Wang: Fossil Mammals of Asia

Chapter 31 Paleodietary Comparisons of Ungulates Between the Late Miocene of China and Pikermi and Samos in Greece NIKOS SOLOUNIAS, GINA SEMPREBON, MATTHEW MIHLBACHLER, AND FLORENT RIVALS

-1— 0— +1—

Extant ungulates occupy a great diversity of Recent habitats such as forests, woodlands, savannas, grasslands, steppe, deserts, steep mountain slopes, and arctic tundra. They also exhibit a diverse array of morphological adaptations that are suited to these varied ecological conditions. Many modern ungulates, particularly ruminants, are descended from lineages that adaptively radiated during the Middle and Late Miocene, a time of dramatic climate cooling during which open grassy and more arid habitats appeared and began to spread. The C4 grasslands had a sudden rise and spread to ecological dominance 8–3 Ma (Edwards et al. 2010). The new Late Miocene open habitats intermingled with more archaic closed habitats and have now largely replaced them in many parts of continents. All of the early mammalian studies (1900–1970s) documenting the opening of habitats are based on the appearance of new taxa of fossil ungulates. Isotopic and tooth microwear and mesowear approaches did not exist then. Paleobotanical studies have also documented patterns regarding these changes (Dorofeyev 1966; Mai 1995; Orgeta 1979). Recently, paleobotanical and isotopic studies have become more focused and have shown the complexity of these paleoecological events (Retallack 2001; Strömberg et al. 2007; Quade, Solounias, and Cerling 1994; Bond 2008; Zachos, Dickens, and Zeebe 2008). It is undisputed that such habitats did open somewhat during the later Miocene, with an onset at around 8 Ma. These new open habitats represented untapped ecological conditions within which many ungulate groups expanded, diversified, and adapted (Bond

49502_1P_03_Wang15012_text14-31.indd 676

2008; Jacobs et al. 1999; Janis, Damuth, and Theodor 2000; Janis 2008). In addition, the evolution of hypsodonty has recently been used as proxy for the degree of aridity or tree cover in various extinct habitats (Fortelius et al. 2003; Eronen et al. 2010a, 2010b). Overall these paleoecological events need more research. The paleobotanical record of the Late Miocene (11– 6 Ma) indicates the persistence of forests and woodlands along with the spreading of grasslands (Axlerod 1975; Mai 1995; Retallack 2001; Bruch et al. 2006; Bruch, Uhl, and Mosbrugger 2007). In other words, habitats during this time interval formed a widespread grassland/savanna/woodland continuum over much of the northern continents. The ecological changes that occurred during the Late Miocene and the complex ecological relationships that ungulates had with these evolving habitats, make the study of ungulate paleobiology both challenging and interesting. During the end of the Miocene and through the Pliocene, the climate of the peri-Mediterranean Sea changed from a warm wet subtropical to a warm, drier, and cooler temperate climate (Dorofeyev 1966; Axelrod 1975; Fortelius et al. 2003; Eronen at al. 2010a, 2010b; Strömberg et al. 2007; Solounias, Rivals, and Semprebon 2010). The Pikermi, Samos, and Maragheh faunas fall at a critical time in the cooling trend of this climatic change and probably were temperate faunas. Considered as a fauna, or chronofauna, the late Miocene of China (hereafter denoted LMC) samples more time depth, but parts of it were near the time of Pikermi and Samos. The modern

6/27/12 12:10 AM

PALEODIETARY COMPARISONS OF UNGULATES BETWEEN CHINA AND PIKERMI AND SAMOS IN GREECE

East African savanna has been used as the primary model for the late Miocene ungulate-rich ecosystems of the northern continents, within which hypsodonty in many ungulates evolved (Abel 1927; Osborn 1910; Matthew 1926; Coppens 1994; Kurtén 1952; 1971; de Bonis et al. 1992). In this model, the woodland component of the savanna was not considered. An emphasis on open grassy savanna plains is apparent in murals and other reconstructions of the time. Other ecologies such as the woodlands of India and of Europe were not considered sufficiently as alternative model biomes for the Late Miocene. Kurtén (1952) represented the ideas and paradigms of many, and he proposed the presence of a vast open and arid region (spanning seven time zones) extending from Greece to China. The periphery of that steppe was hypothesized to have been surrounded by more forested habitats. In Kurtén’s study, particularly referencing his model map, there were three types of faunas: a vast steppe fauna, a peripheral forest fauna, and a transitional fauna bridging the steppe and forest habitats. Pikermi in Greece and Hsin-An-Hsien and Wu-Hsiang-Hsien in the Shanxi Province of China were proposed to be transitional localities, and Samos was part of the steppe (Kurtén 1952). It has been proposed that many of these late Miocene faunas represent a geograph ically widespread extinct biome, the Pikermian Biome, within which the dominant habitats were closed woodlands (Solounias et al. 1999). The physiognomy of the Pikermian Biome was different from the vast expanses of open savanna and grassland in present-day East Africa. In reviewing the paleobotanical record, Solounias and colleagues (Solounias et al. 1999; Solounias, Rivals, and Semprebon 2010) found that the woodland aspect of the savanna is more similar in plant physiognomy to the later Miocene. The Pikermian Biome was widespread during the late Miocene from Greece through Bulgaria, Ukraine, Turkey, Southern Russia, Iran, and east all the way to China (Deng 2006). Taxa from this biome have also been found in southern France, Spain, and North Africa (Thenius 1959). The cohesion of a single Pikermian Biome is most strongly supported by the faunal similarities of the Late Miocene faunas from China to Greece, Spain, and North Africa (Solounias et al. 1999; Solounias, Rivals, and Semprebon 2010; Bernor, Koufos, et al.1996). The biome is characterized by Indarctos, marchairodontids, proboscideans, and chalicotheres. It also includes many species of hyaenids, rhinos, equids, and giraffids unlike those of the African savanna. The bovids are also numerous. Advanced Miocene hypsodont bovids have small masseter muscles; their masticatory morphology resembles extant Tragelaphini (Solounias, Moelleken, and Plavcan 1995). In many

49502_1P_03_Wang15012_text14-31.indd 677

677

cases, the very same species are found in late Miocene of China (LMC) and at Pikermi and Samos. The modern Savanna Biome includes open grassy plains, low-density tree cover woodlands, and densely treed woodlands (Beerling et al. 2005; Judith Harris, pers. comm. 2010). Herbivore dietary interpretations by Solounias, Rivals, and Semprebon (2010) based on analyses of ungulate tooth microwear suggest that Pikermi and Samos were most likely a woodland mosaic with more extensive tree cover with fewer and/or smaller patches of grassy open habitat in comparison to the African open savanna. Pikermian Biome habitats provided a diversity of opportunities for species that depended on browsing as well as species that grazed. Savanna C4 grasses appear to have been subdominant to woody plants and to C3 grasses that would have grown in shaded areas of the woodland, glades, and margins of water. The ungulate components of the late Miocene of China (LMC) Pikermi and Samos faunas were more species rich and more diverse in diet than the ungulates observed in modern African forests, woodlands, or savannas, yet dietarily most similar to the ungulates found in the woodlands of India (based on tooth microwear; unpublished data). It is unlikely that the Pikermi and Samos ungulates inhabited dense forests, because there is no microwear evidence for concentrated fruit browsing. Conversely, widespread grasslands are unlikely because many mixed feeders are present as well as browsers. Koufos et al. (2009) agreed in an additional study of mesowear and microwear for Samos and found an environment of “open bushland with a thick grassy herbaceous layer.” In this study, we use tooth mesowear analysis to compare the paleodietary patterns of the classic Late Miocene faunas of Pikermi and Samos from Greece to those of China. A new scale is used for evaluating mesowear, a technique that uses the sharpness of molar cusp apices as a proxy for dietary abrasion. Details of this mesowear scale can be found in Mihlbachler et al. (2011). Animals with high-abrasion diets, particularly grazers, have high incidences of rounded and blunted cusp apices with low relief, while animals with low-abrasion diets, particularly folivorous browsers, have higher incidences of sharpened cusp apices with high relief. Specifically, this study evaluates similarity in mesowear and in hypsodonty between Pikermi, Samos, and the late Miocene of China (LMC), which record key samples of species of the Pikermian Biome. It tests the integrity of the biome and how the Pikermian Biome compares to the African savanna—the only biome today that rivals these Late Miocene faunas in ungulate diversity. These comparisons are based on mesowear and hypsodonty.

