Moleculas Phylognetis and Bio Geography in Anguid Lizards

8/3/2019 Moleculas Phylognetis and Bio Geography in Anguid Lizards

1/23

Molecular Phylogenetics, tRNA Evolution, and Historical

Biogeography in Anguid Lizards and Related Taxonomic FamiliesJ. Robert Macey,* ,1 James A. Schult e II,* Allan Larson,* Bori s S. Tuni yev,

Nikolai Orlov, and Theod ore J. Papenf uss

* Department of Biology, Box 1137, Washington University, St. Louis, Missouri 63130; Caucasian State Biosphere Reserve,Sochi, Russia; Zoological Institute, Russian Academy of Sciences, St. Petersburg, Russia;andMuseum of Vertebrate Zoology, University of California, Berkeley, California 94720

Received December 24, 1997; revised October 15, 1998

P h y l o g e n e ti c r e l a ti o n sh i p s a m o n g l i z a r d s o f th e fa m i -

i e s An g u i d a e , A n n i e l l id a e , Xe n o sa u r i d a e , a n d S h i n i sa u -

r id a e a r e i n v e s t i g at e d u s i n g 2 00 1 a l i gn e d b a s e s o f

m i t o c h o n d r i a l D N A s e q u e n c e f ro m t h e g e n e s e n c o d i n g

N D 1 ( s u b u n i t o n e o f N A D H d e h y d r o g e n a s e ) , t R N A Il e ,

t R N AGl n , t RN AMe t , N D2 , t RN ATrp , t R N AAl a , t R N AAs n ,

tR N ACy s, t R N ATy r, a n d C OI ( s u b u n i t I o f c y t o c h r o m e c

o x i d a s e ) , p l u s t h e o r i g i n f o r l i g h t - s t r a n d r e p l i c a t i o n

O L) b e t w e e n t h e t R NAAs n a n d t h e t R N ACy s g e n e s . T h e

a l i g n e d se q u e n c e s c o n t a i n 1 01 3 p h y l o g e n e ti c a l l y i n fo r -

m a t i v e c h a r a c t e rs . A w e l l -r e s o l v e d p h y l o g e n e t i c h y -

p o t h e s i s i s o b t a i n e d . B e c a u s e m o n o p h y l y o f t h e f a m i l y

Xe n o sa u r i d a e (S h i n i s a u r u s a n d X e n o s a u r u s) i s sta ti sti -

c a l l y re je c te d , w e r e c o m m e n d p l a c i n g S h i n i s a u r u s i n ase p a r a te fa m i l y , th e S h i n i sa u r i d a e . Th e fa m i l y A n n i e l l i -

d a e a n d t h e a n g u i d s u b f am i l ie s G e rr h on o t i n ae a n d

An g u i n a e e a c h f or m m o n o p h y le t i c g r ou p s r e c e iv i n g

s t a ti s t ic a l s u p p o rt . T h e D i p lo g lo s s i n ae *, w h i c h a p -

p e a r s m o n o p h y l e t i c , i s r e t a i n e d a s a m e t a t a x o n (d e -

n o te d w i th a n a ste r i sk ) b e c a u s e i ts m o n o p h y l y i s sta ti s-

t i c al ly n e i t h e r s u p p o r t e d n o r r e je c t e d . T h e f am i l y

An g u i d a e a p p e a rs m o n o p h y le t i c i n a n a ly s e s o f t h e

D N A s e q u e n ce d a t a , a n d s t a t is t i ca l s u p p o rt f or i t s

m o n o p h y ly i s p r ov i d e d b y r e a n al y s is o f p r e v i ou s l y

p u b li s h e d a l lo z y m ic d a t a . An g u i d l i za r ds a p p e a r t o

h a v e h a d a n o r th e r n o r i g i n i n L a u r a si a . Ta x a c u r r e n tl y

o ca t e d o n G on d w a n a n p l at e s a rr iv e d t h e r e b y d i s -

p e r sa l fr o m th e n o r th i n tw o se p a r a te e v e n ts, o n e fro m

t h e We s t I n d i e s t o S o u t h A m e r i c a a n d a n o t h e r f ro m a

L au r a s ia n p l at e t o Mo ro c c o . B e c a u s e b a s al a n g u i n e

i n e a g e s a r e l o c a t e d i n w e s t e r n E u r a s i a a n d M o r o cc o ,

fo r m a ti o n o f th e A tl a n ti c O c e a n (l a te E o c e n e ) i s i m p l i -

c a te d i n th e se p a r a ti o n o f th e A n g u i n a e fr o m i ts N o r th

A m e r i c a n si ste r ta x o n , th e G e r r h o n o ti n a e . S u b se q u e n t

d i s p e r s a l o f a n g u i n e l i z a r d s t o E a s t A s i a a n d N o r t h

A m e r i c a a p p e a r s t o h a v e f o l l o w e d t h e O l i g o c e n e d r y -

n g o f th e T u r g a i S e a . T h e a l te r n a ti v e h y p o th e si s, th a t

a n g u i n e l i z a rd s o r i g i n a t e d i n N o r t h Am e r i c a a n d d i s -

p e r s e d t o A s i a v i a t h e B e r i n g l a n d b r i d g e w i t h s u b s e -

q u e n t c o l o n i z a ti o n o f E u r o p e a n d M o r o c c o , r e q u i r e s a

p h y l o g e n e t i c t r e e s e v e n s t e p s l o n g e r t h a n t h e m o s t

p a r s i m o n i o u s h y p o t h e s i s . N o r t h Af ri c a n , E u r o p e a n ,

a n d W e st A si a n a n g u i n e s w e r e i so l a te d fr o m o th e r s b y

t h e r ap id u p li ft o f T ib e t i n t h e l at e Ol ig o ce n e t o

Mi o ce n e . P h y l og e n e ti c a n a ly s i s o f e v o lu t i on a ry

c h a n g e s i n th e g e n e e n c o d i n g tRN ACy s s u g g e s t s g r a d u a l

r e d u c t i o n o f d i h y d r o u r i d i n e ( D ) s t e m s b y s u c c e s s i v e

d e l e t i o n o f b a s e s i n s o m e l i n e a g e s . T h i s e v o l u t i o n a r y

p a t t e r n c o n t r a s t s w i t h t h e o n e o b s e r v e d f o r p a r a l l e l

e l i m i n a t i o n o f t h e D - s t e m i n m i t o c h o n d r i a l t R N A s o f

e i g h t o t h e r r e p t i l e g r o u p s , i n w h i c h r e p l i c a t i o n s l i p -

p a g e p r od u c e s d i re c t r e p e at s . An u n u s u a l , e n l a rg e d

TC (T) s te m i s i nfe rr e d fo r t RN ACy s i n m os ts p e c i e s . 1 9 9 9 Ac a d e m i c P r e s s

Key Word s: R e p ti l i a ;S a u r i a ;A n g u i m o r p h a ;A n g u i d a e ;

A n n i e l l i d a e ; S h i n i s a u r i d a e ; X e n o s a u r i d a e ; A s i a ; E u -

r o p e ; M o ro c c o ; N o r t h Am e r i c a ; h i s t o r i c a l b i o g e o g ra -

p h y ; m i t o c h o n d r i a l D N A ; c y s t e i n e t R N A; p h y l o g e n e t i c s

Anguid lizar ds, found predomina tely in th e north ernh e mis p h er e , a r e a n e xcit i n g gr ou p for a mole cu l a rphylogenetic study of biogeographic fragmentation be-tween North America and Eurasia. The anguimorph

family Anguidae contains three subfamilies. The sub-family Gerrhonotinae occurs strictly in North Americaand Central America. The subfamily Diploglossinaer a n g es fr om Mex ico a n d t h e We st I n d ie s t o S ou t hAmerica. The subfam ily Anguinae, compr ising the gen -er a Anguis a n d Ophisaurus, is the only anguid subfam-ily that occurs in the Old World. Anguis is restricted toEurope, whereas Ophisaurus is foun d in easter n NorthAme r ica , e a s t er n As ia , w es t e r n As ia , a n d a d ja ce n tEu rope a nd Morocco.

Some taxonomists consider the anguimorph l izardfamily Anniellidae a fourth subfamily of the Anguidae

(see Gaut hier, 1982). It compr ises two species, Anniellageronim ensis a n d A. pulchra, from the west coast of North America. Three major hypotheses have been

1 To whom correspondence should be addressed. Fax: (314) 935-4432. E-ma il: m [email protected] l.edu.

Molecular Ph ylogenetics an d E volution

Vol. 12, No. 3, August, pp. 250272, 1999

Article ID mpev.1999.0615, available online at http://www.idealibrary.com on

25 0055-7903/99 $30.00

Copyright 1999 by Academic PressAll rights of reproduction in any form reserved.


2/23

considered for the phylogenetic position of the Annielli-dae relative to t he Anguidae: (1) Anniella is the sistergroup to a ll an guid t axa (Good, 1987), (2) Anniella ishe sister group t o the Anguinae (Gaut hier, 1982), an d

(3) Anniella i s t h e s i s t e r t a x o n t o Anguis of Europe(Keqin and Norell, 1998).

The anguimorph family Xenosauridae occurs in theNew World (Xenosaurus) and Old World (Shinisaurus).Some authors consider Shinisaurus a separate mono-ypic family, the Shinisauridae (see Zhao and Adler,

1993). No previous molecular study has examined therelationships of these ta xa.

Amajor question within the Anguinae is the phyloge-netic position of Anguis fragilis, Ophisaurus apodus,a n d O. koellikeri, which occur between t wo extremelya r g e b a r r ie r s t o fa u n a l d is t r ib u t ion s , t h e At l a n t ic

O ce a n a n d t h e Tib et a n P l a t ea u . C u r r e n t t a x on omymplies at least two separate origins of anguid lizards

n this region. Mitochondrial DNA sequences are re-p o r t e d f o r t h e s e t a x a a s w e l l a s f o r t h e E a s t A s i a nO. harti and the North American O. attenuatus a n dO. ventralis.

We e xa mi n e a l l g e n er a w it h i n t h e G er r h on ot i n a eexcept Coloptychon, which is known from only threespecimens. Abronia oaxacae, Ba risia im bricata, Gerrho-notus liocephalu s, a n d Mesaspis moreleti represent th efour tropical genera in our ana lysis. Five species of Elgaria, primarily from the temperate part of NorthAmerica, are examined. Elgaria coerulea, sam pled fromcoastal California, is the most northern species of the

Gerrhonotinae. Two other species a re included fromCalifornia, E. m ulticarinata collected from the east sideof t h e S ie r r a N eva d a a n d E . p a n a m i n t i n a from a nadjacent population in the Inyo Mounta ins. Elgariakingii, sampled from Arizona, occurs along the westcoast of Mexico in the Sierra Madre Occidental oppositeBaja California where E. paucicarinata was obtainedfrom the Sierra de La Laguna. This choice of speciesallows an examination of taxa from both sides of theGulf of California, a region of rifting between tectonicplates.

All genera of the neotropical subfamily Diploglossi-

n a e a r e e x a mi n e d . Ophiodes striatus r e p r e s e n t s t h eonly endemic anguid genus in South America. Celestusenneagrammus from Mexico and Diploglossus biloba-tu s from Costa Rica represent mainland North Ameri-can diploglossines, an d Diploglossus pleei, S auresiaagasepsoides, a n d Wetm orena haetiana represent WestIndian taxa.

Our sampling includes a comprehensive representa-ion of species in the Anniellidae (Anniella geronim en-

si s a n d A. pulchra) and genera of the Xenosauridae(Shinisaurus crocodilurus from China and Xenosaurusgrandis from Mexico).

