ORIGIN of the WHITE SHARK CARCHARODON (LAMNIFORMES: LAMNIDAE) BASED on RECALIBRATION OE the UPPER NEOGENE PISGO FORMATION of PERU by DANA J

Home , Carcharodon, Lamnidae, Lamniformes

[Palaeontology, Vol. 55, Part 6, 2012, pp. 1139-1153]

ORIGIN OF THE WHITE SHARK CARCHARODON (LAMNIFORMES: LAMNIDAE) BASED ON RECALIBRATION OE THE UPPER NEOGENE PISGO FORMATION OF PERU by DANA J. EHRET'^ BRUCE J. MACEADDEN^ DOUGLAS S. JONES^ THOMAS J. DEVRIES^ DAVID A. EOSTER^ andRODOlYO SALAS-GISMONDI^ 'Monmouth University, West Long Branch, NJ, 07764, USA; e-mail: [email protected] ^Florida Museum of Natural History, University of Florida, Gainesville, FL 32611, USA; e-mails: [email protected], [email protected] 'Burke Museum of Natural History and Culture, University of Washington, Seattle, WA 98195, USA; e-mail: [email protected] ^Department of Geological Sciences, University of Florida, Gainesville, FL 32611, USA; e-mail: [email protected] 'Departmento de Paleontología de Vertebrados, Museo de Historia Natural Javier Prado, Universidad Nacional Mayor de San Marcos, Lima 11, Peru; e-mail: [email protected] *Corresponding author.

Typescript received 21 October 2009; accepted in revised form 22 August 2012

Abstract: The taxonomic origin of the white shark, Car- ative of C. hastalis include a mesially slanted third anterior charodon, is a highly debated subject. New fossil evidence (intermediate) tooth. We also provide a recalibration of presented in this study suggests that the genus is derived critical fossil horizons within the Pisco Formation, Peru from the broad-toothed 'mako', Carcharodon (Cosmopolito- using zircon U-Pb dating and strontium-ratio isotopic anal- dus) hastalis, and includes the new species C. hubbelli sp. ysis. The recalibration of the absolute dates suggests that nov. - a taxon that demonstrates a transition between Carcharodon hubbelli sp. nov. is Late Miocene (6-8 Ma) in C. hastalis and Carcharodon carcharías. Specimens from the age. This research revises and elucidates lamnid shark evo- Pisco Formation clearly demonstrate an evolutionary mosaic lution based on the calibration of the Neogene Pisco For- of characters of both recent C. carcharias and fossil C. has- mation. talis. Characters diagnostic to C. carcharias include the presence tooth serrations and a symmetrical first upper anterior Key words: Carcharocles, Cosmopolitodus, geochronology, tooth that is the largest in the tooth row, while those indic- Miocene, Ismus, strontium.

NEOSELACHIANS are well represented in the fossil One of the most debated enigmas within the neosela- record worldwide during the Neogene and most of the fos- chians focuses on the evolution and taxonomic placement of sil material found consists of isolated teeth. Shark teeth are the white shark, Carcharodon carcharias (Linnaeus, 1758), shed by the thousands over an individual's lifetime and within the Lamrdformes. There are two distinct hypotheses have an enameloid crown that acts as a protective layer regarding the evolution of C. carcharias that have been pro- during fossilization. The cartilaginous skeleton of chondri- posed in the literature. The first hypothesis places all large, chthyans does not typically preserve except in rare instances serrated megatoothed sharks within the genus Carcharodon, where calcified vertebral centra, portions of the neurocrani- including C. carcharias as well as those species referred to as um and fin rays have been described (Uyeno et al. 1990; Carcharocles, for example, Carcharocles megalodon (Jordan Shimada 1997; Siverson 1999; Gottfried and Fordyce 2001; and Hannibal, 1923) and its related taxa. Based on this tax- Shimada 2007; Ehret et al. 2009fl). The lack of more com- onomy, the lineage of C. carcharias branched off as smaller plete specimens of most fossil taxa has led to conflicting forms of the megatoothed sharks and co-evolved alongside interpretations about the taxonomy and anatomy of many the truly large taxa, such as C. megalodon (Applegate and species. Such problems have caused much confusion in the Esprnosa-Arrubarrena 1996; Gottfried et al 1996; Purdy nomenclature (including generic and specific names) and 1996; Gottfried and Fordyce 2001; Purdy et al. 2001). terminology (i.e. dental homologies) of fossil neosela- The second hypothesis proposes that C. carcharias evolved chians. from the broad-toothed 'Oxyrhina' {Cosmopolitodus)

© The Palaeontological Association doi: 10.1111/J.1475-4983.2012.01201.X 1139 1140 PALAEONTOLOGY, VOLUME 55 hastalis (Agassiz, 1843-1844) while the megatoothed sharks Hodell and Woodruff 1994, Oslick et al. 1994, Miller and belong to a separate family, the Otodontidae, within the Sugarman 1995, Martin et al. 1999, McArthur et aï. Lamniformes (Casier 1960; Glikman 1964; Cappetta 1987; 2001). We analysed three fossil marine mollusc shells Ward and Bonavia 2001; Nyberg et al. 2006; Ehret et al. from each of five localities to determine the ratio of 2009fl). C. hastalis was originally assigned to the genus *'Sr/*''Sr in the shell calcium carbonate. When compared Oxyrhina and later to Isurus. However, based on affinities with the global seawater reference curve, these data allow with C. cardiarias, Glikman (1964) suggested the reassign- us to estimate the age of the fossil molluscs for each ment of all unserrated forms within the Carcharodon line- locality (Table 1). age to the genus Cosmopolitodus to reflect this relationship For isotopic analyses, we first ground off a portion of (Siverson 1999; Ward and Bonavia 2001). Under this the surface layer of each shell specimen to reduce possible hypothesis C. hastalis and C carcharías represent chrono- contamination. Areas showing chalkiness or other signs of species as C. hastalis is replaced by C. hubhelli and finally diagenetic alteration were avoided. Approximately 0.01- C. carcharias in geological time through a gradation. Fur- 0.03 g of aragonite or low-magnesium calcite powder was thermore, the teeth of C. hastalis do not share all characters recovered from each fossil sample. The powdered samples with Isurus but instead are more similar to C. carcharias were dissolved in 100 /d of 3.5 N HNO3 and then loaded exhibiting triangular crowns that are labiolingually flat- onto cation exchange columns packed with strontium- tened, a flat labial face, a lingual face that is slightly convex selective crown ether resin (Eichrom Technologies, Inc., and a root that is flat and quite high (Cappetta 1987). As Lisle, IL, USA) to separate Sr from other ions (Pin and such, the genus Carcharodon (Smith in Müller and Henle, Bassin 1992). Sr isotope analyses were performed on a 1838) is the senior synonym of Cosmopolitodus (Glikman, Micromass Sector 54 thermal ionization mass spectrome- 1964) the proper genus name for C. hastalis. Fossil material ter equipped with seven Faraday collectors and one Daly collected from the Late Miocene and Early Pliocene of the detector in the Department of Geological Sciences, Uni- Pacific Basin including Peru (specifically the Pisco Forma- versity of Florida. Sr was loaded onto oxidized tungsten tion), Chile, CA, USA, and lapan provide further evidence single filaments and run in triple collector dynamic mode. of this relationship (Muizon and DeVries 1985; Long 1993; Data were acquired at a beam intensity of about 1.5 V for Tanaka and Mori 1996; Stewart 1999, 2002; Yabe 2000; *^Sr, vnth corrections for instrumental discrimination Suarez et al. 2006; Nyberg et al. 2006; Ehret et al. 2009a). made assuming ^''Sr/^^Sr = 0.1194. Errors in measured The recent description of an articulated specimen of "^Sr/^Sr are better than ±0.00002 {2a), based on long- this intermediate form of Carcharodon recorded from the term reproducibility of NIST 987 (^^Sr/'^Sr = 0.71024). Pisco Formation presents important new insights into the Age estimates were determined using the Miocene portion evolutionary history and taxonomy of the genus (Ehret of Look-Up Table Version 4:08/03 associated with the et al. 2009fl). The described specimen was collected from strontium isotopic age model of McArthur et al. (2001). Sud Sacaco (West), a locality referred to the Pliocene Zircons were extracted from samples using standard (c. 4.5 Ma) by Muizon and DeVries (1985). However, iso- crushing, density separation and magnetic separation topic calibration of fossil horizons exposed at Sud Sacaco techniques. The zircons were hand picked, mounted in (West), using strontium and zircon dating, allows us to epoxy plugs along with the reference zircon FC-1 (Paces reassess the age of the Carcharodon specimen. Based on and Miller 1993) and analysed using laser ablation multi- the calibration of these horizons, it is our conclusion that collector inductively coupled plasma mass spectrometry the specimen should be referred to the Late Miocene rather than the Early Pliocene. The purpose of this paper is to present (1) a reassess- TABLE 1. Strontium chemostratigraphic analyses of fossil mar- ment of the ages of some horizons within the Pisco For- ine mollusc shells from the Pisco Formation. mation, (2) formally describe a new species of the white Fossil Horizon Mean 95% CI (Ma) shark and (3) relate changes in age and taxonomy to the Age ''^Sr/^'^Sr estimate (Ma) evolution of the white shark within the Pacific Basin. El Jahuay 0.7089424 7.46 9.03-6.51 Montemar 0.7089468 7.30 8.70-6.45 METHODS AND MATERIALS Sud Sacaco 0.7089659 6.59 10.77-2.50 (West) Numerous studies indicate the Miocene Epoch was char- Sud Sacaco 0.7089978 5.93 6.35-5.47 acterized by rapidly increasing *''Sr/''*Sr in the global (West) Sacaco 0.7090005 5.89 6.76-4.86 ocean; therefore, it is especially amenable to dating and correlating marine sediments using strontium isotope Ages and confidence intervals (CI) determined from McArthur chemostratigraphy (Hodell et al 1991; Miller et al. 1991, et al (2001). EHRET ET AL.: WHITE SHARK EVOLUTION 1141

