Rivista Italìana di Paleontologia e Stratigrafia pagine 135-1'{O Luglio 2O0O TANYSTROPHEU S (ARCHOSAUROMORPHA, PROLACERTIFORMES) REMAINS FROM THE TRIASSIC OF THE NORTHERN FRIULI (NE ITALY) FABIO MARCO DALLA VECCHIA" Receiretl February 28,2Aa0; accepted April 17,2Aa0 Key-uords: Tanystropheus, Carnian, Middle-Late Triassic, Car- nie Aìp.. Northeartcln Itrlt. Riassunto. Vengono descritti i printi resti ossei rinvenuti ne11'I- talja norclorientale scnza dubbio appartenenti al grande rcttile prola- certiformc Tdqtstropheils. Si tratt;r di una vertebra caud:rle prossimale dcl Triassico medio della Val Aupa (Udine, Friuli) e di una vertebra cer- vicale del Carnico di Fusea (Uiline). La vertebr:r cervicalc r:ìppresenta 1a prima segnalazione clel gencre nel Carnico e la sua testimonianza più re cente, se si csclude T. Jossai tlel Norico. Dur;rnte il Triassrco Superio- re Thnystropheu5 1,i1,g1'1 lungo le coste della Tetide nordoccidentale mentre era scompàrso nell'Europa Centrale, dominata da ambienti continentali. Abstrdct. Thc first diagnostic remains of the large prolacerti- form Tanystrophezs arc rcported fron northeastcm ltall'. Thc;- include a proximal caudal vcrtebra from the Mjddlc tiassic of Aupa valley (Udinc, Friuli) and a cervical vcrtebra fron the Carnian of Fusea (Udine). The cervical vertebra rcpresents thc first record of Tanystro- pheus in the Carnian arrd is the geologically 1-oungest occurrence other tlran the Norian Z fossai. Tanystropheus hved along the coasts of the northwestern Tethl-s during thc Late Trjassic r.hile it disappeared in Central Europe u'herc continental cnvironments t-ere prevaìling. lntroduction. During the last ten vears the amount of reptile remaihs found in the Mesozoic formations of northeast- ern Italy increased dramatically (Sirna et al., 1994; Dalla Vecchia, in press; Rieppel & Dalla Vecchia, in press, and references therein; Marco Ar-anzini, Pers. comm.) with the discovery of most of the main reptilian clades, such as Chelonia, Placodontia, Sauropterygia, Ichthyoptery- gia, Prolacertiformes, Crocodylia, Pterosauria, and Dinosauria. Some remains of coastal reptiles from the Middle to Upper Triassic of northern Friuli n'ere described by Dalla Vecchia (1994) and most of the sauropter\-gians, the placodonts and the truly marine rePtiles (ichthy- opterygians) from the Middle-Upper Triassic of NE Italy are described by Rieppel Ec Dalla Vecchia (in press). New specimens were found in two different local- ities and Triassic horizons of the northern part of the Friuli region (Fig. 1). They belong to the prolacertiform Tanystropheu.t v. Me)rer, 1852, and enrich the list of the Mesozoic reptiles found in this eastern area of Southern Alps. Abbreviations: MFSN, Museo Friulano di Storia Naturale, Udine. Systematic PalaeontologY Class RePtilia T,aurenti, 1768 Subclass Diapsida Osborn, 1903 Superdivision Neodiapsida Benton, 1985 Division Archosauromorpha von Huene, 1946 Order Prolacertiformes Camp, 1945 Genus Tanystropbeus H.v. Meyer, 1852 lvpe species: T. conspicuws H.r Meyer' 1852 TanystroPheus sp. Specimen MFSN25761 Description of tbe specimen dnd comparisons. The specimen was completely freed from the matrix. It is a complete, isolated proxirnai caudal vertebra 37 mm long and 52 mm tall of a prolacertiform reptile (Fig.2). Its size and morphology closely corresponds to that of the proximai caudal vertebrae of Tanystropheus figured by Wild (1973, figs. 57-61, pls.7-9;1980, pl. 5). The speci- men is slightly deformed due to compression. The right prezygapophysis, right postzygapoPhysis and right pleurapophysis are clearly more developed that the left ones. This cannot have been caused by compression Museo Paieontolosico Cittadino, Via \ralentinis 1i.