Folivory or fruit/seed predation for Mesopithecus, an earliest colobine from the late Miocene of Eurasia? Gildas Merceron a,*, Jessica Scott b, Robert S. Scott c, Denis Geraads d, Nikolai Spassov e, Peter S. Ungar f a UMR CNRS 5125 PEPS Paleoenvironnements �� & Paleobiosphere �� ` - University Lyon-1, Campus la Doua, GEODE - 2, rue Raphae ��l Dubois, 69622 Villeurbanne, France b Environmental Dynamics Doctoral Program, University of Arkansas, 113 Ozark Hall, Fayetteville, AR 72701, USA c Department of Anthropology and Center for Human Evolutionary Studies, Rutgers University, 131 George Street, RAB 306, New Brunswick, NJ 08901, USA d UPR 2147 CNRS, 75014 Paris, France e National Museum of Natural History, Bulgarian Academy of Sciences, 1, Blvd. Tzar Osvoboditel, 1000 Sofia, Bulgaria f Department of Anthropology, University of Arkansas, Fayetteville, AR, USA a r t i c l e i n f o Article history: Received 25 February 2009 Accepted 19 June 2009 Keywords: Diet Microwear Teeth Neogene Cercopithecidae a b s t r a c t Here we compare dental microwear textures from specimens of the fossil genus Mesopithecus (Cerco- pithecidae, Colobinae) from the late Miocene of Eastern Europe with dental microwear textures from four extant primate species with known dietary differences. Results indicate that the dental microwear textures of Mesopithecus differ from those of extant leaf eaters Alouatta palliata and Trachypithecus cristatus and instead resemble more closely those of the occasional hard-object feeders Cebus apella and Lophocebus albigena. Microwear texture data presented here in combination with results from previous analyses suggest that Mesopithecus was a widespread, opportunistic feeder that often consumed hard seeds. These data are consistent with the hypothesis that early colobines may have preferred hard seeds to leaves. �� 2009 Elsevier Ltd. All rights reserved. Introduction Previous authors (Chivers, 1994 Kay and Davies, 1994) have suggested that folivory in colobine monkeys may have been preceded by ancestral diets rich in seeds. Given that the two earliest known Old World monkeys, Prohylobates and Victoriapithecus both have cercopithecine-like molars, it has been hypothesized that the earliest cercopithecoid feeding habits would not have included specialized adaptations for leaf consumption (Benefit, 2000). While the carbohydrates in sugary fruit flesh are often simple and easy to digest, seeds typically contain polysaccharides similar to those in leaves, which can be a challenge to break down. Thus, seed consumption could have resulted in a gut optimal for the digestion of leaves (Chivers, 1994 Kay and Davies, 1994). This leads to the question ������was the late Miocene Mesopithecus a folivore like many of its Plio-Pleistocene and modern colobine relatives (Rowe, 1996), or did it more often consume fruit seeds like its cercopithecoid forbears?������ Here, we address this question using dental microwear texture analysis. At least four genera of cercopithecoids are known from the Neogene of Europe: Macaca, Paradolichopithecus, Dolichopithecus, and Mesopithecus (Andrews et al., 1996). Of these, Mesopithecus is the earliest. The genus has been reported to date to the early Vallesian (MN-9 11.2���9.5 Ma) at Wissberg in Germany, although the context of this specimen has been called into question (Andrews et al., 1996). Regardless, Mesopithecus at the ������Ravin des Zouaves 5������ locality in Northern Greece (Bonis et al., 1990) has been dated securely to the very base of the Early Turolian mammalian age interval MN-11 (late Miocene, 8.7���7.4 Ma Sen et al., 2000 Koufos, 2006). Cranial and dental features of Mesopithecus have led researchers to argue that the genus may have evolved around the time of the split between the tribes Colobini and Presbytini (Szalay and Delson, 1979 Strasser and Delson, 1987 Bonis et al., 1990), dated to about 10 Ma through a combination of DNA and fossil evidence (Goodman et al., 1998). While the ten extant genera of colobines are found only in Asia and Africa (Rowe,1996), Mesopithecus has been recovered from Western Europe to Central and South Asia during the late Miocene (Szalay and Delson, 1979 Heintz et al., 1981 Rook, 1999) and appears to have survived in Europe until the Early Villafranchian (Pliocene Delson et al., 2005). The Mesopithecus distribution overlapped a wide range of environmental and climatic conditions with * Corresponding author. E-mail address: gildas.merceron@univ-lyon1.fr (G. Merceron). Contents lists available at ScienceDirect Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol 0047-2484/$ ��� see front matter �� 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jhevol.2009.06.009 Journal of Human Evolution 57 (2009) 732���738