—-1 —0 —+1

6/27/12 12:10 AM

678

ZOOGEOGRAPHY AND PALEOECOL OGY

GEOLOGICAL AND BIOSTRATIGRAPHIC SETTING

The Pikermi localities are stratigraphically and geograph ically very confi ned and best approximate a fauna. The localities are concentrated in a small area: 0.2 km 2 with a sedimentary thickness of 40 m. The sediments are relatively horizontal and all samples come from the exposures of a small ravine (Megalo Rema; Gaudry 1862– 1867; Abel 1927). There are no volcanic deposits at Pikermi, and though the radiometric age is not known, Pikermi is considered to be of Turolian age. The Samos samples are mixed from six primary localities. Older collectors did not record locations during their excavations, which resulted in the mixing of fossils from different stratigraphic levels. The bone beds are concentrated on the upper part of the Main Bone Bed Member of the Mytilini Formation. The geographic area is small (2 km 2) (Solounias 1981; Weidmann et al. 1984; Koufos and Nagel 2009). The estimated overall time span for these bone beds is approximately 800,000 yr for the core Samos faunas (our data). An average radiometric age of 7.2 Ma is given for the upper part of these sediments (the Main Bone Beds, Turolian; present study). Fossils were derived from small depression troughs where some bone beds may have taken thousands of years to form. Thus, unlike Pikermi, Samos represents composite faunas. The data from the late Miocene of China (LMC) are preliminary. The samples are mixed from numerous localities and time intervals and defi nitely do not represent a single fauna. The time depth is at least 4 million yr and the geographic area is very large. The late Miocene of China (LMC), as represented by our samples, is a proxy for a chronofauna.

RELATIVE AGE OF PIKERMI BASED ON THE SAMOS SPECIES

Evidence for the relative ages of Pikermi versus Samos comes from the diversity of eight groups of herbivores (see Solounias 1981, Bernor, Solounias, et al. 1996, and Koufos and Nagel 2009 for general faunal lists). A few species listed here are useful in the relative dating of Pikermi, but for some we have no mesowear data. Not all of these species were used in the dietary interpretations that follow.

-1— 0— +1—

1. One species of Aceratherium at Pikermi versus four Chilotherium species at Samos (the Samos taxon cluster is more derived).

49502_1P_03_Wang15012_text14-31.indd 678

2. Two species of hipparionine at Pikermi (Cremohipparion mediterraneum and Hippotherium brachypus) versus six at Samos (Cremohipparion proboscideum, Cremohipparion matthewi, Cremohipparion nikosi, Hippotherium sp., Hipparion dietrichi, and “Hipparion” sp. (Bernor, personal comm. 2010). The numerous horses at Samos follow a “punctuated” event in the Subparatethyan Province recorded in the latest Vallesian levels of Sinap, Turkey (Bernor et al. 2003; Scott et al. 2003), in which several superspecific clades originated across the geographic extent of the Pikermian Biome (Eronen et al. 2009). Thus, Samos is younger than this punctuated event. The Pikermi species Hippotherium brachypus is more primitive than, and could have evolved into, the more derived Samos form of Hippotherium sp. (Bernor, Koufos, et al. 1996). The same is plausible for Pikermi Cremohipparion mediterraneum and the more derived form Cremohipparion proboscideum (Bernor, pers. comm. 2010). The paucity of grazing species such as equids at Pikermi suggests that the habitat was more closed and that Pikermi may be older than Samos (Solounias, Rivals, and Semprebon 2010). 3. Tragoportax rugosifrons and T. curvicornis from Samos can be derived from the more primitive T. amalthea found at Pikermi (Solounias 1981). The presence of Tragoportax amalthea at both localities, however, does suggest that they were close in age. 4. Protragelaphus skouzesi is found both at Pikermi and Samos, but Prostrepsiceros houtumschindleri, a probable descendant of Protragelaphus, is found only on Samos. 5. As with equids, there are more species of Gazella at Samos, which suggests a diversification from the single species found at Pikermi. 6. Oioceros rothii (Pikermi) is fairly primitive and could have evolved into a more derived Samos species (Oioceros wegneri, Samotragus crassicornis, or Prosinotragus kuhlmanni). 7. Palaeoreas is found at both Samos and Pikermi. The taxon is common at Pikermi but rare at Samos. Criotherium argalioides and Parurmiatherium rugosifrons (a rare species whose dentition is not known and hence not included in the mesowear tables) most likely descended from Palaeoreas and are also found at Samos but not at Pikermi. In addition, at Samos, Palaeoreas occurs only at the Stefana and Tholorema locality of Q6. These localities are within the Old Mill Beds (estimated at 8.2 Ma). Th is observation substantiates that Palaeoreas could have been ancestral to Criotherium and Parurmiatherium which are found in the younger horizons (Main Bone Beds). 8. Protoryx carolinae from Pikermi is more primitive and probably ancestral to the two Pachytragus species

6/27/12 12:10 AM


found at Samos (P. laticeps and P. crassicornis; Solounias 1981). In conclusion, although Pikermi is not dated radiometrically, these patterns of closely related species with descendents occurring at Samos but not at Pikermi suggest that Pikermi was slightly older than Samos. The age of Pikermi may be estimated at about 8 Ma.

MATERIALS AND METHODS

Mesowear data for the Late Miocene species from China were obtained primarily from specimens in the collections at the American Museum of Natural History. Additional specimens of giraffids and Hezhengia bohlini were obtained at the Hezheng Paleozoological History Museum. We did not include specimens from the Late Miocene–Pliocene of China housed in the Uppsala collection. For Pikermi and Samos, original specimens and casts of teeth were used from the following institutions: American Museum of Natural History, New York; British Museum of Natural History, London; Carnegie Museum, Pittsburgh; Hessischer Landesmuseum, Darmstadt; The National Museum of Natural History (Smithsonian), The Natural History Museum, London; Naturhistorisches Museum, Basel; Naturhistorisches Museum, Vienna; Musée Géologique, Lausanne; Muséum national d’Histoire naturelle, Paris; Museum of Comparative Zoology, and Peabody Museum at Harvard, Cambridge, Massachusetts; Yale Peabody Museum, New Haven, Connecticut; Museum of Palaeontology and Geology, National and Kapodistrian University of Athens; Geologisch-Paläontologisches Institut, Münster; Paläontologisches Institut der Universität Wien, Vienna; Sammlung für Paläontologie und Historische Geologie, Munich; Senckenberg Forschungsinstitut und Naturmuseum, Frankfurt; and Württenemberische Naturalen-Sammlung, Stuttgart. The extant mesowear data were collected at The American Museum of Natural History, British Museum of Natural History, Carnegie Museum, Museum of Comparative Zoology and Peabody Museum of Harvard, National Museums of Kenya, National Museum of Natural History, and Yale Peabody Museum. The African Savanna Biome includes woodlands and grasslands. Several species found in the African savanna also inhabit the African woodlands—an overlap. Open grasslands are also included in the savanna biome. The African forest species are distinct from the other ecologies. Table 31.1 summarizes the species, number of speci-