Heloderma suspectum a n d Varanus griseus, NewWorld and Old World representatives of the Varanoi-d ea , s er v e a s ou t g r ou p s t o r oot t h e t r e e . P r e v iou s

phylogenetic analyses of morphological data (Estes et

al., 1988; Ma cey et al., 1997a; Schwenk, 1988) cannot

determine whether the Varanoidea or the Xenosauri-

dae is closer to the anguid and anniell id clade; this

question, however, is not a focus of this study.

Phylogenetic relat ionships ar e exam ined u sing 2001

aligned positions (1013 informative) of mitochondrialDNA sequence. The region sequenced extends from

the protein-coding gene, ND1 (subunit one of NADH

dehydrogenase), through the genes encoding tRNAIle ,

tRNAGln , t R N AMe t , N D 2 , t R N ATrp , t R N AAla , t R N AAsn ,

tRNACy s, a n d t R N ATyr to t he protein-coding gene COI

(subun it I of cytochrome c oxidase), and includes the

replication origin for the light strand (O L) between th e

tRNAAsn and the tRNACy s genes.

Pr eviously published allozymic data (Good, 1987,

1988) are reanalyzed and compared with the results

obtained from the new DNA sequence data to provide acompr ehensive assessment of relationships a mong the

ta xa investigated.

The m itochondr ial genomic region sequ enced demon-

strates several unusual characterist ics among squa-

m a t e r ep tiles (Ku m a za wa a n d N is hid a , 1995;

Kumazawa et al., 1996; Macey et al., 1997a,b,c; S eut in

et al ., 1994). Within anguimorph l izards, gene se-

quences encoding tRNACys lack a dihydrouridine (D)

stem and instead conta in a D-arm replacement loop in

Varanus (Macey et al., 1997b). A model involving re pli-

cation slippage has been proposed for the formation of

D -a r m r e p la c eme n t l oop s i n mi t och on d r i a l t R N As(Macey et al., 1997b). Under this model direct r epeats

are expected and the size of the D-arm replacement

loop should be less than 12 bases, the minimum num-

ber of bases normally found between the amino acid-

acceptor (AA) and the anticodon (AC) stems when a

D-stem is present (Macey et al., 1997b). Alternatively,

gr a d u a l r e la xa t ion of p a ir in g a m on g b a se s in t h e

D-stem would not produce either repeats or deletion of

bases. Gra dua l deletion of bases or bas e pair s would not

produce repeats but could produce length variation

that may result in less than 12 bases between the AA-and AC stems. Under a model of gradual deletion of

b a s e s o r b a s e p a i r s , a t R N A t h a t h a s a s i n g l e b a s e

p a i r i n g i n t h e D - s t e m c o u l d r e s u l t , a s h a s b e e n o b -served in th e tRNAAsn gene of the frog, Xenopus laevis(Dirheimer et al., 1995; Roe et al., 1985). Mitochondrial

tRNAs that have a single base pairing in the D-stemhave a tert iary structure dist inct from both standard

tRNAs and tRNAs in which no pairings are observedbetween th e AA- an d AC-stems (Steinber g et al., 1994).

A phylogenetic ana lysis of secondary structure for

tRNACy s w it h i n t h e An g u imor p h a s er v es t o t e st t h e

hypothesis that D-arm replacement loops are formedby replication sl ippage versus gradual relaxation or

deletion of bases with in th e D-stem .

25 1PHYLOGENETICS AND BIOGEOGRAPHY OF ANGUID LIZARDS


3/23

MATERIALS AND METHODS

Specimen Information

Museum numbers and localities for voucher speci-mens from which DNA was extracted, and GenBankaccession numbers are presented below. Acronyms are

CAS for Californ ia Academ y of Sciences, San Fr an cisco;MVZ for Museu m of Vert ebra te Zoology, U niversity ofCalifornia at Berkeley; USNM for United States Na-ional Mu seum , Wash ington, DC; UTA-R for Un iversity

of Texas at Arlington; and ZISP for Zoological Institute,St. Petersburg, Russia. Acronyms followed by a dashRM or TP represent field numbers of the first or sixthau th or, respectively, for un cata logued specimens beingdeposited in the Museum of Vertebra te Zoology. Theacronym followed by a dash SBH represents a fieldnumber of S. Blair Hedges for an uncatalogued speci-me n b ein g d e pos it e d in t h e U n i t ed S t a t e s N a t i on a lMu s eu m. T h e t h r e e p r e vi ou s ly r e p or t e d s e qu e n ce shave been extended by 303 bases to include 101 addi-ional amino a cid posit ions of th e ND1 gene, and the

GenBa nk accessions h ave been u pdat ed accordingly. Heloderma suspectum: no voucher, AF085603, pr ob-

ably Arizona. Varanu s griseus: ZISP 19576, U71334(Macey et al., 1997a), east side of Neph tezavodsk wh ichs 30 km WNW of Deynau (39 15 N 63 11E), Chard-

jou Region, Turkmenistan. S hinisaurus crocodilurus:MVZ 204291, AF085604, China. Xenosaurus grand is:MVZ 137789, U71333 (Macey et al., 1997a), slopes

behind Casa de Miguel Ceron, Cuatlapan, Veracruz,Mexico. An niella geronim ensis: MVZ 134196,AF085605,beach, 3.5 miles W of Colonia Guerrero, Baja CaliforniaNorte, Mexico. A n n i el l a p u l ch r a : MVZ-TP24334,AF085606, SW 1/4 Sec. 23, T. 2 N., R. 2 E., sa nd du ne onN side of ra ilroad t ra cks, 0.2 miles SE J ct of Hwy 4 andBig Break Road, Oakley, Contra Costa Co., California.Celestus enneagrammus: MVZ 191045,AF085607, E lev.2125 m, La J oya, Vera cruz, Mexico. Diploglossus biloba-tus: MVZ 207334, AF085608, 3.3 km E r an ch headqu ar -ers at Moravia on road to indian reservation, Prov.

Car ta go, Costa Rica. Diploglossus pleei: MVZ-TP24475,

AF085609, Bosque de Gua jata ca, Vereda Salome, a p-prox. 7 km airline SW Quebradillas (18 24.5 N 6657.8 W), Puerto Rico. Ophiodes striatus: MVZ 191047,AF085610, E do. Sao P au lo, Br a zil. S auresia agasepsoi-des: USN M-SBH194829, AF085611, Bucan Detwi (1744.0 N 71 30.3 W), Peder na les, Dominican Republic.Wetmorena haetiana: USNM 328858, AF085612, 15.3km S, 6.7 km E (by road) Cabral, Barahona, DominicanRepublic. Barisia imbricata: MVZ 191048, AF085613,3070 m, Mex. Hwy 190, Mexico, Mexico. Gerrhonotusiocephalus: UTA-R-12225, AF085 614, 2377 m , E l Tejo-

cote, Oaxa ca, Mexico. Abronia oaxacae: MVZ 144197,

AF085615, Cerro San Felipe, 20 km NN E of Oaxaca (byHwy 175) to La Cumbre then 4 km NW (by dirt road),Oaxaca, Mexico. Mesaspis m oreleti: MVZ 143472,

AF085616, Elev. 9550 ft., 4.5 km by Road E of TodosSantos, Depto. Huehuetenango, Guatemala. Elgariacoerulea: MVZ-TP24365, AF085617, San Pablo Ridge,Wildcat Canyon Road at Inspirat ion Point, Contr aC os t a C o. , C a l ifor n i a . E . k i n gi i: MVZ-RM1192,AF085618, 10.2 m iles NE of Tanque Verde Road on

Cat alina H wy (Mt. Lemm on Rd.), Pim a Co., Arizona. E .paucicarinata: MVZ 191079, AF085619, La Laguna ,Sierra de La Laguna, Baja California Sur, Mexico. E .multicarinata: MVZ 227733, AF085620, Elev. 5700 ft.,NE 1/4 Sec. 16, T. 13 S., R. 34 E., south fork of OakCreek, 5.0 miles west (airline) of Independence, InyoCo., California. E. panam intina: MVZ 227761, U82692(Macey et al., 1997c), Elev. 2030 m, 10.1 miles E of BigPine on Hwy 168, Inyo Co., California. Ophisauruskoellikeri: MVZ 178120, AF085621, 10.1 km S of Keni-tra (34 16 N 6 36 W) on P-29A, Kenitra, Morocco.

Anguis fragilis: MVZ 219518, AF085622, 2 km SE of

Babukal, also 53 km ENE of Dagomys (43 40

N 3938 E) on road, Krasnodarsky Territory, Russia. Ophis-aurus apodus: CAS 182911, AF085623, Tersko-Kum-skaya Nizmennast, 3 km WNW of Voskresenskaya,which is appr ox. 25 km NNW of Guderm es (43 21 N46 06 E), Schelkovskaya District, Chechenia Autono-m ou s R e pu b li c, R u s s ia . O . h a rt i: MVZ 224111,AF085624, Elev. 9001100 m, Tam Dao (21 27 N 10537 E), Vihn Yen District, Vihn Thu Province, Vietnam.O. attenuatus: MVZ-RM10468, AF085625, 2.4 milessouth of Weldon S prin gs at I-40 on H wy 94, St. Ch ar lesCo., Missouri. O. ventralis: MVZ 137541, AF085626,

Sur f City, Pender Co., North Car olina .

Laboratory Protocols

Genomic DNA was extra cted from liver using th eQiagen QIAam p tissue kit. Amplificat ion of genomicDNA featu red a denat ur at ion a t 94C for 35 s, an neal-ing at 50C for 35 s, and extension at 70C for 150 swith 4 s added to the extension per cycle, for 30 cycles.Nega tive contr ols were ru n for a ll am plificat ions. Ampli-fi ed p r od u ct s w er e p u r ifi ed on 2 .5 % N u s i ev e G TGagar ose gels a nd ream plified un der similar conditions.Reamplified double-stranded products were purified on

2.5% acrylamide gels (Maniatis et al., 1982). Templa teDNA was elut ed from a crylamide pa ssively over 3 da yswith Ma niatis elution buffer (Maniat is et al., 1982).Cycle-sequencing r eactions were run using the Pro-mega fmol DNA-sequencing system with a denatur-ation at 95C for 35 s, annealing at 4560C for 35 s,an d exten sion a t 70C for 1 min for 30 cycles. Sequenc-ing rea ctions wer e ru n on Long Ran ger sequen cing gelsfor 512 h at 3840C.

Amplificat ions from genomic DNA used differentprimer combinations (Table 1): (1) L3002-H4419b, (2)L4160-H4980, (3) L4437-H5934, (4) L3878-H4980, (5)

L3881-H5934, (6) L4221-H5934, and (7) L4437-H6564.B ot h s t r a n d s w er e s eq u en ce d u s in g t h e p r ime r s i nT a b l e 1 . P r i me r n u mb e r s r e f e r t o t h e 3 e n d o n t h e

25 2 MACEY ET AL.


4/23

hum an mitochondrial genome (Anderson et al., 1981),where L and H correspond to light and heavy strands,respectively.

Phylogenetic Analysis

DNA sequences were aligned man ually. Protein-coding sequences were t ra nslated to am ino acids usingMacClade (Maddison an d Maddison, 1992) for confir ma -ion of alignment. Transfer RNA secondary structure

w a s d et e r m in ed m a n u a lly u s in g t h e cr it er ia of Kuma zawa a nd N ishida (1993) to ensur e proper align-men t (Macey an d Verm a, 1997). Positions of ambiguous

alignm ent were excluded from phylogenetic ana lysis(see Results).

