pp.1 73-190 Ciencias da Terra (UNL) Lisboa N" 15 2003 ( 3 Figs.. 2 Tab., 4 PI. 1
The Deinotherium (Proboscidea, Mammalia): an abnormal tusk from Lisbon, the Miocene record in Portugal and the first appearance datum. Evidence from Lisbon, Portugal
Miguel Telles Antunes t'<" & L. Ginsburg (3)
1 - Academia das Cicncias de l.isboa. 2 - Centro de Estudos Geologicos, Faculdade de Cicncias e Tecnologia (UNL), Qu inta da Torre, 2825-114 Caparica. Pornigal . Tel. (35 1) 21 294 8573; [email protected] • lnstnu t de Paleontologic du Museum nat ional d'Histcirc Naturetle, 8 rue Buffon , 75005 Paris, France.
Resume
Mc rs-cles: Deino thcrium; Iere arrivec; fin du Miocene inferieur; debut du Miocene moyen; datation fine; Portugal.
Le contexte stratigraphiq ue et chron ologiquc particulieremc nt favorabl e du Miocene de Lisbonne pennel de mettre en evi dence qu'il y a deux data distincts et succcssifs quant aux premieres immigra tions de Proboscidicns, celie des Gomphotheres et, plus tard, celle des Dcinorheres. L'etude d'unc defense d' age Langhien montre que Deinotherium havaricum etait encore present alors. On a pu preciser la repartitionchronologique de ceue espece.Aprcs une discussion au sujerdu genre Deinotherium etde son amplitude, on presente des considerations sur la presence et repartition des Deinothercs au Portugal, ainsi que sur leur ro1c ccologiquc.
Abstract
Key-word s: [Jeinoiherium; lst arrival; end of Lower Miocene; lower Middle Miocene; accurate dating; Portugal.
An exceptionally favourable stratigraphicand chronologie context concerning the Miocene series in Lisbon allows us to stress that there are two successive data as far as the Proboscideans' immigration into western Europe is concerned: firstly, that of Gomphotheres, and later that of Deinotheres. The study of a Langhian (in age) tusk has shown that Deinotherinm hava ricu m was still present then. The time span of this species could be accurately recognized. A discussion on the genus Deinotherium is presented., as well as its occurrence in Portugal and on its ecologic meaning.
1. Introdu ction Deinotheriu m at the IVb unit would mean (if true, and it is not) that the arri vals ofgomphotheres and deinotheres Th e lowe r Tagus basin in the Li sbon ar ea ha s would have been synchronous. especially good conditions for datin g same very important Some doubts about it wcrc expressed (Antunes, 1960) biological events as the first Immigration ofproboscidcans Nearly all mammalian fossils from Quinta das Pcdrciras into westernmo st Europe. A lte rnating ma rin e a nd (Lumiar) ascribed by Zb yszcwski (1949) to the IVb unit contine ntal le vel s, th e muc h detail ed stratigraphic had instead been collected in thc overlying, Va2 pyrolusite knowledge and fine dat ing (foraminifera, K-Ar, 87 Sr_goSr) rich sands . The typical pyrolusite, black pat ina is c1carly allow us to obtain an accurate chronology of these events shown by the so le Deinotherium fossil (a right D4 germ, (Antunes, 2000 ; Antunes et al. , 2000) (Fig. 1). n" 7920, coil. Geolo gic Museum ofthe Instituto Geo logico The presence of the ftrst gomphotheres in Lisbon's c M ineirol lG M) that was ascribed to the IVb un it IVb unit ("A reias da Quinta do Bacalhau'', Burdi galian) (Zbyszewski, 1949, p.74). The lv b unit yielded there ex had been shown (Zbyszewski, 1949). tremel y rare gornphothere re mnants, all devoid of On the other hand, data concerning Deinotherium were pyrolus ite. Deinoth eres are always much rarer than often confusing as they rely on unaccurately localised gomphothcrcs. Their presence in the IYb had been claimed specimens (Zbyszewski, 1941, 1949). The occurren ce of but on the referred tooth, which wou ~ have been collected
173 CiindaJ da Terre (UNLJ. 15
Qu inta das Ped reiras . This doubtfully-localized specimen is worthle ss as a proofofthe Deinotherium presence in the Vb unit. (b) Qu inta da Farinheira - this was one ofthe richest Vb sites. It yielded a lot ofspecimens that represent the" Hispanotherium fauna". Go mphothc re teeth are plenti ful. Bones and teeth from Qui nta da Farinheira show an iron-rich (goethite and other) crust patinathat includes sand gra ins. One ofus(M.T.A.) could verify this after his field researches and through theclean ing ofspccimens collected there. Or, the only Deinotheriumtooth in the IGM museum labelled as from Quinta da Farinhcira (a right m3, n° 5530) , does not show such a patina nor the usual rus t-colour. 11 is light-coloured , the dentine bei ng white and unstained . It certainly was not collected at Quinta da Farinheira. Owing 10 the very closely similar aspect, theconcernedtooth could well came also from Quinta daBarbacena (Val) as theleft P4 n" 5529, IGM museum. The samc m3 does not proo ve either the presence ofDeinotherium in the Vb unit. Fig. I- Lower Tagus basin (distal part), Lisbon and Setubal Peninsula area (Antunes et al.• 2000). 