49502_1P_03_Wang15012_text14-31.indd 679

679

mens, mesowear average for each species, and the hypsodonty category. Mesowear is a measure of the relative amount of abrasive and attritive wear on teeth as reflected by the shapes of the worn cusp apices and results in an apical shape formed by years of feeding. Because mesowear relies on a macroscopic amount of dental wear, it reflects diet over a much longer time period than microwear and is not susceptible to the “last supper effect” (Fortelius and Solounias 2000). Originally, mesowear had the following subdivisions: high apices or low = occlusal relief; sharp round or blunt = cusp shape (Fortelius and Solounias 2000). The new scale of mesowear used in this study is composed of seven states (0– 6). Observation of many dentitions indicates that there are very few to no teeth that have high and blunt cusps. Thus, this category could be excluded. The same is true with low and sharp cups, which may occur in certain individuals of zebra but are again extremely limited and can be excluded. Thus, the old scale is reduced basically to high relief or low, with high subdivided as sharp or round apices, and low apices subdivided into round or blunt. The new scale can be summarized as follows: high and sharp becomes 0 or 1 (depending on the degree of sharpness); in like fashion, high and rounded becomes 2 or 3 and low and rounded becomes 4 or 5, while low and blunt equals 6. Each number is blunter than the previous (0 is the sharpest and 6 the bluntest). The new scale was built on real Equus molars exhibiting differential mesowear. These teeth were used as a basis for assigning fossil teeth to mesowear stages. These seven cusps were chosen to represent standardized intervals along a mesowear continuum, ranging from sharp to blunt as explained previously. The seven selected Equus molar cusp apices were cast onto an epoxy stage and used as a sort of mesowear “ruler,” such that a fossil tooth is compared to the seven reference cusps and assigned a score ranging from 0 (sharpest) to 6 (bluntest). Figure 31.1 shows the mesowear scale and its application. The mesowear strip is three dimensional, and this greatly facilitates apical cusp comparisons. Th is technique is described more fully by Mihlbachler et al. (2011) and differs somewhat from the original formulation of the mesowear method (Fortelius and Solounias 2000) in which teeth were assigned to two categories, relative sharpness and relief, without any clearly demarcated boundaries between mesowear stages. Cusp sharpness and degree of relief are probably not independent variables. For example, blunt cusps by defi nition have no relief. In this newer system, mesowear is treated as a univariate variable reflecting a continuum from

—-1 —0 —+1

6/27/12 12:10 AM

Table 31 Extant and Extinct Species Studied Using Mesower and Hypsododnty Mesowear scale is from 0 to 6 (sharpest to most blunt); crown height is 1 = barchydont, 2 = mesodont, 3 = hypsodont Species

-1— 0— +1—

Aepyceros melampus Alcelaphus buselaphus Antidorcas marsupialis Ceratotherium simum Connochaetes taurinus Damaliscus lunatus Diceros bicornis Equus burchelli Gazella granti Gazella thomsoni Giraffa camelopardalis Hippotragus equinus Hippotragus niger Kobus ellipsiprymnus Litocranius walleri Ourebia ourebi Redunca redunca Sigmocerus lichtenstanii Syncerus caffer aequinoctialis Taurotragus oryx Tragelaphus imberbis Tragelaphus strepsiceros Boocercus eurycerus Cephalophus dorsalis Cephalophus natalensis Cephalophus niger Hyaemoschus aquaticus cottoni Okapia johnstoni Tragelaphus scriptus Bohlinia attica Ceratotherium pachygnathus Criotherium argalioides Dicerorhinus schleimacheri Gazella gaudryi-capricornis Gazella larger Gazella sp. very small Gazella the U skull Gazella the V skull Graecoryx valenciennesi Hippotherium dietrichi Hippotherium giganteum Hippotherium matthewi Hippotherium primigenium Hippotherium proboscideum Oioceros wegneri Pachytragus crassicornis Pachytragus laticeps

49502_1P_03_Wang15012_text14-31.indd 680

N

Mesowear Score

HYPSODONTY

LOCALITY

17 76 26 24 52 5 34 121 17 146 61 26 20 22 69 128 77 17 31 21 31 7 27 28 6 31 18 8 47 6 13 18 1 26 1 1 1 1 1 11 1 6 14 19 3 23 49

1.2941 3.0789 0.6923 4.6666 3 4 0.11764 4.67768 1.29411 1.20547 0.68852 2.23076 2.6 2.0909 1.42028 1.6875 2.12987 2.58823 2 0.66667 0.8387 2 1.1851 2 2 1.5483 1.66666 0.25 1.02127 2.4 2.92308 1.77777 1 1.2 0 0 0 1 1.4 2.6 0 3.3 2.5 3.33333 2 1.65217 1.26531

3 3 3 3 3 3 2 3 3 3 1 3 3 3 1 3 3 3 3 2 1 1 1 1 1 1 1 1 1 1 3 3 2 2 2 2 2 2 1 3 3 3 3 3 2 2 2

Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna Savanna African forest African forest African forest African forest African forest African forest African forest Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos

6/27/12 12:10 AM

Table 31 (continued ) Species

N

Mesowear Score

HYPSODONTY

Palaeoryx pallasi Palaeotragus coelophrys Palaeotragus rouenii Prostrepciceros houtumschinderi Protragelaphus skouzesi Pseudotragus capricornis Samokeros minotaurus Rongia Samotherium major Samotherium Munich Samotherium sansu lato Sporadotragus parvidens Tragoportax amalthea Tragoportax amalthea – Graecoryx-like Tragoportax amalthea flat front horn Tragoportax amalthea weavy front horn Tragoportax curvicornis Tragoportax rugosifrons Gazella Munich Graecoryx valenciennesi Helladotherium duvernoyi Hippotherium primigenium Oioceros rothii Palaeoreas lindermayeri Palaeoryx pallasi Palaeotragus rouenii Tragoportax amalthea Bohlinia sp. Chilotherium hoberi Dorcadoryx triquetricornis bent braincase early Bos-like Gazella ?dorcadoides Gazella hypsodont Gazella large one fossa deep Gazella-like but more hypso Gazella-like the Samos ones but hypso Gazella paotehensis 3 in a block Gazella smaller than the Samos Gazella type gaudryi-capricornis Gazella very very large Henzengia bohlini Hez Hippotherium dermatorhinum Hippotherium fossatum Hippotherium kreugergi Hippotherium placodus Hippotherium tylodus Honanotherium schlosseri Hez

14 2 13 4 1 1 3 1 1 9 13 21 1 1 1 1 17 4 4 2 14 5 24 2 16 2 2 13 1

1.57143 1 1.5 0.75 1 2 1 2 4 2.88888 0.69231 1.38095 0 2 0 2 1.47059 1.5 1.4 1 2.5 1.33333 1.29167 3 0.875 2 2.5 2.33333 2

2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 3 2 2 2 1 2 1 2 2

Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Samos Pikermi Pikermi Pikermi Pikermi Pikermi Pikermi Pikermi Pikermi Pikermi Miocene of China Miocene of China Miocene of China

1 4 1 2 1 1 3 4 4 7 10 3 3 2 2 3 6

3 0.5 1 1 1.5 1.5 1.66666 0.75 1.26923 2.14286 3.4 3.5 2 3 3 4 3.2

3 2 3 2 2 2 2 2 2 3 3 3 3 3 3 3 2

Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China

LOCALITY

—-1 —0 —+1

(continued)

49502_1P_03_Wang15012_text14-31.indd 681

6/27/12 12:10 AM

682


Table 31 (continued )

-1— 0— +1—

Species

N

Mesowear Score

HYPSODONTY

Honanotherium sp. (schlosseri) Pachytragus largelli small brain bent Palaeoryx sinensis Palaeotragus (very large) Palaeotragus coelophrys Hez Palaeotragus microdon-rouenii Palaeotragus rouenii Palaeotragus sp. (coelophrys) Palaoeoreas lindermayeri Plesiaddax depereti Samotherium (mini species) Samotherium boissieri Hez Samotherium neumayri Hez Samotherium sinense Hez Samotherium sp. (sinense) Schansitherium tafeli Hez Sinoryx bombifrons Sinotragus wimani Urmiatherium intermedium