wooded landscapes in the West transitioning to more open land- scapes in the East (Bonis et al., 1992 Fortelius et al., 1996 Zazzo et al., 2002 Eronen and Rook, 2004 Merceron et al., 2004 Stromberg �� et al., 2007). The closest living relatives of Mesopithecus, the extant colobines, are highly folivorous and arboreal forest species with the exception of Semnopithecus, which is a terrestrial folivore (Dela, 2007 Sayers and Norconk, 2008). Prior reconstructions of Mesopithecus loco- motion have been contradictory. Metric data for the Mesopithecus calcaneus, for example, have been interpreted by various researchers as indicative of either terrestrial or arboreal locomotion (Youlatos, 2003 Escarguel, 2005). Even given this caveat, other morphological features, such as scapular and limb robusticity, all support a semi-terrestrial mode of locomotion (Szalay and Delson, 1979). Therefore, Mesopithecus would have likely foraged both in trees and on the ground, facilitating access to a wide range of dietary options. Most extant colobines are folivorous, with preferences for leaves and leaf buds. Molar microwear and shearing crest analyses suggest Plio-Pleistocene African colobines shared these preferences (El Zaatari et al., 2005 Teaford et al., 2008). However, recent field studies have suggested more complex feeding habits for some present-day colobines such as Rhinopithecus roxellana, with fruits and seeds making up significant portions of the diets (Steenbeek and van Schaik, 2001 Dela, 2007 Guo et al., 2007). Moreover, extant colobines display anatomical specializations of the gastro- intestinal tract, such as a forestomach divided into a proximal presaccus and a saccus that precedes the tubus gastricus and then the pars pylorica (Chivers, 1994 Kay and Davies, 1994 Lambert, 1998). The forestomach plays the role of fermentation chamber and the pH is higher than in the distal stomach chambers. This lower acidity is required to maintain optimal conditions for fermentation. During fermentation, cellulolytic microorganisms reduce poly- saccharides (e.g., cellulose from the leaf cell walls) to simpler sugars, which are more easily hydrolyzed during the transit through the small intestine prior to absorption. Such anatomical features and physiological processes are similar to those seen in ruminants (Lambert, 1998 Zhang, 2003). Although the volume of the forestomach chamber as a whole is correlated with foliage consumption, a closer look suggests that the presaccus size relates to leaf consumption, whereas saccus size may be better correlated with seed consumption (Chivers,1994 Kay and Davies, 1994). Thus, an increase in seed consumption may be an evolutionary link between adaptations for frugivory and folivory (Chivers, 1994 Kay and Davies, 1994). Our present study aims to test this hypothesis. Implications for understanding the evolution of primates are important since if the earliest colobines appear to be more engaged in seed predation than in folivory, it may be that such specialized adaptations for leaf consumption evolved inde- pendently in different tribes of colobines. Because soft tissues are rarely preserved in the fossil record, teeth provide the best means for reconstructing the dietary habits of early fossil colobines. Here we apply dental microwear texture analysis to the study of Mesopithecus. Dental microwear, the study of the microscopic use-wear scars in enamel, reveals direct infor- mation about what an individual ate over a period of time in the past (Rensberger,1978 Walker et al.,1978 Teaford and Oyen,1989 Ungar and Teaford, 1996 Teaford