Phylogenetic trees were estimated using PAUP* betaversion 4.0b1 (Swofford, 1998) with 100 heurist icsear ches feat ur ing ra ndom a ddition of sequen ces. Boot-strap resampling was used to a ssess su pport for indi-vidua l nodes with 1000 bootstr ap replicat es u sing 100heuristic searches with random addition of sequencesper replicate. Decay indices (bra nch su pport ofBremer, 1994) were calculated for all internal brancheso f t h e t r e e u s i n g t w o me t h o d s . F i r s t , 1 0 0 h e u r i s t i csearches with ran dom addition of sequences, which

r e t a in e d s u b op t ima l t r e e s, w er e r u n for n od es w it hdecay indices of 1 to 15. For nodes with decay indicesabove 15, a phylogenetic topology containing the single

n od e in q u es t ion w a s con s t r u ct e d u s in g Ma cC la d e(Maddison and Maddison, 1992) and ana lyzed as a

constra int in PAUP* beta version 4.0b1 (Swofford,

1998) with 100 heurist ic searches featuring ran domaddition of sequences. In th ese searches, trees t ha t did

not contain the imposed constraint were retained. All

searches conducted on allozymic data were exhaustive.Wi lcox on s i gn e d -r a n k s t e s t s (F e l s en s t e i n , 1 9 85 ;

Templeton, 1983) were used to examine statist icalsignificance of the shortest tree relative to alternative

hypotheses. This test asks wh ether the m ost pa rsimoni-

ous tree is significantly shorter than an alternative orwhether their differences in length can be att ributed to

cha nce alone (Larson, 1998). Wilcoxon signed-ranks

tests were conducted both a s one- and t wo-ta iled test s.Felsen stein (1985) showed tha t one-tailed probabilitiesar e close to t he exact probabilities for this test but notalways conservative, whereas the two-tailed test isalwa ys conserva tive. Tests wer e conducted u sing PAUP *beta version 4.0b1 (Swofford, 1998), which incorporatesa corr ection for t ied ran ks.

Alterna tive phylogenetic hypoth eses were tested us-ing th e most p ar simonious p hylogenetic topologies com-

p a t ib le w it h t h e m. To fi n d t h e mos t p a r s imon i ou stree(s) compatible with a particular phylogenetic hy-pothesis, phylogenetic topologies were constructed us-

TABLE 1

P r i m e r s U s e d i n T h i s S t u d y

Huma n position a Gen e Sequ en ce b Reference

L3002 16S 5-TACGACCTCGATGTTGGATCAGG-3 Macey et al., 1997a

L3428 ND1 5-CGAAAAGGCCCAAACATTGTAGG-3 This studyL3878 ND1 5-GCCCCATTTGACCTCACAGAAGG-3 Macey et al., 1998bL3881 ND1 5-TTTGACCTAACAGAAGGAGA-3 Macey et al., 1997aL4160 ND1 5-CGATTCCGATATGACCARCT-3 Kumazawa a nd Nishida, 1993L4178 ND1 5-CAACTAATACACCTACTATGAAA-3 Macey et al., 1997aL4221 t RNAIle 5-AAGGATTACTTTGATAGAGT-3 Macey et al., 1997aH 4419a t RNAMet 5-GGTATGAGCCCAATTGCTT-3 Macey et al., 1997aH 4419b t RNAMet 5-GGTATGAGCCCGATAGCTT-3 Macey et al., 1997aL4437 t RNAMet 5-AAGCTTTCGGGCCCATACC-3 Macey et al., 1997aL4645 ND2 5-ACAGAAGCCGCAACAAAATA-3 Macey et al., 1997aL4882 ND2 5-TGACAAAAACTAGCCCC-3 Schulte et al., 1998H 4980 ND2 5-ATTTTTCGTAGTTGGGTTTGRTT-3 Macey et al., 1997a

L5002 ND2 5-AACCAAACCCAACTACGAAAAAT-3 Macey et al., 1997aH 5540 t RNATrp 5-TTTAGGGCTTTGAAGGC-3 Macey et al., 1997aL5556a t RNATrp 5-AAGAGCCTTCAAAGCCCTAAG-3 Macey et al., 1997a

L5556b t RNATrp 5-GCCTTCAAAGCCCTAAA-3

Macey et al., 1997a

L5617 t RNAAla 5-AAAGTGTCTGAGTTGCATTCAG-3 Macey et al., 1997a

L5638 t RNAAla 5-CTGAATGCAACTCAGACACTTT-3 Macey et al., 1997aH 5692 t RNAAsn 5-TTGGGTGTTTAGCTGTTAA-3 Macey et al., 1997aH 5934a COI 5-AGRGTGCCAATGTCTTTGTGRTT-3 Macey et al., 1997aH 5937a COI 5-GTGCCAATGTCTTTGTG-3 Macey et al., 1997aH 5937b COI 5-AGGGTTCCGATATCTTTRTG-3 This studyH 6564 COI 5-GGGTCTCCTCCTCCAGCTGGGTC-3 Macey et al., 1998a

a Primers are designated by their 3 ends which correspond to the position in the human mitochondrial genome (Anderson et al., 1981) by

convention. H and L designate heavy-strand and light-strand primers, respectively.b Positions with mixed bases are labeled with the standard one-letter code: R G or A.



5/23

FIG. 1. Length-variable regions among the 2101 aligned mitochondrial DNA sequences as used in the phylogenetic analysis. Elevenegions t otaling 100 positions ar e excluded from the an alysis an d ar e under lined. Positions 1360 and 6611600 from th e ND1 and N D2 genes,espectively, are not sh own becau se th is region had no length var iation except for a single codon deletion in Heloderma at positions 211213.

Sequences ar e presented as light-stran d sequence and tRNA secondary stru cture is designated a bove the sequence. Stems are indicated byarrows in the direction encoded: AA, amino acid-acceptor stem; D, dihydrouridine stem; AC, anticodon stem; and T, TC stem. Th e t R NAant icodons are designat ed COD. Asterisks indicate the unpaired 3 tRNA position 73. Periods indicate bases located outside stem regions; 1depicts th e firs t codon position of protein-coding sequ ences. STP indicates stop codons.

25 4


6/23

FIG. 1Continued



7/23

ng MacClade (Maddison and Maddison, 1992) anda n a l yz ed a s con s t r a i n t s u s in g P AU P * b et a ve r s ion4.0b1 (Swofford, 1998) with 100 heuristic searches with

ra ndom a ddition of sequen ces.

Cladistic Analyses of Allozym ic Data

Pr eviously reported allozymic dat a of Good (1987,1988) were coded in two ways for cladistic ana lysis.Although presenceabsence coding of alleles ha s re-ceived considerable criticism for a lack of independenceof alleles and the possibility of no allele being recon-structed for an ancestra l node, it remains t he m ethodha t provides the grea test am ount of resolution. Alter-

natively, combinations of alleles for a particular locusmay be coded as discrete character states (Buth, 1984).

If step mat rices ar e used to connect chara cter sta tes, agreater am oun t of inform at ion can be reta ined (Mabeeand Humphr ies, 1993). In our ana lysis, s tep mat rices

were constructed on the basis of gains and losses of

alleles. F or example, a fixed difference between two

alleles was counted as two steps, one allele lost and

an other gained. If a two-allele polymorphism in onepopulation shares one allele with another monomor-

phic population, a single gain or loss was counted as

one step. Additional polymorphisms were counted inthe same man ner.

In the allozymic data for the Gerrhonotinae (Good,

1987), no outgroup was completely scored. Instead,

alleles found in Abronia, Barisia, a n d Mesaspis t h a t

were present in Elgaria were recorded. When no allele

was r eport ed for a par ticular locus in th e outgroup, we

inferred th at a t least one unique allele was present .

Phylogenetic a nalysis of tRNA stem regions usedMacClade (Maddison and Maddison, 1992). The num-

ber of stem pairings was ordered in these analyses.

FIG. 1Continued

25 6 MACEY ET AL.


8/23

RESULTS

Sequen ces ran ging in size from 2034 to 2061 bases ofmitochondrial DNA for 27 ta xa of an guimorph lizardsar e presen ted a s 2101 aligned positions in Fig. 1.

Authentic Mitochondrial DNAS ev er a l ob se r va t i on s s u gge st t h a t t h e D N A s e-

q u en ce s a n a l yz ed h e r e a r e fr om t h e mi t och on d r ia lgenome an d a re not n uclear-integra ted copies of mito-chondrial genes (see Zhang and Hewitt, 1996). Protein-cod in g ge n es d o n ot h a v e p r e ma t u r e s t op cod on s ,suggesting that these sequences represent functionalcopies th at encode a protein. Tra nsfer-RNA genes ap-pear to code for tRNAs with stable secondary struc-ures, indicating functional genes. The presence of

stra nd bias furth er supports our conclusion th at the 27DNA sequences reported here are from the mitochon-

dria l genome. The sequences report ed here sh ow str ongstr an d bias against gua nine on th e light str an d (G 1114%, A 3036%, T 2 2 29 %, a n d C 2535%),which is cha ra cteristic of the m itochondr ial genome butnot the nuclear genome. See Macey et al . (1997a,c,1998a) for similar strand bias across most squamate-reptile families for the same region of the mitochon-drial genome.

Assessm ent of Homology and S equence Alignm ent

Sequences reported correspond to positions 3874 to5936 on the hu ma n mit ochondr ial genome (Ander son et

al., 1981). Th is sequence conta ins the genes encodingND1 (subunit one of NADH dehydrogenase), tRNAIle ,RNAGln , t R N AMe t, N D 2 , t R N ATrp , t R N AAla , t R N AAsn ,RNACys , tRNATyr , a nd COI (subunit I of cytochr ome c

oxidase), plus th e OL between the tRNAAsn and tRNACy s

genes (Fig. 1). Except for the last couple of codonp os it i on s i n t h e N D 1 a n d N D 2 g en e s a n d a s in g ledeletion of a codon position in the ND1 gene of Helo-derma, no length variation is found in protein-codinggenes, making alignm ent stra ightforwar d. Among pro-ein-coding sequences, only the last few amino acid

positions en coding ND1, t he stop codon, an d n oncoding

s eq u en ce s b et w ee n t h e N D 1 a n d t R N AIle g e n e s a r eexcluded from phylogenetic a na lyses (positions 382407) becau se of consider able length var iat ion. Gaps a replaced in the Heloderma sequence in positions 211213corresponding to codon position 71 of the ND1 genefragment included in this study.

Among tRNA genes, a few loop regions are unalign-able as ar e some n oncoding sequences between genes.Phylogenetic analyses do not include regions encodinghe dihydrouridine (D) and TC (T) loops of th e t RNAIl e

(positions 4214 27, 460471), tRN ATrp (positions 16801689, 17211730), and tRNACy s (positions 19731977,

19351941) genes. Par t of the region encoding t heD-loop of th e t RNATyr gene (positions 20502052) a lso isexcluded.