5 3 13 1 13 8 1 18 1 5 1 7 2 1 2 6 3 7 14

1.8 0 1.46154 2 2.6 2 2 1.72222 2 1 1 3.2 1 2.5 4 2.3 1.66666 2 1.92857

2 2 2 1 1 1 1 1 2 3 2 2 2 2 2 2 3 3 3

sharp cusps with high relief to blunt cusps with zero relief, similar to an approach adopted by Mihlbachler and Solounias (2006). Table 31.1 lists the mesowear data. Crown-height categories (brachydont, mesodont, hypsodont) for the extant species are primarily from Janis (1988). Those for the fossil species were extracted from the NOW database of Eurasian Neogene fossil mammals (Fortelius 2009), available online (http:// www.helsinki.fi/science/now). The dataset was downloaded on May 23, 2009. Crown-height categories were not available for two species, but in those cases we concordantly used measurements collected on museum specimens for Palaeoryx pallasi and Samokeros minotaurus, which are mesodont (hypsodonty index = 0.84 and 0.82, respectively). Some species were assigned a hypsodonty category by visual matching to other known species and without measur ing tooth heights. Statistical tests comparing mesowear data between the faunas were performed on SPSS v14.0. For analyses involving two faunas (e.g., African forest fauna vs. African savanna fauna) we used the nonparametric Mann– Whitney U test. We used Kruskal–Wallis tests when three or more faunas were included in the comparison.

49502_1P_03_Wang15012_text14-31.indd 682

LOCALITY Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China Miocene of China

Figure 31.1 The new mesowear scale, a plastic strip with selected teeth in a sequence from sharpest to most blunt. NS holding the strip at the American Museum with bovids Palaeoryx and Urmiatherium from China.

6/27/12 12:10 AM


683

16

14

Late Miocene of China

Samos

12

Number of species

10

Recent African Savanna

8

Pikermi

6

Recent African forest

4

2

0

4-5 3-4 2- 3 1-2 0-1

4-5 3-4 2-3 1-2 0-1

4-5 3-4 2-3 1-2 0-1

4-5 3-4 2-3 1-2 0-1

4- 5 3-4 2-3 1-2 0-1

average mesowear scores Figure 31.2 All species from Samos, Pikermi, and the Late Miocene of China plotted along with the extant African rainforest species and the African woodland– savanna continuum. The numbers are the mesowear scores.

RESULTS

The fi rst hypothesis is that the distribution of mesowear scores between the three fossil faunas (Samos, Pikermi, and the late Miocene of China; LMC) is the same. The frequency distributions of the average mesowear scores (per species) are plotted in figure 31.2. The overall distributions of the mesowear scores among the three fossil faunas are similar. We were unable to statistically detect significant differences between the average mesowear scores for the three fossil faunas (P = 0.075). However, the small number of species in the Pikermi fauna significantly reduced the power of the test. Significant differences were found between the average mesowear scores of the two larger faunas, Samos and China (p = 0.031), with the late Miocene of China (LMC) having overall higher mesowear scores (i.e., blunter cusps with

49502_1P_03_Wang15012_text14-31.indd 683

lower relief), suggesting slightly more abrasive diets overall. The second hypothesis is that the distributions of mesowear scores in each of the individual fossil areas do not differ from the African savanna. The average mesowear score distribution of the African savanna fauna is intermediate between that of Samos and the late Miocene of China (LMC), and significant differences between the extant fauna and the two fossil faunas were not found (p = 0.158; p = 0.798). The third hypothesis is that the levels of hypsodonty between the three fossil areas are the same. In comparing these areas, the average mesowear score distributions contrast with the distribution of crown height types. Figure 31.3 shows the histograms of the distributions of crown-height types among the three areas. Although the mesowear scores of the African savanna fauna were found

—-1 —0 —+1

6/27/12 12:10 AM

684


brachydont

30

mesodont

Samos

number of species

hypsodont


25


20

15


10

Pikermi 5

0

degree of hypsodonty Figure 31.3 All species from Samos, Pikermi, and the Late Miocene of China sorted according to hypsodonty into three levels: brachydont, mesodont, and hypsodont.

12

brachydont

Samos


Number of species

10

8

mesodont

hypsodont


6


Pikermi

4

2

4–5

5–6

2–3

3–4

0–1

1–2

4–5

5–6

2–3

3–4

0–1

1–2

4–5

5–6

2–3

3–4

0–1

1–2

4–5

5–6

2–3

3–4

0–1

1–2

4–5

5–6

2–3

3–4

0–1

1–2

0

average mesowear scores Figure 31.4 All species from Samos, Pikermi, and the Late Miocene of China plotted along with the extant African rainforest species and the African woodland– savanna continuum. The numbers are the mesowear scores. The hypsodonty is subdivided into three levels: brachydont, mesodont, and hypsodont.

-1— 0— +1—

to be intermediate with respect to the fossil areas, the extant African savanna fauna is markedly more hypsodont overall in comparison to all of the studied Miocene areas. (Samos: P = 0.04, Pikermi: P = 0.018, LMC: P = 0.027). All three fossil faunas are dominated by mesodont taxa, while the African Savanna is dominated by

49502_1P_03_Wang15012_text14-31.indd 684

hypsodont taxa. LMC appears to have slightly more hypsodont taxa in comparison to Samos and Pikermi, although the difference is not significant (Kruskal–Wallis test P = 0.231). Elimination of the small Pikermi fauna does not yield a significant change in the statistical result (P = 0.466).

6/27/12 12:10 AM


The fourth hypothesis is that the levels of hypsodonty within the three fossil areas are not different from the modern African savanna. The fi ft h hypothesis is that there is no correlation between hypsodonty and mesowear scores within any of the three fossil areas. Figure 31.4 plots the distribution of average mesowear scores per crown-height type. If hypsodonty is a reliable proxy for dietary abrasion, higher-crowned species are expected to have higher mesowear scores. Statistically significant differences were found in the average mesowear scores between the crown-height types in the African savanna fauna (P = 0.013) and Samos fauna (P = 0.049). Exclusion of the few brachydont species from the Samos fauna increases the significance (P = 0.019). A weaker relationship between crown-height type and mesowear was found in the LMC (P = 0.112). When brachydont taxa are excluded from the Chinese areas, the test is still insignificant (P = 0.068). These results generally suggest that crown height type and mesowear are weakly related, with hypsodont taxa having blunter cusps overall than brachydont and mesodont taxa. However, a relationship between mesowear and crown-height type is questionable among brachydont and mesodont taxa. Furthermore, among the three fossil faunas, although the mesowear scores for hypsodont taxa are higher on average than those of mesodont taxa, mesodont species show a greater range of mesowear values, in two instances (Samos and LMC) completely overlap the mesowear ranges of the brachydont and hypsodont taxa, and in two instances (Samos, Pikermi) include species with lower (blunter) mesowear scores than any of the hypsodont taxa.

