Regarding the Real Diversity of Glyptodontidae (mammalia, Xenarthra) in the Late Pliocene (chapadmalalan Age/stage) of Argentina

A large diversity of Glyptodontidae has been proposed as characterizing the Chapadmalalan Age (Pliocene). Most of these taxa were recognized on the basis of partial dorsal carapaces and/or caudal tubes, whereas the main diagnostic characteristic is a particular morphology of the exposed surface of the osteoderms. From a biostratigraphic point of view some species are biostratigraphically important. The Upper Chapadmalalan is based on the Paraglyptodon chapadmalensis biozone. Both the re-evaluation of the type and referred materials and new significant findings from the Chapadmalal and El Polvorín Formations indicate that the diversity of Pliocene Glyptodontidae is more limited than previously supposed. The particular morphology of the exposed surface of the osteoderms that characterizes some of the species actually corresponds to a taphonomic alteration, which results in a non-real ornamentation pattern. Thus, the Glyptodontinae P. chapadmalensis must be replaced as a fossil guide because neither this species nor the species included in the genera Urotherium, Trachycalyptus and Lomaphorus are well characterized. Taking into account the diversity of Glyptodontidae for this lapse, the Glyptodontinae are very scarce (a situation that contrasts with its records in the Pleistocene), whereas Eosclerocalyptus, " Plohophorini " (Plohophorus) and Doedicurinae (cf. Eleutherocercus antiquus) are among the most recorded taxa.


INTRODUCTION
Glyptodonts (Xenarthra, Glyptodontidae) constitute an extinct clade of cingulates widely distributed in North and South America (Carlini and Zurita 2010), with records that span from the late Eocene of Patagonia to the early Holocene of the Pampean region of Argentina (Soibelzon et al. 2012).From a taxonomic point of view, most of the genera were recognized in the current territory of Argentina, although the knowledge of glyptodonts from other regions of South and North America has been largely improved over the last decades (Gillette ALFREDO E. ZURITA et al. and Ray 1981, Carlini et al. 2008a, b, Porpino et al. 2010, 2014, Zurita et al. 2011a, 2013).
From a morphological perspective, the Quaternary taxa are, in general, well known, and as a consequence of this, many of the recognized lineages of Glyptodontidae have been characterized by complete (or almost complete) specimens coming from the Pleistocene units, especially from the Pampean region of Argentina (Burmeister 1870-74, Ameghino 1889, Lydekker 1895).This situa tion changes dramatically in what concerns the knowledge of Paleogene and Neogene glyptodonts, which is much more restricted.In fact, several taxa are still recognized on the basis of fragments of the dorsal carapace and/or caudal armor (e.g.Glyptatelinae and "Hoplophorinae" Palaehoplophorini;Hoffstetter 1958, Scillato-Yané 1977).
As a result of new fi eld works carried out in the Chapadmalal and Polvorín Formations (Pliocene) of the surroundings of the localities of Mar del Plata and Olavarría (Buenos Aires province) (Figure 1), several new and more complete remains of Glyptodontidae with strict geographic and stratigraphic control have been exhumed.In the present contribution, we carry out a careful taxonomic revision of some of the traditional species recognized for the Chapadmalalan lapse (Pliocene) and discuss the taxonomic status of other related glyptodonts.This may allow obtaining a more real approach of the diversity of this clade during part of the Pliocene.