The tRNACys gene can be par ticularly problemat ic toalign for ph ylogenetic ana lyses becau se gene sequ encesthat do not encode a D-stem can have stem realignmentin the AA- and T-stems (Macey et al., 1997b). Thepreviously published sequence for Varanus griseus(Macey et al., 1997a,b) appears to be homologous to

other sequences analyzed. Most of the bases from theD-arm replacement loop are placed in the excludedD-loop region in the alignment. In addition, the tRNA-Cy s gene sequences from Elgaria kingii a n d E. paucicari-nata appear to have a T base deleted from the regionencoding the D-stem and a gap is placed at posit ion1970. In th e tRNACys gene, the T-stem in some taxa m aybe extended beyond t he n orma l five pair s. The phyloge-netic analysis includes only the five paired positionsnormally observed. Sauresia h a s a n u n u s u a l t R N ACys

in which a T has been inserted in the region encodingth e AA-stem, forcing t hr ee bases between t he AA- an d

D-stems in the encoded tRNA. A gap is placed in al lother taxa at position 1985.

I n t h e t R N AGln ge n e a d ele t ion of a n A fr om t h eAA-stem at position 486 appears to have occurred inVaranus. T h is d el et i on h a s r e s u lt e d i n a r e a li gn e dT-stem. In the phylogenetic analysis a gap is placed int h e Varanus sequence at position 492, and the Varanussequen ce is aligned to the seconda ry str uctu re observedin the other taxa.

S eq u en ce s b et w ee n t h e t R N ATr p a n d t h e t R N AAla

genes and between the tRNACy s and the tRNATyr genes(positions 17441745 a nd 19921998, r espectively) are

not used in the phylogenetic analyses. The loop regionof the replication origin for the light strand is mostlyun alignable a nd ther efore not u sed (positions 19021912).

Among the 2101 aligned posit ions only 100 sites,constitu ting less th an 5% of the aligned sequences, areexcluded from th e phylogenetic an alyses.

Variation of Stems in tRNA Cys

Tremen dous var iation in stem lengths occur s a mongt h e 2 7 t R N ACy s gene sequences reported here (Fig. 2)

and in Macey et al. (1997b). Both th e D- an d T-stemsshow var iation for t he n um ber of base pairings in st emr e gion s t h a t d evi a t e fr om t h e t y pica l fou r p a ir s i nD-stems and five pairs in T-stems.

In Varanus, tRNACy s is known to lack a D-stem andinstead contains a D-arm replacement loop (Macey etal., 1997b). This structure has been postulated to resultfrom slipped-stra nd mispairing of noncontiguous r e-peats during replication (Fig. 2; Macey et al., 1997b).The number of pairings in the D-stem varies amongother taxa between one and six. Ophisaurus koellikerihas an unusua lly large six-base D-stem that conta ins

two extra pairs. Three other taxa, Xenosaurus a n d t h et wo Anniella species, have five-base D-stems, henceconta ining one extra pair. Most t axa (Heloderma, S au-



9/23

FIG. 2. Potential secondary stru ctures derived from 27 tRNACys gene sequences presented in F ig. 1. A stan dard tRNA with a four-base

D-stem and a five-base T-stem is depicted firs t, where R G or A, Y C or T, and V G, C, or A (after Ku mazawa and Nishida, 1993). Lndicates the three loop regions where length variation is standardly observed. Varanus lack s a D- stem an d in stead co n tain s a D- ar meplacement loop. Bases boxed represent thr ee potential noncontiguous repeats postulat ed t o have r esulted from slipped-stra nd mispairing

dur ing replication (Macey et al., 1997b). Sauresia h as an u n u su a l tR NACys in which an Ah as been inser ted in th e AA-stem, forcing thr ee basesbetween t he AA- an d D-stems inst ead of th e two bases norma lly observed. Note the t remen dous variat ion in sizes of both D- and T-stems. Theposition where an Awas deleted destroying the D-stems in Elgaria kingii a n d E. paucicarinata is indicated.

25 8


10/23

resia, Wetmorena, Barisia, Gerrhonotus, Abronia, Me-saspis, Anguis, a n d a l l Ophisaurus species except O.koellikeri) h a v e t h e n or ma l fou r b a s e p a ir i n gs in t h eD-stem. Eight taxa (Shinisaurus, Celestus, Diploglos-sus bilobatus, D. pleei, Ophiodes, Elgaria coerulea, E.

multicarinata, a n d E. panamintina) have a redu ction ofone pair to produce D-stems composed of three basepairings. Int erestingly, D. pleei a n d E . p a n am i n t i n ahave only two bases in the D-loop, which is less than

h e mi n ima l t h r e e b a s es r e qu i r ed t o r e s t or e t h r ou g hbase substitutions a fourth D-stem base pairing whileret ain ing at leas t one D-loop base.

Only Elgaria kingii a n d E. paucicarinata a r e o b -s e r v e d t o h a v e a s i n g l e b a s e p a i r i n g i n t h e D - s t e m.These taxa have only six bases in the D-loop, which is

le ss t h a n t h e m in im u m of s eve n b a se s r e qu ir e d t or e s t or e t h r ou g h b a s e s u b st i t u t ion s t h r e e a d d it i on a lbase pairings in the D-stem while retaining at least oneD-loop base. F rom compa rison with t he other tRNACys

gene sequences, it appears that a T encoding an Ain thetRNA is deleted, destroying the D-stem. In addition, nodirect repeats or noncontiguous repeats are observed;such r epeats ar e implicat ed, however, in th e form at ionof D-arm replacement loops among eight other lepido-sau rian t axa (Macey et al., 1997b).

Striking variation is observed also in the number ofb a s e p a ir i n gs a mon g T-s t e ms i n t R N ACy s g en e s e -

quences (Fig. 2). Only two taxa, Diploglossus pleei a n dSauresia, h a v e t h e t y p i c a l fi v e b a s e p a i r i n g s i n t h eT-stem. Instead, a surprising situation occurs in which

FIG. 2Continued



11/23

T-s t ems a r e e it h e r r e d u ce d b y o n e b a s e p a ir i n g t oproduce four pairs, or lengthened to as many as eightpairs. The five Elgaria species, Ophisaurus koellikeri,a n d O . h a r ti h a v e T-s t e ms t h a t h a v e l os t a s in g lepairing, producing a T-stem with four base pairings.Heloderma, Varanus, Anniella, Celestus, Ophiodes, a n d

Abronia a l l s h ow a n a d d it i on a l p a ir , p r o du cin g a nncrease in size of the T-stem to six base pairings.

S hin isaurus, Diploglossus bilobatus, Wetm orena, Bari-

sia, Gerrhonotus, Mesaspis, Anguis, Ophisaurus apo-

dus, O. attenuatus, a n d O. ventralis all show two extr abase pairings to produce T-stems seven pairs in length,a n d Xenosaurus has three extra pairs to produce eightbase pairs in the T-stem. Note that with the exceptionof Heloderma, Varanus, a n d Sauresia, a l l t a x a t h a tcontain less than seven base pairings in the T-stemcontain T-loops large enough to produce additionalpairs through base substi tutions that would result in

enlar ged T-stems of seven base pa irs.

Genic Variation

Different levels of DNA substitutiona l variation ar eobserved among the 3 protein-coding genes, 8 tRNA-coding genes, and four noncoding regions (Table 2). All11 genes contain phylogenetically informative charac-ers. The 8 tRNA genes each have phylogeneticallynformative sites in stem and nonstem regions. Each ofh e 3 protein-coding gen es cont ains phylogenetic infor-

ma tion in fir st, second, a nd t hird codon positions. Mostof the variation and phylogenetically informative sites

are from protein-coding regions. Only 23% of variablean d 21% of phylogenetically inform at ive sites a re fromRNA genes a nd n oncoding r egions. Of the 802 ph yloge-

netically inform ative cha ra cters from protein-codingregions, 448 are from third positions of codons. Third-position sites a ccount for slightly less th an ha lf of th ephylogenetically informative sites in the total data set.Only 150 phylogenetically inform at ive sites occur inr e gi on s e n cod in g s t e ms of t R N As , s u gg es t in g t h a tcompensatory substitutions do n ot compromise thephylogenetic an alysis.

Phylogenetic RelationshipsTw o t r e e s of e qu a l le n gt h a r e p r od u ce d fr om t h e

parsimony analysis of the 2001 aligned DNAsequencescontaining 1013 phylogenetically informative base posi-ions (Fig. 3, Table 2). Phylogenetic relationships are

well resolved for most nodes of the tree. All ingroupa xa ar e grouped to the exclusion of the Var an oidea (th e

New World Heloderma and the Old World Varanus)with good support (bootstrap 82%, decay index 12). TheOld World Shinisaurus and the New World Xenosau-rus, often grouped as the Xenosauridae, appear not toform a monophyletic group. Instead, Shinisaurus is

excluded from a monophyletic group containing Xeno-saurus, the Anniellidae, and the Anguidae (bootstrap91%, decay index 18). A monophyletic grouping of

Anniella and the Anguidae receives considerable sup-port (bootstrap 98%, decay index 29). A monophyletic

Anniella (bootstrap 100%, decay index 51) forms thesister t axon to the Anguidae, which appear s m onophy-letic but with wea k su pport (decay index 3).

Within the Anguidae, the Diploglossinae appears

monophyletic with weak support (decay index 4) andfor ms t h e s is t er t a x on t o a cla d e comp os ed of t h eGerrhonotinae and Anguinae (bootstra p 80%, decayindex 6). Monophyly of both the Gerrhonotinae (boot-strap 100%, decay index 22) and Anguinae (bootstrap100%, decay index 40) receives st rong su pport.

In the Diploglossinae, Celestus a n d Diploglossusbilobatus from m ainland Mexico and Centr al Americaform a monophyletic group with good support (boot-str ap 93%, decay index 13). A second clade compr isingWest Indian and South American taxa, Diploglossus

pleei, Ophiodes, Sauresia, a n d Wetmorena (bootstrap

82%, decay index 9) can be recognized in the Diploglos-sinae. Two well-supported groups are observed withinthis clade: (1) the Puerto Rican D. pleei a n d S o u t hAmerican Ophiodes (bootstrap 100%, decay index 47),and (2) the Hispaniolan Sauresia a n d Wetm orena (boot-str ap 100%, decay index 102).

Two clades are recognized in the Gerrhonotinae, oneconta ining t he largely Neotr opical Barisia, Gerrhono-tus, Abronia, a n d Mesaspis (bootstrap 69%, decay index4), and the other composed of the five species from themore temperat e North American genus Elgaria (boot-str ap 100%, decay index 39). In t he t ropical group, only

the sister-taxon relationship of Abronia a n d Mesaspisis recovered with strong support (bootstrap 99%, decayindex 17).Among t he five species ofElgaria, all bra nchesar e well supported. Elgaria coerulea, the most nort herngerrhonotine, is th e sister t axon to a group compr isingthe remaining species (bootstrap 100%, decay index29). The more eastern species, E. kingii, is the sisterta xon to a monophyletic group ofE. paucicarinata, E.multicarinata, a n d E . p a n a m i n t i n a (bootstr ap 94%,decay index 9). Elgaria multicarinata a n d E . p a n a -mintina from Californ ia form a monophyletic group(bootstrap 98%, decay index 8).

I n t h e An g u in a e , mos t r e la t i on s h ip s a r e n ot w el lsupported. Ophisaurus koellikeri from Morocco formsthe sister taxon to the remaining species (decay index1). A monophyletic grouping of the western EurasianOphisaurus apodus a n d Anguis fragilis is well sup-ported (bootstrap 100%, decay index 24) and forms thesister taxon to a weakly supported clade composed oft h e E a s t A s i a n O. harti and the North American O.attenuatus a n d O. ventralis (decay index 1). NorthAmerican Ophisaurus appear monophyletic with mod-era te su pport (bootst ra p 77%, decay index 7).