MESOWEAR PATTERNS OF EXTINCT SPECIES

Various patterns have been observed in the mesowear scores and hypsodonty of the extinct species (table 31.2). The Miocene species distribution in terms of mesowear and hypsodonty is different from the African savanna. Hippotherium is not like the zebra (Equus burchelli). The Miocene Hippotherium species are spread throughout the mesowear score spectrum which is suggestive of a diversity of diets. On the contrary, zebras are grazers. Similarly, the mesowear of Ceratotherium pachygnathus is sharper than that of the modern species Ceratotherium simum. The other two genera of extinct rhinocerotids are Chilotherium and Dicerorhinus. Chilotherium has notably blunter apices than Dicerorhinus. There are many species of Giraffidae, and unlike today, they appear to be diverse in mesowear and were probably similar to bovids in their feeding habits. Giraf-

49502_1P_03_Wang15012_text14-31.indd 685

685

fids are dietarily diverse, as has been shown in previous microwear research (Solounias et al. 2000). Giraffids are spread throughout the mesowear scores and differ from Giraffa camelopardalis, which is a committed browser. Samotherium sensu lato was broadly adapted dietarily. Samotherium spp. are interesting in having the bluntest apices in the samples of both Greece and China. Palaeotragus differs from Samotherium in having sharper apices. Overall, its diet is less abrasive than that of Samotherium. Honanotherium is notably different from Bohlinia in mesowear. Th is result reinforces that the two genera are distinct. The Bohlinia mesowear is very blunt, unlike the modern giraffe (Giraffa), but they are closely related. Helladotherium has sharp apices and strongly differs from the Pleistocene African Sivatherium (both genera are Sivatheriinae). The systematic and hypsodonty relationship of the bovid species to mesowear is complex. Most Miocene bovid species have sharper apices than Samotherium, Honanotherium, and most Hippotherium. They are clearly different in mesowear scores from the majority of Hippotherium. In the modern African savanna, grazing bovids have the same mesowear as the zebra. The only bovids that have extreme bluntness of 3 are Palaeoryx pallasi and Hezhengia. Palaeoryx pallasi and Hezhengia, however, are not more hypsodont than other bovids (e.g., Pseudotragus, Palaeoreas, and Pachytragus). In the modern African savanna, grazer bovids are more hypsodont than many of the mixed feeders. Mesodont bovid taxa such as Tragoportax, Prostrepsiceros, Sporadotragus, and Gazella have sharper scores than the brachydont taxon Graecoryx. There is a single hypsodont species (Pachytragus largelli) with sharp apices. In the modern African savanna, there are no hypsodont bovids with sharp apices. During the late Miocene, we fi nd taxa similar in hypsodonty that differ in mesowear scores. Examples are Sporadotragus and Pseudotragus, Prostrepsiceros and Protragelaphus, Tragoportax and Samokeros, Pachytragus and Sinotragus, Criotherium and Sinotragus, and Urmiatherium and Sinotragus. Similar or conspecific taxa with the same hypsodonty can differ in mesowear scores. Examples are Gazella docradoides and Gazella gaudryi, Tragoportax amalthea from Samos and Tragoportax rugosifrons from Samos, Oioceros rothii and Oioceros wegneri, Pachytragus largelli and Pachytragus laticeps, Hippotherium giganteum and Hippotherium primigenium, and Samotherium neumayri and Samotherium boissieri. Mesodont species of Tragoportax, Oioceros, Palaeoryx, Palaeoreas, and Gazella have blunter apices than Prostrepsiceros, Sporadotragus, and other Tragoportax and Gazella

—-1 —0 —+1

6/27/12 12:10 AM

Table 31.2 Tooth Mesowear and Hypsodonty for the Pikermian Biome Taxa The Miocene species are from two clusters of localities: P from Pikermi and S from Samos (indicated in bold). The Chinese Miocene are not bolded 0 Pachytragus largelli CH

HYPSODONT

Gazella hypsodont CH

1 Pachytragus laticeps S Pachytragus crassicornis S Criotherium argalioides S Plesiaddax depereti CH Sinoryx bombifrons CH Urmiatherium intermedium CH Gazella-like but more hypsodont CH

Hippotherium giganteum S Helladotherium duvernoyi S P

MESODONT

Tragoportax Graecoryx like S Tragoportax amalthea weavy S Prostrepsiceros houtumschindleri S Sporadotragus parvidens S

Gazella U skull S Gazella sp very small CH Gazella dorcadoies CH

Palaeotragus rouenii P BRACHYDONT

Samotherium mini species CH Samotherium neumayri CH Honanotherium sp CH

Palaoeoryx pallasi S Tragoportax rugosifrons S Tragoportax amalthea S Samokeros minotaurus S Oioceros rothii P Gazella gaudryi P S Gazella V skull S Gazella capricornis P Gazella Munich sample P Gazella like Samos but hypsodont CH Gazella more hypsodont CH Gazella gaudryi-capricornis CH Gazella large one fossa deep CH Gazella larger CH Gazella smaller than Samos CH Gazella paotehensis CH Protragelaphus skouzesi P S Palaeoreas lindermayeri P Dicerorhinus schleimacheri S

Palaeotragus rouenii S Palaoeotragus coelophrys S Palaoeotragus coelophrys CH Graecoryx valeciennesi P S

Note: The Miocene species are from two clusters of localities: P from Pikermi and S from Samos (indicated in boldface). The Chinese Miocene species are not in boldface.

-1— 0— +1—

49502_1P_03_Wang15012_text14-31.indd 686

6/27/12 12:10 AM

2 Sinotragus wimani CH

Ceratotherium pachygnathus (P) S Hippotherium dietrichi S Hippotherium primigenium P S Samotherium sensu lato S Samotherium major S Samotherium boissieri S Samotherium colossal CH Samotherium (very large) CH Samotherium sinense CH Schansitherium tafeli CH Palaeoryx sinensis CH Tragoportax amalthea P Tragoportax curvicornis S Pseudotragus capricornis S Oioceros kuhlmanii S Oioceros wegneri S Pseudotragus capricornis S Dorcadoryx triquerticornis CH

3

4

early Bos CH Hippotherium tylodus CH Hippotherium dermatorhinum CH Hippotherium fossatum CH Hippotherium kreugergi CH Hippotherium placodus CH Hipotherium proboscideum S Hippotherium matthewi S Honanotherium schlosseri CH Samotherium boissieri CH

Samotherium Munich S Samotherium sinense CH

Palaeoryx pallasi P

Hezhengia bohlini CH

Gazella gaudryi-capricornis S Gazella very large CH Gazella (very large) CH

Palaeoreas lindermayeri CH

Chilotherium hoberi CH Bohlinia attica P S Bohlinia sp CH Palaeotragus sp (large) CH Palaoeotragus coelophrys CH Palaeotragus micorodon-rouenii CH Palaeotragus rouenii CH

—-1 —0 —+1

49502_1P_03_Wang15012_text14-31.indd 687

6/27/12 12:10 AM

688

-1— 0— +1—


species, which have similar crown height. The most hypsodont bovids are Pachytragus, Criotherium, Sinotragus, Plesiaddax, Sinoryx, and Urmiatherium. These however do not have the same mesowear scores.

combined into one variable is simpler and gives more resolution. The apices of the scale are three dimensional and this enables a more precise comparison with the unknown in question.

DISCUSSION

Paleoecological Approaches

Mesowear, Microwear, and the New Method

Overall three other approaches show a similar pattern for the Pikermian Biome:

Dental wear of herbivorous ungulates is primarily caused by tooth-on-tooth attrition and abrasion from the plants ingested, soil, sand, dust, and other contaminants. These occlusal interactions produce microscopic scars (pits and scratches), frequently described as microwear. As the tooth wears, dental microwear features have a high turnover rate, and microwear reflects what an animal ate in the short term, probably within hours to weeks, depending on the specific dental wear rate (Solounias, Fortelius, and Freeman 1994). In contrast, macroscopic aspects of cusp shape (mesowear) involve substantially more dental wear and therefore reflect diet over a much longer time interval and are not perceptibly subject to daily dietary fluctuation but rather reflect overall diet over the course of months to years (Fortelius and Solounias 2000). Mesowear and microwear reflect aspects of diet on different temporal and spatial scales. Incongruence between these results may not mean that one of these methods is misleading. Rather, incongruence between microwear and mesowear suggests differences between the proximate diet at the time of death and the average diet over the course of seasons (years). Microwear ought to be biased toward the habitats proximate to the location and season of the death of an individual, whereas mesowear ought to be biased toward the overall average diet of an individual. A methodology needs to be developed for an effective comparison of mesowear and microwear on the same samples. The new mesowear scale of seven states of wear in a gradation from 0 to 6 is simple and easy to use. Each apex represents a par ticu lar state but also somewhat sharper states. For example, a score of 3 represents the apical morphology of 3 and all apices that are in between 3 and 2. The state 0 represents quite a sharp tooth but also includes sharper teeth. The fact that we used horse teeth to construct the scale is not relevant. It can be used with any type of selenodont tooth, but we must admit that it is easier to use with perissodactyls than with artiodactyls. We fi nd the plastic strip with the selected apices to be the preferable way to score teeth. The fact that the aspect of high or low relief and relative cusp bluntness have been