(LA-MC-ICP-MS). We used a Nu Plasma mass spectrometer fitted with a U-Pb collector array at the Department À of Geological Sciences, University of Florida. ^^*U and N ^^^U abundances were measured on Faraday collectors and ^°^Pb, ^°*Pb and ^°*Pb abundances on ion counters. The Nu Plasma mass spectrometer is coupled with a New Wave 213 nm ultraviolet laser for ablating 30-60 fim spots within zircon grains. Laser ablation was carried out in the presence of a helium carrier gas, which was mixed with argon gas just prior to introduction to the plasma torch. Isotopic data were acquired during the analyses using Time Resolved Analysis software from Nu Instru- >,Nazca Aguada ments. Before the ablation of each zircon, a 30 s peak / de Lomas zero was determined on the blank He and Ar gases with closed laser shutter. This zero was used for online correc- tion for isobaric interferences, particularly from ^"''Hg. Following blank acquisitions, individual zircons under- went ablation and analysis for c. 30-60 s. The analyses of unknown zircons were bracketed by analysing an FC-1 standard zircon.

GEOLOGICAL SETTING -V-' j Pre-Cenozoic [ j Pisco Basin The shallow-water, marine Pisco Formation of southwest- ern Peru consists of materials that accumulated during Late Miocene to Early Pliocene (c. 11.6-3.6 Ma); it is well FIG. 1. Map of Peru with localities within the Southern known for its wealth of vertebrate fossils including elas- section of the Pisco Formation. mobranchs, teleosts, chelonians, shore birds, seals, dol- phins, whales and aquatic sloths (Hoffstetter 1968; Muizon and DeVries 1985; Muizon and McDonald 1995; Around the region of Sacaco, Muizon and DeVries Muizon et al. 2002; Amiot et al. 2008; Fhret et al. 2009fl). (1985) identified a composite section for the Pisco For- In addition to the presence of complete vertebrate skele- mation in the Sacaco area (which is not fully exposed in tons, the area also contains the partial remains of fossil the region) consisting of five localities: Fl lahuay (also sharks including associated tooth sets, vertebral centra known as Alto Grande), Aguada de Lomas, Montemar, and preserved cartilage of the jaws and neurocrania. The Sud Sacaco (West) and Sacaco (Fig. 2). In addition to the Pisco Formation also preserves the transition from Car- exposed stratigraphy, these localities are also distinguished charodon hastalis to C. carcharías as demonstrated in the by a high diversity of fossilized vertebrate and inverte- shark fossils recovered from localities within the forma- brate species that found in numerous horizons (Marocco tion. Another taxon that exhibits intermediate characters and Muizon 1988). between the two species, previously referred to as Car- Fl Jahuay is designated as the lowermost section of the charodon sp., is also abundant at some localities (particu- Pisco Formation in the Sacaco area (Muizon and DeVries larly Sud Sacaco West and Sacaco) of the region (Muizon 1985). The lowest tuff bed at Fl Jahuay is also the oldest and DeVries 1985; Ehret et al. 2009a, b). in the Sacaco Basin and has been dated to 9.5 Ma (K-Ar The Pisco Formation crops out on the coastal plain of dating, late Miocene; Muizon and DeVries 1985; Muizon southern Peru from the town of Pisco south to Yauca and Bellon 1986). Below this bed, sediments are com- (Fig. 1). The sediments include tuffaceous and diatoma- prised of a rounded boulder conglomerate (Muizon and ceous sandstone and siltstone, ash horizons, bioclastic DeVries 1985). In the 15 m exposed above this tuff, the conglomerate (including the marine mollusc beds sampled section contains two shelly sandstone beds that are mixed in this study) and phosphorite, representing shallow mar- with a bioturbated fossiliferous sandstone (Muizon and ine environments. These preserve cyclic marine transgres- DeVries 1985). This layer had been designated as the El sive and regressive events spanning the Middle Miocene Jahuay vertebrate level (Fig. 2). Fossil selachians of inter- (c. 13 Ma) through the Late Pliocene (c. 4 Ma; Muizon est to this study within the El Jahuay fossil level are iden- and DeVries 1985; DeVries 1998; Amiot et al 2008). tified as C. hastalis and C. megalodon. 1142 PALAEONTOLOGY, VOLUME 55