{, 14074 Monfalcone (Goriz-ia), Italr,: E-mail: fabdalla(rr'tin.it A USTRIA )-'-'-'-'-.-^-- -.-'' 1F- -T=.-=-. .t ) ,a'J 't* a' (r! \ u\*.-. FRIULI ) .r' slo,:.( \ Udine t-ì \í o o J 136 F. M. Dalla Vecchia because compression can deform a body but cannot reduce its size. The skeletal element was probably origi- nal11' 6i5rh"pen because of a rrauma which affected the base of the tail, a part of the body subject to the highest stresses. Despite the fact that the specimen is presen-ed within a conglomerate, representing a relatively high- energy environment of deposition, it did not suffered a prolonged transport. For example, it does not show any darnage to the long and thin pleurapophyses. Only the tip of the left pleurapophysis was posteriorly crushed by a clast and appears spoon-like. The amphicoelous cen- trum (24 mm long) is lareraliy constricted just beiow the pleurapophysis. The ventral profile is also concave (arched) in lateral view and the cenrrum is spool-shaped. The articular surfaces are elliptical because of a slight deformation. There is a narrow longitudinal groove along the ventral side, like in the caudal 6 or 7 o{ T. con- spicuus figured by \(/ild (1973, fig.6Oe). The elongated and'narrow pleurapophyses (the right one is 3Z mm long) are directed laterally, posteriorly and slightly downward. They attach to rhe cenrrum between the basis of the pedicel of the neural arch and the dorsal margin of the centrum. The pleurapophyses are fiat- tened dorsoventrally but the distal tip was probably swollen and rounded. The neural arch is fused to the centrum but a suture is possibly present at the base of the right pedicel. The prezygapophyses are very short and do not project beyond the cenrrum. Their articular surface is eliiptical and faces forward, dorsally and only slightly outward. They are connecred to rhe anterior neural spine by short supraprezygapophysial laminae. Another channel-like srrucrure is developed between the prezygapophyses and just above the small neural chan- nel. Its floor is a short intraprezygapophysial lamina. The postzygapophyses project posterolaterally beyond -rlg. I Location of the fossiliferous localities in Friuli, NE Italy. 1) Dell'Andri Creek, Aupa Valley;2) Fusea. the centrum and are larger than the prezygapophyses. They have elliptical elongated articular surfaces, facing downward and slightly posteriorly and laterally; howev- er, the orientation is somewhat distorted by deforma- tion. Supraposrzlrgapophysial and intrapostzygapo- physial laminae are presenr. The intrapostzygapophysial laminae converge medially and ventrally inside the neu- ral arch, just above the neurai channel. There is a deep infra-postzygapophysial cavity. The neural spine is quad- rangular, wide and slightly inclined backward. Irs cross- section in anteroposterior view is a tall, upside-down tri- angle, and its top is thickened and flat. The surface of the neural spìne i. rugose, rhe anreroventral parr just above the prezygapophyses is blade-like and becomes a prespinal larnina. The centrum is comparatively less eiongated than the centra of the caudai vertebrae o{ T. Iongobardicus (Bassani, 1886) and T. conspicuus. The shape of the pleu- rapophyses is similar ro thar of the pleurapophyses of the anterior caudal vertebrae of T. /ongobardicws (Wild, 1973, p|s. 7-9; 19Sa, pl. 5). The shape of the pleu- rapophyses differs from that of the anterior caudal ver- tebrae of T. conspicuus which are much broader (\7i1d, 1973, {igs. 57-61). There are other characters of MFSN25761 which differ from those of the caudal ver- tebrae of T. conspicwus: the posterior inclination of the neural spine of MFSN25761 is higher, the prezy- gapophyses are shorter and