Rivista Italiana di Paleontologia e Stratigrafia I I I 0"r,""--;--;Ilicembreleee l NOTA BREVE - SHORT NOTE KOGIA PUSILLA FROM THE MIDDLE PLIOCENE OF TUSCANY (ITALY) AND A PHYLOGENETIC ANALYSIS OF THE FAMILY KOGIIDAE (ODONTOCETr, CETACEA) GIOVANNI BIANUCCI 't E \IALTER LANDINI '!'t Receired. JanuarT 22, 1999; accepted JuQ 27, 1999 Key-oords: Cetacea, Kogiidae, Systematics, Phylogeny, Middle Pliocene, Tuscany, Italy. Riasswnto. Viene preso in esrme un crrnio quesi completo di cetaceo odontoceto provenienre dai sedimenti del Pliocene medio di Monte Voltra.io (provincia di Pisa, Toscana). Tale esernplare, in pessato erroneamente attribuito alla famiglia Ziphiidae e descntto come olotipo della specre Hyperoodon pusillus, viene qui riferito al genere Kogla (farniglia Kogiidae). La specie Kogia pusiila è ridescritta e messe a confronto con le specie atruali K. brniceps e K. simus. Da un punto di vista filogenetico, viene ipotízzart una sepa- razione antica (ahneno nel Miocene inferiore) dei Kogiidae e Physe- teridae (Physeteroidea). Il mancato ritrovamento di reperti attribuibili con sicurezza alla famiglia dei Kogiidae in sedimenti piìr antichi del Miocene superiore è probabilmente dovuto alla raritàL di quesri cetecei. Lanalisi delle relazioni filetiche tra i Kogiidae è stata condotta pren- dendo in esame supposte apornorfie relative alla morfologia e/o esten- sione del bacino sopracraniale che alloggia le sacche aeree e l'organo dello spermaceti. Abs*act. A partial skull of an odontocete cetacean from Middle Pliocene sediments of Monte Voltraio (Pisa Province, Tuscany, Italy) is exanrined. f'his fossil, erroneously referred to the fanily Zrphitdte and described in the past as holotype of the species Hyperoodon pusil- /as, is assigned here to the genus Kogia (family Kogiidae). The species Kogia pusilla is redescribed and compared to the living species K. bre- oiceps rnd K. simus. PhylogenericalÌy, an old separation (at least in the Lower Miocene) of Kogiidae and Physeteridae is suggested. The lack of sub- stantiated kogiid records until the Upper Miocene is probably due to the rarity of these cetaceans. Phyletic analysis within the Kogiidae is undertaken and supposed apomorphies in the morphology andlor the extension of the supracranial basin that houses the large air sacs and the spermaceti organ are considered. lntroduction. The Kogiidae are a family of small odontocetes (Cetacea) similar and closely reiated to the extant sperm whale (Physeteridae). !íell-preserved records of Kogi- "1,-1r l nulero 3 idae are very rare: only two almost complere skulls from late Miocene sediments of Mexico (Barnes, 1973,1984) and Peru (Muizon, 1988) have been described previous- ly. Those specimens are the holotl'pe oI rhe extincr gen- era and species Praekogia cedrosensis and Scaphokogia cochlearis, respectively. Other presumed records of kogi- ids are fragmentary and consist essentially of mandibu- lar fragments (Kel1ogg, 1929), isolated teeth (Matsumu- to,1927) and/or auditory bones (Pilleri, 1986a, 1986b, 1987,1988; Pilleri et al., 1989). The skull from the Pliocene of Tuscany described here represents the third known significant record of this family. This specimen s/as referred erroneously to the Ziphiidae in the past. Capellini (1893) assigned it to Placoziphiws and Pilleri (1987) described it as holotype of the species Hyperoodon pusillus. This fossil sku1l, already referred to the genus Kogia by Bianucci (1997) and Bianucci et ai. (1998), is redescribed in detail in this paper. Kogia pusilla represents the ancesror of Kogia bre- viceps and K. simus, the only two living species of this family. Systematic description. Class Mammalia Linnaeus, 1258 Order Cetacea Brisson, 1762 Suborder Odontoceti Flower, 1867 Superfamily P hys e t e r o i de a (Gray,1821,) G111,1.872 Family K o g i i d a e (Gill, 1871) Mlller, 1923 Genus Kogia Gray, L846 'r- Museo di Storia Naturale e del Territorio, via Roma, 103, 5601 1 Calci, Pisa (Italy). E-mail: bianucci@dst.unipi.it 'r''t Dipartimento di Scienze della Terra, via S. Maria, 53,56126 Pisa (Italy). E-mail: landini@dst.unipi.it 446 G. Bianwcci f: W Land.ini Kogia pusilla (Pilleri, 1982) (Fig.1-aa) 1893 Placoziphizs, V Beneden - Capellini, pp. 287-288. 