The primate fossil record in the Iberian Peninsula

During the last decade, new discoveries in several Iberian basins, together with the description of previously unpublished finds, have significantly increased the recorded paleodiversity of fossil Primates (Mammalia: Euarchonta) in the Iberian Peninsula. Here we provide an updated compendium of the primate fossil record in Iberia during the Cenozoic and further summarize the changes in primate paleodiversity through time, which are then analyzed in the light of changing climatic conditions. Thanks to favorable climatic conditions, the highest diversity of Iberian primates was reached during the Eocene, thus reflecting the radiation of both adapoids and omomyoids; only a single plesiadapiform genus is in contrast recorded in the Iberian Peninsula. Near the Eocene-Oligocene boundary, paleoclimatic changes led to a primate diversity crisis and other faunal changes, although two Iberian omomyoids survived the Grande Coupure. From the Middle Miocene onwards, catarrhine primates are recorded in the Iberian Peninsula. During the Middle and Late Miocene, they are represented by pliopithecoids and hominoids, restricted to NE Iberia. The Miocene hominoids from Iberia are of utmost significance for understanding the Eurasian hominoid radiation and its role in the origins of the great-ape-and-human clade. Following the local extinction of these taxa during the early Late Miocene, due to progressively increased seasonality and concomitant changes in plant communities, cercopithecoids are also recorded in the Iberian Peninsula from the latest Miocene through the Plio-Pleistocene, although they finally became locally extinct, whereas hominoids are again represented by fossil humans during the Pleistocene.


Primates
Primates (Mammalia: Euarchonta) are a eutherian order of moderate diversity, with about 300 extant species.They were traditionally divided into two major groups: prosimians and simians (or anthropoids) (e.g., Fleagle, 1999).Most authors, however, currently favor a distinction between strepsirrhines and haplorrhines.Unfortunately, many of the features that enable such division are related to soft anatomy, so that the classification of fossil 'prosimians' as either strepsirrhines or haplorrhines, albeit favored here, is not always straightforward.Crown primates (euprimates, or primates of modern aspect) constitute a monophyletic group including extant strepsirrhines and haplorrhines, together with the extinct taxa more closely related to either of them (see Fleagle, 1999, andHartwig, 2002, for the most recent complete accounts on fossil primates).Morphologically, euprimates are characterized by the possession of a petrosal bulla (the auditory bulla is formed by an extension of the petrosal instead of by a separate entotympanic bone), which is a unique synapomorphy among mammals (Rasmussen, 2002).Other derived diagnostic traits of euprimates, uncommon in other mammals, are the following (e.g., Martin, 1990;Rasmussen, 2002): postorbital bar; orbital convergence, yielding stereoscopic binocular vision; grasping hands and feet, usually with opposable hallux and pollex; flat nails instead of claws on most digits; brains larger than expected on the basis of body size (high encephalization); and a slow life-history profile.The primate total group includes crown primates plus the primate stem lineage, the latter being those taxa preceding the haplorrhine-strepsirrhine divergence, but being more closely related to them than to other euarchontans, i.e. scandentians (tree shrews) and dermopterans (flying lemurs).It is generally thought that primates diverged from other mammals sometime during the Late Cretaceous.Undisputed fossil euprimates are not found until the earliest Eocene, ca.56 Ma (Rasmussen, 2002), and although molecular estimates suggested in the past an ancient divergence data for crown subclades, recent studies suggest that haplorrhines and strepsirrhines diverged close to the Mesozoic/Cenozoic boundary (Steiper and Seiffert, 2012).The origin of euprimates has been related to plesiadapiforms, which show general dental similarities with primates but at the same time lack several euprimate derived features, such as the postorbital bone and the petrosal bulla.The phylogenetic relationships of plesiadapiforms with euprimates remain unresolved, but it seems currently likely that most of them constitute a paraphyletic assemblage of stem primates.

Aims of this study
Most primates are linked to tropical or subtropical forested environments, and currently, modern humans are the only primates naturally inhabiting the Iberian Peninsula.This, however, has not always been the case: during the Paleogene and Neogene, several major primate groups occupied the Iberian Peninsula.In Table 1, we have summarized the systematic scheme employed for primates in this paper down to the superfamily rank, indicating with bold bype those taxa recorded in the Iberian Peninsula.For many years, the study of fossil Iberian primates was led by Miquel Crusafont-Pairó, founder of the Institut Català de Paleontologia Miquel Crusafont (ICP), which currently bears his name.Beginning in the early 1940s, Crusafont and co-workers such as Josep F. de Villalta and Juana M. Golpe-Posse reported new material and described new taxa of "prosimians" and anthropoids alike.It is not surprising, therefore, that Crusafont-Pairó and Golpe-Posse (1974) finally published a synthetic review on the fossil primates from Spain.Since then, some summary syntheses devoted to particular primate groups at the Iberian level have been published (e.g., Moyà Solà et al., 1990;Marigó et al., 2011c), but no detailed and comprehensive synthesis of all Iberian primates has been undertaken since 1974.In the meantime, many new fossil discoveries, together with the restudy of previously available material, have led to a significant increase in the knowledge of their paleodiversity, chronology and paleobiology.Here, we provide an updated synthesis of the fossil primates from the Iberian Peninsula (Spain and Portugal) during the Cenozoic (Paleogene, Neogene and Quaternary), enumerating all the recorded taxa at the species level (when possible), and also providing details on their systematics, chronologic and geographic distribution, paleobiology, and phylogenetic relationships.We further discuss the Iberian primate record in the framework of the evolutionary history of this group as a whole, with particular emphasis on its relationship to paleogeographic and climatic changes through time.form genus, and D. lusitanica is a very primitive species within this genus.It is much smaller than D. magna, somewhat larger than D. gallica, and of similar size or slightly larger than D. provincialis.The lower molars of D. lusitanica differ from those of the other species of the genus in the smaller paraconid and a shorter trigonid when compared to the talonid.The M 1 of D. lusitanica differ from those of D. provincialis and D. gallica in being larger and in having a more subquadrate outline among other features.Moreover, the nannopithex-fold present in the M 1 and M 2 of D. lusitanica is not present in those attributed to D. gallica, and the absence of a postparaconule crista in the upper molars of D. lusitanica makes them different from those of D. provincialis.Main references: Estravís (2000).
Remarks: This genus has been identified in several Spanish and French localities, ranging from the Neustrian (MP8+9) to the Geiseltalian (MP12).The scarce material from Casa Ramón (MP11 or MP12), which consists only of six isolated teeth, precludes erecting a new species, although it is smaller and displays some morphological differences compared to A. roselli.The material from La Coma (MP10), also very scarce, is still under study.Since these two localities are of different study of the available material has revealed that it displays several features distinguishing it from other anchomomyin genera (very buccolingually compressed P 4 , extremely reduced metacone in the M 3 , and the complete premetacristid closing the trigonid basin from the lingual side on the lower molars).
Remarks: The genus Anchomomys is well known from several European sites (Switzerland, France, Spain), ranging in age from the Geiseltalian to the Headonian (MP13-17a, Middle-Late Eocene, 44-37 Ma).The material from Caenes, curently under study, consists of two complete and one fragmentary right upper molars, and two left mandible fragments, one showing a complete P 4 and the other presenting an almost complete P 4 , a partial M 1 and a complete M 2 , and both with the rest of the alveoli from the lower C to the M 3 .In addition, a few isolated teeth from Sant Jaume de Frontanyà 1 can be assigned to the genus Anchomomys, although the scarcity of material precludes a determination to the species level.
Geographic distribution: Exclusively known from the type locality.
Remarks: The material from this locality represents the largest collection of Anchomomys material ever described, including all the permanent teeth except for the incisors.A. frontanyensis resembles A. pygmaeus from Egerkingen (Ef 372) although it has a more reduced hypocone and paraconule, its hypocone is less projected lingually, and its upper molars have a less developed anterocingulum extending further lingually.Despite sharing features with other anchomomyins, A. frontanyensis also shares certain characters with extinct asiadapines and extant lemurs and lorises.Moreover, the presence of A. frontanyensis in the Iberian Peninsula reveals a different lineage from that present at the same time in the rest of Europe.Phylogenetic analyses suggest that anchomomyins might be related to either asiadapines or crown strepsirrhines.
Main references: Moyà-Solà and Köhler (1993b) age, it is very plausible that two different species are represented.
Remarks: First attributed to the Necrolemuridae, it displays a very primitive upper dentition and a derived lower dentition without paraconid and with a very reduced trigonid (Crusafont-Pairó, 1967).
Remarks: The genus Cantius has a wide geographic distribution, including Europe (Spain and France) and North America.It has been identified in Neustrian (MP8+9) and Grauvian (MP10) sites.The material from the two Iberian localities is very scarce, so that an attribution to the species level is not possible until more material is recovered.
Remarks: The attribution of material from Montblanc and Coll de l'Illa to Anchomomyini indet.by Antunes et al. (1997) cannot be confirmed based on the material currently housed at the Institut Català de Paleontologia Miquel Crusafont.Several isolated teeth from Ulldemolins I are also tentatively assigned to the tribe Anchomomyini, until a detailed study enables a more precise determination.