have articular facet facing more forward (i... the facet is subvertical), the intraprezygapophysiai laminae are shorrer and thinner, the neural channel is smaller, and the lateral sides of the centrum are probably more excavated. The size indicates that the vertebra belongs to a rather large individual which, however, was smaller than T. conspicuus. Tanystropheus remains frctm Friult 137 proximal portion and it enllrges fan-wise distally. Although this Lrone ir not actually diagnostic. it clorely resemble. the posteri- or thoracic ribs of T. longob- ardicus (\íild, 1973, fig.35 left). Since it was found in the same block with the caudal vertebra and it belongs to a large reptile of comparable size, it could plausibly belong to the same taxon of the caudal vertebra and possibly to the same individual. Other very fragmentary bones (MFSN25Z63, MFSN 25764, MFSN25765, MFSN 25766) found in the same block are not cornpletely freed from the rock matrix and cannot be identified with certaintl-. Geological setting. The specimens were found by dr. Corrado Rosenfeld in a single isolated block near the conflu- ence of the Dell'Andri creek Fig.2 -MFSN25'61, anterior caudar..ertebra T:\1:.1"1:::T"-111":t^::i' A'rrrr.rior rrcn. B) po\rerior rieu. Cr f ne DrloHe ol tne roao De\-or- dorsal r.icn,, D) r.entral vìer. E) left lat- chians-Studena alta (Aupa Val- eral vien'. The centrum is 2/ mm long ley, Moggio lJdinese, Udine and the wholc vertebra is 52 nm hìgh. province; Fig. 1). The rock con- Legenda: PR:prez,vgapophvsis; PZ: i"irrir.,g the specimens is con_po'tzvq:pophr''ì' glornerate with.r silty-carbon- ate matrix and \\ ith d;rrk gray. The proximal half of a thoracic rib, without the millimetric carbonate clasts, bivalves, and echinoderm articular head (MFSN 25762; Fig. 3), was found in the elements. Only the carbonate Upper Serla Formation same block with MFSN2576|. The proximal portion of (1ate Anisian) and the terrigenous "Torbiditi d'Aupa" MFSN 25762 is anteroposteriorly flatteneci with an (1ate Anisian-early Ladinian) outcrop in the catchment ellipgical cross-section and presents an anterodorsal and basin of the creek (Jadoul & Nicora, 1929). The lime- a posterodorsal longitudinal ridge. The distal portion is stones and dolomitic limestones of the Upper Serla For- elliptical in cross-section, it is less flattened than the mation are decidedly different from the lithology of the l\lFSN25762. p:rrial po.r, ri- or rhoracic rib. Scrle bar : I cm. trio l 138 F. M. Dalla Vecchta bone-bearing block, whrch shows a significativc relriqe- nous content. The "Torbiditi d'Aupa" is a n-rar1-sand- stone basinal, deep-rvater unit. I{or-er.'er, manv other lithostratigraphic units rrnqing fron.r brs:rl Triassic to Norian ourclop .rlong rhe r:rlley and. lor crrmple. 'inri- lar conglon.rerates with crrbonrre-terrigenous compo- nents could be present in other Middle Trirssìc units (e.g. the "Terrigeno Ladinico"; Jadoul Er Nicora, 1929). The block could come frorn the morainal deposits all around the creek and could har.e been transported frorn nearby valleys bv glaciers during Pleistocene timcs. An1'n'ay, a stratigraphic frarnework similar to that of the Aupa valley is found along these vallc1's. Specrmen MFSN25760 Description of the specìmen and comparisons. The specin'ren is the posterior half of a single isolated ceni- cal vertebra, the anterior half of which was weathered away (Fig. a) . The prescrr-ed pert of rhe vertebra, bone .-'l -"','.'l -^l.l i, | )R -- lorr- rnr'l rhn .""