Vertebrate Skolithos burrows, and well-preserved vertebrate remains level • 5-75 Ma+/. 0.5 Ma (Fig. 2; Muizon and DeVries 1985). (Ar-Ar) SAO The third exposed section of the Pisco Formation in the southern Pisco Basin can be found at Montemar (Fig. 2; Muizon and DeVries 1985). Sediments are comprised of coarse and fine sandstone, siltstone and gypsum, Siren i an u Zone along with mixed vertebrate and invertebrate fossils. The p fossil horizon is considered to be uppermost Miocene p (7.3-7.0 Ma) based on the co-occurrence of transitional Carcharodon teeth (Carcharodon hubbelli n. sp. of this e SAS study) and several moUuscan species (Muizon and DeV- 1 ries 1985). This fossil horizon is also noted for the pres- MTM ence of articulated skeletons of baleen whales, penguins M and sloths (Muizon and DeVries 1985). 1 At Sud Sacaco (West), the Pisco Formation is repre- 0 sented by a fossil horizon that is identified by the pres- c ence of fossilized barnacle debris and shell beds that e AGL contain a diverse assemblage of molluscs (e.g. Dosinia n (Scopoli, 1777)) and vertebrates (Fig. 2). These shell beds e are capped by thick vertebrate-bearing tuffs, fining upward cycles of brachiopod-bearing sandstone, which contain serrated C. carcharias-\ike teeth (Muizon and • 7.46 Ma ("Sr-"'Sr) DeVries 1985). Based on the stratigraphy and palaeontol- .8 Ma ogy, Muizon and Bellon (1980) concluded that the shell (K.Ar) i»8.8 Ma; 10+/- 1 Ma beds of the Sud Sacaco (West) fossil horizon were situ- . (K-Ar) (U.Pb) EU ated stratigraphically below the vertebrate-bearing tuff bed dated at 3.9 Ma (i.e. Pliocene), and younger than the Miocene horizons present at El Jahuay and Montemar (Muizon and DeVries 1985). Cross Bedding Pebbly Sandstone Mollusca Finally, the youngest section within the southern I I I ^ I Sacaco Basin is exposed at Sacaco. Sediments are mainly Burrows Fine Sandstone Marine Mammals Mollusc Beds comprised of fine sandstone interbedded with a pebbly Evaporites Coarse Sandstone Selachians Unconformity sandstone layer and two mollusc layers. The horizon also contains interbedded tuffs and cross-bedding is common. TutT Silt.stone Bamacle Dehris This horizon is considered to be above the vertebrate- FIG. 2. Stratigraphie log ofthe Pisco Formation (based on bearing tough bed (3.9 Ma), which is exposed at Sud Sac- Muizon and DeVries 1985). Abbreviations: AGL, Aguada de aco (West), and would indicate an Early Pliocene age Lomas vertebrate horizon; ELJ, El Jahuay vertebrate horizon; (Fig. 2; Muizon and Bellon 1980; Muizon and DeVries MTM, Montemar vertebrate horizon; SAS, Sud Sacaco (West) 1985). Additional sampling of invertebrate fossils and the vertebrate horizon; SAO, Sacaco vertebrate horizon. tuff beds taken from El Jahuay (Alto Grande), Montemar, Sud Sacaco (West) and Sacaco during the summer of At Aguada de Lomas, two tuffs were analysed using K- 2007 were analysed for '''Sr/*'^Sr isotopes and zircon Ar isotopes and dated to 8 and 8.8 Ma, respectively U-Pb and wiU be discussed below. (Fig. 2; Muizon and DeVries 1985; Muizon and Bellon 1986). The sequence of these tuff layers and rounded boulder conglomerates at Aguada de Lomas led Muizon OVERVIEW OF TAXONOMY AND and Bellon (1986) to conclude that the lower section of FOSSIL RECORD OF CARCHARODON this locality might have been produced by similar pro- cesses or even contemporaneous with the El Jahuay sec- As previously discussed, the evolutionary history of Car- tion (Muizon and DeVries 1985). Above these tuff beds, charodon has been a contentious subject. To elucidate our fine sandstone is the dominant lithology for the next current knowledge of the subject, it is important to 80 m of the section (Muizon and Devries 1985). The fos- review the hypotheses that have been previously pro- sil horizon presents at Aguada de Lomas caps these sand- posed. We win then compare and incorporate the mate- stone layers and is comprised of cross-bedded sandstone. rial from Peru into these paradigms. The taxonomy of EHRET ET AL.: WHITE SHARK EVOLUTION 1143 the megatoothed sharks (including Otodus and Carcharo- cles) beyond their relationship to C. carcharias is not within the scope of this paper and will only be addressed to the extent necessary. The first hypothesis places the megatoothed sharks {Oto- dus and Carcharocles) and C. carcharias within the Lamni- dae. The serrated-toothed forms of the megatoothed sharks are assigned to the genus Carcharodon (referred to as Car- \ charocles here), while the unserrated form is recognized as being related and has been given a separate generic name, Otodus (Agassiz, 1833-1844; Applegate and Espinosa-Arru- barrena 1996; Gottfried et al. 1996; Purdy 1996; Purdy et al. 2001). It should be noted that a recent shift in para- digm recognizes Otodus as a chronospecies, thereby placing all species of Carcharocles into the genus Otodus, with the exception of megalodon, which has been referred to the genus Megaselachus (GliJanan, 1964) (Casier 1960; Zhe- lezko and Kozlov 1999; Ward and Bonavia 2001). For the purpose of this discussion, however, we refer all serrated FIG. 3. Carcharocles megalodon tooth, USNM 336204. Scale megatoothed sharks to Carcharocles based on the accepted bar represents 10 mm. taxonomy in the literature at the present time. The grouping of megatoothed sharks with Carcharodon carcharias was based on a number of dental characters C. megalodon and could be misleading, the presence of including: (1) a symmetrical second anterior tooth; (2) the other characters listed above make this identification large third anterior (intermediate) tooth that is inclined more plausible. Furthermore, the reconstruction of dental mesially; (3) upper anterior teeth that have a chevron- patterns of extinct sharks based on isolated teeth or disar- shaped neck on the lingual surface; (4) an ontogenetic ticulated tooth sets is misleading when extant taxa are gradation whereby the coarse serrations of a juvenile used as a template for interpretations (Shimada 2006, C. carcharias shift to fine serrations in the adult, the latter 2007). Finally, the presence of serrations in both the serrations resembling those of C megalodon; (5) morpho- megatoothed and white sharks is a result of homoplasy logical similarity between the teeth of young C. megalodon rather than a diagnostic character uniting the two taxa. A and adult C. carcharias (Gottfried et al. 1996; Gottfried comparison of the serrations shows that they are in fact and Fordyce 2001; Purdy et al. 2001; Ehret et al. 2009a). very different, with the Carcharocles exhibiting very fine Following this hypothesis, C. carcharias is either the result and regular serrations while those of Carcharodon are of dwarfism from a larger megatoothed taxon or evolved coarse and irregular (Fig. 4). via cladogenesis from a large megatoothed shark species The second hypothesis keeps C. carcharias and the sometime during the Palaeocene and coevolved alongside broad-toothed 'makos' within the Lamnidae while the the other species (Gottfried et al. 1996; Martin 1996; Pur- megatoothed sharks are reclassified into their own family, dy 1996; Purdy eio/. 2001). the Otodontidae within the Lamniformes. This hypothesis This taxonomy, however, does not truly reflect the fos- proposes that Carcharodon hastalis gave rise to the ser- sil record. The characters listed above as diagnostic, unit- rated-toothed Carcharodon carcharias during the Late ing the megatoothed sharks and C. carcharias, are a result Miocene or Early Pliocene (Casier 1960; Glikman 1964; of individual variation, interpretation and homoplasy Muizon and DeVries 1985; Cappetta 1987; Ward and Bo- (Hubbell 1996; Shimada 2005). Variation and pathologic navia 2001; Nyberg et al 2006; Ehret et al. 2009fl). This abnormalities within individual teeth can lead to the hypothesis would suggest that the lamnid and otodontid incorrect interpretation and identification of specimens. sharks last shared a common taxon in the Cretaceous As an example, USNM 336204 has been described as lat- (Casier 1960). Casier (1960) proposed that C. hastalis is eral tooth of Carcharodon sp. (Purdy 1996) or C. carcha- ancestral to Carcharodon citing characters of dentition rias (Gottfried and Fordyce 2001) from the Middle (although he did not list these characters) and offering Miocene of the Calvert Formation of Maryland (Fig. 3). as a possible transition between the taxa, Isurus escheri However, re-analysis of this specimen reveals that it is a (Agassiz 1833-1843), which exhibits weak, fine crenula- small C megalodon, based on the presence of a chevron- tions (not true serrations) on the cutting edges of its teeth. shaped neck, the thickness of the crown and the serration It should be noted that Agassiz (1833-1844) separated type. While the serrations are somewhat coarse for the broad-toothed 'makos' into numerous separate 1144 PALAEONTOLOGY, VOLUME 55

FIG. 4. Comparison of serration types. A, Carcharodon hastalis (UF 57267). B, Carcharodon hubbelli (UF 226255). C, hums escheri (UF 245058). D, Carcharocles megalodon (UF 217225). E, Carcharodon carcharias (G. Hubbell collection). Scale bar represents 10 mm. morphotypes, two of the most prominent were originally crowned broad-toothed 'makos'; however, based on the referred to Carcharodon {Oxyrhina) hastalis and Isurus arguments of Leriche (1926) and Ward and Bonavia {Oxyrhina) xiphodon (Agassiz, 1833-1844). His differenti- (2001), we agree that the name /. xiphodon is inappropri- ation of the two taxa was based on C. hastalis having a ate and use C. hastalis for all forms of broad-toothed noticeable flattening of the lingual crown foot and nar- 'makos' at the present time. rower tooth crowns compared to I. xiphodon but admit- The species 1. escheri was originally placed in the genus ted that the latter was too weak of a character for Oxyrhina and later Isurus as a variant of C. hastalis (Agas- identification (Agassiz 1833-1844; Purdy et al 2001). Ler- siz 1833-1844; Leriche 1926; Gasier 1960). Specimens iche (1926) included I. xiphodon in the synonymy of have been reported from the late Middle Miocene C. hastalis and noted the uncertainty of the origin of the through the Early to Middle Pliocene (c. 14-4 Ma) of the types, which have since been lost (Ward and Bonavia northern Atlantic (including Germany, Belgium, the 2001). The species was recognized by Glikman (1964) and Netherlands and Denmark) with materials ranging from Purdy et al. (2001), who attributed specimens of C. has- isolated teeth to an associated set of teeth and vertebrae talis from Peru, North Carolina and Belgium; the new from northern Germany (Agassiz 1833-1843; Leriche species from Peru described herein; and I. escheri from 1926; Van den Bosch et al 1975; Nolf 1988; Mewis and Belgium to /. xiphodon based solely on the broadness of Klug 2006; Mewis 2008). The crenulations on the cutting their crowns. Ward and Bonavia (2001) regarded /. xiph- edges of the teeth are much finer than the serrations of odon as a nomen dubium based on the arguments of Leri- C. carcharias specimens found in the Pacific basin and che (1926), the absence of any type, specimens, and an tend to be along the entire cutting edge of the tooth (Ler- unlikely provenance. Recent morphometric analysis of iche 1926; Van den Bosch et al 1975; Nolf 1988). In broad-toothed 'mako' specimens by Whitenack and Gott- addition to crenulations, some specimens may exhibit 1-3 fried (2010) supports a morphological difference based on pairs of lateral cusplets, which are more pronounced in the broadness of the crown in broad-toothed 'makos'. the lower teeth (Mewis 2008). While this study might accurately separate these species, The crown shape of I. escheri is less dorsoventrally flat- this could also represent sexual dimorphism or ontoge- tened and thinner anterodorsally with a stronger distal netic changes in tooth morphology in one taxon. There inclination than either C. hastalis or Carcharodon (Fig. 5). may be a true delineation between broad- and narrow- This inclination is a result of a marked change in direction EHRET ET AL.: WHITE SHARK EVOLUTION 1145