Iberian localities: Roc de Santa (Barcelona).Remarks: Originally described within the genus Adapis by Filhol (1874), it was subsequently attributed to Leptadapis by Gervais (1876), which after more accurate studies was recently confirmed by Godinot and Couette (2008).The scarce material from Roc de Santa, a right maxilla with P 3 -M 3 and a left mandible with a broken P 3 and the P 4 -M 2 series, was assigned to Adapis cf.magnus by Crusafont-Pairó and Golpe-Posse (1974), given its somewhat smaller size compared to the type material.Nevertheless, subsequent studies have confirmed the ascription of the material from Roc de Santa to this species (Casanovas-Cladellas, 1975;Casanovas, 1998;Antunes et al., 1997).
Iberian localities: Santa Clara (Zamora).Remarks: The only available material is an M 2 that resembles that of A. cf.pygmaeus En-1 specimen from Egerkingen, first attributed to A. stehlini by Gingerich (1977), but later assigned to A. cf.pygmaeus by Godinot (1988).
Remarks: This taxon allowed the description, for the first time, of almost all the dental elements and the anterior dentition for a member of the Anchomomyini.The material attributed to Mazateronodon endemicus is the richest Eocene primate assemblage from the Western Iberian Bioprovince.Mazateronodon differs morphologically and biometrically from all the other anchomomyins, and it is characterized by its extremely buccolingually compressed P 3 and P 4 , with the P 3 larger than the P 4 , its highly imbricated premolars, with overlapping crowns and buccal and lingual cingulids.Some similarities with the material assigned to Anchomomys from Caenes (Duero Basin) and Anchomomys frontanyensis (Pyrenean Basin) strongly suggest an Iberian origin of this taxon.The unique features of Mazateronodon support the idea of endemism in the Eocene faunas of the Duero Basin, and the existente of a Western Iberian Bioprovince, isolated form the rest of Europe during the Eocene due to a marine transgression that connected the Cantabric and the Mediterranean seas.
Remarks: The material consists of two isolated molars of a very small omomyid.The detailed study of these remains, as well as further fieldwork for obtaining more material, may allow the description of a new taxon.
Remarks: The material from Zambrana consists of two well-preserved right mandibular fragments with relatively complete dental series, including P 3 -M 3 and P 4 -M 3 .The overall dental morphology and the presence of enamel wrinkling are similar to those observed in the genera Necrolemur and Microchoerus.The detailed study of this material, now under way, will lead to a more precise determination.This discovery, together with that of Pseudoloris cuestai from Mazaterón (Minwer-Barakat et al., 2012), constitutes the first record of microchoerines in the Western Iberian Bioprovince.
Remarks: This large-sized microchoerine from Sossís was first reported by Crusafont-Pairó (1965b), and subsequently Crusafont-Pairó (1967) briefly described a quite complete sample of this species, including isolated teeth, some dental series and several fragments of mandibles and maxillae.This author further figured a lower dental series including P 3 -M 3 , and an upper dental series with M 1 and M 2 .Crusafont-Pairó (1967) and Crusafont-Pairó and Golpe-Posse (1974) referred these remains to Necrolemur erinaceus, but later works (Antunes et al., 1997;Casanovas, 1998) employed the correct denomination, i.e.Microchoerus erinaceus.Subsequent field work led to the recovery of more abundant material, which is still unpublished.A preliminary observation of this material reveals some biometric and morphologic differences relative to the material from the type locality (Hordle Cliff Crocodile Bed, Hampshire, UK).However, until a detailed study of the population from Sossís is carried out, we prefer mantaining the attribution to M. erinaceus.

Adapis sp.
Iberian localities: Sant Jaume de Frontanyà 3 (Barcelona).Remarks: The genus Adapis has an exclusively European geographic distribution, including France, Switzerland, UK and Spain.It is known from the Robiacian (MP14, Middle Eocene) to the Headonian (MP19, Late Eocene).The material from Sant Jaume de Frontanyà 3 consists of several isolated teeth, currently under study, that may allow a specific determination.
Iberian localities: Mazaterón (Soria).Remarks: A mandibular fragment with the complete molar series from this locality is currently under study.It has been tentatively assigned to Adapis, although the material may be too scarce to be determined to the species level.
Iberian localities: Caenes (Salamanca).Remarks: The genus Microadapis is known from several Swiss and Spanish Middle Eocene localities ranging in age from the MP13 (Geiseltalian) to the MP16 (Robiacian).The material from Caenes consists only of a single upper molar.Until more material is recovered, it is not possible to reach a species determination.
Remarks: Only known from upper teeth, Microchoerus ornatus has been considered doubtfully distinct from M. edwardsi (Hooker, 1986;Gunnell and Rose, 2002).In the Iberian Peninsula, the only material assigned to this species is a fragment of maxilla with P 4 -M 1 from Sant Cugat de Gavadons, which according to Crusafont-Pairó (1967) and Crusafont-Pairó and Golpe-Posse (1974), shares strong similarities with the material from the type locality of M. ornatus (Mormont-Entreroches, Switzerland).

Microchoerus sp.
Iberian localities: Aguatón (Teruel).Remarks: The genus Microchoerus is well represented in Central and Western Europe (Spain, France, UK, Switzerland and Germany), in localities that range in age from the Robiacian (MP16, Middle Eocene) to the Suevian (MP21, Early Oligocene).In the Iberian Peninsula, besides the above-mentioned remains ascribed to M. erinaceus and M. ornatus, a single M 1 from the site of Aguatón was described by Peláez-Campomanes (2000).This tooth, similar in size to those of M. erinaceus, shows morphological and/or metrical differences with all the previously described species of Microchoerus (except for M. ornatus, only known from upper teeth).However, due to the scarcity of material, the author preferred not to describe a new species until more material is available.
Iberian localities: Capella [type locality] (Huesca).Remarks: This species was originally ascribed to the genus Pivetonia by Crusafont-Pairó (1967); however, most authors have considered Pivetonia to be a junior subjective synonym of Pseudoloris (Szalay, 1971;Godinot, 1983;Gunnell and Rose, 2002;Minwer-Barakat et al., 2010, 2012).The only available material from Capella is a mandible fragment with P 3 -M 2 .It differs from other species of Pseudoloris found in the Iberian Peninsula in the smaller size and the presence of a well-individualized tubercular paraconid in the lower molars.
Main references: Crusafont-Pairó (1965b, 1967); Szalay (1971); Crusafont-Pairó andGolpe-Posse (1968, 1974) Remarks: The material of this species represents the most complete sample of the genus Pseudoloris published from the Iberian Peninsula, having enabled the description of almost all the dental elements, including upper and lower incisors, undescribed for most species of this genus.P. pyrenaicus can be distinguished from the other Pseudoloris species by its medium size and by several morphological features, such as the absence of a distinct paraconid and the thickened paracristid in the lower molars, the large hypoconulid in the M 3 , the small hypocone and paraconule in the M 1 and M 2 , and the absence of these two cusps in the M 3 .This species, an intermediate form between P. isabenae and P. parvulus, forms part of an anagenetic lineage that evolved in Europe during the Middle-Late Eocene.
Geographic distribution: Exclusively known from the type locality.
Iberian localities: Mazaterón [type locality] (Soria).Remarks: P. cuestai represents, together with the undetermined Microchoerinae from Zambrana, the only record of this subfamily in the Western Iberian Bioprovince.It can be distinguished from the other Pseudoloris species by its medium size, high and thick paracristid and absence of a distinct paraconid in the lower molars, large hypoconulid in the M 3 , well-developed protocone in the P 3 and P 4 , small hypocone in the upper molars, and especially for its particular lower incisors, with a large, buccolingually enlarged crown and a thick root.It shows several similarities with P. pyrenaicus and P. parvulus, the closest species in age and geographic situation.Nevertheless, there are also clear differences with those species, especially in the anterior dentition.The peculiar morphology of the I 1 , together with the short roots of the lower canine and P 3 , suggests an anterior dentition that differs from that of P. pyrenaicus and P. parvulus, perhaps in relation to dietary differences.
Iberian localities: Sossís (Lleida).Remarks: The material described by Crusafont-Pairó (1965b, 1967) and Crusafont-Pairó and Golpe-Posse (1974) consists of a single mandibular fragment with P 4 -M 3 .However, there is additional, still unpublished material of P. parvulus from Sossís housed at the Institut Català de Paleontologia Miquel Crusafont.Among other features, P. parvulus from Sossís shows a low and sharp paracristid in the lower molars, a compressed hypoconulid in the M 3 , and the hypocone of the upper molars more developed than P. pyrenaicus and P. cuestai.This species has been identified in several Robiacian and Headonian French sites, but the only mention in the Iberian Peninsula corresponds to Sossís.
Remarks: This species was described by Crusafont-Pairó (1967) on the basis of a single isolated lower molar from Sant Cugat de Gavadons, larger than those of P. parvulus, and with a paracristid instead of a tubercular paraconid.Schmidt-Kittler (1977) considered this species as a nomen nudum because of the lack of illustrations of the holotype; however, according to the International Code of Zoological Nomenclature, this reason does not invalidate the species name from a nomenclatural viewpoint.Unfortunately, as pointed out by Köhler and Moyà-Solà (1999), this specimen was lost in the 1970s.Therefore, until more material is available, the taxonomic validity of this nominal taxon cannot be further evaluated, so we consider it a nomen dubium.