."rrr",-ldilU li!tLUldl lllUlU! lJ r-u rilrrr rvrròt artu portion of the centrum is 123 mm long. The part of the centrum preserved as bone is 68 mm long. The middle to anterior part of the centrum, lacking the anterior end carrying the prezygapophyses, is presen'ed only rs natu- ral moid. The tube-shaped centrum is very loq about 1O mm in the middle, and collapsed because ir is hollow inside. The posterior articular end of the centrum, l4 mm high. i. enl.rrged.rnd slightly deflected ventraìly. The processus spinosus of the neural arch projecrs short- ly above the postzygapophyses, is triangular and similar to the lin of rn airplane t:ril. ending cranialll'.rgain,r rhe Fig. ,+ - X,IFSN25Z60, cenicel r-erte- bra, right latcral r,ien-. A) photograph, B; tlrari.inr.'1'he scale blr is ccntirnctric in A and 2 cm in B. Legend.r: Cì - centrum; P$ : processus spinosusi PZ : Ìrost7)- eapophvsis. dorsal mrrgin of the ccntrum. It is similar to the proces- sus spinosus of the cerr.ical vertebra 7 and 1(Wild, i973, pl. 9), Z-8 (\fild, 1923, pl. 11) or 5-Z (Wild,1973, pl. 16) of T. longobardicus. The neural:rrch is exrremelv lo$t:.r feature of the cervical vertebrae oÍ Tanystropheus. In f;ct, rccording to Pever & Kuhn-Schnvder (1955, p.591, translated from Frcnch) in the cervical vertebrae of Thnystropheur "..the neural arch is not rised above the r.ertebral centrum, and the neurai channel seems ro pàss inside the latter". This character can be observed in the cervicals of T. conspìcuus figured in Wìld (1973, see the r.'ertebral cross sections in figs. 43, 11, 51). Thc rìght postzvgapophysis is relativeiv srrong and long. project- ;-,. *.^ll h.-,^-l ,L^ ^osterior end of the centrum. The specimen MFSN25760, with its elonsated tubular aspect, and very lon neural arch, resemble s most closeh' the middle to posterior cen.ical r.errebrae of Tanystro, pheus longobarc{icus (\lild, 1923" pls.9, 11, 13, 16). GeoÌogica/ setting. Thu specimen n-rs found in the Fusea site (Tolmezzo, Udine; Fig. 1). I collected the bone ar rhe top surface of the laver E (Fig. 5) of this site n'hich is particularly rich in fossii ve rtebrates. Abundant disarticulated elemenrs ol Notbct- srturus sp., Ìt/. cf. giganteus Múnster, 1834 and of a cya- modontoid placodont (Dalla Vecchia, 1994; Rieppel & Dalla Vecchia, in press), together with Acyodus-like teeth and other fish teeth, and plants were found on the ex, posed upper part of layer E. Scattered bones and armour elements of cyamodontid placodonts, bones and teeth of l"lothosaurus sp. and fish teeth are also found in the mid- dle part of the layer and in the layers A, D and F. In layer D, skull roof bones of a dipnoan fish were also found. Tanystropheus remains from Friult 139 1O cm - 25760 M FSN AAA AAA 't, L A' A A A ,AA 4A A.A AA Fig. 5 - Stratigraphìc column of the Fusea site, lower Carnian. Leg- cnda: 1) black limestonc (n-achestones-packstones), 2) con- glomcrate wìth prei'ailìng carbonate clasts, 3) s'hite to ' light-gral' dolomitic 1ìmcstone, poorlv bedded and with lenses of intraformational breccia. The arrow- point to the posìtion of the vertebra N{FSN25Z60. The stratigraphic sectiotr contr;ning the bone- bearing layers is just at the transition between the top of a thick carbonate platform sequence and well-bedded black limestones hundreds of metres thick. The carbonate platform sequence is referred to as Dolomia Cassiana by Carulli et al. (I995, p. 76) and as Dolomia dello Schlern by Prsa (in Braga et a\.,1971'), ^nd dated as late Ladinian-middle Carnian in both PaPers. It is the iast carbonate platform before the deposition of the Dolomia Principale. The thick sequence of "black limestones" is the basal portion of the so called "Raibl Group" (Pisa in Braga er al., I97I). This basal Part is considered middle Carnian in age by Pisa (in Braga et al., [_lll v{,1lo"ooì E-a;l lo oil 1971) because of the presence oÎ Myopboria leefersteini (Mùnster, 1845) and