characters were better than those combining dental and nondental characters (Shimada 2005). Shimada did acknowledge, however, that patterns indicated that dental characters are important for at least some phylogenetic signal. A review of the analysis by Mewis (2008) reveals that three of the synapomorphies uniting /. escheri and C. carcharias relate to the presence of serrations, which have evolved numerous times in the Lamniformes, while two others are dependent on correct tooth position assignment of disarticulated specimens. Additionally, the restricted fossil distribution of /. escheri to the Northern Atlantic and Mediterranean does not support the earliest Pliocene occurrences of C. carcharias in the Pacific, while appearances of the white shark are slightly later in the Atlantic. In contrast to I. escheri, the evolution of the white shark in the Pacific during the Late Miocene and Early Pliocene (c. 10^ Ma) is well documented in the fossil records of Peru, Japan, California, Australia and Chile (Muizon and DeVries 1985; Kemp 1991; Long 1993; Tanaka and Mori 1996; Stewart 1999, 2002; Yabe 2000; Nyberg et al 2006; Ehret et al 2009fl). Based on the rich fossil record and the rejection of I. escheri as a sister taxon to C. carcharias, we agree more with the second origin hypothesis outlined above but propose a Pacific Basin origin for the genus Carcharodon with C. hastalis as a closely related taxon. FIG. 5. Isurus escheri from the Delden Member (Early Furthermore, we designate a new species of Carcharodon Pliocene), the Netherlands. A-B, Upper Anterior tooth, UF from the Pacific basin that represents an intermediate 245058. A, Labial view. B, Lingual view. C-D, Upper Lateral form between C. hastalis and C. carcharias. tooth, UF 245059. C, Labial view. D, Lingual view. Scale bar represents 10 mm. Institutional abbreviations. UF> Florida Museum of Natural His- tory (FLMNH) University of Florida; USNM, United States of the distal cutting edge of the tooth crown and appears National Museum; MUSM, Museo de Historia Natural Javier to be more pronounced in lateral teeth. The roots of the Prado, Universidad Nacional Mayor de San Marcos, Lima, Peru. upper teeth also appear to be more lobate and angular than either C. hastalis or Carcharodon, which have very Terminology and anatomical abbreviations. Following Cappetta square, rectangular roots (Ehret et al 2009a). The teeth of (1987) and Siverson (1999), specifically we define intermediate teeth as those that form on the intermediate bar, vary in number I. escheri appear to have stronger affinities with a more and are less than one half the crown heights of the anterior narrow-toothed form of C. hastalis, whereas the vertebral teeth. centra reported from Germany appear to have the typical amphicoelous lamniform morphology that is not diagnostic for refined taxonomic assessment above the ordinal level (Mewis and Klug 2006; Mewis 2008). SYSTEMATIC PALAEONTOLOGY

While the latter hypothesis more accurately portrays Class CHONDRICHTHYES Huxley, 1880 the evolutionary history of Carcharodon, the designation Subclass ELASMOBRANCHII Bonaparte, 1838 of /. escheri as a sister taxon to C. carcharias based on cla- Order LAMNIFORMES Berg, 1958 distic analysis of dental characters by Mewis (2008) could Family LAMNIDAE Müller and Henle, 1838 be misleading. Shimada (2005) has shown that cladistic analysis combining both dental and nondental characters Genus CARCHARODON Smith in Müller and Henle, 1838 of extant and fossil shark taxa generates considerable phylogenetic noise. Comparisons of consensus trees using Remarks. Glikman (1964, p. 104) placed the genera Car- dental characters versus trees combining dental and all charodon and Cosmopolitodus within the Carcharodontidae other morphological characters result in different topolo- (Gill, 1893; mis-cited as 1892 in Glikman) based on the gies. Tree statistics for the cladogram just using nondental 1146 PALAEONTOLOGY, VOLUME 55 follovñng characters: teeth broad and blade-like, crowns of Holotype. UF 226255 (Fig. 6), an articulated dentition including upper lateral teeth dorsoventraUy fiattened, neck very small, 222 teeth, 45 vertebral centra, portions of the left and right Mec- roots short, base of the tail with keels. The idea of the Gar- kel's cartilages and palatoquadrates, and neurocranium. charodontidae was also raised separately by Whitley (1940) and supported by Applegate and Espinosa-Arrubarrena Type locality. Sud Sacaco (West) (15°33'S, 74°46'W), 5 km east (1996), who cited anatomical evidence including fin posi- of Lomas, Arequipa Region, Peru. Nearly 20-40 cm above the vertebrate-bearing tuff bed, Pisco Formation; Upper Miocene. tions (but did not explicitly list the characters) and the presence of serrations and lateral denticles as synapomorphies. Referred specimens. Upper teeth (UF 245052-245057; Fig. 9). However, the characters of the Garcharodontidae listed by These teeth were recovered from the type locality (Sud Sacaco Glikman (1964) are not synapomorphies specific to Car- West), above the vertebrate-bearing tuff bed. charodon. Furthermore, the separation of the Otodontidae from the Lamnidae and the presence of a weakly crenulated Diagnosis. A lamnid shark that differs from all other I. escherí do not restrict serrations and lateral cusplets to lamnids by the following unique combination of dental the Garcharodontidae, and therefore, the family is not characteristics: (1) An A3 tooth distally inclined; (2) an supported. Glikman (1964) also differentiated the genus Al tooth that is the largest tooth in the dentition and is CosmopoUtodus from Carcharodon based on the absence of symmetrical; (3) an A2 tooth that is slightly larger than serrations and lateral cusplets. the a2 tooth; (4) serrations present but not fuUy devel- oped.

Carcharodon hubbelli sp. nov. Description. The mandibular arch of UF 226255 is partially pre- Figure 6 served, making it not possible to distinguish some features of the jaws. The left and right palatoquadrates are preserved anteri- Derivation of name. Named in honour of Dr. Gordon Hubbell orly along with the symphysis. The symphysis of both palatoqua- of Gainesville, Florida, in recognition of the substantial contribu- drates is square and deep. The upper dental buUae are also tion he has made to the field of shark palaeontology. preserved in both palaquadrates. Within these bullae, the upper

FIG. 6. Carcharodon huhhelli, UF 226255 (holotype). Scale bar represents 10 cm. EHRET ET AL.: WHITE SHARK EVOLUTION 1147 anterior teeth are found within a hoUow, vvith an intermediate tooth series present for each tooth position. The teeth are weakly bar, or a labiolingual constriction of the cartilage, present serrated, with anterior teeth averaging more than 30 serrations between the third anterior tooth and the first lateral tooth. The per side for anterior to no serrations for the distal-most laterals. distal portions of both palatoquadrates are not preserved, and Of the teeth that do have serrations, there is an average of 8-12 the medial and lateral quadratomandibular joints are not dis- serrations per centimetre on both edges of each tooth. Basal ser- cernable due to the dorsoventral flattening of the specimen. rations on the teeth are larger than the other serrations on the The Meckel's cartilages are much deeper than the palatoqua- edges. drates. The posterior portions of both are highly fragmented, There are three anterior teeth present in each palatoquadrate. making the original shape impossible to discern. The lower sym- The Al is symmetrical, and it is the largest in the tooth in the physis is preserved; however, it is less deep than the symphysis dentition. The A2 is also symmetrical and is slightly larger than of the palatoquadrates. The Meckel's cartilages are situated much the a2, while the A3 is distaUy inclined. The upper lateral teeth lower than the palatoquadrates, suggesting a subterminal mouth. are also inclined distally (with LI having the greatest inclination) Portions of the neurocranium are preserved; however, the and get progressively smaller towards the distal edge of the jaw. structure is not discernable due to the dorsoventral flattening of The roots of the teeth are relatively flat and rectangular in shape. the specimen. The occipital hemicentrum consists of the poster- In the Meckel's cartilages, there are three anterior teeth pres- ior half of a calcified double-cone centrum, which articulates ent. Overall, the lower teeth have crowns that are more slender with the anteriormost centrum of the vertebral column. Other than those of the uppers. There is very little inclination seen in portions of sheet-like cartilage are preserved within the gape of any of the lower teeth. As in the uppers, the teeth of the Mec- the jaws; however, it is not identifiable. Based on the position kel's cartilages are progressively reduced distally. The roots of and preservation, they most likely represent the basal plate of the lower teeth have a deep basal concavity and are thicker than the neurocranium. Other openings within these fragments of those of the upper teeth, being somewhat lobate. cartilage could represent foramina; however, due to the preserva- There are 45 vertebral centra of UF 226255 preserved in the tion, they probably represent areas of weathering. holotype. These centra are laterally compressed and are com- The dentition of C. huhbelli is represented by 222 teeth posed of two calcified cones connected by radiating calcified located on the palatoquadrates and Meckel's cartilages. The teeth lamellae within the intermedalia (Fig. 8). These centra are as- are flattened labiolingually, and triangular in shape (Fig. 7). The terospondylic. The articular surfaces are concave and show clear, crowns have a slight convex curve lingually. There are five to six calcified lamellae as well as pits in the centre that represent the