Remarks: P. godinoti is known only by six complete teeth (including an upper series of P 2 -M 1 designated as the holotype) and two fragments, recovered from the Oligocene localities of Santpedor-2 and Fonollosa-13.The M 1 has a reduced trigon basin, a large and well-individualized hypocone and a prominent metastyle, and the M 1 shows a distinct paraconid.This is the only record of the genus Pseudoloris after the Eocene-Oligocene boundary.

Pseudoloris sp.
Iberian localities: Casa Ramón (Huesca).Remarks: The genus Pseudoloris is known from several sites in France, UK, Germany and Spain, ranging in age from the Geiseltalian (Middle Eocene, MP11-12) to the Suevian (Early Oligocene, MP22).The Geiseltalian Spanish site of Casa Ramón (Peláez-Campomanes, 1995) has yielded two fragments of M 1 and M 3 and a complete M 2 ascribed to Pseudoloris sp., being similar in size but morphologically different compared to those of P. parvulus.The M 2 from Casa Ramón shows poorly-developed metaconule crests, and small paraconule and metaconule.The scarce material from this Spanish site, assigned to the MP11 or MP12, represents the earliest record of the genus Pseudoloris.
Iberian localities: Roc de Santa (Barcelona).Remarks: Known from several Late Eocene sites from France, N. antiquus has only been mentioned in the Iberian Peninsula from the locality of Roc de Santa.The material consists of a single, heavily worn M 3 described by Casanovas-Cladellas (1975).The scarcity and poor preservation of the material makes such specific assignation doubtful, but until a more detailed study is carried out, we prefer maintaining the above-mentioned determination.
Remarks: The genus Necrolemur is known from several French, Swiss and Spanish sites ranging in age from the Geiseltalian (MP13, Middle Eocene) to the Headonian (MP20, Late Eocene).There are two citations of the genus Necrolemur without specific determination in the Iberian Peninsula.On the one hand, Crusafont-Pairó (1967) and Crusafont-Pairó and Golpe-Posse (1974) assigned to this genus a single, large and very worn M 2 from Sant Cugat de Gavadons.There is no available illustration of this specimen, whose attribution cannot be confirmed based on the material from the Institut Català de Paleontologia Miquel Crusafont.The generic ascription of this tooth is therefore doubtful, since another large-sized microchoerine, ascribed by the same authors to Microchoerus ornatus, was identified in this locality; it seems probable that these remains belong to the same taxon.On the other hand, Necrolemur sp. has also been cited from the Robiacian levels of Sant Jaume de Frontanyà 1 and 2 (Busquets et al., 1992;Antunes et al., 1997).The material consists of several isolated teeth and a mandible with P 4 -M 3 of a small microchoerine.A detailed study, currently under way, will allow a more precise determination of this material.
Infraorder Simiiformes Hoffstetter, 1974(=Anthropoidea Mivart, 1864) Remarks: Hoffstetter (1974) proposed Simiiformes as a replacement name for Anthropoidea, apparently on the basis of insufficient grounds (Delson, 1977), although it has several advantages such as avoiding the standard superfamily ending in -oidea for a taxon above the family-group level (see Delson, 1977, andHoffstetter, 1982, for further details).Both Simiiformes and Anthropoidea have been recently used by several authors with the same meaning (e.g., Groves, 2001 andFleagle, 1999, respectively;but  Iberian localities: Abocador de Can Mata/C3-B2.Remarks: The oldest record of Iberian pliopithecids corresponds to an isolated M 3 from ACM/C3-B2 (Alba et al., 2012a), which has an estimated age of 12.0 Ma (MN7, late Aragonian, Middle Miocene).This tooth displays a very peculiar occlusal morphology, which enables it to be distinguished from all other pliopithecids for which this dental posi-tion is known (including Barberapithecus, see below, but not Pliopithecus canmatensis).It is also possible that this primate tooth does not belong to the Pliopithecidae but to some other group of stem catarrhines previously not recorded in Eurasia.However, on paleobiogeographic grounds, an attribution to the Pliopithecidae was favored by Alba et al. (2012a).
Remarks: This species is known from several mandibles, dentognathic fragments and isolated teeth that overall record most of the permanent dentition as well as the dP 4 .These remains correspond to at least 8 individuals from 5 different localities of the Abocador de Can Mata series; although an age range of 11.7-11.6 was originally reported for this species (Alba et al., 2010d), subsequent analyses have redated it to 11.8-11.7 Ma (Alba et al., 2012a).The pliopithecine morphology and the small size of the dental remains from these several localities from Abocador de Can Mata is most similar to that of Pliopithecus antiquus from the French localities of Sansan (MN6) and La Grive (MN7+8), but dental proportions as well as several occlusal details (less peripheral position of the protoconid and more median position of the hypoconulid, more mesial position of the buccal cuspids compared to the lingual ones, narrower but distinct mesial fovea, higher trigonid, and more extensive buccal cingulid) justify a taxonomic distinction at the species level for the ACM material (Alba et al., 2010d).
Iberian localities: Sant Quirze (Barcelona).Remarks: Pliopithecids are recorded in the late Aragonian (MN8, ca.11.8-11.2Ma) locality of Sant Quirze by an isolated M 2 described by Harrison et al. (2002).Both in size and morphology, this tooth is consistent with an attribution to Pliopithecus platyodon, otherwise unknown from the Iberian Peninsula.However, the scarcity of the available material precludes evaluating whether it might represent a different species.In any case, an attribution to P. canmantensis can be easily discarded on the larger size of the Sant Quirze molar.
Remarks: This taxon is only known from isolated dental remains recovered at the locality of Castell de Barberà, initially attributed to the Aragonian, but currently attributable to the latest Aragonian or, most likely, the earliest Vallesian.The find was initially reported more than three decades ago (Crusafont-Pairó, 1978).Most of the remains, including upper and lower teeth, correspond to a single, female adult individual.They were partly described and initially attributed to Pliopithecus sp. by Crusafont-Pairó and Golpe-Posse (1981), who nevertheless noted dental similarities with Plesiopliopithecus lockeri, currently attributed to the Crouzeliini.These authors noted that this material was likely to represent a new pliopithecid taxon, but refrained from formally erecting it.Later, Ginsburg (1986) attributed the Castell de Barberà remains to Crouzeliinae nov., but subsequently the taxon was considered a pliopithecine similar or attributable to Pliopithecus antiquus by several authors (Moyà Solà et al., 1990;Andrews et al., 1996;Harrison et al., 2002).Both Begun (2002a) and Alba et al. (2010d) considered it to be a new pliopithecid taxon, and most recently Alba and Moyà-Solà (2012a) erected a new genus and species for it within the Crouzeliinae.Besides the holotype individual, an isolated P 3 and male C 1 from the type locality are also attributed to this taxon (Alba and Moyà-Solà et al., 2012a).The latter was formerly considered a female upper canine of the hominoid Hispanopithecus laietanus (Harrison, 1991;Golpe Posse, 1993), but on morphological grounds it is best attributable to the Pliopithecidae (Begun, 2002a;Alba et al., 2010d;Alba and Moyà-Solà, 2012a).Like crouzeliins, Barberapithecus is characterized by buccolingually-compressed cusps, welldeveloped and sharp crests, and large and well-defined occlusal foveae.Its dental morphology, however, might have evolved from a pliopithecine ancestor similar to Pliopithecus canmatensis, which leaves open the possibility that, as currently conceived, the crouzeliines are polyphyletic (Alba and Moyà-Solà, 2012a).
Remarks: This taxon is known on the basis of a fragmentary mandible with right P 3 -M 3 and left P 3 -M 2 (holotype), as well as an isolated P 3 probably from the same individual.This material was initially published by Golpe-Posse (1982b), who attributed it to Pliopithecus sp. in spite of noting several differential features that, according to her, might justify the erection of a new taxon.This was confirmed by Moyà Solà et al. (1990) and Andrews et al. (1996), although the new genus and species were not erected until Moyà-Solà et al. (2001) described it and attributed it to the Crouzeliinae.Subsequently, Begun (2002a) considered it to be a Pliopithecinae, but this has been subsequently dismissed by other authors (Harrison et al., 2002;Alba et al., 2010d;Alba and Moyà-Solà, 2012a).
Remarks: The sample from the late Early Pleistocene site of Quibas (ca.1.3-1.0Ma) includes dentognathic remains, isolated teeth and some postcranial bone fragments (Alba et al., 2010c).Dental morphometric comparisons indicate that this sample is morphologically similar to M. s. florentina from the type locality (Upper Valdarno, Italy), further reinforcing the view that this subspecies is morphologically distinct from the extant nominotypical subspecies, by displaying on average absolutely longer upper molars, relatively wider upper molars, longer M 3 compared to the M 2 , absolutely longer M 1 and M 3 , and relatively narrower M 3 (Alba et al., 2011c).