Phylogenetic relationships among the Anguidae, An-

niellidae, an d Xenosau rida e resolved from r ean alysis ofallozymic data (Good, 1987, 1988) are largely th e sa mewhether an alyzed with allelic combinat ions as char ac-

26 0 MACEY ET AL.


12/23

e r s t a t e s u s in g s t e p ma t r i ce s or p r e se n ce /a b s en cecoding of alleles (Fig. 4). In the survey of higher-levelaxa (Good, 1987), a topology is acquired that is com-

pletely concorda nt with the DNA sequence da ta (Figs.4 A a n d 4 B). W h en t h e d a t a a r e cod e d u s i n g a l le liccombinations as character states and analyzed withstep ma tr ices, only t wo loci provide phylogenetic infor-ma tion. Monophyly of th e Anguida e (Celestus, Elgaria,a n d Ophisaurus) is supported with a bootstrap value of

79% and a decay index of 2, and monophyly of a groupcomposed of th e Gerr honotinae (Elgaria) and Anguinae(Ophisaurus) is supported by a bootstrap value of 72%an d a decay index of 1. The sam e phylogenetic relat ion-ships r esult from presen ce/absen ce coding of individua lalleles. The 20 informative alleles in the latter analysisprovide considerably better support. Monophyly of theAnguidae is supported with a bootstrap value of 99%a n d a d e ca y in d e x of 9 , a n d mon op h yl y of a gr ou pcontaining the Gerrhonotinae and Anguinae is sup-ported by a bootstrap value of 96% and a decay indexof 3.

The phylogenetic relationships between the five E l-garia species resolved from reanalysis of allozymic data(Good, 1988) are in conflict with the result of the DNA

ana lysis (Figs. 4C a nd 4D). When the data are codedu s in g a l le li c c omb in a t i on s a s ch a r a c t er s t a t e s a n danalyzed with step matrices, five loci provide phyloge-netic inform at ion. This a na lysis pr oduces two equa llymost parsimonious trees in which the phylogeneticrelationships of E. coerulea a n d E. multicarinata r e-ma in unr esolved. Monophyly of a group conta ining E .

paucicarinata , E. k ingii, a n d E. panamintina is weaklysupported (bootstrap 66%, decay index 1). Monophyly

of a group containing E. kingii a n d E . p a n am i n t i n areceives bett er support (bootstr ap 80%, decay index 2).A similar phylogenetic tree is resolved from presence/absence coding of individual alleles. The 23 informativealleles in th e lat ter an alysis pr ovide considerably bet-ter support . In this analysis, E. coerulea is excludedfrom a monophyletic group composed of E. multicari-nata, E . paucicarinata, E. k ingii, a n d E. panamintina(bootstrap 84%, decay index 3). Monophyly of a groupcontaining E. paucicarinata, E. kingii, a n d E . p a n a -m i n t i n a is not well supported (bootstrap 69%, decayindex 1) but the grouping of E. kingii a n d E . p a n a -

m i n t i n a receives good support (bootstrap 98%, decayindex 6). Disagreement between analyses of the DNAsequence data and allozymic data for Elgaria species

TABLE 2

D i str i b u ti o n o f P h y l o g e n e ti c a l l y In fo r m a ti v e a n d V a r i a b l e P o si ti o n s

ND1 Codonposit ion s t RNAIle a tRNAGln tRNAMet

ND2 Codonpositions

1st 2n d 3r d St em Non -st em St em Non -st em St em Non -st em 1st 2n d 3r d

n for m a t ive sit es 4 4 22 118 19 6 18 10 13 6 188 96 321Va r ia ble sit es 59 34 125 25 7 28 16 19 9 236 150 340

Noncoding b

region 1

tRNATrp a tRNAAla

Noncoding b

region 2

tRNAAsn

St em Non -st em St em Non -st em St em Non -st em

n for m a t ive sit es 1 23 2 19 8 1 13 7Va r ia ble sit es 2 28 5 30 12 1 16 10

Noncoding b

region 3

tRNACys a tR NATyr c

Noncoding b

region 4

COI Codon positions

St em Non -st em St em Non -st em 1st 2n d 3r d

n for m a t ive sit es 26 7 19 13 3 1 9Va r ia ble sit es 2 30 8 29 15 1 5 2 9

Total

P rot ein -codin g codon posit ion s t RN A codin gNoncoding

regionsAll aligned

sequence1st 2n d 3r d St em Non -st em

n for m a t ive sit es 235 119 448 150 59 2 1013Va r ia ble sit es 300 186 474 205 82 6 1253

a Not including D- and T-loops wh ich wer e excluded from t he a nalyses.b Noncoding region 1 includes the ND2 stop codon and sequences between the ND2 and the tRNA Trp genes. Noncoding r egion 2 is between

h e tR NAAla an d th e tR NAAsn genes. Noncoding region 3 is between t he t RNAAsn gene and the OL . Noncoding region 4 is between t he t RNATyr

and t he COI genes.

c Not including par t of the D-loop, which was excluded from th e an alyses.



13/23

occurs regarding th e r elative positions ofE. k ingii, E.m ulticarinata, E. panam intina, a n d E. pau cicarinata.

The ph ylogenetic results provide a n ar ea cladogramfor t he H olar ctic region. To confir m t hese r esults a nd t oest support for the origins of clades foun d in separa te

historical regions, t he Wilcoxon signed-ran ks t est (Fel-

senstein, 1985; Templeton, 1983) is applied (Table 3).(1). Current taxonomy places Shinisaurus a n d Xeno-

saurus in a single family, th e Xenosauridae. When thewo shortest trees overall (A12 in Appendix 1) showing

a nonmonophyletic Xenosauridae are compared to theshortest alternative trees (B14 in Appendix 1) show-ng a monophyletic Xenosauridae, this al ternative is

rejected in favor of the overall shortest tr ees by eitherhe one-tailed or the two-tailed test (test 1 in Table 3).

T h is r e su lt s u gge st s t h a t t h e Xe n os a u r id a e is n otmonophyletic.

(2). The two overall shortest trees from analysis of

he DNA sequence data (Fig. 3) show that the Annielli-dae, Anguida e, Diploglossina e, Gerrh onotina e, an d An-guina e each form monophyletic groups. In a ddition, theoverall shortest tree from analysis of allozymic data

also shows m onophyly of the Anguidae. When the twoovera ll shortest tr ees (A12 in Append ix 1) from a na ly-sis of the DNA sequence data, which show a mono-p h yl et i c A n n ie lli da e , a r e comp a r e d t o t h e s h or t e s talternative trees (C12 in Appendix 1) having a non-monophyleticAnn iellidae, this alt ern at ive is rejected infavor of the overall shortest trees by the two-tailed test(test 2 in Table 3). Because the decay index on thebranch leading to a monophyletic Anguidae is only 3from ana lysis of the DNA sequence data , t his bra nchcan not receive stat istical support from th e Wilcoxonsigned-ranks test which requires at least 4 unopposed

char acter s to be significan t (Felsenst ein, 1985).When the overall shortest al lozymic tr ee (Fig. 4B,

allele pres ence/absence coded), which sh ows a monophy-

F IG . 3 . Phylogenetic relationships based on DNA sequences.Strict consensus of two equally most parsimonious trees produced

from ana lysis of th e 2001 aligned (1013 phylogenetically informa tive)positions. The tr ee has a length of 5452 steps an d a consist ency indexof 0.394. Bootstrap values ar e present ed a bove branches and decayndices below bra nches.

FIG. 4. Phylogenetic trees from analyses of allozymic data fromthe literature (Good, 1987, 1988). Bootstrap values are presented

above branches and decay indices below branches. (A) The mostparsimonious tree from our ana lysis of Goods (1987) dat a usingallelic comb in ation s as ch ar acter states an d an aly zed with stepma tr ices. Two of the 22 loci ar e ph ylogenetically inform at ive. The tr eehas a length of 91 steps and a consistency index of 0.100. (B) The mostpar simonious tr ee from our an alysis of Goods (1987) data coded bythe presence or absence of the 72 (20 informative) individual alleles.

The tr ee has a length of 76 steps a nd a consist ency index of 0.921. (C)Str ict consensus t ree of two equally most pa rsim onious t rees from ouran alysis of Goods (1988) data using allelic combinations as chara cters t a t e s a n d a n a l yz ed w it h s t e p m a t r i ce s . F i ve of t h e 1 8 l oci a r ephylogenetically informative. The tree has a length of 88 steps and aconsist ency index of 0.955. (D) The most par simonious tr ee from our

an alysis of Goods (1988) da ta coded by th e pres ence or abs ence of th e67 (23 informative) individual alleles. The tree has a length of 75steps an d a consistency index of 0.867.

26 2 MACEY ET AL.


14/23

TABLE 3

R e su l ts o f Wi l c o x o n S i g n e d -R a n k s Te sts

Alt er n a t ive h ypot h eses t est ed Tr eesa Nb Zc P d

1. Mon oph yly of Xen osa u r ida e A1 vs B1A1 vs B2

17 314 9

1.791.97

0.0370*0.0244**

A1 vs B3 93 2.59 0.0048**

A1 vs B4 119 2.14 0.0164**A2 vs B1 148 2.00 0.0230**A2 vs B2 173 1.84 0.0330*A2 vs B3 115 2.27 0.0115**A2 vs B4 93 2.59 0.0048**

2. Non m on oph yly of An n iellida e A1 vs C1

A1 vs C2

18 8

20 7

3.63

3.49

0.0002**

0.0003**A2 vs C1 159 3.97 0.0001**A2 vs C2 180 3.71 0.0001**

3. Nonmonophyly ofAnguidae e Fig. 4B vs D1 13 2.50 0.0063**4. Non mon oph yly of Diploglossin a e A1 vs E 1

A2 vs E1

12 5

93

0.35

0.41

0.3619

0.34165. Non mon oph yly of Ger r h on ot in a e A1 vs F 1

A1 vs F 27268

2.592.67

0.0067**0.0055**

A2 vs F 1 36 3.67 0.0001**A2 vs F 2 36 3.67 0.0001**

6. Non mon oph yly of An gu in a e A1 vs G1A1 vs G2

18 718 8

2.822.83

0.0043**0.0054**

A1 vs G3 193 2.73 0.0031**A2 vs G1 156 3.09 0.0010**A2 vs G2 159 3.04 0.0012**A2 vs G3 169 3.00 0.0014**

7. Anniella a s t h e sist er t a xon t o t h e An gu in a e A1 vs H 1A2 vs H1

11989

1.381.59

0.08460.0559

8. Anniella a n d A ngui s for m a m on oph ylet ic gr ou p A1 vs I1A1 vs I2A2 vs I1A2 vs I2

22 820 420 423 5

6.016.296.365.90

0.0001**0.0001**0.0001**0.0001**

9. Ophiodes a n d Ophisaurus koellikeri for m a m on oph ylet ic gr ou p A1 vs J 1A2 vs J1

28 930 8

8.798.62

0.0001**0.0001**

10 . Diploglossus pleei, S auresia, a n d Wetm orena form a m onophyleticgroup

A1 vs K1A1 vs K2A2 vs K1A2 vs K2

81114111

81

5.674.604.845.67

0.0001**0.0001**0.0001**0.0001**

11. Ophisaurus koellikeri, Anguis fragilis a n d Ophisaurus apodus form a

monophyletic group

A1 vs L1

A1 vs L2A1 vs L3A2 vs L1A2 vs L2

A2 vs L3

86

62315731

61

0.74

0.871.260.931.26

0.90

0.2291

0.19270.10440.17690.1044

0.185112 . A ngui s a n d Ophisaurus apodus a r e n ot sist er t a xa A1 vs M1

A2 vs M1

80

112

2.68

2.21

0.0037**

0.0135**13 . Elgaria kingii a n d E . panam i nt i na form a m onophyletic sister group to

E. paucicarinata

A1 vs N1A2 vs N1

3668

4.333.15

0.0001**0.0001**

14 . Elgaria mu lticarinatae a n d E . panam i nt i na form a m onophyletic sistergroup to E. paucicarinata

Fig. 4D vs O1 11 2.71 0.0036**

a See Appendix 1 for phylogenetic topologies used in tests.b Num ber of chara cters differing in minimu m nu mbers of cha nges on paired t opologies.c Norma l appr oximat ion for Wilcoxon signed-ra nks test .d Asterisk indicate a significant difference between the overall shortest tree and an alternative tree. One asterisk denotes significance using

he one-tailed pr obability only an d two ast erisks denote significance u sing t he two-tailed pr obability for the Wilcoxon signed-ran ks test.One-tailed probabilities are shown and two-tailed probabilities are double these values. A significant result means that the alternativehypothesis as stated can be rejected.

e Tests using allozymic data; all other tests are done on DNAsequence data.



