" group: P. excelsus, P. spectabilis, P. tenebrosus, and P. spurcus, which exhibit orange-red coloration on their abdominal region.Given this evidence, it seems unlikely that this new taxon is nested within the palluma group.A more careful examination of the material and the collection of additional specimens should help to resolve this question.
The new species described in this investigation have distinctive color patterns, and all (except P. excelsus) are allopatric.Phymaturus tenebrosus and P. spectabilis exhibit an interesting polymorphism in pattern, but they do not resemble other species of the genus.Some specimens of P. tenebrosus have a brown background coloration with irregular black spots, which becomes increasingly melanistic in intermediate specimens reaching the completely black pattern.A few specimens of P. spectabilis have a brown background coloration (without typical white occellation) with irregularly located small markings on the dorsolateral region of the trunk The only species living syntopically are P. excelsus and P. spurcus, yet no intermediate individuals were found at Ojo de Agua.Rocky formations in western Argentina and Patagonia are disjunct, forming small patches (of a few kilometers of area)that isolates these populations of lizards, which are very restricted in their microhabitat preferences (restricted to rocky outcrops).Other terminals included in this study (as "cf.") are the subjects of current studies associated with determining their taxonomic status.

Phylogenetic conclusions
Using different coding methods for analyzing polimorphisms can affect results and can yield to different topologies.Systematists analyzing morphological information should not avoid using polymorphic or/and continuous characters and should pay more attention in methods they apply (not just one as we can see in most of published analysis).Studying congruence between our results and others constructed from different (independent) evidence (e.g., DNA) could provide tests of the accuracy and performance of these different approaches to understanding morphological variation.Another way of evaluating these competing hypothesis would be applying reconcilied trees (Page, 1994;Page & Charleston, 1998) or other alternative methods for studying the biogeographic history of a group of organisms, those hypothesis that implies fewer hypothesis of events (dispersals, extinction, etc.) would be the preferred ones.
The palluma and patagonicus groups -The palluma group is strongly supported in this analysis (always over 90% of jacknife values), so there is no doubt about its monophyly.Characters supporting the palluma group are not seen in the other two genera of Liolaemidae.There are three additional characters that could be considered apomorphies for this group (that should be revisited in future studies): all members of the group exhibit a longitudinal central band in the dorsum of sligthly enlarged scales that are larger than those along the flanks.This character is not exhibited by species of the patagonicus group.Scales of the tail are spinier than those of the patagonicus group and have surface rugosities, which are absent in members of the patagonicus group.Scales of the tail of Phymaturus patagonicus are not more spinose than many species of Liolaemus, a closer observation of scale shape and arrangements reveals that the difference between the genera is not the size of spines but their shape (subquadrangular/rectangular in Phymaturus, round to lanceolate in Liolaemus) and their disposition, forming more conspicuous rings around the tail perimeter in Phymaturus.Characters states given by Etheridge (1995) for the patagonicus group are shared with Liolaemus or exhibit variation within that genus, so it is not surprising to find that this group lacks support in our analysis or was even found to be paraphyletic in some analyses (see Figures 4 and 5).
Relationships within the palluma group -In all analysis performed northern species of the palluma group (P.punae, P. cf. punae LR, P. antofagastensis, P. cf. antofagastensis SC, and P. mallimaccii) formed a monophyletic group, and in all topologies P. cf. palluma (EP) is nested within the group or as the sister taxon of the group.Phymaturus dorsimaculatus is the basal species of the group.Cei & Videla (2003) described a pattern of transverse black bars over the dorsum of specimens of P. verdugo, we were not able to examine those specimens, but this character suggests a relationship with P. dorsimaculatus.In some analysis Phymaturus cf.palluma (CH) falls as sister taxon to P. dorsimaculatus or occupied a basal position in the palluma group.Both species have ringed tails and their distribution are proxi-mate.However, larger samples of P. cf. palluma (CH) are needed for study.The position of P. cf. palluma (ME), P. cf. palluma (PA), and P. cf. palluma (LB) are uncertain, in some analysis they form a monophyletic group, yet in others they are not closely related and represent subsequent diverging branches in the evolution of the group previous to the origin of the northern subclade.