1987 Hyperoodon pusillus Pilleri, pp. 35-36, fig. 11, pl. 13. 1997 Kogia pusilla (Pillerí) - Bianucci, pp. 172-1,73, tig. 6b. 1998 Kogia p wsilla (P rllerl) - Bianucci et al., pp. 1,24, 126, 127, Íigs. 3t1,22 . Emended diagnosis of species. A species of the genus Kogla that differs from K simus and K. brmicEs in having a more elongated rostrum, a less anteriorly extended antorbital processr a smaller lacrimal, a slighrly more prounonced dorsal asymmerry and a lesser elevation of the posterior region of the neurocranium. It differs from !r"?:"r*Ot in irs sn'raller size and in having a more narrow sagittal Holotype. IGF1540V distorted and incomplete skull, lacking part of the right side and of the ventral portion of the braincase. Audi- tory bones, teeth, rnandible and postcranial skeleton are not presewed. The specirnen s.as donated by Lawley in L877 to the Museo di Paleon- tologia of the University of Florence, where it is preseroed today. Type locality. La Rocca locality in Monte Voltraio hill, near Volterre, province of Pisa, Tuscany (Italy). Age. Middle Pliocene: Globorotalia crassaformis crassaformis subzone of Iaccarino El Salvatorini (1982), approximately between 3 M-A and 2.6 MA (Bianucci et al., 1998). Description. The condylobasal iength of the only known but incomplete skull is estimated to be about 270 mnr. This skull size is similar ro rhar ol Kogia simus and it is slightly smaller than that of an adult of Kogia bred- ceps. In fact, according ro Ross (1979), the condylobasal length of 24 skulls of Kogia simws range between 201 mm and 323 mm whiie the condylobasal length of 22 skulls of K. breviceps range berween 237 mm and 467 mm (less than 300 mm are young animals). Compared to the living species, the rostrum has a simiìar triangular shape from a dorsal view; but differs in being rnore elongate (Fig. a). Moreover, it has a well- developed dorsal concrrvity and a venrral convexiry rhar u.'e harve not observed in extant species. Post morten-r deformation may have accenruared these last characters in the fossil specimen. The cerebral skull has a sernicircular outline in dorsal view. In this respecr the skull is similar to thar of K. breoiceps and K simus but differs from the anrero- posteriorly elongated skulls of both Praekogia and Scaphokogia. Nevertheless, this skull differs from those of extant Kogia spp. in having a more pronounced dor- sal asyrnmetry (due to the displacement of the external nares tosrard the left side and to the displacement of the sagitt:ìl crest tor,'ard the right side), and in hrving small- er elevation of the posterior region of the neurocranrum. The first of these two fearures may have been empha- sized by Dost mortem deformation of the skull. In dorsal view (Figs. 1,a1, la2) the premaxillae, the maxillae and probably rhe vonìer (covered by marrix) extend as far as the apex of the rosrrum, as in living species. A simiiar exrension of the rosrral bones was observed also in Scaphokogia by Muizon (1988). Muizon (1991) considered this feature an apomorphy of the Kogiidae. Two asymmetricai superior nares are present at the base of the rostrum as in all Physeteroidea (Physe- teridae and Kogirdae). In particular, the left naris is 14 mm wide and 10 mm long whiie the right is 6 mm wide and 10 mm long. The nasals are missing, as in the other Kogiidae. The ieft premaxilla exrends posteriorly as far as the postero-dorsal margin of the skull and its medial margin joins the right maxilla ro form a very promlnenr sagittal crest, as in living species. This crest is bent left in its anterior porrion. It is delirnited by vertical walls and is very narros/ ar the verrex (width: 13.3 mm), as in K. simus. The width of