15/23


16/23

etic Anguidae, is compa red t o the shortest altern ativeree (D1 in Appendix 1) having a nonmonophyletic

Anguidae, this alternative is rejected using the allelepresence/absence coded data (Good, 1987) in favor ofhe overall shortest tree by the two-tailed test (test 2 in

Table 3).

When the two overall shortest trees (A12 in Appen-dix 1) from ana lysis of the DNA sequence data , whichshow a monophyletic Diploglossinae, ar e compa red tohe shortest altern at ive tr ee (E1 in Appendix 1) show-ng a nonmonophyletic Diploglossinae, this alternative

cannot be rejected in favor of the overall shortest trees(test 4 in Table 3). When the two overall shortest trees(A12 in Appendix 1) from ana lysis of the DNA se-q u en ce d a t a , w h ich s h ow a mon op h yl et i c G e r r h o-notinae, are compared to the shortest alternative trees(F12 in Appendix 1) showing a nonm onophyletic Ger-rhonotinae, this alternative is rejected in favor of the

overall shortest trees by the two-tailed test (test 5 inTable 3). When the two overall shortest trees (A12 inAppendix 1) from an alysis of th e DNA sequence da ta ,which show a monophyletic Anguinae, are compared tohe shortest al ternative tr ees (G13 in Appendix 1)

showing a nonmonophyletic Anguinae, this alternatives rejected in favor of the overall shortest trees by thewo-tailed test (test 6 in Table 3). Hence, the Annielli-

dae,Anguidae, Gerrhonotinae, and Anguinae each formmonophyletic groups with sta tistical support, but sta tis-ical support is lacking for monophyly of the Diploglos-

sinae.(3). Th e ph ylogenet ic position ofAnniella is a point ofd is a gr e e men t . Wh e n t h e t w o o ve r a ll s h or t e s t t r e e s(A12 in Appendix 1), which show Anniella as th e sisteraxon to the Anguidae, are compared to the shortest

alternative tree (H1 in Appendix 1) showing Anniellaas the sister taxon to the Anguinae, this al ternativecannot be rejected in favor of the overall shortest trees,but it is close to significance using the one-tailed test(test 7 in Table 3). When the two overall shortest trees(A12 in Appendix 1), which show Anniella as th e sisteraxon to the Anguidae, are compared to the shortest

alternative t rees (I12 in Appendix 1) showing thehypothesis of Keqin and Norell (1998) that Anniella ishe sister ta xon to Anguis, th is altern at ive is rejected in

favor of the overall shortest tr ees using t he two-ta iledest (test 8 in Table 3). Anniella is unlikely to representhe sister taxon to either Anguinae or Anguis.

(4). Only Ophisau rus k oellikeri from Morocco an dOphiodes from South America are endemic to Gond-wana n continen ts. When th e two overall short est tr ees(A12 in Appendix 1), in which O. koellikeri a n dOphiodes do not form a monophyletic group, ar e com-par ed to the shortest altern at ive tr ee (J 1 in Appendix 1)

showing them a s sister ta xa, this altern at ive is rejectedn favor of the overall shortest t rees using the two-ailed test (test 9 in Table 3). Taxa inhabiting Gond-

wanan plates therefore do not share a common Gond-wana n origin.

(5). West Indian taxa appear not to form a monophy-letic group. When the two overall shortest trees (A12in Appendix 1), in which West Indian taxa (Diploglos-sus pleei, S auresia, a n d Wetm orena ) d o n o t for m a

mon op h yl et i c g r ou p , a r e comp a r e d t o t h e s h or t e s ta l t er n a t i ve t r e es (K 1 2 in Ap p en d ix 1 ) s h ow in g amonophyletic grouping of th ese ta xa, th is alter na tive isrejected in favor of the overall short est t rees u sing thet w o-t a i le d t e st (t e st 1 0 i n Ta b le 3 ). T h es e r e s u lt sindicate that the South American genus Ophiodes isderived from th e West In dies.

(6). The three Old World taxa that occur between theAtlantic Ocean and the Tibetan Plateau, Ophisauruskoellikeri, Anguis fragilis, a n d O. apodu s, ar e found notto form a monophyletic group. When the overall short-est trees (A12 in Appendix 1), in which Ophisaurus

koellikeri, Anguis fragilis, a n d O. apodus do not form amon op h yl et i c g r ou p , a r e comp a r e d t o t h e s h or t e s talternative trees (L13 in Appendix 1) showing thesespecies as a monophyletic group, this alternative costsseven steps but can not be rejected in favor of the overa llshortest trees (test 11 in Table 3). The most parsimoni-ous t rees suggest tha t t he history of anguine lizards inwestern Eurasia and Morocco is older than anguinehistory in North America, contra ry to th e alternativehypothesis grouping taxa from western Eurasia andMorocco as closest relatives. Anguis fragilis a n d O.apodus for m a mon op h yle t ic g r ou p . W h en t h e t w o

overall shortest t rees (A12 in Appendix 1), whichgroup A. fragilis a n d O. apodus a s s i s t e r t a x a , a r ecompa red to the sh ortest altern ative tree (M1 in Appen-dix 1) in which these species a re not sister taxa, t hisalternative is rejected in favor of the overall shortesttrees using the two-tailed test (test 12 in Table 3). Ascurrently recognized, Ophisaurus is n ot monophyletic.

(7). The only point of disagreement between thean alyses of the DNA sequence da ta an d a llozymic datais the rela tive placemen t ofElgaria kingii, E. mu lticari-nata, E. panam intina, a n d E. paucicarinata. When theovera ll shortest DNA tr ees (A12 in Appendix 1), which

show E. m ulticarinata a n d E. panamintina as a mono-phyletic sister group to E. pau cicarinata, ar e compa redt o t h e s h or t e s t a l t er n a t i ve t r e e (N 1 in Ap p en d ix 1 )showing E. kingii a n d E. panamintina as a monophy-letic sister gr oup t o E. pau cicarinata , this altern at ive isr e je ct e d b y t h e D N A se qu e n ce d a t a i n fa v or of t h eoverall short est t rees using t he t wo-ta iled test (test 13in Table 3). When the shortest allozymic tree (Fig. 4D,allele presence/absence coded) showing E. kin gii a n d E .

panamintina a s a mon op h yl et i c s is t e r g r ou p t o E .paucicarinata is compared to the shortest al ternativetree (O1 in Appendix 1) showing E. multicarinata a n d

E. panamintina as a monophyletic sister group to E .paucicarinata, this a lterna tive is rejected by th e allelepresence/absence coded data (Good, 1988) in favor of



17/23

he overall shortest tree using th e two-tailed test (test14 in Table 3). The DNA sequence an d allozymic dat aare in conflict with regard to the relative grouping ofhe species Elgaria kingii, E. multicarinata, E. pana-

mintina, a n d E. paucicarinata. Two explanations forhis discorda nce can be given. F irst, it is possible th at

he mitochondrial genome h as under gone lineage sort -ng (Pamilo and Nei, 1988) and is misleading phyloge-

netically. Alterna tively, the allozymic data of Good(1988) ma y be misleading because of small sa mple size;h e sam ple size for E. pau cicarinata is four individuals,

E. kingii is two individuals, and for E. panamintina isonly one individual. These sample sizes are not ad-equate for evaluating occurrence of alleles in a popula-ion or a species. Further work is needed to confirm our

results inferred from mitochondrial DNA sequences,but we suggest that this hypothesis of phylogeny is themore reliable estimat e.

DISCUSSION

Phylogeny and Biogeography of Extant Form s and the

Fossil Record

Anguid lizards a re inferred t o have originat ed in th enorthern hemisphere. Our data considered in l ight of biogeographic and paleontological evidence clearly re-ect a Gondwan an origin for the Anguidae. No a nguid

fossils are known from tectonic regions of Gondwananorigin. Only two t axa occur exclusively on separa te

Gondwanan continents, Ophiodes in South Americaa n d Ophisaurus koellikeri in Morocco (Fig. 5). A sist ergroup relationship between these taxa is statisticallyrejected, indicating t hat they do n ot share a commonGondwanan origin. The molecular phylogenetic analy-sis places these taxa in different clades of Laurasianorigin in the northern hemisphere. Ophiodes is nestedwithin West Indian diploglossines and monophyly ofWest Indian diploglossines is sta tistically r ejected,ndicating that Ophiodes descends from a lineage that

origina ted in th e West In dies and subsequently movedo South America. Ophisaurus koellikeri appears t o be

h e s is t er t a xon t o t h e r e m a in in g m e m be r s of t h eAnguinae. All other anguines occur in Eur ope, WestAsia, Ea st Asia, or Nort h Amer ica, which consist pr ima r-ly of Laura sian plates. Becau se other a nguids, ann iel-i ds , a n d Xenosaurus occu r i n N or t h Ame r ica , t h e

Anguinae is n ested within a clade of norther n form s ofLau ra sian origin (Fig. 5).

The opening of the Atlantic Ocean in the late Eocene[50 m illion years before present (MYBP)] may ha veproduced the divergence between the Anguinae and theGerrhonotinae, which are sister taxa. The location ofhe most basa l an guine lineages in Morocco and west-

ern Eura sia supports this explana tion. Miocene cli-matic changes and montane uplifting in North Americama y ha ve separat ed th e two major clades of gerrh ono-

tines, one primarily tropical and the other primarilytemperate.

Phylogenetic relationships (Fig. 3) are surprisinglywell resolved for branches ranging from the late Creta-ceous (9575 MYBP) to the P leistocene (1.5 MYBP) an dprovide insights for interpreting biogeographic andpaleontological data. Two groups of fossil lizards occur-ring in Eu rope have been referr ed to the Anguidae. Theextinct Glyptosaurinae dates to th e late Cretaceous

(9575 MYBP; Gauthier, 1982), and its phylogeneticposition is not well under stood. Fossils in Eu rope fromtwo later periods ma y be related to extant form s withinthe Anguinae. The earliest fossils from the middleEocene (4050 MYBP) ar e either grouped with themodern Old World anguines (Meszoely and Haubold,1975) or considered the sister l ineage to al l modernanguines (Gauthier, 1982). If these fossils are groupedwith a modern Old World angu ine lineage (Fig. 3), th eyplace modern anguine l ineages in Europe during theEocene pr ior to th e Oligocene drying of th e Tur gai Sea .