Relationships within the "patagonicus" group -Within this "group" relationships are problematic.Few of the internal nodes were well supported and the competing hypothesis indicate that much additional work is needed.Relationships between P. spurcus and P. spectabilis, like those found between P. tenebrosus and P. zapalensis, were paired in several analyses, but not in every case (see frequency bins analysis, Figure 6).Even without specific analysis of levels of polymorphism and degree of overlap of continuous characters across the genus, preliminary observations suggest that the patagonicus group is more evident and that fact can affect more the morphological analysis than in the palluma group.Perhaps in the patagonicus complex radiation of independent lineages was fast and more recent, not giving sufficient time for the accumulation of morphological differentiation.Overall, we found more divergent patterns of morphological variation within the patagonicus group than in the palluma group (see Figures 3 and 4), yet we found less differentiation in other types of characters (squamation, skeleton, etc.).
Sampling limitations -This first morphological approach to recovering the phylogeny of Phymaturus is based on a heterogeneous assemblage of evidence.First, it is obvious that the alpha-taxonomy of the genus is less studied and known than in many groups of Liolaemus, so it seems likely that many more terminal taxa should be included in future analyses(including the recently described species P. verdugo and P. calcogaster).Second, we had incomplete information from allozymes (only for 27% of terminals) and skeletons (only 59% of terminals).In some cases we had only a small number of specimens (i.e., P. cf. palluma LB, P. cf. palluma CH; P. nevadoi, only known from three specimens), which can affect results of the phylogenetic analysis (Hillis, 1998).One of our goals for future studies will be increase these samples both in characters and taxa.
Comparisons to a previous revision of the genus -The main goals of this investigation were proposing hypothesis about the history and phylogenetic relationships within the genus Phymaturus based on derived shared characters.The most extensive recent study on Phymaturus was conducted by Pereyra (1992a), he studied 35 meristic, 52 morphometric, 19 enzymatic, and 4 chrommosomal characters for six especies including in his clustering analysis five of them.Because Pereyra's (1992a) study was conducted using phenetic criteria it is difficult to compare his results with ours, yet some observations are warranted.Even when a phenetic analysis only show degree of similarity without discriminating similarities originated by common ancestry from homoplasy or plesiomorphy, Pereyra (1992a) made phylogenetic conclusions including a proposed common ancestry for P. antofagastensis, P. mallimaccii, and P. payunae (pg.99), saying that conclusion is consistent with phenograms obtained.He adds to his conclusions that the rest of "flagellifer group" species as derived from close ancestor to the P. antofagastensis-P.mallimaccii node which is contradictory with the previous sentence.His separated phenetic analysis of meristic, morphometric, chromosomal and biochemical data show incongruent results among them.The meristic analysis of Pereyra (1992a) found payunae as the most distant phenetic species ((((P.antofagastensis-P.punae) (P.cf.palluma ME P. cf. palluma CH)) P. mallimaccii) P. payunae) (fig.15, pg.36) while its morphometric data-based phenogram shows ((((antofagastensis cf.palluma ME) P. mallimaccii) P. punae) P. payunae) cf.palluma CH) with P. palluma from Chile as the most distant phenetic species (fig.24, pg.47).Pereyra's chrommosomal analysis found the node P. payunae-P.cf.palluma CH (fig.38, pg.61) linked to ((P.antofagastensis-P.mallimaccii) P. cf. palluma ME) and his analysis of enzymes found P. payunae-P.cf.palluma ME (fig.39, pg.64) linked to ((P.antofagastensis-P.mallimaccii) P. cf. palluma CH).Thus, his own results as was shown above do not allow conclusions as those proposed.Beyond systematic or/ and phylogenetic conclusions that circumstantially we can agree or disagree, Pereyra's (1992a) study brings valuable information that we already re-analyzed in the context of our cladistic analysis (see methods section).Morrone (1994Morrone ( , 1996) ) analyzed relationships among provinces of the Andean subregion (corresponding to southern South America below 30° South latitude) based mainly on arthropod fauna.He found subantarctic and central Chilean provinces as sister taxa and more closely related to Puna province.In our analysis, P. cf. palluma (EP) inhabiting Cordillera andina and Valle Central (circa Santiago latitude) is related to spe-cies of Prepuna and Puna, whereas P. cf. palluma (CH) (southern populations) and P. cf. palluma (ME) distributed in the Argentine province of Mendoza (eastern Andean mountains) are related to other "palluma" of Payún and central Neuquén.Recent analysis of the chiliensis group of Liolaemus (Lobo, in press) found species groups living mainly at Valle Central (L.gravenhorsti and L. robertmertensi) more related to the alticolor group (Puna) than to any other group.

Biogeographic Implications
An obvious conclusion of this study is that more taxonomic and phylogenetic studies of Phymaturus are needed.We are confronted with a group that is substantially more species rich than originally thought and because their distribution and peculiar natural and life history (saxicolous, herbivorous, and viviparous) offers a very interesting and tempting challenge to herpetologists, both systematic and ecological.