Remarks: Although the macaque dentognathic remains from the Vallparadís Section (Cal Guardiola and Vallparadís Estació, ca.1.0-0.78Ma) differ metrically from the extant, nominotypical subspecies, its taxonomic attribution will remain tentative until the taxonomic distinction between M. s. florentina and the Middle Pleistocene M. s. pliocena is further clarified (Alba et al., 2008).The specimens from Canal Negre, previously attributed to M. s. florentina by Guillén-Castejón (2010), have an uncertain age, and a more detailed morphometric study would be required in order to more conclusively ascertain its taxonomic attribution to this subspecies.A dP 4 from the Early Pleistocene locality (Middle Villafranchian, MN17) of La Puebla de Valverde, initially attributed to Macaca sp.(Heintz et al., 1971), might belong to this taxon, although subsequently Moyà Solà et al.Remarks: The genus Macaca is first recorded in Europe in the latest Miocene locality of Almenara-Casablanca M (Köhler et al., 2000), after which it is known across the continent throughout the Pliocene and Pleistocene.Although there is consensus that fossil European macaques belong to the Macaca sylvanus lineage (Szalay and Delson, 1979;Alba et al., 2011c), of African origin, the scarce material from Almenara-Casablanca M does not enable an attribution to the species level.All other Iberian citations of Macaca sp. are most likely attributable to M. sylvanus, since this is the only fossil European macaque species recognized besides the endemic M. majori from Sardinia.
Macaca sylvanus (Linnaeus, 1758) Geographic distribution: W. Eurasia and N. Africa.Chronological range: Pliocene and Pleistocene.Remarks: This species, the Barbary macaque, currently presents a disjunct distribution in Algeria and Morocco, and is also represented by the Gibraltar population, which has an anthropocorous origin (Fooden, 2007).This species was also widely distributed throughout the Plio-Pleistocene in Western Eurasia, where three distinct subspecies are customarily recognized (Szalay and Delson, 1979;Delson, 1980): M. s. prisca, from the earlier Pliocene; M. s. florentina, from the late Pliocene to the early Pleistocene; and M. s. pliocena, from the Middle and perhaps Late Pleistocene.
Remarks: With the exception of Solana del Zamborino (Martín Penela, 1983, 1988), the material from the abovementioned localities has not been described in detail, and the published citations did not include an attribution to the species level except for Guillén-Castejón (2010; see below).On chronological grounds, the macaque remains from the Middle Pleistocene localities of Cueva Negra, Solana del Zamborino and Gruta da Aroeira probably belong to M. s. pliocena.The remains from Canal Negre I, of uncertain age, have been attributed to both M. s. florentina and Macaca cf.sylvanus prisca by Guillén-Castejón (2010), mainly on the basis of size.However, size does not seem to be a reliable criterion for distinguishing among the several extinct subspecies of M. sylvanus (Rook et al., 2001), and given that no detailed description or measurements are provided by Guillén-Castejón (2010), the remains from Canal Negre are best left unassigned at the subspecies level.attributed to P. arvernensis by Guillén-Castejón ( 2010), but given the lack of a detailed description and uncertainties in the attribution of other Iberian Paradolichopithecus remains, the teeth from this locality are only tentatively attributed to that species here.
Geographic distribution: Africa, India, Spain and perhaps Israel.
Remarks: Here we follow the distinction in T. oswaldi of three distinct subspecies (Jablonski, 2002): late Early to Late Pleistocene remains of T. oswaldi are attributed to T. oswaldi leakeyi, whereas the remains from the late Early or Middle Pleistocene of India are attributed to T. oswaldi delsoni, morphologically more similar to the latter, and remains from the Late Pliocene to the Early Pleistocene of Africa are attributed to the nominotypical subspecies (Delson, 1993;Jablonski, 2002).The Iberian remains of Theropithecus, unambiguously recorded at Cueva Victoria by dental remains (Gibert et al., 1995), were attributed to T. o. oswaldi, although it is uncertain that it belongs to the same subspecies recorded in Africa (E.Delson, pers. com.).A phalanx from the same locality, originally attributed to humans (Gibert Clols and Pons Moyà, 1985;Gibert et al., 1985;Gibert and Pérez-Pérez, 1989), was later reassigned to Theropithecus (Martínez-Navarro et al., 2005); although no consensus has been reached in this regard (see discussion in Gibert et al., 2008;Martínez-Navarro et al., 2008), we tentatively favor the former interpretation.
Main references: Remarks: The presence of the genus Mesopithecus in the Iberian Peninsula was first reported by Montoya et al. (2006) as Mesopithecus sp., on the basis of the Turolian remains from Venta del Moro (5.8-5.5 Ma).An ongoing study of the Remarks: The mandible from Lezetxiki was tentatively attributed to M. s. pliocena by Castaños et al. (2011) merely on chronological grounds.
Remarks: A fragmentary incisor from this Early Pleistocene locality (ca.1.8 Ma) was tentatively assigned to the genus Paradolichopithecus by Moyà Solà et al. (1990).Additional dental material would be required to discount an alternative attribution to the genus Macaca.
Remarks: The Late Pliocene remains from Moreda 1a (3.5-3.2Ma) and Cova Bonica (3.2-2.6 Ma) represent de oldest records of this genus in the Iberian Peninsula.According to Moyà Solà et al. (1990), they may correspond to a different species than P. arvernensis.Additional studies, however, are required in order to clarify the taxonomic status of these remains.In the past, several authors also cited the presence of Macaca at Cova Bonica (e.g., Delson, 1971;Crusafont-Pairó andGolpe-Posse, 1974, 1984), whereas, according to Moyà Solà et al. (1990), only Paradolichopithecus is recorded there.The detailed study of the material from Moreda and Cova Bonica, currently in progress, will enable a decision as to whether Macaca is present in either of these localities and hopefully also provide a determination of the Paradolichopithecus remains to the species level.Iberian localities: La Puebla de Valverde (Teruel); Canal Negre I (Barcelona).
Remarks: The Paradolichopithecus material from La Puebla de Valverde was attributed to P. cf. arvernensis by Aguirre and Soto (1978).We provisionally maintain such an attribution here, although noting that the species attribution might be further refined after the study of the Paradolichopithecus remains from Moreda 1a and Cova Bonica, currently under study (see comments above).Several isolated dental remains from Canal Negre, of uncertain age, have been also available dental remains from this locality will hopefully determine whether these remains correspond to the type species M. pentelicus (undoubtedly known from MN11-MN13 of Europe, Iran and Afghanistan) or to M. monspessulanus (from MN14-MN17 and perhaps also MN13 of Europe) (Andrews et al., 1996;Koufos, 2009).The presence of Mesopithecus in the Iberian Peninsula is further confirmed by the remains of uncertain age from Canal Negre reported by Guillén-Castejón (2010).The latter author attributed these remains to both M. pentelicus and Mesopithecus sp., but given age uncertainties and the lack of detailed metrical comparisons, here these remains are provisionally left unassigned to the species level.
Remarks: With an estimated age of 12.3-12.2Ma (MN7, late Aragonian), the two partial upper postcanine series of a single individual from ACM/C1-E* constitute the oldest hominoid record from the Vallès-Penedès Basin (Casanovas-Vilar et al., 2011;Alba et al., 2012a).The description and study of these remains, currently under way, will hopefully enable to ascertain whether they correspond to any of the other dryopithecin taxa recorded at Abocador de Can Mata (Pierolapithecus, Anoiapithecus, Dryopithecus) or to another, otherwise unrecorded hominid taxon.
Remarks: Several upper teeth of a single individual from ACM/C4-Cp and a mandibular fragment from ACM/BCV4, with similar estimated ages of 11.9 Ma (late Aragonian), are currently under study; attributed to Hominidae indet.by Casanovas-Vilar et al. (2011), their dental morphology indicates that they can be attributed to the Dryopithecinae, and most likely, any of the three dryopithecine genera recorded at Abocador de Can Mata (Pierolapithecus, Anoiapithecus, Dryopithecus).An isolated upper molar from Hostalets de Pierola indeterminate, initially mistaken for a suid, was attributed by van der Made and Ribot (1999) to Dryopithecus.Given the greater diversity of hominoids evidence in the area of Hostalets during the last decade, Casanovas-Vilar et al. (2011) merely attributed it to the Hominidae, although on the basis of its occlusal morphology, an attribution to the Dryopithecinae is warranted.A female lower canine from the same area, in particular from the locality of Can Mata 1 (ca.11.2 Ma, MN8, late Aragonian), was tentatively attributed to Hispanopithecus laietanus by Crusafont-Pairó and Golpe-Posse (1973a), and subsequently Golpe Posse (1993) reafirmed such attribution.Again, given the unexpected diversity of dryopithecines in this area, and the lack of homologous material for the three above-mentioned genera, Casanovas-Vilar et al. (2011) attributed this canine to Hominidae indet., although an assignment to the Dryopithecinae is most likely.Finally, a fragment of molar crown attests the presence of dryopithecines in the locality of Sant Quirze (11.8-11.2Ma, MN8, late Aragonian); attributed to H. laietanus in the past (Golpe Posse, 1993), and most recently to Hominidae indet.(Casanovas-Vilar et al., 2011), the preserved morphology precludes an attribution to the genus level.