the sagirtal crest is a diagnostic character for the distinction of the two living specles (Handley, 1966). Ross (1979) observed that in 26 skulls of K. simus the minimum wrdth of this crest ranges from 5 to 20.5 mm while in 22 skulls ol K. breaiceps this n-idth ranges from 14.5 to 49 n-rm (from 23 to 49 mm exclud- ing five young specimens). The right premaxilla forms a small elevated basin on the neurocranium, extending from the sagirr.ll crest on the left and to its lateral margin on rhe right. A sìrr- ilar depression is present in the living species of Kogia where the spermaceri organ (Kernan & Shulte, 1918; Raven & Gregory, 1933; Schenkkan Ec Purves, 1973) resides. In contrasr, in Praekogia, rhe lareral rìlargin of the right premaxilla is not protuberant and the r.igbr maxilla and the right prerl-raxilla forrl a single craniai fossa that probably resulted in a more developed sper- naceri organ rhan n KogiLl. Scaphokogia shows an even more dissimilar dorsal cranill architecture than Kogia.In fact, the skull of Scaphokogta h:rs no sagitt:r1 cresr ar all (the facial crest is displaced near rhe left m:.rrgin of the neurocraniunr) rnd con\eqLrenrly it is likely rlr.rr the spermaceti organ also invaded the left side of cerebral skull (Muizon, 1988). The antorbital notcl-res penerrate the supracranial basin as in irving species o[ Kogia and in Praeleogia,but nor in Scapbokogia and in the other Physeteroidea. This character is considered by Muizon (1988) to be an apo- morphy of the Kogiinae (Kogia and Praelectgia) re1:rted to the large antero-lateral development of the supracrxnial basin. Kogta pusilla (Pilleri, 1982) holotype; La Rocca (lGF1540V). a1,a2, dorsal view; b1,b2, vcntral view sagittal crest; Vo, vomer. Iocaìity, Monte Voltraio (province of Pisa, -l'uscar-ry). Almosr complete skull cf, cranial fossa; Fr, fronral; L;r, lacrirnal; Mx, maxilla; n, naris; Pn-x, prerlaxillae; sc, Fig. 1 Kogia pusilla from Middle Pliocene of Tuscany 448 G. Bianucci & V/. Landini 10 cm The antorbital process is narrower than in living species and does not extend anteriorly on the proximal portion of the rostrum. As a consequence, the cranial f ossae (antero-laterally delimited by the antorbital processes) are less developed antero-posteriorly in this fossil species. The anterior portion of left cranial fossa is almost completely covered dorsally by the lateral and sagittal crests, both being bent towards the fossa. Posteriorly, the cranial fossae are slightly more elevated than in K brear.ceps. This feature makes K. pwsilla very different from K. simws, in which the posterior part of the craniai fossae rises strongly (Handley, 1966; Ross, 1979). The iateral view (Figs. 2ay 2a2) shows a larger antero-posterior exposure of the rostrum than in the liv- ing species. This condition is due to both the greater elongation of the rostrum and the smaller anterior extension of the rrntorbital processes (in K. brer:iceps and K. sìwws the antorbital processes cover the proximal part of the rostrum) . The anterior portion of the lacrimal is smaller than in the living species and is comma-shaped as in the Phy- seteridae, whereas it is triangular in the other Kogiidae. There is no evident lacrimal-maxilla suture. Figs" 2a2 and 3 present the probabie course of this suture: the pos- tero-dorsal process of the lacrimal is large and posteri- Fig.2 - Kogta pusilla (Pilieri, 1982) holotype; La Rocca localitS Monte Voltraio (province of Pisa,'Iuscanv). Almost con- plete s kull (I GF 15 40V). a1,a2, laterel view; b1,b2, pos- terior view. Fr, frontal; La, . lacrimal; Mx, naxilla; Oc, occipital; sc, sagittai crest. orly exìended, to a greater extent than in the other Kogi- idae. Muizon (1988) noted that the posrero-dorsal process of the lacrimal is not present in the Physeteridae except in a young specimen of Awlophyseúer, and the presence of this process in the Kogiidae is considered as a plesiomorphic character. The supraorbital process is thinner