et al., 2008). Texture analysis, a recent method of dental microwear analysis (Ungar et al., 2003, 2007, 2008 Scott et al., 2005, 2006, 2008 Merceron et al., 2005, 2006), offers repeatable and quantitative characterizations of 3D surfaces. This new method is a promising tool for investigating subtle variations in the diets of extinct mammals (for comparisons with previous methods, see Estebaranz et al. [2008] and Ungar et al. [2008]). Materials and methods Materials A total of 17 adult individuals of Mesopithecus were included in this study: 12 specimens of Mesopithecus from Greece (Pikermi) and 5 from Bulgaria (3 from Kalimantsi, 2 from Hadjidimovo-1 Table 1 Appendix 1). These localities have been dated to the late Early Turolian (Hadjidimovo-1) and Middle Turolian (Pikermi and upper level of Kalimantsi Spassov, 2002 Kostopoulos et al., 2003 Spassov et al., 2006). The Pikermi material is housed in the Labo- ratoire de Paleobiodiversite �� �� at the Museum �� National d���Histoire Naturelle de Paris (France), and the Bulgarian fossils are stored in the Assenovgrad Museum, a paleontological division of the National Museum of Natural History, Sofia (Bulgaria). All of the specimens from Pikermi and two from Kalimantsi belong to Mesopithecus pentelicus (Zapfe, 1991 Koufos et al., 2003). The two specimens from Hadjidimovo-1 (HD 339 and HD 340) and one individual from Kalimantsi (K-348) are referred to Meso- pithecus aff. M. delsoni (Koufos et al., 2003). In this paper we distinguish these specimens at the genus level only, given ques- tions about the distinctiveness of these two species (Delson, 1994 Andrews et al., 1996). Data for these fossil specimens are compared with an extant baseline series, including two New World monkey species (Cebus apella and Alouatta palliata) and two Old World monkey species (Trachypithecus cristatus and Lophocebus albigena Scott et al., 2006). These four species are appropriate for compari- sons since they plot at opposite ends of the microwear spectrum, with the two species that feed on soft, tough foliage (A. palliata and T. crisatus) separated from the other two (C. apella and L. albigena) that at least occasionally consume hard and brittle items. Cebus apella diets are highly variable by location and season. However, capuchins often forage on fruit flesh, seeds, and palm fronds. Leaves are rarely more than 10% of the diet of most capuchin populations (Galetti and Pedroni, 1994 Rowe, 1996). Lophocebus albigena also prefers fruit flesh, bark, and seeds, with seasonal variation in availability of each of these (Chalmers, 1968 Rowe, 1996). In contrast, A. palliata and T. cristatus tend to be more folivorous and often consume other tough, fibrous foods (Estrada, 1984 Bro- toisworo and Dirgayusa, 1991 Rowe, 1996). Table 1 Descriptive statistics for complexity (Asfc), anisotropy (epLsar), scale of maximum complexity (Smc), texture fill volume (Tfv), and heterogeneity (HAsfc)a n Asfc EpLsar Smc Tfv Hasfc Trachypithecus cristata 12 m 0.734 4.788 10 3 1.223 9532.21 0.612 s.d. 0.660 2.602 10 3 2.529 5687.17 0.286 s.e.m. 0.191 0.751 10 3 0.730 1641.74 0.082 Alouatta palliata 11 m 0.360 5.816 10 3 53.427 2610.83 0.692 s.d. 0.183 2.140 10 3 175.287 3225.63 0.383 s.e.m. 0.055 0.645 10 3 52.851 972.56 0.115 Lophocebus albigena 15 m 1.769 3.776 10 3 28.417 11388.39 0.535 s.d. 1.740 1.962 10 3 52.586 3389.60 0.276 s.e.m. 0.449 0.507 10 3 13.578 875.19 0.071 Cebus apella 13 m 5.466 3.667 10 3 31.902 9682.57 0.786 s.d. 6.304 1.855 10 3 91.169 4923.61 0.334 s.e.m. 1.749 0.515 10 3 25.286 1365.56 0.093 Mesopithecus (All populations) 17 m 1.866 3.082 10 3 2.781 8223.95 0.517 s.d. 1.291 1.326 10 3 10.209 3785.69 0.188 s.e.m. 0.313 0.322 10 3 2.476 918.16 0.045 Mesopithecus from Bulgaria 5 m 2.751 2.823 10 3 0.227 10789.41 0.516 s.d. 1.659 1.389 10 3 0.118 2124.70 0.239 s.e.m. 0.742 0.621 10 3 0.053 950.19 0.106 Mesopithecus from Greece 12 m 1.497 3.189 10 3 3.845 7155.01 0.518 s.d. 0.958 1.347 10 3 12.141 3868.59 0.174 s.e.m. 0.276 0.389 10 3 3.505 1116.76 0.050 a m �� mean s.d. �� standard deviation s.e.m. �� standard error of the mean. G. Merceron et al. / Journal of Human Evolution 57 (2009) 732���738 733