Main references: Crusafont-Pairó and Hürzeler (1969); Crusafont-Pairó andGolpe-Posse (1973a, 1974) Remarks: This nominal taxon is based on two lower molars of a single individual (Villalta Comella andCrusafont Pairó, 1941, 1944;Pickford, 2012) from the locality of Can Vila, of uncertain stratigraphic provenance (Moyà-Solà et al., 2004;Casanovas-Vilar et al., 2011).Initially attributed to Dryopithecus fontani by Villalta Comella and Crusafont Pairó (1941), the same authors subsequently used them to erect a new taxon (Villalta Comella and Crusafont Pairó, 1944), which was subsequently synonymized with Hispanopithecus laietanus.According to the description of the site location provided by Villalta Comella and Crusafont Pairó (1941), the type locality must have been situated within or very near the area of Abocador de Can Mata, with an estimated age of 12.5-11.5Ma (Casanovas-Vilar et al., 2011).However, according to the recollections of Jordi Martinell (pers.comm.to DMA), the classical locality of Can Vila would have been situated in the Barranc de Can Vila, in a lower stratigraphic position than BCV1 (the type locality of Pierolapithecus catalaunicus), in which case an estimated age of 12.1-11.9Ma would be attributable to the former.Even though an attribution to the genus Sivapithecus can be readily dismissed, "Sivapithecus" occidentalis remains a nomenclaturally valid taxon, whose taxonomic affinities remain to be ascertained.This is complicated by the presence of several dryopithecine genera at Abocador de Can Mata around 11.9-11.8Ma, and especially by the lack of lower teeth attributable to Pierolapithecus catalaunicus.At the species level, "S." occidentalis is potentially a senior synonym of P. catalaunicus, in which case, the former would take precedence.However, this potential synonymy cannot be resolved on the basis of currently available material, so we concur with other authors (Moyà-Solà et al., 2004;Casanovas-Vilar et al., 2011) that this nominal taxon is best considered a nomen dubium.Most recently, Pickford (2012) synonymized "S." occidentalis with Neopithecus brancoi (Schlosser, 1901); the latter, however, has been also considered a nomen dubium by Casanovas-Vilar et al. (2011).Additional material is required to solve this taxonomic and nomenclatural conundrum.
Remarks: A partial face of an adult male individual from ACM/C3-Ae (11.9 Ma, MN7 or MN8, late Aragonian) was attributed to Dryopithecus fontani by Moyà-Solà et al. (2009a) on the basis of several craniodental features that do not fit with other Vallès-Penedès hominoids.If this attribution is correct, this is the only specimen representing the facial morphology of the type species of Dryopithecus, otherwise known from several mandibles and a few isolated upper teeth.On the basis of the Abocador de Can Mata specimen, D. fontani displays a mosaic of derived great-ape features, coupled with several primitive hominoid retentions and several gorilla-like features of uncertain phylogenetic interpretation (Moyà-Solà et al., 2009a).An upper male canine from ACM/ C4-Ap (11.9 Ma, MN7 or MN8, late Aragonian), figured but incorrectly attributed by Pickford (2012) to Hispanopithecus laietanus, is in fact attributable to Dryopithecus fontani (Alba and Moyà-Solà, 2012b).Remarks: A partial (proximal) femur from ACM/C3-Az, with an estimated age of 11.9 Ma (MN7 or MN8, late Aragonian), and an estimated body mass around 44 kg, has been tentatively attributed to D. fontani mainly on the basis of its large size (Moyà-Solà et al., 2009a).This specimen displays several modern hominoid-like characteristics, coupled with pronograde-related features, suggesting a greater emphasis on quadrupedalism than in Hispanopithecus laietanus (Moyà-Solà et al., 2009a).With regard to Castell de Barberà (ca.11.2-10.5Ma), the presence of hominoids at this locality was originally reported on the basis of a purported female upper canine (e.g., Golpe Posse, 1993), which in fact belongs to a male specimen of Barberapithecus (Begun, 2002a;Alba and Moyà-Solà, 2012a).However, the undoubted presence of hominoids at Castell de Barberà is documented by two (proximal and distal) pollical phalanges as well as a partial humeral diaphysis (Moyà Solà et al., 1990;Alba et al., 2011a;Almécija et al., 2011Almécija et al., , 2012)).The humeral diaphysis displays several modern hominoid-like features, most closely resembling the more complete and slender (presumably female) humerus of D. fontani from Saint Gaudens in France (Alba et al., 2011a).The phalanges more closely resemble those of other Miocene apes than those of extant great apes, indicating a significant degree of thumb-assisted, powerful grasping during above-branch quadrupedalism and cautious climbing (Almécija et al., 2012).
Remarks: The holotype and currently only known individual of this taxon includes a partial face and skeleton with Remarks: A. brevirostris has been described on the basis of a partial face and associated mandible from ACM/C3-Aj (Moyà-Solà et al., 2009b), although unpublished dental remains from a second individual have been recovered from the same locality.As for Pierolapithecus, doubts have been raised regarding the distinct taxonomic status of Anoiapithecus relative to Dryopithecus (Begun, 2009;Begun et al., 2012).In fact, however, Anoiapithecus can be distinguished from both Pierolapithecus and Dryopithecus by the strikingly orthognathous facial profile of the former (Moyà-Solà et al., 2009b), which cannot be merely discounted on the basis of purported distortion (contra Begun et al., 2012).Moreover, Anoiapithecus further differs from the other dryopithecins by several other craniodental features, most notably, in the case of Pierolapithecus, by the presence of a frontal sinus (Moyà-Solà et al., 2009b;Pérez de los Ríos et al., 2010, 2012).Nevertheless, the three dryopithecin genera display several hominid facial synapomorphies, suggesting that they are more closely related to crown hominids than to kenyapithecines from the earlier Middle Miocene (Moyà-Solà et al., 2009b;Casanovas-Vilar et al., 2011).Geographic distribution: Only known from several localities of the Vallès-Penedès Basin (NE Spain).
Remarks: Most of the dentognathic remains attributed to H. crusafonti come from Can Poncic 1.Although initial finds were assigned to Hispanopithecus laietanus, after the recovery of more abundant material, Dryopithecus crusafonti was distinguished (Begun, 1992); it was recently transferred to the genus Hispanopithecus by Moyà-Solà et al. (2009a).After its description, this taxon has been accepted (Begun, 2002b;Casanovas-Vilar et al., 2011;Pickford, 2012), questioned (Andrews et al., 1996), or formally synonymized with H. laietanus (Ribot et al., 1996).On the basis of dental differences between the Can Poncic 1 sample and those of H. laietanus from later Vallesian localities (Begun, 1992), a distinct species status is favored here, following Alba et al. (2012c).At the same time, several shared-derived dental features suggest that they belong to a single genus, even though the lack of cranial material from Can Poncic makes the generic attribution to Hispanopithecus somewhat provisional, and some authors maintain its original assignment to Dryopithecus (Pickford, 2012).The partial mandible from Teuleria del Firal (Vidal, 1913a, b;Woodward, 1914), traditionally attributed to Dryopithecus fontani (Harrison, 1991;Andrews et al., 1996;Ribot et al., 1996), is also included here in the hypodigm of H. crusafonti (Begun, 1992;Casanovas-Vilar et al., 2011;Alba et al., 2012c), given some dental differences relative to D. fontani from France and several resemblances to the few lower teeth available from Can Poncic 1.
Iberian localities: Sima del Elefante TE9 (Burgos).Remarks: The partial mandible recovered from the Early Pleistocene site (ca.1.2 Ma) of Sima del Elefante (Atapuerca) was initially assigned to H. antecessor by Carbonell et al. (2008), but most recently it has been argued that it might correspond to a different taxon (Bermúdez de Castro et al., 2011).On this basis, we prefer to leave unassigned to the species level the remains from this locality, which represent the oldest osteological record of the genus Homo in the Iberian Pensinsula.Slightly older stone tool assemblages from Fuente Nueva-3 and Barranco León-5 in Orce (Guadix-Baza Basin) further record the early presence of humans in southern Iberia (Martínez-Navarro et al., 1997;Oms et al., 2000), although the even older (1.4 Ma; Duval et al., 2011), purported human remains from Venta Micena (e.g., Gibert et al., 2002), in the same region, are not currently accepted as of H. laietanus are known from several localities, although the largest sample comes from the early Vallesian site of Can Llobateres 1 (Begun et al., 1990;Harrison, 1991;Golpe Posse 1993;Alba et al. 2012b), with an estimated age of 9.7 Ma.The intial hominoid dental finds from Can Llobateres 1 were attributed to H. laietanus, but subsequent finds resulted in the naming of two additional taxa (Dryopithecus piveteaui and Rahonapithecus sabadellensis), which are currently considered nomina nuda (e.g., Golpe Posse, 1993; but see Pickford, 2012).During the last decades, the dental sample from this locality has been generally attributed to a single species (e.g., Begun et al., 1990;Harrison, 1991;Alba et al., 2012c), with some authors setting apart an upper canine (Crusafont-Pairó and Golpe-Posse, 1973a;Golpe Posse, 1993) that nevertheless can be also attributed to the same taxon.With regard to the cranial anatomy of H. laietanus, it is known by a partial cranium from Can Llobateres 2 (Moyà-Solà and Köhler, 1993aKöhler, , 1995;;Köhler et al., 2001a, b), with an estimated age of 9.6 Ma.This taxon displays several cranial features indicative of hominid status, being in some regards more derived towards the modern great-ape condition than the Middle Miocene dryopithecins.Finally, besides some isolated bones from Can Llobateres 1 and the unpublished remains from EDAR (Checa Soler and Rius Font, 2003), the postcranial anatomy of H. laietanus is best known from the partial skeleton (comprising about 60 elements) from Can Llobateres 2 (Moyà-Solà and Köhler, 1996;Köhler et al., 2001a;Almécija et al., 2007) and the more incomplete partial skeleton from Can Feu (Alba et al., 2011b(Alba et al., , 2012d)).Several features of these skeletons indicate the possession of an orthograde body plan with suspensory adaptations (Moyà-Solà and Köhler, 1996;Köhler et al., 2002;Almécija et al., 2007;Alba et al., 2010a;Susanna et al., 2011;Pina et al., 2012a,b), although some manual features and other anatomical characteristics indicate the simultaneous retention of features functionally-related to above-branch quadrupedalism (Almécija et al., 2007;Alba et al., 2010aAlba et al., , 2012d)).