AI A2 LI L2

al a2 a3 11 12 13 14 15 16 17 18 FIG. 7. Functional tooth series of Carcharodon hubbelli, UF 226255 (holotype). Scale bar represents 5 cm. 1148 PALAEONTOLOGY, VOLUME 55

A B

FIG. 8. Vertebral centrum of Carcharodon hubbelli, UF 226255 (holotype). Scale bar represents 10 mm. notochordal constricture (Ridewood 1921; Gottfried and Fordyce DISCUSSION 2001). In UF 226255, the occipital hemicentrum followed by the first 45 centra, which get larger in size in ascending order. Geology and stratigraphy. The new absolute ages obtained from molluscs recovered at many of the localities of the Remarks. Carcharodon hubbelli is an intermediate form Pisco Formation suggest that deposition of the strata, between Carcharodon hastalis and Carcharodon carcharias now referred to the Pisco Fm, occurred earlier than and demonstrates a mosaic of characters from both spe- thought previously. The results have direct bearing on the cies. Tooth crowns of C. hubbelli are convex and curve age of the holotype for C. hubbelli sp. nov. and evolution lingually similar to the crowns of C. hastalis. The serra- of white sharks in the Pacific basin. tions of C. hubbelli are enlarged basally but weaker overall Previous dating of the tuff bed at Alto Grande within than those of C. carcharias (but stronger than those seen the El Jahuay vertebrate horizon resulted in an age of in I. escheri) and appear to be intermediate between the 9.5 Ma based on K-Ar dating (Muizon and Bellon 1986; unserrated C hastalis and the coarsely serrated C. carcha- Muizon and DeVries 1985). Analysis of 55 zircon grains rias (Fig. 4). The A2 is symmetrical and is slightly larger collected from the tuff bed at Alto Grande yielded 54 that than the a2, which are characters also seen in C. carcha- gave ages that were Neoproterozoic through Cretaceous, rias (Compagno 2001). Whereas the A3 is distally which are obviously inherited or mixed detritus. One inclined, a character is found in C. hastalis. Therefore, the grain gave a concordant crystallization age of 10 ± 1 Ma, taxa may be useful as chronospecies, because they are which could also be inherited and should be treated as an diagnostically distinguishable by dental characteristics and upper limit for the horizon. The strontium ages of the are beneficial for biostratigraphy. Specimens collected in shell layers overlying the tuff bed yielded an age range of Upper Miocene deposits throughout the Pacific basin 9.03-6.51 Ma with a mean of 7.46 Ma (Fig. 2), which are exhibit morphological gradation of dental characteristics younger than those for the ash layer. However, the upper from Carcharodon hastalis to C. hubbelli, and finally limits of these ages are within the range for the other dat- C. carcharias in the Early Pliocene. The observable transi- ing methods making these age estimates congruent with tion of species through geologic time is denoted here by previously published dates. the designation of C hastalis and C. hubbelli to the genus Carcharodon (Smith in Müller and Henle, 1838), which is Mollusc samples analysed from the Montemar horizon given precedence over Cosmopolitodus (Glikman 1964). yield ''^Sr/'^Sr dates of 8.70-6.45 Ma with a mean of Furthermore, the diagnosis of Cosmopolitodus by Glikman 7.30 Ma. These ages are congruent with the estimation of is misleading since some Late Miocene C. hastalis teeth Muizon and DeVries (1985) based on the vertebrate and may have basal serrations. invertebrate faunas (Fig. 2). While the resolution of the EHRET ET AL.: WHITE SHARK EVOLUTION 1149

FIG. 9. Individual upper teeth demonstrating the gradation of serrations from the Pisco Formation, Peru. Upper teeth from left to right, UF 245052-245057. A-F, labial view; G-L, lingual view. Scale bar represents 10 mm.

Strontium data is not more precise than the estimate layer, a shell bed containing abundant Ophiomorpha based on the molluscan fauna, it does confirm a Late (Lundgren, 1891) burrows was sampled for strontium Miocene age (Table 1). A Late Miocene (c. 8.7-6.5 Ma) and yielded an age range of 6.35-5.47 Ma, with a mean age based on the correlation of gypsum beds with Aguada of 5.93 Ma (Fig. 2). Previous methods of calibrating the de Lomas, fossiliferous sandy siltstone with Sud Sacaco Sud Sacaco (West) level include correlation of molluscs (West) and calibration of molluscan fauna was postulated with white shark {Carcharodon) fossus. Furthermore, the by Muizon and DeVries (1985). reassignment of the shark fossils from C. carcharias Both geochemical methods, zircon U-Pb and strontium (Early-Middle Pliocene) to C. hubbelli (Late Miocene- isotopic data, yield ages for the Sud Sacaco (West) level that Early Pliocene) also correlates with the older age for the are 1-2 million years older than reported previously by horizon. Muizon and DeVries (1985) and Tsuchi et al (1988). Thus, For the Sacaco horizon, mollusc samples were collected making a change of age for the Sud Sacaco (West) horizon near the Museo de Sacaco and yielded *''Sr/**Sr dates of from the Early Pliocene to the Late Miocene (Gradstein 6.76—4.86 Ma, with a mean of 5.89 Ma. This age is much et al 2004). At Sud Sacaco (West), a thick Dosinia bed was older than the underlying tuff bed dated at 3.9 Ma by sampled for strontium (Fig. 2). ^'^Sr/^^Sr dates for the bed Muizon and Bellon (1980). Additional unpublished Ar-Ar ranged from 10.77 to 2.50 Ma with a mean of 6.59 Ma data from the Sacaco ash bed located on the floor of east- (Table 1). Additionally, the vertebrate-bearing tuff dis- ern side of Quebrada Sacaco (locality DV 514-2 Snee) cussed in Muizon and DeVries (1985) was located and sam- below the unconformity that separates Sacaco shell beds pled for zircon data. In total, 35 zircon grains were from upper portion of sequence on the hillsides of Sacaco analysed with a majority being inherited or mixed detrital yielded an age of 5.75 ± 0.05 Ma (Lawrence Snee, USGS, grains. Of the Oligocène and Miocene zircons identified, sample collected 1987). These new Sr and Ar data are in the youngest grain is 7 ± 1 Ma and is probably the youn- disagreement with and much older than the 3.9 Ma age gest age of the depositional range for the tuff. It should also of Muizon and Bellon (1980). Therefore, the horizons be noted that the in situ holotype of Carcharodon hubbelli exposed at the Sacaco locality appear to be older than (UF 226255) was recovered 20-40 cm above the base of the previously recorded. vertebrate-bearing layer (Ehret et al 2009a). On the east side of Sud Sacaco (West) horizon, Taxonomy and evolution of Carcharodon. The taxonomic approximately 10 m above the vertebrate-bearing tuff evolution and placement of Carcharodon is a complex 1150 PALAEONTOLOGY, VOLUME 55