than in both Iiving species and Praekogia. The lateral view shows also an elevation of the posterior portion of the cranium, smaller than rn living species and in Praebogta. In posterior view (Figs. 2bi.2b') the sagirral crest is high, narro'w and delimited laterally by vertical wal1s. In ventral view (Figs. 1b1, 1b2) the skull is badly damaged and most of the ventral surface of the neuro- craniun is not preserved. ln particular, only the left frontal, the left lacrimal and a small porrion of the left squamosal are preserved. The medial groove of the ros- trum allows the vomer to be observed. Along the right preserved margin of the rostrum is a sma1l deep furrow but without distinct alveo1i. In this respect K. pusilla is similar to the living species and differs îrom Scaphokogia (with distinct but shallow alveoli). Comparisons. A peculiar kogiid feature of the only skuli known of l{ogia pusilla is the lack of both nasals. The presence o[ a supracranial basin and the strong asymmetry of the skul1 are other characrers typical of all the Physeterida (Kogiidae and Physeteridae). Kogia pusilla is similar to K. breoiceps and K simus in the pecu- liar outline of the skull in dorsal view (the cranium is semicircular and the rostrum is triangular), and in hav- ing a vertical sagittal cresr on the cranium (Fig. a). Kogia pusilla from Middle Pliocene of Tuscany 449 Fig. 3 - Kogia .pusilla (Pilleri, 1982) holotype; La Rocca localitv, Monte Voltraio (province of Prsa, Tuscany). Neurocrani- um ìn lateral view of skul.l (IGF1540V). Fq fronral; l.a, lacrimal; Mx, rnaxilla; Oc, occrpital; porp, posrorbital p roce \s i \q. : qur mo 'eì; r f. temporal fossa. Moreover, as rhe extant species, K. pwsilla shows a srnall basin on the cerebral porrion of the right premaxilla that extends as far as the postero-dorsal margin of the skull. K. pusilla is more similar to K. simus than K. breuiccps in the relatively small size of the skull and in the narrow sagittal crest at the vertex. Fig. 4 - Comparisons of the skull in dorsal view of: a, Kogia pusilla (restoration of the skull based on holotype, IGF 1540V); b, Kogia brniceps (Laboratoire de Anatomie Comparée de Paris, 1833-483); c, Kogia simus (Accademia dei Fisiocritici di Siena). ,1 cm, 450 G. Bianucci G \Y Landint K. pwsilla differs from both K. brersiceps and K. simws in its longer rostrum, smaller lacrimal and perhaps grear.er asymmetry of its skull in dorsal view. Phylogenetic relationships. The genus Kogia shows evident affinities with Physeter. Common apomorphies of these ts/o genera concern particularly the facial anatomy (Heyning, 1989) and the concave shape of the dorsal surface of the skull (scaphidiomorphy) for housing the large air sacs and the spermaceti organ. Consequently, the singie bones of the skull suff ered large modifications, in parricular, a marked asymmetry and the reduction of the right exter- nal nares. Even if there is consensus on the close affinities of Kogia and Physeter, precise systematic relationships between these two genera are debated. Some authors (Caldwell & Caldwell, 1989; Fraser & Purves, 1960; PHYSETERIDAE QUATERNARY 1.8 MA Handley, 1966;Heyning, 1989; Simpson, 1945) consider that Kogia and Physeter belong to the same family Phy- seteridae. Others (Barnes, 1,973; Kaxrya, 1.973; Miller, 1.923; Muizon, 1991) think these rwo genera have such distinctive characters that their separation in two fami- lies (Kogiidae and Physeteridae) is justified. This iast hypothesis was strengrhened by recenr studies on cetacean mitochondrial DNA (Arnason et al., 1.993; Milinkovitch et aL.,1.994) that confirm a large divergence between Physeter and Kogia. We believe that the fossil and phylogeneric evi- dence reveal that the separation of these two groups may be relatively old. In particular, all fossil skulls of Physe- teridae show a marked enlargement of the posterior por- tion of the right premaxilla, an apomorphy absent in Kogia and in fossil taxa