the common ancestor of both modern humans and Neandertals.Such a phylogenetic hypothesis is currently uncertain, although the validity of the species, definitely identified only from Atapuerca, is accepted by some researchers (Stringer, 2012).Its distinctive taxonomic status, either at the species or subspecies level, might ultimately be vindicated in the future by the find of additional remains.Currently, however, the available remains from TD6 are relatively scarce, fragmentary and/or subadult, and one of its purported most diagnostic features (the development of the canine fossa) is much more obvious in the juvenile than in the adult specimen, this feature further being variable in other populations attributed to H. heidelbergensis (Rightmire, 2008).Resolving the taxonomy of fossil European hominins is of course outside the scope of this paper.However, given recent taxonomic analyses favoring the distinction of a single, Afro-European species H. heidelbergensis s.l.(Mounier et al., 2009;see Stringer, 2012, for a recent review on this species), we provisionally prefer to attribute the Gran Dolina hominins to the latter taxon.The somewhat younger population from Sima de los Huesos, on the other hand, display a mosaic of morphologic features indicating that they are already derived towards Neandertals (Martinón-Torres et al., 2012), so that their removal from H. heidelbergensis and their inclusion into H. neanderthalensis might be justified (Stringer, 2012).Following the same criterion of H. heidelbergensis understood in a broad sense, the Middle Pleistocene remains from Cueva Negra del Estrecho de Quípar (Walker et al., 2010), Galeria Pesada (Trinkaus et al., 2003) and Cova de Mollet (Maroto et al., 1987) are here also attributed to H. heidelbergensis s.l.
Remarks: The classification of latest Early and Middle Pleistocene humans is still one of the most hotly debated topics in paleoanthropology and remains largely unresolved (e.g., Rightmire, 2007;2008;Hublin, 2009;Harvati et al., 2010).Traditionally, they were lumped, together with Neandertals, into a single informal category termed 'archaic' H. sapiens.In this way, the phylogenetic affinities of such taxa with later Homo and their mosaic morphology, showing a combination of plesiomorphic (Homo erectus-like) traits with some more derived features, were recognized.However, such grade classification failed to address the alpha taxonomy of these specimens (Harvati et al., 2010).Currently, there are two main views on the phylogeny and taxonomy of these forms (Rightmire, 2007;Hublin, 2009).One considers that there is a single species H. heidelbergensis s.l., which would be ancestral to both Homo neanderthalensis and modern H. sapiens, being distributed in Europe, Africa and probably Asia (e.g., Rightmire, 1998Rightmire, , 2007Rightmire, , 2008;;Mounier et al., 2009).Alternatively, several authors (e.g., Arsuaga et al., 1997) have argued that European Middle Pleistocene humans display several Neandertal derived features, thereby indicating that these forms should be classified into H. neanderthalensis, or alternatively that H. heidelbergensis should be restricted to European forms (a chronospecies ancestral to Neandertals, i.e. H. heidelbergensis s.s.), their African counterparts being attributed to a different species, H. rhodesiensis (see also Stringer, 2012).
The hominin remains from Sima de los Huesos (0.53 Ma; but see Stringer, 2012, regarding dating uncertainties) were initially attributed to Homo heidelbergensis by Arsuaga et al. (1993), although most recently Martinón-Torres et al. (2012) argued that the dental remains from this locality display more derived Neandertal traits than the typical H. heidelbergensis from Mauer or Aragó.We however think that an attribution to H. heidelbergensis is still warranted, especially when it is taken into account that this taxon is here considered in a broad sense.Similarly, the older hominin remains from Gran Dolina TD6 (ca.0.88-0.78Ma) were also initially assigned to an archaic form of H. heidelbergensis by Carbonell et al. (1995), but later Bermúdez de Castro et al. (1997) erected a new species, Homo antecessor, arguing that it would be different localities, are reported (27 from the Eocene, 3 from the Oligocene, 32 from the Miocene, 4 from the Pliocene, 51 from the Pleistocene and one with an undetermined age).This information is summarized in Figures 3 and 4, which show the ranges of the different Iberian primate taxa in relation to climatic evolution during the Cenozoic.In the following section, we discuss primate diversity dynamics in the Iberian Peninsula through time, further emphasizing its interplay with global and regional paleogeographic and climatic events.

Discussion
4.1.Iberian primate diversity in the Paleogene (Fig. 3) In Europe, the earliest occurrence of plesiadapiforms is dated to the Late Paleocene, as documented by the record of Berruvius and some members of the Saxonellidae and Plesiadapidae in France and Germany (Silcox, 2001;Bloch et al., 2007).In contrast, the record of primates (including plesiadapiforms) in the Iberian Peninsula does not go back beyond the Early Eocene.This lack of evidence for Paleocene primates in Iberia may be due to the scarcity of Paleocene continental sites: mammal remains from this epoch have been only identified in the localities of Claret, Tendruy, Palau, Fontllonga and Campo (Gheerbrant et al., 1997;López-Martínez and Peláez-Campomanes, 1999;Peláez-Campomanes et al., 2000).Anyway, the Eocene record of plesiadapiforms in this area is sparse, being restricted to the paromomyoid Arcius, known from the locality of Silveirinha (Mondego Basin, Portugal; Estravís, 2000) and recently identified also in Masia de l'Hereuet (Àger sub-basin, Spain; Marigó et al., 2012a).In the former locality, eight isolated teeth enabled description of the species A. zbyszewskii, whereas the four molars recovered from the latter site did not allow a specific determination.
The record of euprimates during the Iberian Eocene shows a much greater diversity, being represented by two distinct groups of Paleogene "prosimians": adapoids and omomyoids (8 and 3 genera, respectively, apart from several without generic ascription, which in some cases may lead to the description of new taxa after further studies).The record of adapoids begins, together with that of plesiadapiforms, in the Neustrian, with the presence of the genus Donrussellia in the Mondego Basin.In the Grauvian, the genera Cantius and Agerinia are well represented in several localities from the Pyrenean and Ebro basins, Agerinia persisting into the Geiseltalian.The Robiacian records a great diversity of adapoids, represented in the Pyrenean and Duero basins by the genera Anchomomys, Mazateronodon, Microadapis, Adapis and Leptadapis, the latter two being also identified in Headonian levels.With respect to omomyoids, the genus Pseudoloris has its oldest record in the Iberian locality of Casa Ramón (Geiseltalian).It is also well represented, together with Necrolemur, in the Robiacian and Headonian, while Microchoerus is recognized in several Headonian sites.Such da Oliveira (Santarém); Gruta Nova da Columbeira (Leiria); Lapa da Rainha (Lisboa).
Remarks: There are more than 30 Iberian localities where osteological remains of H. neanderthalensis have been found, although no partial or complete skeletons have been unearthed up to date.Several of these localities, situated in southwestern Iberia, record the latest Neandertals from Europe (Hublin et al., 1995), with a chronology of 35 ka (Trinkaus, 2007).
Remarks: There are numerous Iberian localities with fossil remains of H. sapiens, so that only the oldest citations are provided above.On the basis of osteological remains, H. sapiens is first recorded at the Iberian Peninsula in Lagar Velho (Duarte et al., 1999) and Malladetes (Arsuaga et al., 2002), and slightly later at Gruta do Caldeirão (Trinkaus et al., 2001).These remains of modern humans postdate the last Iberian Neandertals, but this is not the case of the archeological remains associated to the former.Thus, although the attribution of archeological remains to a particular human species is questionable when not accompanied by osteological remains, the recent dating of Paleolithic cave art from Spainpresumably made by H. sapiens-to 40.8 ka suggests that modern humans arrived to the Iberian Peninsula ca.41.5 ka, like in other areas of Western Europe (Pike et al., 2012).The find of an almost complete skeleton of a child in the Lagar Velho cave caused a lot of controversy, because Duarte et al. (1999) initially concluded that it displayed modern human and Neandertal traits simultaneously.Subsequently, however, other authors have pointed out that the Lagar Velho child is clearly H. sapiens and does not display Neandertal characters (Tattersall and Schwartz, 1999).