2. A d iscu ssion on th e gen us during one of Zbyszewski's visits to the sandpits (not more than half an hour, nor more than once a month). The taxonomic status of the first european deinotheres Thi s hypothesis is utterly unlikely. has been discussed. A revision ofthe problem is underway The first appeara nce ofDeinotherium in Lisbon must (Ginsburg, in press). We accept the following poi nts: - the th erefore be ascribed to th e Va 2 un it , a lt houg h distinction be tween Deinotherium and Prodein otherium Zbyszewski ( 1973, p.I OI) did not acknowledge it. is not j ustified ; - the so le early european spec ies is Hence the first appe ara nces of gomphotheres and Deinotherium bavaricum H. von Meyer, 1834. deinothe res, by that order, are not synchronous but Gohlich ( t 999 ),after Hanis (1973), report theeuropean successive events (G insburg & Antunes, 1967; Antunes, Deinotheriidae to the genera Deinotherium (type species 1990). D. giganreum Kaup, 1829) and Prodinotherium (type The presence of Deinotherium in Lisbon's next. Vb species P. hungaricum Ehi k, 1930). unit (" Areias do Vale de Chelas") is not so clear. Th is P. hungaricum is represented but by thetype specimen intcnsely exploited unit yielded hund reds ofgomphotherc (one mandib le and some remains from the fore limb from teeth along with many other verteb rate fossils. Kotyheza), besides a P3 fro m Kirald, Hun gary. This Deinotherium was identified for the first time by dein othere is thc same size as D. cuvieri. Ehik asc ribed Zbyszews ki (194 1) on a fragment of left maxillary with these specimens to a new spec ies on the following reasons: the last molars (M2 and M3); the spec imen is vaguely said to have bee n found at Chameca, near Lisbon (Vb) I) p3 - the hypoconi d is not linked to the entoconid, (ibid.). Our late Colleague Zbyszewski explained that it as it is in D. hobleyi; wa s not purchased at the sandpit, as usual; somebody 2) mandible - the posterior foramen mcnta1eis located had noticed that a vertebrate fossil was kept by another undcrthe middle part of the p4, whereas inD. hobfeyi person and persuaded the owner to offe r it to the IGM it is located under the diastheme between p3 and p4; museum. The specimen shows the typical pyrolusi te 3) mandible - the anterior forame n mentale is under patina from Qui nta das Pedrciras, ncar Lumiar (Val). As the toponym Cham cca is an abridged vers ion ofChameca the midd lc part ofp3 whi lst in D. hobfeyi it is under do Lumiar, a village north-cast ofLumiar, confusion is the anterior border of the same tooth. almost certain. Hence the specimen does not proo ve the presence ofDeinotherium in the Vb (even ifit has bee n Ehik segreg ates a new genus because of MclV (the included in a list given by Bergounioux et al., 1953, p. 16). sole available metapodial) proportions are different from Zbyszewski (1973, p.101) recorded Deinotherium those of the D. giganteum MclV from Pikermi [cf from two Vb sites, (a) Azi nhaga do Pinhal and (b) Qu inta Ga udry, 1862, pl. XXV, fig.3; Dietrich, 1916). The MclV da Farinheira: from Pikermi is short and stout, the same bone from (a) Azinhaga do Pinhal- there is no reference on this Hungary be ing distinctly more slender. site, not even in Zbyszew ski's synthesis in Bergounioux As far as P. hungaricum speci fic charac ters are et al. (1953, p. IS, fig. B). A fragment of left maxillary concerned, the p3 structure cannot be upholded because with M2-M3 (n" 5528, IGM museum ) wa s collected the metacon id-hypoconid connec tion widel y varies from (according to its labe l) at Quinta da Casinha - Azinhaga one specimen to another. qn fou r Deinotherium p3 from do Pinhal. It also shows the typical pyrolusite pat ina from Pontlevoy (Mu seum de Pans collections), three show this
174 Ciincias da Terra (UNL}. I's connection as a small tran verse crest, but it entirely lacks Deinatherium bavaricum the small deinothere from the on the fourth specimen. Furthermore, the right p3 from " faluns" of th e Anjou . This mean s th at th e genus the Deinoth erium giganreum type spec imen shows this Prodinotherium cannot be accepted on the basis of the connecting crest that is entirely absent in the left p3. dental criteria proposed by Harris. The skull and post The posi tions of the mcntalc fora mina in the two skull characters put forward by Harr is are no t valid species under comparison are too close for them to have because he (maybe a priori) has included in the same a meaningful taxonom ic value. gen us both the spec ies bavaricum (from Europe) and As far as the different MclV proportions (that justified hobley i (from Africa), the latter bein g the only spec ies he the segregation of two genera) are concerned, Tobien co mpares to D. giganleum. Hence his conclusions abo ut ( 1962) described a Deinotherium giganteum hand from the segregation of the european forms in two genera are Howenegg (near Eppe lsheim) whose MclV proportions worthless. are the same as in the hungarian deinothere. We may Can these two genera be distingui shed after skeletal conclude that the genus and spec ies proposed by Ehik elemen ts? It would beneeded to compare the Deinotherium cannot be accepted on this auth or's criteria. gigantissimum from Romania to that of Deinotherium lIarris ( 1973) restored the genus, however naming it ba varicum from Lan genau (di scovered in 1976 and as Prodeinotherium inste ad of Prodinotherium, He mounted at the Stuttgart Museum). However this has not included in it th e s pec ies hob ley i (as cribed to yet been do ne. Deinotherium by Ehik). Harris presented a tab le to stress A similar case is known as far as Carn ivora arc con the differences between the two genera. As for dent ition, cerned. It is the ease ofPseudaelurus (Heinzma nn, 1973; he presents 3 characters (tab le I). Ginsburg, 1983). This genus includes a single early (MN3) species, Pseudaelurus turnauensis, from which a some Tab le I what larger form, R lorteti, derived later (MN4); from the latter derived (still later in MN4 or MN4b) P. romieviensis. P"'i"",lwri_ DriNHlwri"", and from th at a fourt h, s till lar ger spe cie s, P.