FIGURE 1 .
FIGURE 1. (A) Dorsal view of the holotype of Phymaturus dorsimaculatus sp.nov.(B) Ventral view of the same specimen.Snout-vent length = 92.6 mm.(C) Dorsal view of the holotype of Phymaturus excelsus sp.nov.(D) Ventral view of the same specimen.Snout-vent length = 82.5 mm.
-Male.SVL 87.2 mm.Head length 16.2 mm.Head width 15.5 mm.Head height (at parietal) 9.2 mm.Axilla-groin 46.5 mm (53.3% of Snoutvent length).Tail length (complete, not regenerated) 81.1 mm (0.93 times SVL).Body moderately wide, trunk width: 33.3 mm (0.38% of SVL).Twenty-two smooth dorsal head scales.Seven, four, and seven scale organs in each postrostral.Nasal bordered by eight scales, not in contact with rostral.Canthal separated from nasal by two scales.Loreal region flat.Eight enlarged supralabial scales with seventh upturned posteriorly but not contacting subocular.Seven enlarged infralabials.Auditory meatus oval with five projecting scales on the anterior margin.Auricular scale absent.Nine convex, juxtaposed temporals.Rostral undivided.Mental subpentagonal, in contact with four scales.Interparietal bordered by seven scales.Frontal region without an azygous scale.Supraorbital semicircles incomplete posteriorly.No distinctly enlarged supraoculars.Nine imbricate flat superciliaries.Subocular unfragmented, separated from supralabials by one row of lorilabials.Nine lorilabials, the eight to ninth contacting subocular.Preocular separated from lorilabial row by two scales.Scales of throat round, flat, and juxtaposed.Eighty-one gulars between auditory meata.Lateral nuchal folds well developed, with granular scales over longitudinal fold.Antehumeral pocket well developed.Eighty-one scales between auditory meatus and shoulder.In ventral view, gular fold not well developed and posterior gular folds present with their anterior margins with enlarged scales on their borders.Dorsal scales round, smooth, juxtaposed.Thirty-nine dorsal scales along midline of the trunk in a distance equivalent to head length.Scales around midbody: 202.Mid-dorsal scales not enlarged in comparison to those on flanks.Ventral scales larger than dorsals.Ventral scales between mental and precloacal pores: 170.Nine precloacal pores.Brachial and antebrachial scales smooth with rounded posterior margins.Supracarpals laminar, round, smooth.Subdigital lamellae of fingers with 3-5 keels (more conspicuous in proximal lamellae).Number of subdigital lamellae of fingers I: 11; II: 16; III: 22; IV: 24; V: 15.Claws moderately long.Supradigital lamellae convex, imbricate.Infracarpals and infratarsals with round margins and 2-3 obtuse keels.Supracarpals and supratarsals smooth, with round posterior margins.Subdigital lamellae of toes I: 13; II: 17; III: 18; IV: 23; V: 21.

FIGURE 8 .
FIGURE 8. Majority-rule consensus tree of the four original runs.Numbered nodes are common to all topologies and their corresponding apomorphies are indicated in the text.
cladogram published by Flores & Roig-Juñent (2001), Roig-Juñent & Flores (2001) and Roig-Juñent et al., (2002), the relationship between Monte Austral and Patagonia Central is suggested in three of the four trees: ((somuncurensis (spectabilis spurcus)) in two runs, the any instance and missing analysis, and the relationship between patagonicus and indistinctus in the scaled analysis.The relationship between Payunia and Patagonia Occidental is suggested here by tenebrosus and zapalensis in three runs, any instance, scaled and missing analysis.The relationship between Cordillera Andina Valle Central as basal of
Figure 8 a majority rule consensus tree of the four original analysis show species of Prepuna and Puna areas of the palluma group forming a monophyletic group with cf.palluma (EP) inhabiting Valle Central and Cordillera Andina as the sister taxon.Phymaturus dorsimaculatus found in Patagonia Occidental is sister taxon of cf.palluma (CH) a Chilean species distributed not far from this locality.Phymaturus cf.palluma (LB) distributed at southern latitude as dorsimaculatus and cf.palluma (CH) is the subsequent basal species.Western central Argentina species as cf.palluma (ME) and cf.palluma (PA) are more related to the cf.palluma (EP) and the northern clade than any other.In the patagonicus group there exist a repeated pattern where Monte Austral was related to the patagonian areas, at Node 5 somuncurensis basal species of this clade, and cf.patagonicus (EC) basal to the other clade.