Results
An updated summary of Iberian primate taxa and localities from the Cenozoic is provided here, based on the revision of more than 200 bibliographic references.A total of 25 primate genera, corresponding to 7 families (one of plesiadapiforms, 2 of strepsirrhines and 4 of haplorrhines), from up to 116 During the Middle and Late Eocene, the mammal faunas from the western and central Iberian basins were clearly different from those of the Pyrenean basins, the latter being similar to those from Central Europe (France and Switzerland).This faunal differentiation, mainly observed in perissodactyls, artiodactyls and rodents, led to the definition of the Western Iberian Bioprovince (Cuesta, 1991), which included several western and central Iberian basins (Duero, Almazán and Oviedo Basins) characterized by Eocene mammal faunas different from those of the Pyrenees and the rest of Europe.Subsequent studies of these mammalian groups have further reinforced this contention (see Badiola et al., 2009, and references therein).The recent description of the adapiform Mazateronodon endemicus and the microchoerine Pseudoloris cuestai (Marigó et al., 2010;Minwer-Barakat et al., 2012) abundance and diversity of both adapoids and omomyoids parallels the diversity peak that primates experienced during the Eocene in the Northern hemisphere (e.g., Fleagle, 1999).The Eocene represents the warmest epoch of the Tertiary, especially during the Paleocene/Eocene Thermal Maximum and the subsequent sustained period of global warmth known as the Early Eocene Climatic Optimum (Zachos et al., 2001).Warm temperatures and not-marked seasonality favored the development of tropical forests with a predominance of evergreen thermophilic elements in latitudes as far north as 50º (Collison and Hooker, 2003).Although variable in different regions and time intervals, these dense forests would have provided suitable habitats for the radiation of these groups of prosimians.Mazateronodon endemicus and Anchomomys frontanyensis, and the omomyoids Pseudoloris pyrenaicus and P. cuestai (Marigó et al., 2010(Marigó et al., , 2011a;;Minwer-Barakat et al., 2010, 2012).Moreover, the first citation of plesiadapiforms in Spain has been reported (Marigó et al., 2012a), raising to two the mentions of this group in the Iberian Peninsula.Therefore, it seems highly probable that the diversity of Paleogene (especially Eocene) primates in the Iberian Peninsula recorded by currently available remains is even higher than recognized, and that the knowledge of this group of mammals will grow in the next few years as more detailed studies are carried out.

Primate Iberian diversity in the Miocene and
Plio-Pleistocene (Fig. 4) In the Iberian Peninsula, simians (anthropoids) are not recorded until well within the Miocene (late Aragonian), being represented by pliopithecoids (a Eurasian clade of stem catarrhines) as well as hominoids (apes and humans), whereas cercopithecoids (Old World monkeys) are recorded from the latest Miocene onwards.
Pliopithecoids from the Miocene of Eurasia are currently considered to be stem catarrhines on the basis of the possession of only two premolars (e.g.Andrews et al., 1996;Begun, 2002a), although it is uncertain whether they are more derived than African propliopithecoids and other stem catarrhines (Begun, 2002a).Once considered to be phylogenetically related to hylobatids, in fact they retain several primitive features indicating that they precede the cercopithecoid-hominoid split.In any case, they were the first anthropoids that dispersed into Eurasia, where they experienced an evolutionary radiation, in a continent previously devoid of other anthropoids, from the Early Miocene onwards (Begun, 2002a;Harrison, 2005).Pliopithecoids are first recorded by dionysopithecins from the Early Miocene of Asia (Harrison and Gu, 1999), whereas pliopithecines and crouzeliines are recorded in the Middle and Late Miocene of both Europe and Asia (Begun, 2002a;Harrison, 2005).In the Iberian Peninsula, pliopithecoids are exclusively known from the Vallès-Penedès Basin (NE Spain), where they are recorded by both pliopithecines (the genus Pliopithecus) and crouzeliines (the genera Barberapithecus and Egarapithecus).This restricted geographic distribution is probably attributable to the peculiar physiographic and paleoecological conditions of the Vallès-Penedès Basin, which by the Middle and early Late Miocene was more similar to France and other Central European countries than to the inner Iberian Basins (Agustí et al., 1984;Agustí, 1990;Casanovas-Vilar and Agustí, 2007;Casanovas-Vilar et al., 2008).Their oldest record in Iberia (ca.12.0 Ma; Alba et al., 2012a), however, is quite late compared to France and Central Europe (where finds occur from MN5), perhaps due to the lack of adequate outcrops from the early Middle Miocene in the Vallès-Penedès Basin.The best known pliopithecine from Iberia is Pliopithecus canmatensis, from the late Middle Miocene of several localities from the Abocador from the site of Mazaterón demonstrates that primates from the Western Iberian Bioprovince were also clearly different from their counterparts in the Pyrenean basins, thus reinforcing the endemic nature of the mammal faunas from this bioprovince.
The Eocene-Oligocene boundary was marked by a major faunal turnover that dramatically affected the continental vertebrate faunas and involved the extinction of a large number of Eocene mammals.This global event, named the "Grande Coupure" by Stehlin (1910), was mainly related to the climatic deterioration that took place at the Eocene-Oligocene transition, which led to a marked loss of diversity in both the marine and continental realms (Prothero, 1994;Hooker et al., 2004).Primates were not an exception, and the number of species of both adapoids and omomyoids, which had been highly succesful during the Eocene, drastically decreased in the Early Oligocene.Adapiforms persisted until the Miocene, although their record is very scarce and geographically restricted to Southeastern Asia (Gebo, 2002), whereas in Europe there is no mention of this group after the Eocene-Oligocene boundary.Regarding omomyoids, there are some brief appearances in the Oligocene and the earliest Miocene of North America (MacDonald, 1963;Albright, 2005) and Egypt (Simons et al., 1986).In Europe, however, after the Grande Coupure the omomyoid record is restricted to two finds from the Early Oligocene of Spain.Thus, Pseudoloris was able to survive the Eocene-Oligocene faunal turnover, as evidenced by Pseudoloris godinoti from the Suevian sites of Fonollosa and Santpedor in the Ebro Basin (Köhler and Moyà-Solà, 1999).Also, a single lower molar from the Early Oligocene of Aguatón evidences the persistence of Microchoerus in the Calatayud-Teruel Basin (Peláez-Campomanes, 2000).Therefore, at least two (Pseudoloris and Microchoerus) of the three microchoerine genera that inhabited the Iberian Peninsula during the Eocene further persisted, although rarely at best, into the Early Oligocene, constituting the only record of this group in Europe-the German fissure fillings of Ehrenstein and Harrlingen, containing microchoerine remains and originally assigned to the Early Oligocene (Schmidt-Kittler, 1971, 1977), were later considered to be Late Eocene in age (BiochroM'97, 1997).This might be related to maintenance of warmer temperatures in the Iberian Peninsula, compared to more northern areas of Europe, which would have favored the persistence of tropical to subtropical forested areas allowing the survival of some omomyoids until the Early Oligocene.
In summary, the Paleogene primate record in the Iberian Peninsula shows a great diversity of both adapoids and omomyoids in the Eocene, followed by a marked drop in the number of species in the Oligocene, in agreement with other areas of the Northern hemisphere (e.g., Fleagle, 1999).However, it is remarkable that the diversity of Eocene primates from the Iberian Peninsula has notably increased in recent years.For instance, in the last three years four new primate taxa have been described from the Iberian Eocene: the adapoids ing the Early and Middle Miocene (Begun, 2007;Harrison, 2010).From the beginning of the Middle Miocene, hominoids are also recorded in Eurasia, where they diversified into a plethora of genera, most of which are considered members of the great-ape-and-human clade (Hominidae) (Begun, 2002b(Begun, , 2007;;Kelley, 2002;Casanovas-Vilar et al., 2011).Some authors have favored a Eurasian origin of hominids and their early divergence during the middle Miocene into pongines and hominines, followed by a later hominine dispersal event back into Africa by the late Miocene (Begun, 2002b(Begun, , 2007(Begun, , 2009;;Begun et al., 2012).However, given the incomplete nature of the primate fossil record and the probable multidirectional nature of hominoid intercontinental dispersal and range extension events during the Middle Miocene, the place of origin of hominids cannot be currently resolved.The oldest record of Eurasian hominoids corresponds to the latest Early Miocene, just before the Langhian transgression (ca.16.3 de Can Mata series (Alba et al., 2010d).Crouzeliines, in turn, are recorded by the anapithecins Barberapithecus huerzeleri, from the latest Middle or earliest Late Miocene of Castell de Barberà (Alba and Moyà-Solà, 2012), and Egarapithecus narcisoi, from the Late Miocene of Torrent de Febulines (Moyà-Solà et al., 2001).With an estimated age of ca. 9 Ma, Egarapithecus represents the last record of a Miocene primate in Iberia before the occurrence of cercopithecoids.
With regard to hominoids, like pliopithecoids and probably due to the same paleoecological reasons, in the Iberian Neogene they are restricted to the late Middle and early Late Miocene, further displaying a similar geographic distribution, restricted to the Vallès-Penedès Basin and to the small, intra-Pyrenean Seu d'Urgell Basin (Casanovas-Vilar et al., 2011).This group, including extant apes and humans, originated around the Oligocene/Miocene boundary in Africa, where they experienced a remarkable evolutionary radiation dur- towards cooling and increased seasonality initiated after the Mid-Miocene Climatic Optimum (Zachos et al., 2001) apparently led to the crossing of some paleoenvironmental threshold, to which hominoids and several other mammals did not adapt (Casanovas-Vilar et al., 2011).This is further supported by the paleoenvironmental reconstruction of Can Llobateres 1 (Marmi et al., 2012), which records one of the latest occurrences of the hispanopithecin Hispanopithecus laietanus (9.6 Ma) before its last apperance datum at ca. 9.5 Ma (Casanovas-Vilar et al., 2011).The vegetation from Can Llobateres 1 is consistent with a subtropical to warm-temperate climate, but evidence from other Vallès-Penedès sites indicates that such habitats would have been restricted to lowland humid areas, with warm-temperate mixed forests with a significant proportion of deciduous elements growing elsewhere (Marmi et al., 2012).Following global climatic changes (Zachos et al., 2001), the progressive disappearance of tropical and subtropical plants-providing a continuous supply of fruits all year long-, and the concomitant increasing dominance by deciduous trees, may have resulted in the lack of adequate trophic resources for both hominoids and pliopithecoids during the unfavorable season, ultimately leading to their extinction not only in Iberia but also in the rest of Europe and most of Asia.