49502_1P_03_Wang15012_text14-31.indd 688

1. An isotopically C3-dominated ecosystem was found in various areas of the Miocene of Greece and Turkey, including Pikermi and Samos, which was composed of mostly woodland (Quade, Solounias, and Cerling 1994; Quade et al. 1995; Cerling et al. 1997). 2. A recent dental microwear study of the Samos and Pikermi faunas suggests that the Samos and Pikermi faunas occupied woodland habitats with limited areas of open savannas (Solounias, Rivals, and Semprebon 2010). Overall, the ungulates were mixed feeders, suggesting seasonality or a complex woodland environment with some grasses. Many giraffid, bovid, and hippotheres were mixed feeders. The grazers were primarily giraffids and hippotheres. Most of the browsers were Antilopini. The discovery of seasonality from microwear is novel, and in some ways better than paleobotanical inference. Most Miocene plant taxa span such wide latitude that it is difficult to decipher seasonality form genera. 3. Paleobotanical data from various localities during the Late Miocene of the Mediterranean are also in agreement with C3 vegetation and woodland environments (Axelrod 1975; see Solounias, Rivals, and Semprebon 2010 for a summary). Under the woodland scenario, the presence of C3 grasses in the woodland clearings, meadows, and water margins was possible. The presence of C3 grassland needs to be proven but could not occur under present-day carbon dioxide conditions. The Late Miocene is also marked with the onset of C4 grasses (Edwards et al. 2010), which were limited and did not necessarily grow in the areas encompassed under this investigation. The differences of the overall mesowear scores between the species studied from Greece and China are minor, suggesting that the distribution and range of diets of the ungulates are similar in terms of abrasion (see figure 31.2). The mesowear scores of the Chinese fauna (LMC) are slightly higher in comparison to the Greek faunas, possibly suggesting slightly more open or arid conditions. Th is small difference could be due to sam-

6/27/12 12:10 AM


pling, and a broader study of fossils from China may change this result. The prevailing signal from mesowear suggests paleoecological uniformity throughout a vast area (eight time zones, Spain to China) during the Late Miocene. Th is result, which is congruent with the concept of a single Pikermian Biome, is not dissimilar to what a basic comparison of faunas would suggest; as discussed in this chapter, many ungulate species unify the Pikermian faunas from Spain to China (see table 31.1). The smaller number of species for the Pikermi fauna is consistent with the suggestion that it is older (about 8 Ma instead of 7.2 Ma), because it might have had fewer taxa in the newly developing biome. By “developing biome,” we mean that the Pikermi fauna samples species from the onset of the Pikermian Biome. Richer faunas derive from slightly younger faunas such as Samos, the core of Maragheh and the main Shanxi and Gansu faunas. We hypothesize that from 8 Ma to 7.2 Ma, there was a major adaptive radiation of ungulate species. Pikermi, however, is fundamentally similar in mesowear scores to Samos and China. The differences between the overall mesowear scores between the three fossil samples and the modern African savanna are minor, suggesting overall similar levels of abrasion (see figures 31.2 and 31.3). Th is similarity in mesowear is limited and needs to be evaluated by comparisons to other modern faunas such as those from the Indian and European forests. Currently, such data are not available. The modern African forest is notably different from the Pikermian Biome. Individual Species

Preliminary comparisons solely of the average mesowear scores can be conveniently subdivided into three: the bluntest apices with score averages 3 and 4, an intermediate range with averages 1 and 2, and the sharpest apices scoring near 0. At Samos and at the LMC, the bluntest dentitions (mesowear score averages of about 3 and 4) are from the Hippotherium species, the bovid Hezhengia, and the giraffids Honanotherium and Samotherium. The modern African savanna with comparable mesowear score averages includes the alcelaphine bovids (Alcelaphus and Connochaetes), the grazing rhinocerotid Ceratotherium, and the zebra Equus burchelli. Clearly, the two clusters of taxa are derived from different kinds of species; similar mesowear is found in such different species. The hippotheres differ from zebras but are both Equidae. The Alcelaphini, as a bovid tribe, are grazing hypsodont species, but there is no such a grazing tribe during the Miocene. Sys-

49502_1P_03_Wang15012_text14-31.indd 689

689

tematic mesowear comparisons need to be developed further. The intermediate range in the late Miocene includes several hippotheres, giraffids, and rhinocerotids. In addition, there are numerous Gazella species and many other bovids. The giraffids are numerous. The rhinocerotid Ceratotherium pachygnathus (an archaic form of Ceratotherium) and Dicerorhinus fall in this range. In the modern African savanna, this range is fi lled primarily by Gazella, the gazelle-like Aepyceros, and various bovids (but of different genera than the Miocene such as Hippotragus species and Syncerus). Syncerus is in the tribe Bovini, which, like the Alcelaphini, includes many more species than in the Miocene past. During the Late Miocene, Samokeros is a Bovini precursor, and in China a more Bos-like species was present (not yet described). In the two clusters, taxa are again derived primarily from different systematic groups. Modern Gazella has the same mesowear with the extinct species of Gazella. In the middle range, hippotheres are particularly abundant. In contrast, the modern zebra falls in the group with the bluntest apices. Equids from the intermediate range of mesowear scores are absent in the modern African savanna. In the low range of mesowear scores (0–1) are several Gazella species and other bovids, hippotheres, and giraffids. The giraffids are of the more archaic type (Palaeotragus rouenii and Helladotherium). Early Caprini (Pachytragus and Sporadotragus) have mesowear similar to Antilopini. The Gazella species are present, but they also have a wide range of mesowear scores. In the modern African savanna, this mesowear range (0–1) is occupied by Tragelaphus species, the giraffe, Diceros, and the Antilopini Antidorcas. The findings show that similar or different mesowear— and presumably diets—can occur between species that are systematically close. For example, the giraffid Samotherium strongly differs in mesowear from the recent giraffe. Similarly, a mesodont Samotherium can have mesowear similar to a Hippotherium, which is significantly more hypsodont. The mesowear of Miocene Ceratotherium is different from the present day C. simum.

Overall Picture and Comparisons

In contrast to the similar mesowear score distributions, the hypsodonty of the modern African savanna is radically greater. In this case, the difference is clearly influenced phylogenetically. The notable hypsodonty difference is attributed to the presence of several extant species

—-1 —0 —+1

6/27/12 12:10 AM

690

-1— 0— +1—


of the tribes Alcelaphini, Reduncini, Hippotragini, as well as Syncerus (Bovini) and the rhinoceros Ceratotherium simum. During the Miocene, none of these highly hypsodont taxa were present. In contrast, the pool of available species in the Miocene of Eurasia was mostly mesodont Bovidae (archaic Antilopini, Caprini, and Boselaphini). Thus, the recent African savanna species are more hypsodont, but it is interesting that their mesowear is very similar to that of the Miocene localities. Despite having relatively fewer hypsodont species, Pikermian Biome paleodiets (in terms of mesowear) appear to have been similar to those of the modern African savanna, that is to say, they exhibit a range of mesowear consistent with a similar mixture of browsers, mixed feeders, and grazers. Hypsodonty is a morphological characteristic that has evolved within many lineages of herbivorous mammals, including ungulates, rodents, and other extinct groups. Yet we know of no instance where hypsodonty has been secondarily lost. It has been suggested (Feranec 2003; Mihlbachler and Solounias 2006; MacFadden 2008) that while hypsodonty might be an adaptation to high-abrasion diets such as grazing, it does not constrain a species to an abrasive diet. Rather, it expands niche breath by increasing the amount of abrasion a species can tolerate. Our data support this hypothesis to some degree. For example, Hippotherium spans from 0 to 3 in mesowear scores, although the hypsodonty of these taxa is about the same. Similarly, Samotherium spans from 2 to 4 in mesowear values, Pachytragus from 0 to 1, and Gazella from 0 to 2. Similarly, the mesowear scores for these fossil localities and the recent African savanna overlap considerably. Despite a weak correlation, hypsodonty is a poor predictor of mesowear using the present data. Given the lack of phylogenetic evidence for crownheight character reversal, the prevalence of hypsodonty is expected to increase through time, as long as episodes of selection for hypsodonty repeatedly occur. However, the increasing frequency of hypsodonty does not require a prevailing overall dietary or associated climatic trend such as opening habitat. Even if climate from time to time reverses to cooler wetter phases, hypsodonty levels will not record this. Th is suggests that relative hypsodonty levels may be informative of paleoecology within a narrow window of time, among ecologically linked regions (such as Samos and China in the Miocene), that can potentially draw from a similar taxonomic pool. For instance, the Chinese Miocene fauna (LMC) is slightly more hypsodont overall in comparison to that of Samos. Th is small difference is proportional to the small difference in average mesowear scores. The LMC is both