CLypeina besici (Pantic, 1966). The layer where MFSN25760 was found is at the very base of these "black limestones". For references on the geology of the surroundings of Fusea, see Dalla Vecchia (1994). The comparison with the well-studied stratigraphy of Dolomites suggests that the stratigraphic position of the section is most prob;ably in the upper part of the lower Carnian, between Car2 and Car3 depositionai sequence (De Zanche et al., 1,993). This dating is also supported by Carulli et al. (1995, p.76). The tentative of biostrati- graphical correlation of the fossiliferous section by paly- nomorphs w.as not successful (Guido Roghi, pers. comm.), and the rare foraminifers indicate only a possi- bly late Ladinian to earll. Carnian age (presence ol Tro- cholina cf. cordeoolica, Sandro Venturini, pers. comm.). Conclusions. Most of the fossil record of Tanystropbeus comes from the Muschelkalk of Europe (southern Spain, France, Germany, Slesia,Transylvania) (.i.e. T. antiquus v. FIuene, 1908, Z conspicuus and Z. sp.), the Muschelkalk of Israel (Peyer, 1955), the upper Anisian-Anisian/ Ladinian Grenzbitumenzone (= Besano Formation) of Tessin (Southern Switzerland) and Lombardy (NV Ital;') (Z longobardicws), and the lower Ladinian Lower Meride Limestone of Tessin (7. meridensis \fild, 1980)' A fragmentary cervical vertebra is reported from the Middle Triassic of Saudi Arabia (Vickers-Rich et al.' 1999). A partial, single tooth attributed rc T. cf. meri- densis (Wrld, 1980) was also found in the lowermost Keuper (upper Ladinian) of SV Germany. A partial dor- sal centrurn tentatively identified as Tanystropheus 6y \flild (1980, p. 14, fig. 1O) comes from the "Tufi a Pachi- cardie", upper Ladinian of Alto Adige/Sùd T;rol (N Italy) . The youngest record ol Tanystropbeus is fron-r the upper Norian Argilliti di Riva di Solto of Lombardy with the sn-ral1 Z/ossal \flild, 1980 (\fild, 1980; Benton, 1994). T. fossai is represented only by a segment of four vertebrae from the cervical vertebral column. The speci- men belongs to a very small individual, the vertebrae are not longer than l5 mm. ]t is rery similar to a segrncnt oI the n-riddle tail of a basal pterosaur, and pterosaurs have been found in the same formation. Hor''ever, some structures (e.g. the rised facets for the articulation of the cervical ribs) support the identification as cervical verte- brae of Tanystropheus. The specimens MFSN2576A and MFSN25761 here described are the first remains unambiguously attributable rc Tanystrophews fron northeastern Itaiy. The caudal vertebra from Aupa valley partly differs from the proximai caudal vertebrae of Tanystrophezzs described in literature but the material is too incomplete to war- 2 ^O QgoQ Ipo o oc ogoC)oooo oo o oooo a OooÒoooò OoooÒooo sOÒOOoOo-O oO Og 006 ^ O O O ó o oo "oo oo^oó oÒ aOOO"Ooo "O O O oO O O o OOooo(2oo 140 F. M. Dalla Veccbia rant the erection of a new species. The specimen from Fusea is the first Carnian evidence of Tanystropbeus and, other than the Norian T. fossai, is the geologically youngest occurrence of the genus. The absenc e of Tany- stropbeus in the Central Europe during the Late Triassic (Keuper) is most probably related ro the reduction of coastal environinenrs and the predominance of conti- nental environments not suited for the way of life of this reptile. The genus survived during the Carnian and the Norian along the coasrs of rhe western Tethys. The tubular vertebra MFSN25Z60 confirms the phylogenetic trend of the cervical vertebrae of Tanystro- pheus toward a more and more elongation, hypothesized by \lild (1980, p. 25-26, fig. 1a). Acknouleclgenents. I thank Dr. Carlo Morandini and Dr. Giuseppe Muscìo, respcc- tivell' director and curaror