p...... lIy I... k rnroo!llyl.. p....01'1",poss>o"muoolyl•• quadridentatus. As it also occurred with the dcinotheres, cach new species did not elim inate the species from which M , ~ ..lIh ..~II d~ lop~ M... .. lIh rod...,... ptlN lm. l....ph p""lmrtMloph o. ....m. nl. l...... n.m.n\.O lI"" it derived: it persisted. As an example, the same level yie lded at La Gri ve-S ai nt- A fban Pseudaelu rus T...b _ rty _tl..1 T..b Ione:'::::;b.be .... ~rood .....11• .;.. . . tum auensis, P. loneu and P. quadridentarus. Nobody ever felt the need to ascribe these different species to different genera. At most, viret ( 195 1) proposed subgenus names The P3-4 mcsostylc may be present or not in the two for each spec ies. genera. Th is is further corroborated by the coe xistence of P. The postmetaloph ornamentation exists in both genera, tumauensis, R lorteti and P. romieviensis in Lisbon 's Vb although it may be more or less important. It may be added unit (Antunes, 2000, tab le I , p. 292) . that Deinotherium material (Museum de Paris collections) At last, Tobien ( 1988) s tudied th e Montredon fro m Pontlevoy and " Fa luns de I ' Anjou"( M N5, Dein otherium and named the small euro pean spec ies as D. bavaricum ), from Samatan (later age, D. giganleum ) Deinatherium bavaricum... as well as that from Montredon described byTobien (1988) show exactly the reverse situation: the always weak mcsostyle is clearly more distinct in the MN5 spec imens 3. An ab nor ma l Deinotherium t us k Iro m th e earl)' than in D. giganleum . This is also the case as far as the middl e Miocene or Lisbon postmetaloph ornamentation is conce rned. These two characters do not seem stable nor convincing enough to A previou sly unreported spec imen from Lisbon (Vb separate two genera. unit) is described and classified . As for the third character (tusks), in ouropinion it seems deceptive on the two skulls whe re tusks remain in situ: - ttie D. gig onteum type skull from Eppe lsheim , and Systematics - the P. bavaricum from Langenau depicted by Class MAMMALIA Linnaeus, 1758 Heizmann et al. (1996, fig.8). Order PROBOSCIDEA llIiger, 1811 The tusk 's size di fference is quite weak , and in both Family Deinotheriidae Bonaparte, 1845 cases each tusk is arc hed un til the rear part of the Genus Deinarherium Kaup, 1829 symphisis (against Harris' viewpoint). (Type-species: Deinotherium giganteum Kaup, 1829) Furthermore. one of us (L. G ) collected at Chevilly the type locality ofDeinot herium cuvieri - a 50 cm long tusk (C HE 99) arched in a way more close ly similar to Deinotherium bavaricum H. v. Meyer 183 1 that of' the D. g iganrcum type than to the dcinothere from (Text. fig. 2; PI. I, figs 1-3) Lan gcnau. Heimia nn ( 1992) gives Dein omerium as genus name Site: Qu inta Grande, near Charnec,do Lum iar (North of for the Languenau deinothere and Ginsburg (2000) called Lisbon).
175 Cuncios do Terra (UNL), 15
Srratlgr a phy a nd a ge : Vb unit ("Areias do Vale de On the other hand. the distal part of Deinotherium Chelas") from Lisbon's Miocene series. MN5 Mammal tusks usually presents an axis that is stra ight and a rou nd unit. Db iberian Mammal-uni t. Ca . 15.9-16.1 Ma. Stage: section, whereas the specimen under study shows anarched Langhian . axis, a median surface that is somewhat flatter than the lateral surface . as well as a distinctl y ell iptical sect ion :\I aterial & dimension s (mm): a left tusk. Total length. (whose great axis is more or less antero-posterio r). Th e 300. Proximal transverse & anterior-pos terior diameters Quinta Gra nde tusk is obviously a path ologic tooth, both (A, fig.!), 59 x 7 1.5. Pro ximal section at the level where very small (its correspond ing Chev illy tusk is one and a the tooth gets out ofits alveo lus (B , fig.I), transverse & hal ftimes larger), with an abnormal axis and an abnormal anter ior-posterior diameters. 55.5 x 68. Distalmost meas section, and presenting uttetly uncommon growth striae urable section (C, flg.l), transverse & anterior-pos terior and ename l stnps. d iame ters, 42 x 53.