The same global changes that, in most of Eurasia, drove the extinction of pliopithecoids and hominoids, probably favored the dispersal and diversification of monkeys.By the Early and Middle Miocene, cercopithecoids were already distinct but much less diverse than hominoids, but from the Late Miocene onwards, their disparity, geographic distribution and paleodiversity increased dramatically (Szalay and Delson, 1979;Jablonski, 2002;Eronen and Rook, 2004;Jablonski and Frost, 2010).This might be related to different dietary adaptations between both groups (Andrews, 1981) and/or to differences in life-history parameters between them (Jablonski and Kelley, 1997).Such factors likely conditioned differently the evolutionary responses of cercopithecoids and hominoids to the changing climatic conditions initiated by the Middle Miocene, but most strongly expressed during the latest Miocene and the Pliocene.Thus, both colobines and cercopithecines possess specific dietary adaptations for more efficient folivory and/or consumption of unripe fruits compared to hominoids (Andrews, 1981), thereby enabling the former to more easily thrive when the ripe fruits habitually consumed by most hominoids are scarce.Moreover, hominoids display a low life-history profile, suitable for the tropical to subtropical forested conditions with low seasonality of the African Early Miocene, providing an abundant provision of high-quality food resources; in contrast, the faster life history of cercopithecoids is more suitable for the highly seasonal environments that prevailed in Europe by the latest Miocene, thus enabling them to thrive in a wider range of habitats compared to hominoids (Jablonski and Kelley, 1997;Eronen and Rook, 2004).Hence, many latest Miocene and Pliocene cercopithecoid lineages occupied less heavily for-Ma), but undoubted kenyapithecines such as Griphopithecus (most likely being stem hominids) are not recorded until the early Middle Miocene (ca.14 Ma) (Casanovas-Vilar et al., 2011;contra Andrews and Kelley, 2007, who favor an older dating).By this time, roughly coinciding with the following Mid-Miocene Climatic Optimum (Zachos et al., 2001), the kenyapithecines of African origin apparently expanded their range into eastern Europe and Turkey, and apparently gave rise to the Eurasian radiation of hominids (Moyà-Solà et al., 2009b;Casanovas-Vilar et al., 2011;Alba, 2012).
Whereas Miocene hominoids from Asia are generally regarded as pongines (i.e., members of the orangutan clade) (Kelley, 2002;Begun, 2007), the Miocene hominoids from Europe, here included into the Dryopithecinae, have been variously considered to be stem hominids (e.g., Casanovas-Vilar et al., 2011), pongines (e.g., Agustí et al., 1996) or hominines (e.g., Begun, 2009;Begun et al., 2012).Despite these phylogenetic disagreements, the discoveries of Miocene hominoids from Spain that have taken place during the last two decades have provided a wealth of new information that provides several insights into hominoid evolution.The fact that the pongine Sivapithecus is first recorded at ca. 13.0-12.5Ma in Asia (Kunimatsu et al., 2004), whereas almost simultaneously (ca.12.3-12.2) a hominid of uncertain affinities is recorded at the other end of Eurasia in the Vallès-Penedès Basin, suggests that the initial diversification of Eurasian hominoids took place somewhat earlier, probably between 14-13 Ma (Casanovas-Vilar et al., 2011).At the same time, the cranial and dentognathic characteristics of the later, Middle Miocene Spanish dryopithecins, suggest they descended from a thick-enameled kenyapithecine ancestor (Moyà-Solà et al., 2009b;Alba et al., 2010b;Casanovas-Vilar et al., 2011).The global climatic trend towards progressive cooling and increased seasonality initiated during the middle Miocene (Zachos et al., 2001;Kovar-Eder, 2003;Mosbrugger et al., 2005) apparently favored the adaptive radiation of Eurasian hominoids (Casanovas-Vilar et al., 2011), as reflected by their widened geographic range and increased taxonomic diversity, probably as a result of the new selection pressures posed by changing ecological conditions towards more heterogeneous biotopes with increased seasonality (Begun et al., 2012).Soon after the early/late Vallesian transition (ca.9.6 Ma), however, hominoids became extinct in Western and Central Europe, with the exception of the insular endemic Oreopithecus, which survived until ca.7 Ma in Italy (Casanovas-Vilar et al., 2011).
In the Iberian Peninsula, dryopithecines are last recorded by Hispanopithecus at ca. 9.5 Ma, slightly predating the last occurrence of pliopithecoids ca.9.0 Ma.The extinction of European hominoids has been linked to the vegetation changes that took place during the late Vallesian (Fortelius and Hokkanen, 2001;Agustí et al., 2003), which further resulted in the extinction of many other mammalian taxa (i.e., the Vallesian Crisis; see Agustí and Moyà-Solà, 1990;Agustí et al., 1999;Casanovas-Vilar et al., 2010).The climatic trend different dispersal events of early humans into the Iberian Peninsula and Western Eurasia in general took place during the Pleistocene (Bermúdez de Castro and Martinón-Torres, 2013), thus being still uncertain (see discussion in Madurell-Malapeira et al., 2012) whether humans continuously occupied the Iberian Peninsula between their early record at ca. 1.2 Ma and their subsequent record at ca. 0.88-0.78Ma (Carbonell et al., 1995;Bermúdez de Castro, 1997).
ested and more seasonal habitats than the preceding hominoids.Two distinct families of cercopithecoids are usually distinguished: crown cercopithecoids are classified into a single family, Cercopithecidae, with two distinct subfamilies (Colobinae and Cercopithecinae), distinguished at the family level by some authors (e.g., Jablonski, 2002); whereas stem cercopithecoids are classified into another, extinct family (the Victoriapithecidae).In the Iberian Peninsula, both cercopithecines and colobines are recorded.Colobines are represented by Mesopithecus in the latest Miocene, and by its presumed descendant, Dolichopithecus, during the Pliocene.Cercopithecines, in turn, are recorded by Macaca, from the latest Miocene onwards, as well as by Paradolichopithecus from the Late Pliocene and Early Pleistocene, and Theropithecus from the Early Pleistocene.
Until recently, the colobine Mesopithecus had not been recorded in the Iberian Peninsula, but recent finds (Montoya et al., 2006;Guillén Castejón, 2010) have shown that this was merely an artifact of the record.At Venta del Moro, Mesopithecus is recorded ca.5.8-5.5 Ma (Montoya et al., 2002(Montoya et al., , 2006)), thus conclusively indicating that Mesopithecus dispersed into westernmost Europe by at least the latest Turolian (MN13).Similarly, Macaca is recorded by ca.6.1-5.3Ma at Almenara-Casablanca M (Köhler et al., 2000).The dispersal of these semi-terrestrial and oportunistic genera into the Iberian Peninsula must be understood within the framework of the significant faunal turnovers that took place in Europe by this time (van der Made, 2006;Agustí et al., 2006), apparently before the first evaporitic cycle of the Messinian Salinity Crisis.This event implied the clousure of the Gibraltar strait, ultimately leading to the total (or almost complete) desiccation of the Mediterranean between 5.5-5.3Ma (Hsü et al., 1973).At the same time, this created an African-European connection that allowed the dispersal of certain mammal taxa including hippopotamids, gerbils, crested rats and likely cercopithecines as well (Agustí et al., 2006).
Subsequent dispersal and range extension events took place into the Iberian Peninsula during the Pleistocene, as reflected by the record of the gelada Theropithecus at ca. 1.0 Ma in Cueva Victoria (Gibert et al., 1995), and that of early Homo from ca. 1.2 Ma (Carbonell et al., 2008;Bermúdez de Castro et al., 2011, 2012b) onwards.Throughout the remaining Pleistocene, humans and macaques were apparently the only primates that managed to survive in the Iberian Peninsula and the rest of Europe, being (for different biological reasons) well suited to survive under the changing paleoenvironmental conditions (with marked glacial cycles) of the Pleistocene (Eronen and Rook, 2004).In the case of European macaques, they apparently extended their geographical range much farther northwards during the interglacials, and retreated into their Mediterranean refugia during glacials, although they ultimately became locality extinct due to their inability to tolerate the most severe stage of the last glaciation (Fooden, 2007), being last recorded in Iberia by ca.80 ka (Castaños et al., 2011).With regard to humans, there is the possibility that ; Antunes et al. (1997); Marigó et al. (2011a, b, c); Minwer-Barakat et al. (2011a); Roig and Moyà-Solà (2011); Moyà-Solà et al. (2012).

Fig. 3 .
Fig. 3.-Range chart of Paleogene Iberian primate taxa.Uncertain attributions (cf., indet.)are shown in dark gray, whereas nomina nuda and nomina dubia are not considered.The global oxygen isotope record, together with the development of ice-sheets and major climatic/physiographic events, are also included.Age boundaries for epochs, stages, European Land Mammal Ages (ELMA) and Paleogene Mammal Zones (MP) are after Luterbacher et al. (2004).Climatic data are taken from Zachos et al. (2001).
see Rosenberger, 2006, for a different option), and hence are considered synonyms here.The International Code of Zoological Nomenclature does not regulate these taxonomic ranks, and hence the use of one or the other remains arbitrary.