relared to this genus. As observed by Muizon (1991) this feature is relatively less emphasized in living Pbyseter than in fossil physeterids. This derived character is also present in the most ancient known skull of physeterid, assigned to Diaphorocetus Kogia pusilla KOGIIDAE PLIOCENE MIOCENE OLIGOCENE Fig. 5 - Phylogeny of the Kogiidae. Dotted lines indicate no fossil or fragmentary record. Kogia pusilla from Mid,dle Pliocene of Tuscany 451 K. breviceps K. simus K. pusilla Praekogia Scaphokogia Diaphorocetus Squalodon Fig. 6 - Maxinun-parsimony cladogram (cosistency index : 100) conputed from the data in Tab. 1 using Henning 86, ver- sion 1.5 (Farris, 1988). powcheti (Early Miocene from Argentine; Moreno, 1892; Lydekker,1894) The primitive state of Kogiidae, regarding the shape of right premaxillae, suggesrs rheir origin is no younger than the beginning of rhe Miocene (Fig. 5). The lack of of kogiid records prior ro the Late Miocene may be due to their low frequency in the ancient sea. Flowewer, some isolated auditory bones and reerh from the Lower Miocene of Piedmont (North Italy) and Ba1- tringhen (Austria) and Middle Miocene of Visiano (North Italy) were referred ro this family by Piileri (1986b, 1988) and Pilleri et al. (1989). Nevertheless, because these elements (periotics and teeth) are rhe only remains of kogiids found in the sediments, their conclu- sion must be viewed with caution. As in Kogra, some periotics show an articular surface for the tympanic flat and parallel to the general plane of the periotic and not ventrally oriented; nevertheless, on the whole rhey dif- fer substantially from Kogia periotics and show also affinities with those of some fossil physeterids. The dis- tinction of the two families on rhe basis of rhese struc- tures is problematic, considering also that skulls of fos- sil physeterid and kogiid genera with preserved auditory bones are rare. The isolated teerh are referred to Mioho- gia elongata (emended of Miokogia elongatum), a genus and species created on rhe basis of three teeth from Bal- ringen (Piileri, 1986b). In fact, all teeth referred to this presumed kogiid are similar in shape and size to those of the physeterid Scaldicetws bellunensls frorn the Early Miocene of Belluno (North Italy) (Dal Piaz, 1977).In accordance with orher authors (Barnes, 1976; Muizon, 1988) we attribure a 1ow systematic value to isolated teeth and therefore consider nr.tmina dubia borh senus and species names of Miokogia elongata. Considering other described kogiid fossils, only two skulls were referred correctly to this family. One, from the Late Mìocene of Mexico (Barnes, 1973,1984), was named Praekogia ced.rosensis, and the other, from the Late Miocene of Peru (Muizon, 1988) was named Scaphokogia cochlearis. Praekogia and Scaphokogia share wrth Kogia the loss of nasal bones, an apomorphy of the family Kogiidae. Nevertheless, the rwo fossil genera dif- fer from Kogta by having a different architecture of the supracranial basin. In parricular, in Praekogia the basin [or the spermaceri orgrn exrend\ ro rhe entire right por- tion of the dorsal surface of the cerebral skull that is formed by the premaxilla and maxilla. As observed by Barnes (1.973), this genus lacks a distinct right maxillary fossa that is present in Kogia. The basin for the sperma- ceti organ in Scaphokogia takes up rhe entire dorsal sur- face of neurocranium and only a large cranial fossa is present. Because of this peculiar shape of the supracra- nial basin and other peculiar characrers, Scapholeogia was assigned to the subfamrly Scaphokogiinae by Muizon (1988). Apparently Scaphokogia shows affinities with Pbyseter in having a large cranial basin for the spermaceri organ. Nevertheless, rhe cranial basìn of Scapbokogta does not extend on the rostrum as in Pbyseter. A cranial basin more anteriorlyextended than rn Scapholeogia is present also in Kogia and Praekogia, both having the dorsal depression that exceeds the external nares. More- Taxon Character I 2 7 + 5 6 7 8 9 121110 l3 I4 Siua,lodon Diaphorocetus 0 0 1 0 1 2 2 2 2 2 00 ; t;UI ,T- >:0 00 -00 f-00 0 00 00 , : o _f T01 r01 T01 0 0 0 i. 1 1 000 ò o-- nnln OO:i ll110 111 11. 