Description : thi s tooth is simple and long. Section is 3. The Deinmh erium record in Portugal subelliptic. The proximal (root's) extremit y shows: (a) the beginnings of an axial ho le (shaped us an upside-down As it has been shown (Introduction). a d iscussion on cone) that corresponds to the roo t; (b) the section of the the occurrence o f Deinotherium in Portugal (and especially tooth. much larger than the section ncar its tip; (c) a bone on the time span of Deinotherium bavaricum in the Lisbon fragment adhering o n the tooth surface (this fragment's area) is interesting (Table 2). ex tremity indi cates the line where tooth gels out of the Deinotherium bavaricum is mainly represented in the mandibu lar bone). On the medial side. the tooth is flatter Lisbon area by 13 specimens from the Miocene Va2 unit. than on lateral side. It shows a large number of growing If account is taken of a fragm ent (MTA coil.) from Q uinta striae tha t arc fine, parallel and slightly sinuous (especially do Pombeiro ncar Chclas (Va2) , the tot al amount is 14. on the lateral surface); there arc also longitud inal. parallel The same spec ies have been recogni zed in the Vb unit striae (spaced from 2 to 6 nun).Three remnants of ename l {Langhian, early middle Miocene) after the tusk desc ribed strips are pre served between longitud inal striae at the here. It still occurs a little later in Langhi an at Qu intanelas proximal part ofthe antero-Iateral surface . These enamel (a small basin NW from Lisbo n) (Zbyszew ski, 1952 ). strip remnants measurements (mm) are as follows: Length Afte rward s , th e important Se rra vall ia n (middle x proximal width x d ista l width = 28 x 4.5 x 5.5; 24 x 6 x Miocene) transgression is related to mo ister cond itions . 5; 22 .5 x 8 x 8.5. Total (reconstructed ) length would be although less wa rm than before (text-fig. 3). That allowed ca. 400. the same genus to reappear in the inner part ofthe low er Tagu s bas in at Formi ga, Aza mbuja (later part o f middl e Relationships and d ifferences: The tusk under study can Miocene) (Antunes et 01. , 197 1): it is represented there not beconfused with a gomphothere 's one. Themandibu lar by Deinotherium giganteum, tusks o f the gornphothcres are straig ht or nearly so. and D. giganteum still persisted at Azambujeira (midd le show a distinctly diffe rent section - the lower tusks are level), in Upper Vallcsia n (Antunes, 1984). ventrally plate. slightly convex dorsally. and laterally Both occurrences are related to less-warm but rather thinner. The gomphotherc's upper tusks arc nearly elliptic moist environments. proximally. but very different d istally - entirely flat at the ventra l side and dorsa lly arched ; a broad enamel band is present on the ventral surface ncar the distal extremity 4. Co mpar tsons wlth otbes-eu ropea n a reas (observing the tusk in proximal direction. that band turns progressively to the external side); when the band is In France, the first Deinotherium occur in the "Sables vertica l (and proximal) it can be came reduced and de I'Orlcanais'' (dated MN4b)at Chevilly (Loiret).Cuvier dissociated into fine strips that rem ind the small enamel ( 1822, chapter: Animaux fossiles voisins des Tap irs) has strips on the Q uinta Grande tusk. However the tooth under shown 3 teeth (1822. t. II, pl. IV. corresponding to the study is elliptic in section ove r its whole length and shows text page 222 ): a general arched shape that is not kno wn in gompbothcre's - fig. I: ml (Museum de Paris' collec tion number: tus ks . This sha pe is indee d id entical to tha t from CHE 19); dcinotherc's tusks whenever found in situ on the mandible - fig. 2, M3 (CH E 22) ; (Eppe lshcim. Languenau, Omo). - fig. 3, dp4 (CIiE 20). Furthermore , this deinothere tusk somewhat recalls the dista l part of an upper morse 's tusk as far as size is Kau p ( 1832) estab lished the species Deinotherium conce rned, but it is otherw ise distinctly differen t in shape, cuvieri on these remnants. However; the same species falls in its section. and in geolog ic age. Indeed the Odobeninae in synonymy (as L.G veri fied) o n (or maybe bett er said. appeared much later, on ly in Pleistocene tim es, according after) Deinotheriumbavaricum (H. von Meyer. 183 1). The to Mit chell (1968) . They descend from the Otariidae last spec ies wa s described after some remnants fro m Imagotariinae (Hames, 1979); these on ly appear in Middle Georgensmu nd that have been destroyed. with the sole Miocene. tha t is to say. d istinctly later than the depos ition exception o f a p3 kept at tlJ,c Vienna (A ustria) Museum. o f Lisbon's lower Middle Miocene Vb unit. Georgensmu nd has been reported to MN6 by Abusch-
176 Ctmcias da Terra (UNL). 15
_ ,lJeinoi kerium hava.ricum _
lj'uinla Grande ( CharnecB. Oo J um iar ) .... I1N5 _ Porl u:r ?-
Fig. 2 - Lateral view and proximal section; exceptionally, the surface presents remnants oflongitudinal enaplel strips as well as some fragments ofmandibular bone.
177 Ciencias da Terra (UNL), 15
Table 2
Miocene Deinotherium teeth from Portugal and measurements (mm). (Collections: IGM, Instituto Geologico e Mineiro; 1ST, Instituto Superior Tecnico, Lisbon; MTA, M. T. Antunes).