49502_1P_03_Wang15012_text14-31.indd 690

slightly more hypsodont with slightly higher (blunter) mesowear scores. These differences are congruent and proportional. In contrast, the modern African savanna fauna and the Miocene faunas have similar mesowear levels, but the modern African savanna is disproportionately hypsodont in comparison. We suggest the high degree of overall hypsodonty is less a function of recent climate and more a consequence of 8 myr of cumulative evolution for hypsodonty in numerous ungulate lineages. Comparisons of hypsodonty levels of taxonomic groups or faunas from widely different time periods are complicated by the additional phylogenetic inertia characterizing the younger taxa and faunas. In conclusion, we fi nd, as in previous studies (Fortelius and Solounias 2000; Mihlbachler and Solounias 2006; Rivals, Mihlbachler, and Solounias 2007), that mesowear is a useful tool for comparisons of groups of taxa. The ecological cohesion of the Pikermian Biome is supported by mesowear. Combining the results of this mesowear analysis with an earlier microwear analysis suggests a rich mosaic with woodland but including a limited area of open savanna. We also fi nd the hypsodonty of this biome to be distributed in harmony with the mesowear and with evolutionary levels for hypsodonty.

ACKNOWLEDGEMENTS

We thank the committee of the Neogene Terrestrial Mammalian Biostratigraphy and Chronology Symposium in China. We also thank the curators of natural history museums in Paris, London, Vienna, Lausanne, Basel, Bern, New York, Frankfurt, Munich, Münster, and Stuttgart. Th is study was supported by funds from three previous National Science Foundation grants (one specifically for Samos, NSF dissertation improvement grant EAR 76- 00515); the others for microwear of ruminants and equids, BSR 8605172 1985–86 and IBN 9628263 1995–96, respectively. Tooth mesowear was recorded in 2009 on specimens covered by these grants. We thank Judith Harris for inspiration and numerous discussions and collaborations. We also thank Louis Abatis, Peter Andrews, Athanassios Athanassiou, Raymond Bernor, Shao-kun Chen, Eric Delson, Vera Eisenmann, Burkart Engesser, Mikael Fortelius, Jens Franzen, Henry Galiano, Alan Gentry, Ursula Göhlich, Judith Harris, Elmar Heizmann, Kurt Heissig, Thomas Keiser, Ursula Menkveld, Doris Nagel, Malcolm McKenna, Clemens Oekentorp, David Pilbeam, Karl Rausen, Peter Robinson, Socrates

6/27/12 12:10 AM


Roussiakis, George Schaller, Shan-qin Chen, Shao-kun Chen, Fritz Steininger, Pascal Tassy, Richard Tedford, George Theodorou, Herbert Thomas, John Van Couvering, Alan Walker, Xiaoming Wang, Marc Weidmann, and Eileen Westwig.

REFERENCES Abel, O. 1927. Lebensbilder aus der Tierwelt der Vorzeit, 2nd ed. Stuttgart: Jena Verlag Gustav Fisher. Axelrod, I. D. 1975. Evolution and biogeography of the Madrean Tethyan sclerophyll vegetation. Annals of the Missouri Botanical Garden 62:280–334. Beerling, D. J. and C. P. Osborne. 2005. The origin of the savanna biome. Global Change in Biology 12:2023–2031. Bernor, R. L., G. D. Koufos, M. O.Woodburne, and M. Fortelius. 1996. The evolutionary history and biochronology of Eu ropean and Southwest Asian late Miocene and Pliocene hipparionine horses. In The Evolution of the Western Eurasian Neogene Mammal Faunas, ed. R.L. Bernor, V. Fahlbusch, and H.-W. Mittmann, pp. 307–347. New York: Columbia University Press. Bernor, R. L., N. Solounias, C. C. Swisher III, and J. A. Van Couvering. 1996. The correlation of the classical “Pikermian” mammal faunas—Maragheh, Samos, and Pikermi—with the Eu ropean MN unit system. In Th e Evolution of the Western Eurasian Neogene Mammal Faunas, ed. R.L. Bernor, V. Fahlbusch, and H.-W. Mittmann, pp. 137–154. New York: Columbia University Press. Bernor, R. L., R. S. Scott, M. Fortelius, J. Kappelman, and S. Sen. 2003. Equidae (Perissodactyla). In The Geology and Paleontology of the Miocene Sinap Formation, Turkey, ed. M. Fortelius, J. Kappelman, S. Sen, and R.L. Bernor, pp. 220–281. New York: Columbia University Press. Bond, W. J. 2008. What limits trees in C 4 grasslands and savannas? Annual Review of Ecology and Systematics 39:641– 659. Bruch, A. A., T. Utescher, V. Mosburugger, I. Gabrielyan, and D.A. Ivanov. 2006. Late Miocene climate in the circum-Alpine realm— a quantitative analysis of terrestrial paleofloras. Palaeogeography, Palaeoclimatology, Palaeoecology 238:270–280. Bruch, A. A., D. Uhl, and V. Mosbrugger. 2007. Miocene climate in Europe—patterns and evolution: a fi rst synthesis of NECLIME. Palaeogeography, Palaeoclimatology, Palaeoecology 253:1–7. Cerling, T. E., J. M. Harris, B. J. MacFadden, M. G. Leakey, J. Quade, V. Eisenmann, J. R. Ehleringer. 1997. Global change through the Miocene/Pliocene boundary. Nature 389:153–158. Coppens, Y. 1994. The east side story: the origin of humankind. Scientific American 270:88–95. de Bonis, L., G. Bouvrain, D. Geraads, and G. Koufos. 1992. Diversity and paleoecology of Greek late Miocene mammalian faunas. Palaeogeography, Palaeoclimatology, Palaeoecology 91:99–121. Deng, T. 2006. Paleoecological comparison between late Miocene localities of China and Greece based on Hipparion faunas. Geodiversitas 28:499–516. Dorofeyev, P. I. 1966. Flora of the Hipparion epoch. International Geological Reviews 8:1109–1117.