1 ;--T--" ^-^UU olo ::0 00 0 ll 0 Scaphobogia 0 Praekogia 1 1 1 1 K"4o p"tr\ Kogia simus Kogia breoiceps 1 1 1 Data matrix used in the cladistic analysis. 't= missing characters. Characrer states: 1) Supracranial basin: O, no; 1, yes; 2) Strong rsym- metry: 0, no; 1, yes; 3) Lack of nasals: 0, two nasalsl 1, one nasal; 2, nasals absent; 4) VomeS premaxillae and maxillae reach the apex of rostrum; 5) Enlargement of posterior portion of right premaxillae: O, no; 1, yes; 6) Antorbital notch penetrating rhe cranial fossae: O, no; 1, yes, 7) Supracranial basin formed by only one hemispherical depression posteriorly developed and anteriorly closed (see Muizon, 1988, 1.991.,for more detail on the bone modifications): 0, no; 1, yes; 8) Craniurn half-circle shaped: O, no, 1, yesl 9) Rostrum triangular: 0, no; 1, yes; 10) Small basin on the cerebral portion of right premaxilla: O, no; 1, yes; 11) Very prominent sagittal crest: O, no; 1, yes; 12) Very large antorbital process and lacrimal; 0, no; 1, yes; 13) Rostrum verv shorr: 0, no; 1, yes; 14) Relarively broad sagittal crest: 0, no; 1, yes. Tab. 1 452 G.Bianucci&WLandini over, in these last two genera, the cranial basin takes up the antorbital processes and consequently the antorbital notches penetrate the cranial fossae. This last character was considered as an apomorphy of the Kogiinae (Kogia and Praekogia) by Muizon (1988). Considering interrelationships among the three species of the genus Kogia, apparently K. brez,iceps and K. simws are more derived than K. pusilla in having more anterioriy extended antorbitai processes (and conse- quently more elongated cranial fossae) and a shorter rostrum. A parsimonious interpretation would suggest that the wide sagittal crest of K. brer.ticeps is an apomor- phy, considering the primitive status of the narrow crest observed in K. pusilla and K. simus. The characters discussed belovr are used to create a maximum-parsimony cladogram of the relationships of Kogia pwsilla using Henning 86, version 1.5 (Farris, 1988) (Tab. 1, Fig. 6). Regarding other fossils referred to the family Kogiidae in the past, we agree with Barnes (1973) and Muizon (1988) on the attribution of Kogiopsis floridanus from the Miocene of Florida (Kellogg, 1929) as a prob- able physeterid and in considering the name of the species Kogia prisca (Miocene?, from Japan; Matsumoto, llt1l/ ) a nomen au0tum. Pilleri (1986a, 1982) identifies the living species Kogia simus and K, breoiceps in the Pliocene of Tuscany on the basis of isolated teeth and periotics. \fe consider these fossils and other unpublished specimens as belonging to the Kogiidae and most probably to the gentts Kogia. Nevertheless, we do not agree with Pilleri in referring these f ossils to the two living species because the periotics show different characters and the teeth are not diagnostics at this systematic level. It is possible that most of these specimens belong to the fos- sil species Kogia pusilla. Fossil kogiids were reported recently from Pliocene sediments of the Yorktown Formation (Vhit- more, 1994) and northern Appennines (Cigala Fulgosi, 1996), but they were not described in detail. Conclusions. The assignation of the specimen examined here to the Kogiidae confirms the presence of this farnily in the Pliocene of Tuscanl', already documented by Pilleri (1986a) on the basis of isolared teeth and periotics. Other auditory bones from these sediments are under study by the authors. K. pwsilla is a primitive species of the genus Kogla that differs significantly from living K. breaiceps and K. simus in having a smaller antorbital process and a more elongated rostrum. The phyletic relationships between the Kogiidae and the Physeteridae were examined and an origin of the Kogiidae older (at least Early Miocene) than that sug- gested by the paleontological data (Late Miocene) is suggesreq. In accordance with Muizon (1988, 1991), the Kogiid,ae are separated into two subfamilies on the basis of different architecture of the supracranial basin: the Kogiinae (Kogia and Praekogia) and the Scaphokogiinae (Scaphokogia). The morphology of rhe supracranial basin of Kogia pusilla is similar to that o[ K. breaiceps and K. simws eyen if the basin of the fossil species differs in having a smaller anterior exrension. Acknouledgements. 