Species Locality Agel unit! MNI Max. Length Max. Specimen1Collection Dating (Ma) width Deinotherium bavaricwn Quinta das Uppermost not measured not measured Right])l germ IIGM n" Pedreiras (reported Burdigalian/Va2I 7920 by error to the MN 4/17.8 - 16.4 earlier rvs unit) Ma ··~tij·i'TiGMiio··5529""··"····· ···~t~·d~··········..···--······....···· "'id~""'"'''''''''''''''''''''''''''''''''''''''' "'~~t'~;;~;d"" "'~~i";;;'~~;;d"" Barbacena ..Ri'g'hi'M:i'TMiA: · ~~.~ ·id~ · ·6i':S ·· 4'9:0 ··..· .. Pedreiras ..~ft..Mi"j'iGM·~o·5658 id: ·..····· · id~ ·..·..·..·· ·~~t·~;;~';d· ·~~i..;;;-;;:;;~;;d .. ..~ft..~~'iiGM:·if Q;rl~t~·d~ ·· ·id~·(~-;;·V;;Z) ·· · . 5528 (Zby.1941) Pedreiras (ascribed by error to Quinta M 2 da Casinha- 63 62 M 3 Azinhaga do Pinhal, 63 59 ...... ~..~!>...... Left maxillary1IGM If Quinta das id. not measured not measured 5659 with ~_M3 Pedreiras ..~ft..Mi"j'MiA ·id:·.._·· id~ 643 ·· 64:6 ···.. ::!4.~~:M~:z:~~::::::::::::::::::::::::::: :)~;:::::::::::::::::::::::::::::::::::::::::::::::::::: 3~:;::::::::::::::::::::::::::::::::::::::::::::::: :::~I~::::::::::::::::::::::::::::: :::~;?:::~:::::::::::::::::::::::: Left hemimandJ MTA Quinta das
M2 Pedreiras (60) >53 ..~L ?~.:~ s..s..:~ . Mandible1IGM id. id. (Zby., 1973) left right P3 41 41 31 31 P4 49 49 41 41 ~ 6060 ~~ ~ ~~ ~~ ..~J ?~ .7.~ s..? ?.?. . Complete, left tusk 1 id. id, ca. 565 (total) MTA x - 59.8 - 91.9 (proximal) ...... Jpr.?~~L .. ..g~..~~.. g~~~!...~!~ .!~.: }~.: }~.~ k!:~ . Right M31 IGMn" 5530 According to the Id., most probably not measured not measured label, from QUnta Va2 da Farinheira (Vb); but certainly from elsewhere, maybe Quintada ...... •...... _...... •..•...... Barbacena __ . _ _ __ - . Nearly complete left Quinta Grande Langhianl Vbl 300 or - 400 - 59 tusk (Charneca do early MN ':t 16.4 - as (proximal) Lumiar) 153 Ma reconstructed ...... ~..?!:?Jpr..~.~.:L . Mandible 1IGM Quintanelas lower Middle (Zby.1952, p.77) MiocJMN5 Uncomplete right M3 (other teeth lost) ->54 - >52 Deinotherium gigantewn1 Casais da Formiga, upper Middle IGM near Azambuja MiocJMN7-8 (Antunes et 01., 1971) Left p3 >63 >65 ..~.~~..~: ~~ ~.~ . 1ST (Alberdi et 01., 1978; Azambujeira, lower Upper Antunes, 1984) middle level Miocene1MN10 uncomplete m1 not measured not measured
178 Ciencias do Terra (UNL). 15
Siewert (1983) and by De Bruij n et al., (1992); however, ~ A m~:.~~ , AmbienlGl marlnhol it is most clearly MN5 (Ginsb urg, 1999). M. nIItigratlil T In Germany, the oldest Deinotherium is the comp lete Ii f _-. skeleton from Lange nau, MN4 (Stuttgart Mu seum). ~ 0 Let us add that Blai nv ille (1845) has dep icted .;:0 ,0 • 0 0 (Deinotherium, pI. III) a hemimandible (certa inly a left one, 10 0. " 0 but reversed in the figu re) from Chevilly; that figure was "~ l-" TT v'"' present ed again in 1958 by Vaufrey (Traite de T1 ..., Paleontologic J. Piveteau, t.VI (2), p. 249 , fig. 6 1). We 12 :;0 ,z> ~ "', regret that the species bavaricum (named by von Meyer 13 .5? ~ without any figu re in a letter that was pu blished soo n .'C" • 51 "'. after) has the priority over cuvieri, created by Kaup and " 0 E '" ve 15 u established on pub lished (a nd per fectl y we ll drawn) z 0 16 ·w ""j'" Vb specimens. U L1 0 17 Vo' As far as we can ascertain, the first appearance of :>: .r , Deino therium in Germany and France closely agrees with s 16 ., 'Vb available data from Portu gal. I. I N' There is no con tradicting ecologic ev idence from the Si ~ 20 concerned european areas. .E .. 21 E
22 S. Ecology i 23 ~
It is not just by chance that Georges Cuvie r, in 1822, asc ribed isolated molars ofthe later-named Deinotherium Fig. 3 - Temperature and moisture evolutionduring Miocene to a "tapir de taille gigantesque". Such molars are typi (data from lower Tagus basin), modified from Pais, 1999). ca lly loph odon t and brac hyo dont, with distinct crests; unde r these viewpoints there are real likeli nesses towards the tapirs. Hence, ecolog ic compar isons with the extant tapirs (that are typically tropical, humid forest dwellers) interesting assemblage of (highly characteristic for the are logical. Lower-Middle Miocen e transition) thermophilous snakes Other data seem to support these interpretation. Indeed that comprise a large boa (Bavariaboa sp.) and large the evi dence from Lisbon is highl y meaningful because "oriental vipers"(Vipera sp.) (Szyndlar, 2000). there are ma ny palaeo ntologie and geo logic data that Go mphotheres prevail among proboscideans and may allow qu ite accura te env ironment recon stituti on s: indicate the occ urrence of open spaces, but Deinotherium Temperature and moisture control is possible