subject that has been compounded by the lack of more numerous times throughout the evolution of the sela- complete specimens and high levels of homoplasy in the chians (Casier 1960; Cappetta 1987; Frazzetta 1988). tooth morphology of many species. While a C. hastalis- Based on the living mako sharks, their tooth shape and origin hypothesis has gained popularity in recent years, overall body size, it is hypothesized that species within the timing and transition from one taxon to the other the genus Isurus were piscivorous (Applegate and Espin- have remained unresolved. It is generally accepted that osa-Arrubarrena 1996; Purdy et al 2001; Aguilera and de C. carcharias is directly related to Carcharodon hastalis Aguilera 2004). The evolution of serrations or crenula- and not the megatoothed sharks as previously discussed. tions would be advantageous for competition with other Isurus escheri from the Middle Miocene through Early piscivorous shark taxa (C. hastalis, Galeocerdo (Müller Pliocene of the Northern Atlantic has been suggested as a and Henle, 1837)), Hemipristis (Agassiz, 1833-1843) and possible sister taxon to C. carcharias, based mainly on the Carcharhinus (Blainville, 1816). Therefore, it is not sur- presence of crenulations on the cutting edges of the teeth prising that more than one form would acquire serra- (Fig. 4). However, these fossils are restricted to the tions. Northern Atlantic and Mediterranean and become extinct The evolution of the white shark in the Pacific Basin is just prior to or as a result of the appearance of C. carcha- validated by the presence of weakly to moderately serrated rias in the Atlantic by the mid-Pliocene. This species teeth in the fossil deposits from the Late Miocene and likely evolved from an Atlantic population of slim- Early Pliocene of North and South America, Asia and toothed C. hastalis or hastalis-like taxon in the early Mio- Australia (Muizon and DeVries 1985; Kemp 1991; Long cene. Meanwhile, materials from the Miocene Pacific 1993; Tanaka and Mori 1996; Yabe 2000; Stewart 1999, Basin clearly demonstrate an intermediate form between 2002; Nyberg et al 2006; Ehret et al 2009a). While many C hastalis and C. carcharias that is significantly different of these specimens from the Pacific represent different from I. escheri. The discovery and description of C. hub- degrees of evolution between C. hastalis and C. carcharias, belli from the Late Miocene of Peru and the recalibration it is not possible to separate these based on isolated teeth. of the Pisco Formation demonstrate the presence of ser- A complete tooth set, exhibiting more definitive charac- rated forms in the Miocene and early Pliocene of the ters, would be required to differentiate potentially differ- Pacific, while there is general lack of specimens from the ent forms. Therefore, these teeth are assigned to the Atlantic at the same period. species Carcharodon hubbelli; previous identifications as The distinction of C. hubbelli from /. escheri is vali- /. escheri (Kemp 1993), Carcharodon sp. (Nyberg et al dated by the morphological differences exhibited in UF 2006; Ehret et al 2009a) or C. carcharias (Muizon and 226255 and specimens of I. escheri examined from the DeVries 1985; Long 1993; Tanaka and Mori 1996) should Delden Member (Early Pliocene), the Netherlands (UF be amended to reflect this new assignment. 245058 and UF 245059) and the description of an associ- Additional research on C hubbelli has shed light on the ated specimen from Germany (Mewis and Klug 2006; palaeobiology of this species (Ehret et al. 2009a, b). Incre- Mewis 2008). 7. escheri has been identified from the Mid- mental growth analyses of the vertebral centra of UF dle Miocene (c. 14 Ma) to at least the Early Pliocene 226255 have revealed annular growth patterns that relate (c. 5-4 Ma) throughout Northern Europe. Van den Bosch to the life history of this specimen. Growth rings visible (1978, 1980) differentiates I. escheri from the late Miocene using X-radiography appear to represent annual periodic- as weakly crenulate and early Pliocene teeth as strongly ity based on the calibration of oxygen and carbon isotope crenulate types. However, this difference was not quanti- analysis within the rings. Counts of the grovrth rings fied or figured in either of the studies and is inconsistent using the X-radiographs provide an age estimate of at with our Early Pliocene samples that demonstrate weak least 20 years. The overall length of the specimen was crenulations (Fig. 5). C. hubbelli teeth show a progressive estimated using the averages of both crown height and increase in the number and overall coarseness of serra- vertebral centrum diameter regressions from previously tions on their cutting edges from the Late Miocene to the published studies of C. carcharias. Resulting data provide Early Pliocene that are distinctively different and appear total length estimates between 4.89 and 5.09 m. for the to evolve separately from /. escheri (Figs 4, 9). Addition- specimen. Based on growth curves of modern C. carcha- ally, the in situ dentition of UF 226255 demonstrates a rias, UF 226255 appears to have been growing at a slower mixture of characters expressed in both C. hastalis and rate than white sharks today. C. carcharias, supporting our revised C. /¡asfaZis-origin Further palaeobiological information about C. hubbelli hypothesis. includes a partial mysticete whale mandible from the SAS Isurus escheri, on the other hand, appears to be a sepa- layer containing a partial tooth crown MUSM 1470, rate taxon more closely related to a narrow-crowned housed in the collection of the Museo de Historia Natural C. hastalis or /wsfa/is-like taxon. Tooth serrations (or in Javier Prado, Universidad Nacional Mayor de San Mar- this case crenulations) have evolved independently cos, Lima, Peru (MUSM), described by Ehret et al EHRET ET AL.: WHITE SHARK EVOLUTION 1151

(2009^). This specimen represents direct evidence of feed- Editor. Adriana López-Arbarello ing behaviour of the species in the Late Miocene. The presence of tooth serrations and MUSM 1470 provides definitive proof that C. hubbelli was adapted for taking REFERENCES marine mammals as prey as early as c. 6.5 Ma. AGASSIZ, L. J. R. 1833-1844. Recherches sur les poisons fossiles. Text (5 vols; L, xlk -I- 188 pp., II xii + 310 -I- 366 pp.. Ill CONCLUSIONS viii -I- 390 pp., IV xvi -I- 296 pp., V xii + 122 + 160 pp.) and Atlas. Imprimerie de Petitpierre, Neuchâtel. AGUILERA, O. and AGUILERA, D. R. de 2004. Giant- The recalibration of fossil horizons within the Pisco For- toothed white sharks and wide-toothed mako (Lamnidae) mation finds ages are older than previously published from the Venezuela Neogene: their role in the Caribbean, shal- (Muizon and Bellon 1980; Muizon and DeVries 1985; low-water fish assemblage. Caribbean Journal of Science, 40, Muizon and Bellon 1986). While these changes are not 368-382. exceptionally large, it does directly relate to the evolution- AMIOT, R., GÖHLICH, U. B., LÉCUYER, C, MUIZON, ary history of the genus Carcharodon. The discovery and C. DE, CAPPETTA, H., FOUREL, F., HERAN, M. A. and description of an outstanding specimen from the Pisco M ART I NE AU, F. 2008. Oxygen isotope composition of Formation further elucidate the taxonomy and palaeobi- phosphate from Middle Miocene-Early Pliocene marine verte- ology of the white sharks. The hypothesis that 1. escherí is brates of Peru. Palaeogeography, Palaeoclimatology, Palaeoecolo- a sister taxon of C. carcharías is refuted based on the gy, 264, 85-92. Miocene and Pliocene distribution of Carcharodon fossils APPLEGATE, S. P. and ESPINOSA-ARRUBARRENA, L. from the Pacific Basin and tooth morphology. The genus 1996. The fossil history of Carcharodon and its possible ances- name Carcharodon is proposed for the species hastalis, tor, Cretolamna: a study in tooth identification. 19-36. In KLIMLEY, A. P. and AINLEY, D. G. (eds). Great white hubbelli and carcharías based on dental characters shared sharks: the biology of Carcharodon carcharías. Academic Press, between the taxa discussed above, and our interpretation San Diego, CA, 515 pp. of the C. hastalís-hubbellí-carcharías transition as an BERG, L. S. 1958. System der rezenten und fossilen Fischartigen example of chronospecies. Palaeobiological information und Fische. Deutsche Verlag Wissenschaften, Berlin, 310 pp. from UF 226255 reveals that this specimen grew at a rate BLAINVILLE, H. M. D. de 1816. Prodrome d'une distribu- comparatively slower than modern white sharks. MUSM tion systématique du regne animal. Bulletin des Sciences pars la 1470 confirms that the diet of C. hubbelli was at least par- Société Philomathique de Paris, 8, 105-124. tially comprised of marine mammals as early as the Late BONAPARTE, C. L. 1838. Selachorum tabula analytica. Nuovi Miocene. Gontinued research of these specimens and Annali della Science Naturali, Bologna, 2, 195-214. newly discovered materials from the Pisco Formation and BOSCH, M. VAN DEN 1978. On shark teeth and scales from the other fossil localities will only further advance our knowl- Netherlands and the biostratigraphy of the Tertiary of the edge of the fossil lamnid sharks. eastern part of the country. Mededelingen van de Werkgroep voor Tertiaire en Kwartaire Geologie, 15, 129-135. 1980. Elasmobranch associations in Tertiary and Quaternary Acknowledgements. This research is supported by National Sci- deposits ofthe Netherlands (Vertebrata, Pisces), 2. Paleogene of ence Foundation grants EAR 0418042 and 0735554. We espe- the eastern and northern part of the Netherlands, Neogene in cially thank Gordon Hubbell from Jaws International, the eastern part ofthe Netherlands. Mededelingen van de Werkg- Gainesville, Florida, for donating UF 226255 to the Florida roep voor Tertiaire en Kwartaire Geologie, 17, 65-70. Museum of Natural History as well as access to his collection CADÈE, M. and JANSSEN, A. W. 1975. Lithostrati- and his wealth of knowledge of fossil sharks. We thank M. Stuc- graphical and biostratigraphical subdivision of Tertiary depos- chi from the Museo de Historia Natural of Lima, Peru, who its (Oligocene-Pliocene) in the Winterswijk-Almelo region provided assistance in the field in 2007. We also thank M. Siver- (eastern par ofthe Netherlands). Scripta Geológica, 29, 1-167. son from the Western Australian Museum, Welshpool, Western CAPPETTA, H. 1987. Chondrichthyes 2. Mesozoic and Ceno- Australia, and J. Bourdon of http://www.elasmo.com for discus- zoic elasmobranchii. In Handbook of paleoichthyology. Vol. 3B. sions and constructive suggestions for the improvement of this Gustav Fischer Verlag, Stuttgart, Germany, 193 pp. manuscript. Thanks to E. Mavrodiev from the Florida Museum CASIER, E. 1960. Note sur la collection des poissons Paléocé- of Natural History for assistance with Russian translations. We nes et Eocenes de l'Enclave de Cabinda (Congo). Annales du also appreciate the comments from J. Bloch, R. Hulbert and J. Musée Royal du Congo Belge A III, 1, 1-47. Bourque from the Florida Museum of Natural History and two COMPAGNO, L. J. V. 2001. Sharks of the world: an anno- anonymous reviewers towards the improvement of this manu- tated and illustrated catalogue of shark species known to date. script. This is University of Florida Contribution to Paleobiology Volume 2: Bullhead, mackerel, and carpet sharks (Heterodont- 627. Any opinions, findings, conclusions or recommendations iformes, Lamniformes, and Orelectolobiformes). Food and expressed in this paper are those of the author and do not nec- Agriculture Organization Species Catalogue for Fishery Purposes, essarily reflect the views of the National Science Foundation. 1, 269 pp. 1152 PALAEONTOLOGY, VOLUME 55