of the Museo Friulano di Storia Naturale of Udine for thcir pernission to study the material. Thanks to Dr. Cor- rado Rosenfeld, rvhose collaboratìon has always bccn fundamental for the progress of the vertebrate paleontology in thc Frjuli region, to Dr. Sandro Venturini for the biostratigraphic and the nicrof:rcies analysìs and to Dr. Guido Roghi for the research of palynomorphs in the sam- ples frorn the Fusea site. I am grateful to Mr. Maurizio Tentor who realizcd the thin scctions of the samples at the laboratoires of the Museo Paleontologico Cittadino of Monfalcone (Gorizia). Thanks to Dr. Stefania Nosottì, Dr. Silvio Rencsto and Dr. Rupert Wild for thc ìnformations and the discussion and to Dr. Anna p;rganoni for the access to the collections of the Museo Civico di Scienze Natur;rli of Bersamo. Fina1l1', I thank Dr. Stefania Nosotti and Dr. Olivier Rieppel for the criricel rer:ew ol rhe m.rnrscripr. REFE,RENCES Benton M.J. (1991) - Late Triassic to Middle Jurassic exrinc- tlons among continental tetrapods: tesring the prttern. In Fraser N. C. & Sues H.-D. (Eds.) - 1n the shadow of the Dinosaurs - Early Mesozoic Tetrapods, pp. 366-397, Cambridge. Braga Gp., Carloni G.C., Colantoni P, Corsi M., Cremonini G., Frascari F., LocateÌli D., Monesi A., pisa G., Sassi F.P, Selli R., Vai G.8., Zirpoli G. (1971) - Note illusrra- tive della carta geoloeica d'Italia alla scala 1:1OO.OOO: Fogli 4c-13 - Monte Cavallino-Ampezzo. Sen. Geol. Italia, pp. 1-108, Roma. Carulli G.B., Longo Salvador G., Ponton M. (1995) - Lc unità ladino-c:Lrniche nella Carnia centro-occidentale. Ann. Unio. Ferrara, Sci. Terra, v. 5 (sr.rppl.), pp. 75-81, Ferrara. Dalla Vecchia F.M. (1994) - Reptile remains from the Middle- Upper Triassic of the Carnic and Julian Alps (Friuli - Venezia Giulia, Northeasrern Italy). Gortama - Atti Museo Friul. St. Nat., v 15 (1993), pp. 19-66, Udine. Dalla Vecchia F.M. (in press) - I reperti ossei dei tetrapodi con- tinentali paleozoici e mesozoici d'Italia. 1z Leonardi G. . & Mietto P (Eds.) - I dinosauri di Rovereto. Le piste giurassiche dei Lavini di Marco (Trentino, Italia) nel contesto dei dinosauri d'Italia. Museo Civico di Rovere- to & Museo Tridentino di Scienze Naturali. De Zanche V, Gianolla P, Mietto P, Siorpaes C. & Vail pR. (1993) - Triassic sequence stratigraphy in the Dolomites (ItaIy). Mem. Sci. Geol., v. 45, pp. 1-23, Padova. Jadotrl F. & Nicora A,., (1979) - Lasserto stratigrafico-paleo- gcografico del Trias n'redio-superiore della Val Aupa (Carnià orientale). Riv. Ital. Paleont. Strat.,v.85, pp. 1- 30, Milano. Pe1'gr g. (1955) - Demonstration von tias-Vertebraren aus Palàstina. Eclogae Geologicae Helvetiae, v. '18, pp. 486- 490, Basel. Peyer B. & Kuhn-Schnyder E. (1955) - Squamates du Trias. In Pivetau J (Ed ) - Traité de Paléontologie, v. 5, pp. 528- 605, Paris. Rieppcl O. & Dalle Vecchia F.M. (in press) - Marine Reptiles from thc Triassic of the Tre Venezie :lrear northeasrern kely. Fielcliana, Chrcago. Sirna G., Dalla Vecchia F.M., Muscio G. & Piccoli G. (1994) - Catalogue of Paleozoic and Mesozoic Verrebrares and Vertebrate localities of the Tre Venezie area (North East- ern Italv). Mem. Scí. Geol.,v.46, pp.255-281, Padova. Vickers-Rich P, Rich T.H., Rieppel O., Thulborn R.A. & McClure H.A. (1999) - A Middie Triassic Vertebrate fauna from the Jilh Formation, Saudi Arabia. N. /ú. Geol. Palàont. Abh., v.213 (2), pp.2A1-232, Stuttgrrt. \i/ild R. (1973) - Die Triasfauna der Tcssiner Kalkalpen XXIII. Thnysnopheus longobardicus (Bassani). ScLttoeiz. paLaont. Abh.,v.95, pp. 1-162, Basel. Sfild R. (1980) - Die Triasfauna der Tessiner Kaikaipen XXIV Neue Funde von TanystropDezzs (Reptilìa, Sciuamata). Schweiz. palàont. Abb.,:,. 102, pp. 1-,{3, Basel.