for marine bava ricum is not rare and suggests a nearby forest and non marine levels as well (pollen &spores, foraminifera, developm ent (the whole environment con ditions would mo lluscs, fis hes, rept iles, clay min er alogy, 0 and C be excellen t for this) (PI. 4). iso topes) [see Antunes (2000, table 3, according to J. Pais)] (F ig. 3). b) in Vb times - palaeobotanical and othe r ev idence On the other hand, clearenvironmentdifferences between clearly shows that climate became much drier, even if still Va2 and Vb stratigraphic units have long been recognized. subtropical. Open, savannah areas prevail ed and we re the Successive and more and more acc urate reconstitutions habitat ofprevailing cursorial, rather hypsodont-teethed were produ ced (Antunes, 1965, 1969, 1984; Antunes & rhinoc eros as an other im migrant from Asia: Pais, 1983) . There are some basic facts: Hisp anotherium. a) in Va2 times - warm, tropical to subtropical, moist Primary forest to bush dwellers as Glinds among rodents co nditions prevailed. The presence (among others) oft he (that prevailed before) became scarce and were (for the pelecypod gen us Placuna as we ll as the prevalence of first time) largely superseded by Cricetids. Deinotherium stenotherm, warm wate r s harks (a nurse- sha rk , bavaricum became exceedingly rare. On the contrary, Ginglymostoma; lemon shark, Negaprion; Hemipristis; gomphotheres were plentiful. tiger shark, Galeocerdo) and large barracudas (Sphyraena olisiponensis ) leave no doubt as to environmental condition s rather similar to Senegal' s today - as far as 6. Conclusions sea waters are concenied. As for fresh- or estuarine waters, common Lates and Summing up as far as the occurrence ofDeinotherium catfishes are at least sub tropical and poi nt out to sizable bavaricum is concerned in Portugal - takin g into account freshwater contribution s. Reptilian fauna poi nts out the the chronologie data (Antunes, 2000, p. 294 , tab . 2) as same way: an association of common, typically tropical we ll as th e concerned Mammal -units MN and the crocodilians (large Tomistoma and Gavialis) with a corresponding Iberian biochronology units (Me in, 200 0), monitor li zard (lberovaranus ) and an espec ia ll y this study leads to the following conc lusions.
179 Ciimclas till Terra (UNL), 15
I. Discussion of all evi dence from the lower Tagu s 5.The Deinotherium bavaricum recorded time span Bas in concerning the Lisbon area does not confirm the in Portugal from its first appearance is from ca. 17 Ma presence ofDeinotherium in the TVb local stratigraphic (late Burdigalian) to ca. 15.5 Ma (Langhian), or from the uni t, the firs t one whe re the re is evidence ofthe arrival of end of MN3 to M 5. gomphotheres in Portugal (datum) at 18-18.2 Ma. 8.Deinotherium bavaricum was collected in the small 2. Deinotherium bavaricum is rather common in the Qu intanelas basin in associ ation with elements of the Va2 unit, the main locality being Quinta das Pedreiras. Hispanotherium fauna, thus confirming the Vb evidence.
3. There are indeed two success ive, diachronic data 9. Ranty in the Vb unit is related to environmental for Proboscideans' immigration in Europe: that of changes from a climate optimal event (tropical to subtropical, gomphotheres and later that of deinotheres. The first humid prevailing conditions) that corresponds to the Va2 appearances are: unit to much drier conditions that correspond to the Vb - that of gomphotheres in the IVb unit , ca. 18 -18.2 (ac co mpa nied by forest recession and sa vanna h Ma, MN3 -MN4 limit, NB Iberian unit ; dev elopment). - that ofDeinotherium in the Va2 unit at ca. 17 - 17.2 Ma, MN4: C Iber. unit. 10. It is obvious that deinotheres and gomphotheres had very di fferent environmenta l requi rements. This situation seems general for western Europe Deinothe res certainly needed developped, rich forest areas at least. whose development required a high degree of humidity. Lisbon ' evidence distinctly corroborates this interpretation. 4. The Deino therium bavaricum occ urrence in the Vb unit (ca . 15.9 - 16.1 Ma; MN5 ; Db Ibcr. unit) , amo ng the II . The genus is re presente d much la ter by "Hispanotherium fau na" canno t be dem onstrated on Deinotherium giganteum in the inner part of the lower prev ious data. We can ascertain on the evidence of the Tagus basin at Formiga (later part of Middl e Miocene); ab no rma l tus k described here tha t Deinotherium and atAzambujeira, middl e level (Upper Vallesian) . Both bavaricum indeed occurs in the Vb unit but is occurrences are related to less-warm but rathe r moist exceedingly rare . enviro nments.