49502_1P_03_Wang15012_text14-31.indd 691

691

Edwards, E. J., C. P. Osborne, C. A. E. Strömberg, and S. T. Smith, and C 4 grasses consortium. 2010. The origins of C 4 grasslands: Integrating Evolutionary and ecosystem science. Science 328: 587–591. Eronen, J. T., M. M. Ataabadi, A. Micheels, A. Karme, R. L. Bernor, and M. Fortelius. 2009. Distribution history and climatic controls of the Late Miocene Pikermian chronofauna. Proceedings of the National Academy of Sciences 106:11867–11871. Eronen, J. T., K. Puolamäki, L. Liu, K. Lintulaakso, J. Damuth, C. Janis, and M. Fortelius. 2010a. Precipitation and large herbivorous mammals II: Application to fossil data. Evolutionary Ecology Research 12:235–248. Eronen, J. T., K. Puolamäki, L. Liu, K. Lintulaakso, J. Damuth, C. Janis, and M. Fortelius. 2010b. Precipitation and large herbivorous mammals I: Estimates from present-day communities. Evolutionary Ecology Research 12:217–233. Feranec, R. S. 2003. Stable isotopes, hypsodonty, and the paleodiet of Hemiauchenia (Mammalia: Camelidae): A morphological specialization creating ecological generalization. Paleobiology 29:230–242. Fortelius, M. (coordinator). 2009. Neogene of the Old World Database of Fossil Mammals (NOW). University of Helsinki. http:// www.helsinki.fi/science/now/. Fortelius, M., J. Eronen, L.P. Liu, D. Pushkina, A. Tesakov, I. Vislobokova, and Z.Q. Zhang. 2003. Continental-scale hypsodonty patterns, climatic paleobiogeography and dispersal of Eurasian Neogene large mammal herbivores. Deinsea 10:1–11. Fortelius, M., J. Eronen, L. Liu, D. Pushkina, A. Tesakov, I. Vislobokova, and Z. Zhang. 2006. Late Miocene and Pliocene large land mammals and climatic changes in Eurasia. Palaeogeography, Palaeoclimatology, Palaeoecology 238:219–227. Fortelius, M. and N. Solounias. 2000. Functional characterization of ungulate molars using the abrasion-attrition wear gradient: A new method for reconstructing paleodiets. American Museum Novitates 3301:1–36. Gaudry, A. 1862–1867. Animaux fossiles et géologie de l’Attique. Paris, Martinet. Jacobs, B. F., J. D. Kingston, and L. L. Jacobs. 1999. The origin of grass-dominated ecosystems. Annals of the Missouri Botanical Garden 86:590– 643. Janis, M. C. 1988. An estimation of tooth volume and hypsodonty indices in ungulate mammals and the correlation of these factors with dietary preferences. In Teeth Revisited: Proceedings of the VII international symposium on dental morphology, ed. D. E. Russel, J. P. Santorio, and D. Signogneu-Russel, pp. 367–387. Paris: Muséum national de Histoire Naturelle. Janis, C. M. 2008. An evolutionary history of browsing and grazing ungulates. In The Ecology of Browsing and Grazing, ed. I. J. Gordon and H.H.T. Prins, pp. 21– 45. New York: Springer. Janis, C. M., J. Damuth, and J. M. Theodor. 2000. Miocene ungulates and terrestrial primary productivity: Where have all the browsers gone? Proceedings of the National Academy of Sciences 97:7899–7904. Koufos, G. D. and D. Nagel, eds. 2009. The Late Miocene Mammal Faunas of Samos. Vienna: Beiträge zur Paläontologie, Band 31. Koufos, G. D., D. Kostopoulos, and G. Merceron 2009. Palaeoecology–Palaeogeography. In The Late Miocene Mammal

—-1 —0 —+1

6/27/12 12:10 AM

692


Faunas of Samos, ed. G. D. Koufos and D. Nagel pp. 409– 427. Beiträge zur Paläontologie 31. Kurtén, B. 1952. The Chinese Hipparion fauna. Commentationes Biologicae Societatis Scientiarum Fennicae 13:1– 82. Kurtén, B. 1971. The Age of Mammals. New York: Columbia University Press. Mai, D. H. 1995. Tertiäre Vegetationsgeschichte Europas, Methoden und Ergebnisse. Jena: Gustav Fischer Verlag. MacFadden, B. J. 2008. Geographic variation in diets of ancient populations of 5-million-year-old (early Pliocene) horses from southern North America. Palaeogeography, Palaeoclimatology, Palaeoecology 266:83–94 Matthew, D.W. 1926. The evolution of the horse. Quarterly Review of Biology 1:139–185. Mihlbachler, M. C., F. Rivals, N. Solounias, and G. Semprebon. 2011. Dietary change and evolution of horses in North America. Science 331:1178–1181. Mihlbachler, M. C. and N. Solounias. 2006. Coevolution of tooth crown height and diet in oreodonts (Merycoidontidae, Artiodactyla) examined with phylogenetically independent contrasts. Journal of Mammalian Evolution 13:11–36. Orgeta, M. 1979. Erste Ergebnisse einer palynologischen Unterschuchung der Lignite von Pikermi/Attica. Annales Géologiques des pays Helléniques 2:909–921. Osborn, H. F. 1910. The Age of Mammals. New York: Macmillan Co. Quade, J., T. E. Cerling, P. Andrews, and B. Alpagut. 1995. Paleodietary reconstruction of Miocene faunas from Paşalar, Turkey, using stable carbon and oxygen isotopes of fossil tooth enamel. Journal of Human Evolution 28:373–384. Quade, J., N. Solounias, and T. E. Cerling. 1994. Stable isotopic evidence from paleosol carbonates and fossil teeth in Greece for forest or woodlands over the past 11 Ma. Palaeogeography, Palaeoclimatology, Palaeoecology 108:41–53. Retalack, G.J. 2001. Cenozoic expansion of grasslands and climatic cooling. Journal of Geology 109:407– 426. Rivals F., M. C. Mihlbachler, and N. Solounias. 2007. Effect of ontogenetic-age distribution in fossil samples on the interpreta-

tion of ungulate paleodiets using the mesowear method. Journal of Vertebrate Paleontology 27:763–767. Scott, R. S., M. Fortelius, K. Huttunen, and M. Armour- Chelu. 2003. Abundance of “Hipparion.” In Geology and Paleontology of the Miocene Sinap Formation, Turkey, ed. M. Fortelius, J. Kappelman, S. Sen, and R.L. Bernor, pp. 380–397. New York: Columbia University Press, Solounias, N. 1981. The Turolian fauna from the island of Samos, Greece. With special emphasis on the hyaenids and the bovids. Contributions to Vertebrate Evolution 6:1–232. Solounias, N., M. Fortelius, and P. Freeman. 1994. Molar wear rates in ruminants: A new approach. Annales Zoologici Fennici 31:219–227. Solounias, N., W. S. McGraw, L.-A. C. Hayek, and L. Werdelin. 2000. The paleodiet of the Giraffidae. In Antelopes, Deer, and Relatives, ed. E. S. Vrba and G. B. Schaller, pp. 84–95. New Haven: Yale University Press, Solounias, N., S. M. C. Moelleken, and J. M. Plavcan. 1995. Predicting the diet of extinct bovids using masseteric morphology. Journal of Vertebrate Paleontology 15:795– 805. Solounias, N., M. Plavcan, J. Quade, and L. Witmer. 1999. The Pikermian Biome and the savanna myth. In Evolution of the Neogene Terrestrial Ecosystems in Europe, ed. J. Agustí, P. Andrews, and L. Rook, pp. 427– 444. Cambridge, Cambridge University Press. Solounias, N., F. Rivals, and G. Semprebon. 2010. Dietary interpretation and paleoecology of herbivores from Pikermi and Samos (late Miocene of Greece). Paleobiology 36:113–136. Strömberg, C. A. E., L. Werdelin, E. M. Friis, and G. Saraç. 2007. The spread of grass-dominated habitats in Turkey and surrounding areas during the Cenozoic: Phytolith evidence. Palaeogeography, Palaeoclimatology, Palaeoecology 250:18– 49. Thenius, E. 1959. Tertiär. Stuttgart: Ferdinand Enke Verlag. Weidmann, M., N. Solounias, R. E. Drake, and G. H. Curtis. 1984. Neogene stratigraphy of the Mytilinii Basin, Samos Island, Greece. Géobios 17:477– 490. Zachos, J.C., G. R. Dickens, and R.E. Zeebe. 2008. An early Cenozoic perspective on green house warming and carbon-cycle dynamics. Nature 451:279–283.

-1— 0— +1—

49502_1P_03_Wang15012_text14-31.indd 692

6/27/12 12:10 AM