'We are very grateful to the following persons for grrnting us access to cetacean collections and for their assistance during our vrsirs at the museums: P. Mazza and F. Cozzini (Museo di Geologia e Pale- ontologia, Università di Firenze); D. Robineau (Laboratoire de Anatomie Comparée de Paris); F. Cancelli (Accademia dei Fisiocririci di Siena). \7e also wish to thank C. de Muizon (Museurn Natinal d'Histoire Naturelle de Paris) for critically reading the rnanuscnpr. The photographic plates were prepared by the euthors and M. Gini (Dipartimento di Scienze della Terra, f Inir.ersità di Pisa) RE,FERENCES A.r,rror Ú., Grllb..g A. 6c \X/idegren B. (1993) - Ceraceen mitochondrial DNA control region: sequences of all extapt baleen whales and two sperm r.vhales..species. Mol. Biol. Eool.,v. 10 (5), pp. 960-97A, Chicago. Barnes L.G. (1973) - Praekogia cedrosensis, a new genus and species of fossil pygmy sperrn whales from Isla Cedros, Baja California, Mexico. Contr. Sci., nat. Hist. Mus. Los Angeles C., v. 247, pp. 1-20, Los Angeles. Barnes L.G. (1976) - Outline of eastern Noth Pacific fossil cetacean assemblages. Syst. Zool., v. 25 (a), pp.321-343, New Haven. Barnes L.G. (1984) - Fossil odontocetes (Mammalia: Cetacea) from the Almejas Formation, Isla Cedros, Mexico. Pale- obios,v. 42, pp. 1-46, Berkeley. Bianucci G. (1997) - The Odontoceti (Mammalia, Cetacea) from Italian Pliocene. The Ziphiidac. Palaeont. Ital., v. 84, pp. 1,63-192, Pisa. Bianucci G., Sarti G., Catanzariti R. & Santini U. (1998) - Middle Pliocene cetaceans from Monte Voltraio (Tus- cany, ItaÌy). Biostratigraphical, paieoecological and pale- oclimatic observations. Rio. Ital. Pal. Strat., v 104, pp. 1.23-13A, Milano. Capellini G. (1893) - Nuovi resti di Ztftoidi in Calabria e in Toscana. Atti. R. Accad. Lincei, Rend., v. 2 (7), pp. 283- 288, Roma. Cigala Fulgosi F. (1996) - Rere oceanic deep weter squaloid sharks from the Lower Pliocene of the Northern Apen- nines (Parma Province, Italy). Boll. Soc. Paleont. Ital.,v. 3a (3), pp. 3A1-322, Modena. Caldwell D.K. & Caldwell M.C. (1989) - Pyg-y and dwarf sperm whales. In: Ridgway S. H. & Harrison R. (Eds.) - Handbook of Marine Mammals, vol. 4: River dolphins Kogia pusilk from Middle Pliocene of Tuscany 453 and the larger toothed whales, pp.349-430, Acad. Press, San Diego, California. Dal Piaz G. 0977) - Gli odontoceti del Miocene bellunese. Parti Quinta-Decima. Cyrtodelphis, Acrodelphis, Pro- todelpbis, Ziphíodelpbis, Scaldicetus. Mem. Ist. Geol. Unia. Padooa, Allegato v. a (1916) oÍ 122 pp., Padova. Farris J.S. (1988) - Hennig 86. Hennig 86 Reference version Fordyce R.E, & Barnes L.G. (1994) - The evolutionary histo- ry of whales and dolphins. Annu. Rev. Earth Planet. Sci., v. 22, pp. 419-455, Palo Alto, California. Fraser F. C. & Purves PE. (1960) - Hearing in cetaceans. Evo- lution of the accessory air sacs and rhe structure and function of the outer and middle ear in Recent cetaceans. Bull. Britisb Mus. (Nat. Hist.), Zool., v. 7 (1), pp. l-140. L.rndon. Iaccarino S. & Salvatorini, G. (1982) - A framework of plank- tonic foraminiferal biostratigraphy for Early Miocene ro Late Pliocene Mediterranean area. Paleont. Strat. Eaol.. v. 2, pp. 115-125, Roma. Handley C.O. Jr. (1,966) - A synopsis of the genus Kogla (pyg-y sperm whales). In Norris K.S. (Ed.) - \Whales, dolphins, and Porpoises. Univ. Calif. Press, pp. 62-69, Berkeley and Los Angeles. Heyning