References Abusch-Siewcrt, S. (1983) - GebiBmorphologisehe Untersuchungen an eurasiatisehen Anehitherien (Equidae, Mammalia) unter bcsonderer Beriicksichtigung der Fundstelle Sandel shauscn. Courier Forschungsinstitut Senkenberg. Frank furt an Mein, 62: 1-36J. Alberdi, M. T.; Antunes, M. T.; Sondaar, P. Y. & Zbyszewski, G.(1978) - Les Hipparion du Portugal. Ciencias do Terra (UNL) , Lisboa, 4: 129-156,1 0 fig., 2 pl. Antunes, M.Telles (1965) - Notessur la geologic et la paleontologic du Miocene de Lisbonne. I - Stratigraphieet faunes terrestres. Bolet tm Sociedade Geologica Portugal. XIII: 257-276. Antunes, M. Telles (1960) - Notes sur la geologic el la paleontologic du Miocene de Lisbonne. V -Un Schizotheriine du genre Phyllotitlon (ChaJicotherioidea, Perissodaelyla) dans lHelvetien V-Bde Chameea do Lumiar. Remarques ecologiques sur la faune de mammiferes. Boletim Sociedade Geologica Portugal. Lisboa, XVI: 159-178. Antunes, M.Telles (1969)- Mamiferos nilo marinhos do Miocenico de Lisboa: ecologia e estratigrafia (Nota preliminar). Boletlm Sociedade Geologica Portugal. Lisboa, XVII: 75-85.
Antunes. M. Telles (1984) - Essai de synthese sur les Mammiferes du Miocene du Portugal. Voilime d 'Hommage au geologue G. Zbys zewski. Ed. Recherche sur les Civilisations, Paris: 301-323.
Antunes. M. Telles(1990) - The Proboseideans data, age and paleogeography: evidence from the Miocene of Lisbon. III E. II. Lindsay et 01. (Edit.) European Mamm al Chronology, Plenum Press: 253-262.
Antunes, M. Telles - Miocene mammal s from Lisbonandgeologic age. Ashowcase formarin e-continental corre lations. Ciencias do Terra (UNL). Lisboa, 14: 343-348.
Antunes, M. Telles; Legoinha, P.;Cunha, P. & Pais, J. (2000) - High resolution stratigraphy and Miocene facies correlation in Lisbon and Setubal Peninsula (Lower Tagus basin, Portugal. Ciencias do Terra (UNL). Lisboa, 14: 183-190.
Antunes,M.Telles & Pais, J. (1983) - Climate during MioceneinPortugaland itsevolution. Paleobiologie continentate. Montpellier, XIV(2): 75-89.
180 Ciincias do. Terra rUNL). 15
Antunes, M. Telles; Veiga-Ferreira, O. da & Zbysz ewski, G (1971) - Mamiferos do Miocenico supenor do areeiro da Forrniga (Aza mbuja). Baletim Sac. Por tuguesa Ciencias Naturais. Lisboa, XIII, 2' serie: 25-31, 3 fig., 4 pI.
Barnes, L. G (1979) - Fossil Enaliarctinae Pinnipeds (Mammalia: Otar iidae) from Pyramid Hill, Kern County, California. Contributions ill Science. Natural History Museum Ang eles County. 318: 1-41.
Bergounioux, F.M.; Zbyszewski, G & Crouzel, F.(1953)· Les Mastodontes rniocenes duPortugal.Memoires Services Geologiques Portugal, Lisboa, N. S., I , 139 p. .
Bru ijn, II. de; Dasms, R.; Daxner-Hock, G; Fahlbusch, V.; Ginsburg L.; Mein, P. & Morales, J. (1992 - Report of the RCMNS working group on fossil Mammal s, Reisensburg 1990. Newsletters an Stratigraphy. Berlin, 26 (2/3): 65-118.
Dietrich, W. O. (19 16) - Ober die Hand und den FuGvon Dinotheri um. Zeitschrift del' Deutschen Geologischen Gesellschaft, B. Monatberichte. Berlin, 68 (1-3): 44-53.
Ehik. J. (1930)- Prodinotherium hungaricum nov. gen., nov. sp. Geologia Hungarica. Ser. Palaeont., Budapest. 6: 1-24.
Gaud ry, A. (1862) - Animas fossiles et Geologie de FAttique. 466 p. E. Sav y ed. Paris.
Ginsburg L. (1983) - Sur les rnoda lites d' evolut ion du genre neogene Pseudaetums Gervais (Felidae , Carnivora, Mamma lia). Colloque CNRS, Paris, 330: 131-136.
Ginsburg L. (2000) - Chronologie des de pots miocenes du Blesois ala Bretagne. Symbiose, Orleans, N. S., 2: 3-16, 11 fig.
Ginsburg L. & Antun es, M. Telles (1967) - Considerations sur les Mastodontes du Burdigalien de Lisbonne et des sables de I' Orleanais (France). Revista Fac. Cienc.Lisboa, 2' serie - C - XIV(2): 135-150, 1 fig.,4 pI.
Gohlich, U. B.(1999) - 77,e Miocene land Mammals in Europe. 13. Order Proboscidea Verlag Friedrich Pfeil edit. MOnchen: 157-168.
Harris, J.M. (1973) - Prodeinotherium from Jebel Zelten, Libya. Bulletin British Museum (Natural History), Geology, London, 23(5): 285-348.