Our ph ylogenetic tr ee and int erpr eta tion of hist orical

events suggest rapid separation of the western Eur-a s ia n a n g u in e li n ea g es fr om e a s t er n E u r a s i a n a n dNorth American a nguines by the Tibetan uplift du ring

F IG . 5 . Ar ea clad og r am for an g u id l izar d s an d r elated tax a.Taxonomy is shown to the right. Note deep divergences in NorthAmer ica with a d isp er sal ev en t fr om th e West I n d ies t o So u thAmerica and a nested position for Old World anguines. The shortest

estimate of phylogeny suggests that the formation of the AtlanticOcean sep ar ated th e Ger r h on o tin ae an d An g u in ae. Fo llowin g th eOligocene drying of th e Turgai Sea , dispersa l of an guine lizards fromEu r o p e to East Asia an d acr o ss th e B er in g lan d b r id g e to No r thAmerica was possible, but would have been interrupted rapidly byt h e u p li ft of t h e T ib et a n P l a t ea u . I t cos t s s ev en s t ep s on ou r

phylogenetic estimate to construct a topology compatible with an-guine ta xa originat ing in Nort h America an d crossing the Bering land

bridge to East Asia, with cont inued disper sal t o Europe following theOligocene drying of the Turgai Sea .

26 6 MACEY ET AL.


18/23

h e OligoceneMiocene. Within th e Anguina e, fossilsh a t a r e a s s i g n e d t o t h e AnguisOphisaurus apodus

cl a de a p p ea r i n E u r op e fi r s t i n t h e la t e O li goce n e(2530 MYBP; Gaut hier, 1982). This dat e coincideswith the connection of Europe with AsiaAmericafollowing drying of the Turgai Sea (Briggs, 1987). At

his time a continuous land connection was availablefrom Eur ope through Asia to North America. At thissam e time (30 MYBP), the fir st ph ase of uplifting of th eTibetan Plateau was coming to a close with the plateaureaching an average elevation of 3000 m (Dewey et al.,1989). The second phase of uplifting maintained thiselevation of 3000 m unt il 10 MYBP wh en fau lting an du p li ft i n g of t h e T ib et a n P l a t e a u e xce ed e d e r os ion(S h a ck le t on a n d C h a n g, 1 98 8), r e s u lt i n g i n a t h i r dphase of uplifting to produce an average elevation of5000 m (Dewey et al., 1989). By th e late Oligocene toearly Miocene, taxa in Europe and western Asia prob-

ably were isolated from t axa in easter n Asia an d NorthAmerica after sharing a brief connection immediatelyfollowing th e dr ying of the Tur gai Sea .

T h e a n c e s t r a l a n g u i n e l i n e a g e ma y h a v e e n t e r e dE u r op e fr om N or t h Ame r ica i n t h e E oce n e p r ior t oform at ion of the North Atlantic and then expanded itsdistribution southwar d to North Africa and eastwardnto West Asia. The drying of the Turgai Sea in the

O lig oce n e cou l d h a v e p er mi t t ed t a x a t o mi gr a t e t oeastern Asia and then back to North America via theBering land bridge. The Oligocene-to-Miocene upliftingof Ti be t w ou l d h a v e for me d a b a r r ie r t o mi gr a t i on

between eastern and western Eurasia shortly after thedrying of the Turgai Sea. This scenario predicts thatNorth African, European, and West Asian anguineswould not necessarily form a monophyletic group buthat North American anguines should be a monophy-etic group. This scenar io is compa tible with both our

phylogenetic hypothesis an d the hypoth esis th at E oceneEur opean fossils of th e Anguina e ar e affiliated witheither the modern North African lineage or the Euro-pean and West Asian lineage (Meszoely and Haubold,1 97 5). T h e E u r o pe a n fos s il a n g u in e s of t h e l a t eOligoceneMiocene are assigned to the AnguisOphis-

aurus apodus clade (Gauthier, 1982). The first fossilappearance ofOphisaurus in North America occurs inh e late Miocene of Sa ska tchewan (Holma n, 1970). This

observation is consistent with a post-Oligocene arrivalof Ophisaurus in North America by dispersal from th eBering lan d bridge across Canada to its curr ent distr i-bution in sout heast ern North America.

I n a n a l t er n a t i ve s ce n a r io, t h e An g u in a e a r os e inN or t h Ame r ica a n d s p r ea d t o e a s t e r n A sia v ia t h eBering land bridge prior to the Oligocene drying of theT u r g a i S e a . W h e n t h e T u r g a i S e a w a s d r y , w e s t e r nEur asia would h ave been invaded an d quickly blocked

o t h e e a s t b y t h e u p li ft i n g of Tib et (O lig oce n e t oMiocene). This scenar io predicts tha t North Americananguines would not necessarily form a monophyletic

g r ou p b u t t h a t N or t h Afr i ca n , E u r op ea n , a n d We stAsian an guines would. Ph ylogenetic predictions of thiss e con d s ce n a r io a r e n ot comp a t i ble w it h ou r mos tparsimonious tree. This second scenario also requiresth at Eocene E uropean fossil anguines ar e phylogeneti-cally outside a group conta ining OligoceneMiocene

Eu ropean fossil an guines and all exta nt an guines fromNorth Africa, Europe, and West Asia. A tree showingt h e N or t h Afr i ca n a n d w es t e r n E u r a s i a n a n g u in e sfor mi n g a mon op h yle t ic g r ou p a s p r e di ct e d b y t h esecond scenario was not statistically rejected but wascostly (seven extra steps required). Both hypothesess u gge st t h a t t h e d r yin g o f t h e Tu r ga i S ea a n d t h eformation of Tibet were instrumental in shaping cur-rent biogeographic patt erns.

DNA Sequence Divergence and the Fossil Record

Rate of m olecular evolution for the mitochondrial

DNA region sequenced here has been estimat ed foragamid lizards, bufonid frogs, and fishes (Berminghamet al ., 1997; Macey et al., 1998a,b) as 0.650.69%change per l ineage per mill ion years. If this rate isappr oxima tely corr ect for an guid, ann iellid, a nd xeno-saur id lizards, then t axa sam pled here ar e extra ordinar-ily old. Relatively few divergences are under 10 millionyears (between Anniella species; between Sauresia a n dWetmorena; between Abronia a n d Mesaspis; among

Elgaria species; between Anguis a n d Ophisaurus apo-dus; an d between Ophisaurus attenuatu s a n d O. ventra-lis; Table 4). Of these taxa, only divergences among

Elgaria species ar e less th an 7 m illion years. After 10million years, mitochondrial DNA is expected to sa tu-ra te (Moritz et al., 1987); hence, a linea r r elationship ofnu cleotide substitut ions a nd time is not a nticipated.

The branching event separating Ophisaurus apodusa n d Anguis a p p ea r s t o b e a p p r oxi ma t el y 9 MYB P,which may be an underestimate if some substitutionalsat ur at ion h as occurr ed. The fossil record for t hese t axais difficult to interpret because small Ophisaurus apo-d u s-like specimens can be confused with Anguis-likespecimens (Gauthier, 1982). Note that the divergenceb et w ee n t h e s e t a x a a n d Ophisaurus koellikeri of Mo-

rocco is great er t ha n 10 MYBP.The fossil record is consistent with our in ter pret at ion

of ve r y ol d d iv er g en ce s a mon g t h e ma j or l in e a ge s(Gaut hier, 1982). Fossils of New a nd Old World xenosa u-rids and anguids are known from the late Cretaceous(9575 MYBP). Th e t axa Anniellidae, Anguina e, Diplo-glossinae, and Gerrhonotinae all are known from atleast th e ear ly Eocene (5055 MYBP).

Within Elgaria, t h e m ole cu la r d a t a e st im a t e t h edivergence between the northern E. coerulea a n d t h eremaining taxa at 6.6 MYBP. The Gulf of Californiaformed 56 MYBP (reviewed in Murphy, 1983) and

could have separated E. kingii from the clade of west-ern species ( E. paucicarinata, E. multicarinata, a n d E .

panamintina ); th e DNA sequences est ima te 4.0 MYBP,



19/23

which is slightly less than expected. Alternatively, theform at ion of th e Mojave Desert could ha ve separa ted E .kingii from the western clade. Continued aridization ofhe Baja California peninsula (3.4 MYBP) could have

separated E. paucicarinata from E. multicarinata a n dE. panamintina, and a Pleistocene (1.5 MYBP) diver-

gence for E. multicarinata a n d E . p a n am i n t i n a m a yhave occurred across the Owens Valley of California.This result is consistent with our molecular calibrationa n d w it h cu r r e n t h y pot h e s es of P l ioce n e d r yi n g of western North America that continued into the Pleis-ocene (Axelrod, 1979).

Taxonomic Recommendations

Two considerat ions should be addr essed when ma k-ng ta xonomic cha nges t o preserve m onophyly. First , ish e evidence for nonm onophyly of cur ren tly r ecognized

groups sta tistically robust an d second, h ow disru ptive

s th e proposed taxonomic chan ge?Among higher taxa, the overall most parsimonious

rees from analysis of DNA sequence data depict asmonophyletic groups th e Anniellidae, Anguidae, Diplo-glossinae, Gerrhonotinae, and Anguinae but not theXenosauridae (Xenosaurus a n d Shinisaurus). The liz-ard family Xenosauridae as currently recognized con-a ins two subfam ilies, th e Shinisaur inae an d Xenosau -

r i n a e. B eca u s e m on op h yly of t h e Xe n os a u r id a e(Shinisaurus a n d Xenosaurus) is statistically rejected,we propose to recognize as separate lizard families theShinisauridae (genus Shinisaurus) and Xenosauridae

(genus Xenosaurus). This taxonomic change affects asingle species, Shinisaurus crocodilurus, and therefores not consider ed disru ptive.

The Anniellidae, Gerrh onotinae, a nd Anguinae eachreceive sta tist ical su pport a s monophyletic groups fromana lysis of DNA sequence data. Recognition of theizard family Anniell idae has been a topic of debate

(Gau th ier, 1982; Good, 1987; Keqin an d Norell, 1998).The phylogenetic analyses of DNA sequences and allo-zymic dat a place the Ann iellidae, wh ich includes onlywo species, as the sister taxon to th e Anguidae, an d

monophyly of the Anguidae r eceives stat istical su pport

from analysis of allozymic data. The Anniellidae ap-pears not to be the sister taxon to either Anguis or thean guines a s previously pr oposed (Gaut hier, 1982; Ke-qin and Norell, 1998). Because the family Anniellidaehas been recognized for a long t ime and is currentlyused in popular field guides as well as the scientificitera tur e, placing this t axon in th e Anguidae would be

disruptive. Hence, we recommend continued recogni-ion of the Anniellidae.

Sta tistical support was n ot obta ined for m onophyly ofhe Diploglossinae*, which therefore is retained as a

metataxon (Estes et al., 1988; Gauthier et al., 1988)

denoted with an aster isk, indicating t ha t m onophyly isneither stat istically support ed n or rejected (Schu lte etal., 1998). The genus Diploglossus is not monophyletic

but a more detailed sampling is needed before stabletaxonomic changes can be made.