DEVRIES, T. J. 1998. Oligocène deposition and Cenozoic LINNAEUS, C. 1758. Systema naturae. Tenth Edition. Larentii sequence boundaries in the Pisco Basin (Peru). Journal of Salvii, Stockholm, 824 pp. South American Earth Sciences, 11, 217-231. LONG, D. J. 1993. Late Miocene and Early Pliocene fish assem- EHRET, D. J., HUBBELL, G. and MACFADDEN, B. 1. blages from the north central coast of Chile. Tertiary Research, 2009fl. Exceptional preservation of the white shark Carchar- 14, 117-126. odon (Lamniformes, Lamnidae) from the early Pliocene of LUNDGREN, B. 1891. Studier öfver fossilförande losa block. Peru. Journal of Vertebrate Paleontology, 29, 1-13. Geologiska Föreningen i Stockholm Förhandlinger, 13, 111-121. MACFADDEN, B. I. and SALAS-GISMONDI, R. MAROCCO, R. and MUIZON, C. de 1988. Los vertebrados 2009fe. Caught in the act: trophic interactions between a 4- del Neogeno de la Costa Sur del Peru: Ambiente Sedimentario million-year-old white shark (Carcharodon) and mysticete y Condiciones de Fosilización. Bulletin Institut Francais Etudes whale from Peru. Palaios, 24, 329-333. AndinesXVII, 2, 105-117. FRAZZETTA, T. H. 1988. The mechanics of cutting and the MARTIN, A. F. 1996. Systematics ofthe Lamnidae and origi- form of shark teeth (Chondrichthyes, Elasmobranchii). Zoo- nation time of mitochondrial DNA sequences. 49-53. In morphology, 108, 93-107. KLIMLEY, A. P. and AINLEY, D. G. (eds). Great white GILL, T. R. 1893. Families and subfamilies of fishes. Memoirs of sharks: the biology of Carcharodon carcharias. Academic Press, the National Academy of Sciences, Washington, DC, 6, 125-138. San Diego, CA, 515 pp. GLIKMAN, L. S. 1964. Sharks ofthe Paleogene their stratigraphie MARTIN, E. E., SHACKLETON, N. J., ZACHOS, J. C. significance. Nakua Press, Moscow, 229 pp. [In Russian]. and FLOWER, B. P. 1999. Orbitally-tuned Sr isotope chem- GOTTFRIED, M. D. and FORDYCE, R. E. 2001. An associ- ostratigraphy for the late middle to late Miocene. Paleoceanog- ated specimen of Carcharodon angustidens (Chondrichthyes, raphy, 14, 74-83. Lamnidae) from the Late Oligocène of New Zealand, with MCARTHUR, J. M., HOWARTH, R. J. and BAILEY, T. R. comments on Carcharodon interrelationships. Journal of Verte- 2001. Strontium isotope stratigraphy: LOWESS Version 3: best brate Paleontology, 21, 730-739. fit to the marine Sr-isotope curve for 0-509 Ma and accompa- COMPAGNO, L. I. V. and BOWMAN, S. C. 1996. Size nying look-up table for deriving numerical age. The Journal of and skeletal anatomy of the giant megatooth shark Carchar- Geology, 109, 155-170. odon megalodon. 55-89. In KLIMLEY, A. P. and AINLEY, MBWIS, H. 2008. New aspects on the evolution of Carchar- D. G. (eds). Great white sharks: the biology of Carcharodon car- odon carcharias (Lamniformes: Lamnidae) - A phylogenetic charias. Academic Press, San Diego, CA, 515 pp. analysis including a partly associated specimen of Carcharodon GRADSTEIN, F. M., OGG, J. G. and SMITH, A. G. 2004. A escheri from Groß Pompau (Schhleswig Holstein/Germany). geologic time scale. Cambridge University Press, Cambridge, Unpublished Masters Thesis, Freie Universität Berlin, Institut 500 pp. für Geowissenschaften, Berlin, 113 pp. HANNIBAL, D. S. and JORDAN, H. 1923. Fossil sharks and and KLUG, S. 2006. Revision of Miocene mackerel sharks rays of the Pacific slope of North America. Bulletin of the (Chondrichthyes, Lamniformes), with special reference to Southern California Academy of Sciences, 22, 27-63. "Isurus" escheri from northern Germany. Journal of Vertebrate MUELLER, P. A. and GARRIDO, J. R. 1991. Variations Paleontology, 26 (Suppl), 99A. in the strontium isotopic composition of sea water during the MILLER, K. G. and SUGARMAN, P. J. 1995. Correlating Neogene. Geology, 19, 24-27. Miocene sequences in onshore New Jersey boreholes (ODP HOFFSTETTER, R. 1968. Un gisement de vertébrés tertiaires á Leg 150 X) with global á'^0 and Maryland outcrops. Geology, Sacaco (Sud-Pérou), témoin nèogéne d'une migration de faunes 23, 747-750. australes au long de la côte occidentale sudaméricaine. Comptes FEIGENSON, M. D., WRIGHT, J. D. and CLEM- Rendus de l'Académie des Sciences, Serie D, 267, 1273-1276. ENT, B. M. 1991. Miocene isotope reference section. Deep HUBBELL, G. 1996. Using tooth structure to détermine the Sea Drilling Project site 608: an evaluation of isotope and bio- evolutionary history ofthe white shark. 9-18. In KLIMLEY, stratigraphic resolution. Paleoceanography, 6, 33—52. A. P. and AINLEY, D. G. (eds). Great white sharks: the biol- MUIZON, C. de and BELLON, H. 1980. L'âge mio-pliocéne ogy of Carcharodon carcharias. Academic Press, San Diego, CA, de la Formation Pisco (Pérou). Comptes Rendus Hebdomad- 515 pp. aires des Séances de l'Académie des Sciences, Paris, Série D, 290, JORDAN, H. and HANNIBAL, D. S. 1994. Variations in the 1063-1066. strontium isotopic ratio of seawater during the Miocene: 1986. Nouvelles données sur l'âge de la Formation Pi- stratigraphie and geochemical implications. Paleoceanography, sco (Pérou). Comptes Rendus Hebdomadaires des Séances de 9, 405-426. l'Académie des Sciences, Paris, Série D, 303, 1401-1404. KEMP, N. R. 1991. Chondrichthyans in the Cretaceous and and DEVRIES, T. J. 1985. Geology and paleontology of Tertiary of Australia. 497-568. In VICKERS-RICH, P., late Cenozoic marine deposits in the Sacaco area (Peru). Geo- MONAGHAN, J. M., BAIRD, R. F. and RICH, T. H. logische Rundschau, 74, 547—563. (eds). Vertebrate paleontology of Australasia. Pioneer Design and MCDONALD, H. G. 1995. An aquatic sloth from Studio, Victoria, Australia, 1437 pp. the Pliocene of Peru. Nature, 375, 224-227. LERICHE, M. 1926. Les poissons Néogènes de la Belgique. DOMNING, D. P. and KETTEN, D. R. 2002. Odoben- Mémoires du Museé Royal du d'Histoire Naturelle de Belgique, ocetops peruvianus, the walrus-convergent delphinoid (Mam- 32, 367^72. malia: Cetácea) from the early Pliocene of Peru. 223-262. In EHRET ET AL.: WHITE SHARK EVOLUTION 1153