J.E. (1989) - Comparative facial anatomy of beaked whales (Ziphiidae) and a systematic revision among the families of extant Odontoceti. Contr. Sci., ndt. Hist. Mus. Los Angeles C., v. 405, pp. 1-64, Los Angeles. Kasuya T. (1973) - Systematic considerations of Recent toothed whales based on the morphology of rympano- periotic bone. Scl. Rep. Whal. Res. Inst., v 25, pp. 1-103, Tokyo. Kellogg R. (1929) - A new fossil toothed whale from Florida. Amer. Mws. Nooit., v. 389, pp. 1-10, NewYork. KernanJ.D. 8r Shulte H. von\il (1918) - Memoranda upon rhe anatomy of the respiratory tract, foregut, and thoracic viscera of a foetal Kogia breviceps. Bull. Am. Mus. Nat. Hist., v. 38, pp. 231,-267, New York. Lydekker R. (1894) - Contributions to a knowledge of the fos- sil vertebrates of Argentina. Pan II. Cetacean skuÌls frorn Patagonia. Ann. Mus. La Plata, v. 2, pp. 1-14, Buenos Aires. Matsumuto H. (1927) - On some fossil cetaceans of Japan. Scl. Repx. Toh. Imp. Unio., 2nd Ser. (Geol.), v. 10 (1), pp. 17-27, Sendai. Milinkovitch M.C., Meyer A. Ec Powell J.R. (1994) - Phylo- geney of all major groups of cetaceans based on DNA sequences from three mitochondrial genes. Mol. Biol. Eool., v. 1 1 (6), pp. 936-948, Chicago. Miller G.S. Jr. (1923) - The telescoping of the cetacean skull. Smiths. Misc. Coll.,v.76 (5), pp. 1.-71, Baltimore. Moreno F.P (1892) - Lijeros apuntes sobre dos géneros de Cetóceos fósiles de la República Argentina. Rer.t. Mus. La Plata, v. 3, pp. 393-400, Buenos Aires. Muizon C. de (1988) - Les Vertébrés de Ìa Formation Pisco (Pérou). Troisième partie: Les Odontocètes (Cetacea, Mammalia) du Miocène. Inst. Fr. Etud. Andine.r, mem. 78. pp. l-244. Paris. Muizon C. de (1991) - A new Ziphudae (Cetacer) from the Early Miocene of \Vashington Srate (Usa) and phyloge- netic analysis of the major groups of odontocetes. Bull. Mus. ndtn. Hist. nat.,v. 12 (3-4'5, pp.279-326,Pa,ris. Pilleri G. (1986a) - Pygroy sperm whales (Kogia) in the Italian Pliocene. Inuest. on Cetacea, v 18, pp. 133-154, Vam- mala Pilleri G. (1986b) - The taxonomic status of "Shizodelphis elongatus" : Miokogia elongatus (Probst, 1BB6) (Cetacea, Physeteridae). Inaest. on Cetacea, v. 18, pp. 155-162, Vammala. Pilleri G. (1987) - The Cetacea of the Italian Pliocene with a descriptive catalogue of the species in the Florence Museum of Paleontology. V of 1eo pp., printed by Vam- -.r.- v:..:.-.:- î1, Vammala.lrrdrrrr r\tt.,dPdrrtu \ Pilleri G. (1988) - Pre- and Post-Messinian Cetacea in the Mediterranean Tethys and the Messinian "saLinity cri- sis". In Pilleri, G. (Ed.) - Contributions to the Paleon- tology of sorne Tethyan Cetacea and Sirenia, pp. 19-4A, printed by Vammalan Kirjapaino Oy, Vammala. Pilleri G., M. Gihr & C Kraus (1989) - Odontoceti (Mam- malia: Cetacea) from the Lower Miocene of Rosignano, Piedmont, North Italy. Inrtest. on Cetacea, v. 22, pp. 1,89-291, Vammala. Raven H.C. & Gregory \(K. (1933) - The spermaceti orgen and nasal passages of the sperm wl-raie (Physeter catodon) and other odontocetes. Amer. Mus. Nc,,oit., v. 677, pp. 1-18, New York. Ross G.J.B. (1979) - Records of pygrny and dwarf sperm whales, genus Kogia, frorn southern Africa, with bio- logical notes and some comparisons. Ann, Cape Proo. Mus. (nat. Hìst.),v. 11 (14), pp.259-327, Grahamstown. Simpson G.G. (1945) - The principles of classification and a classification of mammals. Bull. Am. Mus. Nat. Hisr., v. 85, pp. v-xvi, 1-150, New York. Schenkkan E.J. & Purves PE. (1973) - The compertative anaromv of the nrsal rract and rhe [unction of rhe sner-_"_'" " '/ màceti organ in rhe Physeteridae (Mamrralia, Odonto- cetr). Bijdragen tot de Dierkwnde, v. 43, pp. 93-112, Amsterdan. \lhitmore F.C.Jr. (199a) - Neogene climatic change and emer- gence of the modern whale fauna of the Norrh Atlantic Ocean. In Berta A. & Deméré T.A. (Eds.) - Contribu- tions in marine mammals paleontology honoring Frank C. \ú/hitmore Jr. Proc. San Diego Soc. Nat. Hist., v. 29, pp. 233-227, San Diego, California.