Within the Anguinae, Ophisaurus is not monophy-letic, an d st at istical su pport is obtained for the group-ing of Ophisaurus apodus with Anguis fragilis r a t h e rth an with th e oth er species ofOphisaurus. Because old

generic names exist, Hyalosaurus for O. koellikeri a n dPseudopus for O. apodus, two options are presented.O n e op t ion w ou l d b e t o ch a n g e O. koellikeri fromOphisaurus t o Hyalosaurus a n d O. apodu s from Ophis-aurus t o Pseudopus. If these changes were made, theremaining Ophisaurus would st i ll be considered ametataxon because monophyly of t his group is sup-ported only by a decay index of 1 and statistically isneither supported n or rejected. This chan ge would n otb e d is r u p t iv e u n l es s ot h e r Ophisaurus species arefoun d not t o form a monophyletic group, ther eby requir-ing more taxonomic changes. Alternatively, all taxa in

the Anguinae could be referred to the genus Anguis,which would provide a long-lasting sta ble t axonomy.Because few sp ecies a re involved, we favor recognitionof a single genus, Anguis.

Evolution of tRNACys

Trem endous var iat ion occurs a mong species in poten-tial stem sizes of both D- and T-stem regions of DNAsequen ces encoding tRN ACys (Fig. 2). Taxa basal on ourph ylogeny (Heloderm a, Varan us, a n d Shinisaurus) h a v ezero, three or four base pairings in the D-stem (Fig. 6).

Xenosaurus a n d Anniella demonstra te enlarged stems

of five base pairings. All other ta xa except two groupshave the normal four base pairings or a sl ightly re-duced stem of three base pairings. An en larged stem ofsix pairin gs occurs in Ophisau rus k oellikeri, and varia-tion occurs in Elgaria for eith er t hree ba se pairings orone base pairing. Because E. kingii a n d E. paucicari-nata h a v e a s i n g l e D - s t e m p a i r a n d E. coerulea, E.multicarinata, a n d E . p a n a m i n t i n a h a v e t h r e e b a s epairings in the D-stem, an equivocal r econstr uction ispresented in Fig. 6. Two interpretations are possible:(1) two base pa irings wer e lost to produce a st em of onebase pairing in the ancestor ofE. kingii, E. paucicari-

nata, E. multicarinata, a n d E. panamintina a n d t h e ntwo base pairings were regained in the ancestor of E .multicarinata a n d E . p a n a m i n t i n a to produce threebase pairings or (2) two base pairings were lost indepen-d e n t ly i n E . k i n gi i a n d E. paucicarinata. Paralleldeletion of a single base t ha t dest roys two base pa iringsin the D-stems of both E. kingii a n d E. paucicarinata(Figs. 1 and 2) seems m ore likely than reinsertion of abase in exactly the same place following its deletion,th ereby favoring the second h ypoth esis.

Most species have enlarged T-stems (Fig. 2). Thebasal condition appears to be a stem size of six to eight

base pairings instead of the normal five base pairings(Fig. 6). In th e Anniellidae an d Anguidae t he an cestra lcondition is a n enlarged stem of six base pairings.

26 8 MACEY ET AL.


20/23

Among diploglossines, two independent losses are ob-s er v ed t o p r od u ce n or ma l fi ve -b a s e s t e ms, a n d t w ondependent gains are observed to produce enlarged

seven-base stems. All tropical gerrhonotines have en-

arged stems of six or seven base pairings whereas allElgaria species have reduced stems of four base pair-ngs. The sa me situ at ion is observed among the Angui-

nae; species exhibit either enlarged stems of sevenpair ings or r educed stem s of four pair ings.

The loss in Elgaria of base pairings in D-stems of RNACys differs from the eight independent losses previ-

ously observed among lepidosaurian reptiles (Macey etal., 1997b). In Elgaria, a sin gle base pa iring is observedn th e D-stem of tRN ACy s whereas no such base pa irings

rema ined in the eight other independent losses. Tra ns-fe r R N As w it h s in g le b a s e p a i r s i n t h e D -s t e m a r e

h ou g h t t o for m a t e r t ia r y s t r u ct u r e d iffe r en t fr omRNAs with D-arm replacement loops (Steinberg et al.,

1 99 4). B eca u s e t h e l in e a ge a n ce st r a l t o Elgaria isassociated with D-stem size r eduction from four tohree base pairings, a gradual process of deletion ismp lica t e d i n t h e for ma t i on of t h e u n u s u a l t R N ACy s

observed in E. kingii a n d E. pau cicarinata. In addition,no repeats that would indicate slippage events duringreplication (Macey et al., 1997b) ar e observed.

The large T-stems ar e u nique. Mitochondrial tRNAshave been shown to lose stem regions, but this is thefirst observation of massive stem increase. In addition,

after the T-stem increased in size, decreases in size areobserved to occur in parallel. Interestingly, return to as m a lle r T-s t em is a s s oci a t ed w it h r e du ct i on of

the D-stem in t he Elgaria species. Size changes in thetwo stems could be related, but a more detailed sam-pling of taxa should be used before any major conclu-sions are made about possible correlated evolution of

these stems in tRNACy s among anguid, anniellid, xeno-sau rid, and shinisaurid lizards.

ACKNOWLEDGMENTS

We a re grat eful to t he following people for collecting specimens orproviding t issue sam ples: Natalia B. Anan jeva, St ephen D. Busack,J o h n E . C a dl e, J o n a t h a n C a m pb el l, D a vi d C . C a n n a t e ll a , P a u lChippindale, Carla Cicero, David M. Darda, Paul Elias, David A.

Good, S. Blair Hedges, Robin Lawson, Robert L. Seib, Stanley K.Sessions, H. Bradley Shaffer, Richard Thomas, and David B. Wake.Kraig Adler provided helpful comments on an earlier version of themanuscript. This work was supported by grants from the National

Science F oundat ion (predoctoral fellowship to J .R.M.; DEB-9318642to J .B. Losos, K. de Qu eiroz, a nd A.L.; DEB-9726064 to A.L., J .R.M.,

and T.J.P.), National Geographic Society (4110-89 and 4872-93 toT.J.P. and J.R.M.), Russian Foundation for Basic Research (N97-04-5 0 09 3 to Natalia B . An an jeva) , an d th e C alifor n ia Acad emy of Sciences.

APPENDIX 1

Alternative hypotheses used in Wilcoxon signed-r a n k s t e s t s (F e l se n s t e in , 1 98 5; Te mp le t on , 1 98 3).Lengths of trees and consistency indices (CI) (Swofford,1998) are given in par entheses. Num bers refer to the

following taxa: (1) Heloderm a suspectum , (2) Varanusgriseus, (3) S hinisaurus crocodilurus, (4) Xenosaurusgrandis, (5) Ann iella geronim ensis, (6) Anniella pul-

FIG. 6. Evolution of stem r egions in tRNACys as inferred from DNA sequences derived from anguid lizards and related taxa. Stem size innumber of pairs is ma pped on t he shortest estimate of phylogeny. Light patt erns r epresent few or n o pairings a nd black represents a lar genumber of pairings.



21/23

chra, (7) Celestus enneagrammus, (8) Diploglossus bilo-batus, (9) Diploglossus pleei, (10) Ophiodes striatus,(11) S auresia agasepsoides, (12) Wetmorena haetiana,(13) Barisia imbricata, (14) Gerrhonotus liocephalu s,(15) Abronia oaxacae, (16) Mesaspis m oreleti, (17) E l-garia coerulea, (18) Elgaria kingii, (19) Elgaria pauci-

carinata, (20)Elgaria m ulticarinata, (21)Elgaria pana -mintina, (22) Ophisaurus koellikeri, (23)Anguis fragilis,(24) Ophisaurus apodus, (25) Ophisaurus harti, (26)Ophisaurus attenu atus, and (27) Ophisaurus ventralis.

The two overall m ost parsimonious trees using th eDNA sequen ce dat a (length 5452 steps a nd CI of 0.394):A1. (1, (2, (3, (4, ((5, 6), (((7, 8), ((9, 10), (11, 12))), ((((13,14), (15, 16)), (17, (18, (19, (20, 21))))), (22, ((23, 24), (25,(26, 27))))))))))). A2. (1, (2, (3, (4, ((5, 6), (((7, 8), ((9, 10),(11, 12))), (((13, (14, (15, 16))), (17, (18, (19, (20, 21))))),(22, ((23, 24), (25, (26, 27))))))))))).

The most parsimonious trees derived by constraining

Shinisaurus a n d Xenosaurus to form a monophyleticgroup using the DNA sequence data (length of 5477steps and a CI of 0.392): B1. (1, (2, ((3, 4), ((5, 6), (((7, 8),((9, 1 0), (11, 12))), (((13, (14, (15, 16))), (17, (18, (19, (20,21))))), (((22, 25), (23, 24)), (26, 27)))))))). B2. (1, (2, ((3,4), ((5, 6), (((7, 8), ((9, 10), (11, 12))), ((((13, 14), (15, 16)),(17, (18, (19, (20, 21))))), (((22, 25), (23, 24)), (26,27)))))))). B3. (1, (2, ((3, 4), ((5, 6), (((7, 8), ((9, 10), (11,12))), ((((13, 14), (15, 16)), (17, (18, (19, (20, 21))))), (22,((23, 24), (25, (26, 27)))))))))). B4. (1, (2, ((3, 4), ((5, 6),(((7, 8), ((9, 10), (11, 12))), (((13, (14, (15, 16))), (17, (18,(19, (20, 21))))), (22, ((23, 24), (25, (26, 27)))))))))).

The most parsimonious trees derived by constrainingAnniella not to form a monophyletic group using theDNA sequence data (length of 5503 steps and a CI of0.390): C1. (1, (2, (3, (4, ((5, ((7, 8), (((9, 10), (11, 12)),(((13, (14, (15, 16))), (17, (18, (19, (20, 21))))), (((22, 25),(23, 24)), (26, 27)))))), 6))))). C2. (1, (2, (3, (4, ((5, (6, (7,8))), (((9, 10), (11, 12)), (((13, (14, (15, 16))), (17, (18, (19,(20, 21))))), (((22, 25), (23, 24)), (26, 27))))))))).

The m ost par simonious tree derived by constr aininghe Anguidae not to form a monophyletic group usinghe allozymic data of Good (1987; length of 85 steps and

a CI of 0.824): D1. (Xenosaurus, ((Anniella, (Ophisau-

rus, Elgaria)), Celestus)).The m ost par simonious tree derived by constr aining

he Diploglossinae not to form a monophyletic groupusing the DNA sequence dat a (length of 5456 steps anda CI of 0.394): E 1. (1, (2, (3, (4, ((5, 6), ((7, 8), (((9, 10),(11, 12)), (((13, (14, (15, 16))), (17, (18, (19, (20, 21))))),(((22, 25), (23, 24)), (26, 27)))))))))).

The most parsimonious trees derived by constraininghe Gerrhonotinae not to form a monophyletic group

using the DNA sequence dat a (length of 5474 steps anda CI of 0.392): F1. (1, (2, (3, (4, ((5, 6), (((7, 8), ((9, 10),(11, 12))), (((13, (14, (15, 16))), (22, ((23, 24), (25, (26,

27))))), (17, (18, (19, (20, 21))))))))))). F2. (1, (2, (3, (4, ((5,6), (((7, 8), ((9, 10), (11, 12))), ((13, (14, (15, 16))), ((17,(18, (19, (20, 21)))), (22, ((23, 24), (25, (26, 27)))))))))))).

The most parsimonious trees derived by constrainingthe Anguinae not to form a monophyletic group usingth e DNA sequence dat a (length of 5492 steps an d a CIof 0.391): G1. (1, (2, (3, (4, ((((((5, 6), (7, 8)), ((9, 1 0), (11,12))), ((13, (14, (15, 16))), (17, (18, (19, (20, 21)))))), ((22,25), (26, 27))), (23, 24)))))). G2. (1, (2, (

Moleculas Phylognetis and Bio Geography in Anguid Lizards

Documents