EMRY, R. J. (ed.). Cenozoic mammals of land and sea: trib- dentitions of extinct sharks. Journal of Fossil Research, 38, utes to the career of Clayton E. Ray. Smithsonian Contributions 141-145. to Paleobiology no. 93, 372 pp. 2007. Skeletal and dental anatomy of lamniform shark, Cre- MÜLLER, J. and HENLE, F. G. J. 1837. Ueber die Gattungen talamna appendiculata, firom Upper Cretaceous Niobrara der Plagiostomen. Archiv für Naturgeschichte, Berlin, 3, 394- Cbalk of Kansas. Journal of Vertebrate Paleontology, 27, 584— 401. 602. 1838. On the generic characters of cartilaginous fishes. SIVERSON, M. 1999. A new large lamniform shark from tbe Magazine of Natural History, 2, 33-37, 88-91. uppermost Gearle Siltstone (Cenomanian, Late Cretaceous) of NOLF, D. 1988. Dents de requins et de raies du Tertiaire de la Western Australia. Transactions of the Royal Society of Edin- Belgique. Edition de ¡'Institute Royal des Sciences Naturelles de burgh: Earth Sciences, 90, 49-66. Belgique, 184 pp. STEWART, J. D. 1999. Correlation of stratigraphie position with NYBERG, K. G., CIAMPAGLIO, C. N. and WRAY, G. A. Isurus-Carcharodon tootb serration size in the Capistrano For- 2006. Tracing the ancestry of the great white shark, Carchar- mation, and its implications for the ancestry of Garcharodon odon carcharias, using morphometric analyses of fossil teeth. carcharias. Journal of Vertebrate Paleontology, 19 (Suppl), 78A. lournal of Vertebrate Paleontology, 26, 806-814. 2002. The first paleomagnetic framework for tbe Isurus has- OSLICK, J. S., MILLER, K. G. and FEIGENSON, M. D. talis-Garcharodon transition in tbe Pacific Basin: the Purisama 1994. Oligocene-Miocene strontium isotopes: stratigraphie Formation, Gentral California. Journal of Vertebrate Paleontol- revisions and correlations to an inferred glacioeustatic record. ogy, 22 (Suppl), lllA. Paleoceanography, 9, 427-443. SUAREZ, M. E., ENGINAS, A. and WARD, D. 2006. An Early PAGES, J. B. and MILLER, J. D. 1993. Precise U-Pb ages of Miocene elasmobrancb fauna fi-om tbe Navidad Formation, Duluth Complex and related mafic intrusions, northeastern Central Ghile, South America. Cainozoic Research, 4, 3-18. Minnesota: geochronological insights to physical, petrogenetic, TANAKA, T. and MORI, S. 1996. Fossil elasmobrancbs fi-om paleomagnetic, and tectonomagmatic process associated with tbe Oiso Formation (Late Miocene) in tbe western part of the 1.1 Ga Midcontinent Rift System. Journal of Geophysical Kanagawa Prefecture. Bulletin of the Hiratsuka City Museum, Research, 98, 13997-14013. 19, 67-71. PIN, C. and BASSIN, C. 1992. Evaluation of a Sr-specific TSUCHI, R., SHUTO, T., TAKAYAMA, T., FUJIYOSHI, extraction Chromatographie method for isotopic analyses in A., KOIZUMI, I., IBARAKI, M., RANGEL, Z. C and geological materials. Analytica Chimica Acta, 269, 249—255. ALDANA, A. M. 1988. Fundamental data on Cenozoic bio- PURDY, R. W. 1996. Paleoecology of fossil white sharks. 67- stratigraphy of the Pacific coast of Peru. 45-70. /«TSUCHI, 78. In KLIMLEY, A. P. and AINLEY, D. G. (eds). Great R. (ed.). Reparts of Andean studies. Special Volume 2. Kofune white sharks: the biology of Carcharodon carcharias. Academic Printing, Shizuoka, Japan, 108 pp. Press, San Diego, GA, 515 pp. UYENO, T., KONDO, Y. and INOUE, K. 1990. A nearly SCHNEIDER, V. P., APPLEGATE, S. P., MCLE- complete tootb set and several vertebrae of tbe lamnid sbark LLAN, J. H., MEYER, R. L. and SLAUGHTER, B. H. Isurus hastalis from tbe Pliocene of Cbiba, Japan. Journal of 2001. The Neogene sharks, rays, and bony fishes from Lee the Natural History Museum and Institute, Chiba, 3, 15-20. Greek Mine, Aurora, Nortb Carolina. 71—202. In RAY, G. E. WARD, D. J. and BONAVIA, C. G. 2001. Additions to, and a and BOHASKA, D. J. (eds). Geology and paleontology of the review of, the Miocene shark and ray fauna of Malta. Central Lee Creek Mine, North Carolina, IU. Smithsonian Institution Mediterranean Naturalist, 3, 131-146. Press, ?????, 365 pp. WHITENACK, L. B. and GOTTFRIED, M. D. 2010. A RIDEWOOD, W. G. 1921. On the calcification of the vertebral morpbometric approach for addressing tootb-based species centra in sharks and rays. Philosophical Transactions of the delimitation in fossil Mako sharks, Isurus (Elasmobrancbii, Royal Society of London, Series B, 210, 311-407. Lamniformes). Journal of Vertebrate Paleontology, 30, 17—25. SCOPOLI, G. A. 1777. Introductio ad Historiam Naturdem, si- WHITLEY, G. P. 1940. Tbe fishes of Australia. Part I. Tbe stens genera Lapidum, Pantarum et Animalium, hactenus de- sbarks, rays, devilfish, and otber primitive fisbes of Australia tecta, characteribus essentialibus donata in tribus divisa, subinde and New Zealand. In Australian zoological handbook. Royal ad leges naturae. Gerle, Pragae, 506 pp. Zoological Society of New Soutb Wales, Sydney, 280 pp. SHIMADA, K. 1997. Skeletal anatomy of tbe late Cretaceous YABE, H. 2000. Teeth of an extinct great wbite sbark, Carchar- lamniform shark, Cretoxyrhina mantelli, from the Niobrara odon sp., from tbe Neogene Senhata Formation, Miura Group, Chalk in Kansas. Journal of Vertebrate Paleontology, 17, 642- Cbiba Prefecture, Japan. Tertiary Research, 20, 95-105. 652. ZHELEZKO.V. and KOZLOV, V. 1999. Elasmobrancbii and 2005. Phylogeny of lamniform sharks (Chondricbthyes: El- Paleogene biostratigrapby of Transurals and Central Asia. asmobranchii) and the contribution of dental characters to Materials on stratigraphy palaeontology of the Urals, Vol. 3. lamniform systematics. Paleontological Research, 9, 55-72. Russian Academy of Sciences Urals Brancb Uralian Regional 2006. (date of imprint 2005). Types of tooth sets in tbe Interdepartment Stratigrapbical Commission, Yekaterinburg, fossil record of sharks, and comments on reconstructing 324 pp. Copyright of Palaeontology is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.