Heinzmann, E. P. J.; ( 1973) - Die Carnivoren des Steinheimer Beckens. B. Ursidae, Felidae. Viverridae undsowie Erganzungen und Nachtrage zu der Mustelid ae. Palaeontographica. Stuttgart, supplement-Band VIII: 1-95.
Heinzmann, E.P. J. (1992) -Das Tertiar in SOdwestdeutschland. Stuugarter Beitriige zur Naturk unde. Stuttgart, C, 33: 1-61.
Heinzmann, E. P. J., Duranthon, F. & Tassy,P. (1996) -Miozane Grolssaugc tiere. Stuttga rter Beitrdge Zlir Na turkunde. Stuttgart, C,39: 1-60.
Kaup, J.(1832) . Description d 'ossernentsfoss iles inconnusjusqu 'Qpresent qui se trouvent au Musee grand-ducal de Darms tadt . fasc.l, Deinotherium: 1-16. Me in, P.(2000) -La biochrono logie des Mammi feres neogenes d' Europe L' echelle MN, son application a la succession des faunes du Portu gal. Ciencias do. Terra (UNL). Lisboa, 14: 335-34 2.
Meyer, H. von (1834) . Diefossilen Zahne W id Knochen Wid ihreAblagerung in der Gegen VOIl Georgensmund ill Beyern: 1-126.
Meyer. H. von (1831) . Mitteilungen an geheimen Rath von Leonhard. Jahrbuch fii T Mineralogi e,Geognosie, Geo logie und Petren faktenkunde, Stuttga rt, 2: 296-297.
Mitchell, E. (1968) - The Mio-Pliocene Pinniped Imagotoria. JOIIl'1lal ofthe Fisheries Research Board of Callada, 25 (9): 1843- 1900.
Pais, J.(1999) - Historia geo logica da Peninsula de Setubal nos ultimos 20 milhoes de anos. Guia de visita de camp o integrada. Programa Ciencia Viva, Geologia no Verno. Centro Estudos Geologicos. Monte de Caparica, 21 p.
Szyndla r, Z. (2000) - The snakes (Reptilia, Serpentes) ofthe Miocene of Portugal. Ciencias do. Terra (UNL), Lisboa, 14: 359-364, 2 fig., 1 tab.
Tobien , H. (1962) - Ober Carpus und Tarsus von Deinotherium giga nteum Kaup (Mamm., Proboscidea). Paldontologische Zeitschrift, Stuttgart: 231-2 38. Tob ien, II. (1988) - Contributi on al'etude du gisement miocene superieur de Montredon (Herault). Les grands Mammi feres. 7 • Les Proboscidiens Deinotheriidae. Palaeavertebrata. Memoire extraordinaire, Montpellier:135-175.
Vaufrey, R.(1958) - Proboscidea Etude systematique. Troise de Paleontologle dir. J. Piveteau, VI(2): 203-295, Masson edit. Paris.
Zbyszews ki, G. (1941 ) - Note sur la decouverte du genre «DINOTHERIUM» au Portugal. Comuntc. ServoGeol. Portugal, XXII: 39-44, I pI.
Zbysz ewski, G (1949) - Les Vertebres du Burdigalien superieur de Lisbonne. ServoGeol. Portugal. Mem ., 77p., 22 pI.
Zbyszewski, G (1952) -Les mammiferes rniocenes de Quintanelas (Sabugo). Comunlc. ServoGeol. Portugal, XXXIll: 65-82, 10 pI.
Zby szewski , G (1973) · Note sur une mandibule de Deinotherium de ['Helverien de Lisbonne. Comunic. ServoGeol . Portugal. LVI: 101-105, 1 pI.
181 Ciencias da Terra (UNL). 15
Plate 1
Deinotherium bavaricu m from Quinta Grande, left incisor
Fig. I - (a) Median view; at the proximal part there are some remnants ofthe mandibular bone; (b) lateral view, showing remains ofenamel strips in its proximal part (deta il fig. 2). x 1/2.Seale, 10 em.
Fig. 2 - Latera l view, deta il to show remains ofenamel strips. Natural size. Scale, 5 em
Fig. 3 - (a) Prox imal view to show the begi nnings of the axia l hole (sha ped as an upside-down cone); (b) distal extremity. Natural size.
182 Cisncias PLAT E I 10 em 5 cm 183 Ciincias cia Terra (UNLJ. 15 Plate 2 Deinotherium bavaricum from Quinta das Pedreiras (Va2 unit, M 4), fossils with the typical black, pyrolusite gangue - M.T.A. collection Fig. 1-3 - Complete left tusk median (I), anterior (2) and lateral (3) views. Scale, 10 em. 184 Ciincias cia Terra rUNL). 15 PLATE 2 185 Ciencias da Terra (UNL ). 15 Plate 3 Deinotherium bavaricum from Quinta das Pedreiras (Va2 unit, MN4), fossils with the typical black, pyrolu site gangue - M.T.A. collection Fig. 1-2 -Incomplete left hemimand ible with M2 (part) and M3, lateral (4) and occlusal (5) views. Scale, 10 em. 186 Ciencias da Terra fUNL), 15 PLATE 3 10 em 1 10 em 2 187 Ciencias da Terra ( UNL ). 15 Plate 4 Deinotherium bavaricum, a reconstitution. Coloured after a sketch by M.T.Antunes. 188 Ciencias da Terra (UNL), 15 PLATE 4 189