U. simplex (the type species) was described and characterized by Castellanos (1926) on the basis of a partial caudal tube (MACN Pv 5813) (Fig. 3a), originally collected by Carlos and Florentino Ameghino in the surroundings of Chapadmalal, Buenos Aires province, Argentina, in 1908.In the same contribution, Castellanos (1926: 268) tentatively assigned fi ve osteoderms of the dorsal carapace (MACN Pv 6288) (formerly assigned by F. Ameghino to "Sclerocalyptus sp?") originally interpreted by the same author as belonging to the genus Lomaphorus (see Castellanos, 1926: 268).The main characteristics mentioned for U. simplex include some morphological details of the exposed surface of the osteoderms of the caudal tube and dorsal carapace, which are clearly rugose and uniformly perforated by numerous small foramina, showing some resemblance to that observed in the Pleistocene genus Neuryurus (Zurita et al. 2006).In some osteoderms, it is possible to observe a central fi gure surrounded by a series of small radiating foramina.The contact area and articulation between adjacent osteoderms is evident and, in some cases, the osteoderms are not fi rmly united, a characteristic that is present in some juvenile specimens of Eosclerocalyptus C. Ameghino and Neosclerocalyptus Paula Couto (see Zurita 2007b).According to Castellanos (1926: 269), the caudal tube shows only one distal lateral fi gure on each side.The reexamination of the holotype indicates that the caudal tube is mostly restored, and thus its real morphology is diffi cult to observe.Furthermore, a comparison with new and more complete materials exhumed from the El Polvorín and Chapadmalal Formations, in addition to materials from "Araucanian" levels (late Miocene-Pliocene) of Northwestern Argentina, undoubtedly indicates that the main characteristics proposed for Urotherium are also present in some caudal tubes and dorsal carapaces of already well known taxa, especially in Eosclerocalyptus (Xen 30, Mam-63, MMP 4842, MFCA 676) and Plohophorus fi guratus Ameghino (MMP 4823; Xen 72) (Figs. 4a,a´,b,b´,c).The particular morphology characterizing the referred specimens of Urotherium is produced by a taphonomic alteration at the level of the exposed surface of the osteoderms (Fig. 4b´ and 5).Our observations show that this non-real pattern is mainly observable in the lineages of glyptodonts with "rosette" ornamentation (e.g.Plohophorus and Eosclerocalyptus).In addition, the presence of only one lateral fi gure on each side of the caudal tube (mentioned as a diagnostic characteristic of U. simplex) is also shared by other related glyptodonts such as the Neosclerocalyptini Neosclerocalyptus (Zurita 2007a).On the other hand, the linear measurements given by Castellanos (1926: 263) are within the ranges of Eosclerocalyptus and Neosclerocalyptus (see Zurita 2007a).In summary, new specimens preserving dorsal carapaces and/or caudal tubes and showing the two types of ornamentation (i.e."Urotherium" and Eosclerocalyptus patterns) are quite common in Pliocene levels (Figs. 4a,a´,a´´ and 5d).Notably, and supporting our taxonomic hypothesis, none of the materials referred to Pliocene Urotherium include the skull (except U. antiquus, a species coming from the Monte Hermoso Formation, see below).This is because, when the skull is found associated with the dorsal carapace, the specimen is assigned to already known taxa, like Eosclerocalyptus and Plohophorus.
Although it is not the main purpose of this contribution, the revision of the holotype of U. simile (MFCA 727), a species recognized by Castellanos (1948) from "Araucanian" levels (late Miocene-early Pliocene) in Santa María, Valle de Yocavil, Catamarca province, Argentina, indicates that it corresponds to a juvenile specimen, a situation observed by Castellanos himself (1948: 6).The materials include a very badly preserved caudal tube, a large number of non-articulated osteoderms, and a fragment of the right hemimandible with two molariforms (MFCA 593) (Fig. 3b).The morphology of these structures completely coincides with that of a juvenile glyptodont (Fig. 4d,f), as observed in Zurita (2007b) (e.g.MFCA 676) and Zurita et al. (2009Zurita et al. ( , 2011c) (e.g.PVE-F 85 and MNAP-V 6146a) and thus shows no diagnostic characteristic.In fact, the almost only characteristic mentioned by Castellanos (1948) to identify this species (i.e. the smaller size of the caudal tube compared to U. simplex) could correspond to the early ontogenetic stage of this specimen.
With regard to U. interundatum, the holotype is currently lost (MCNP-; see Mones 1986a, Scillato-Yané et al. 2013) but a calcotype is preserved (MLP M-229).As pointed out by Scillato-Yané et al. ( 2013), the fossil is restricted to a single osteoderm of the dorsal carapace.Although some characteristics suggest a Doedicurinae resemblance, the scarcity and bad state of preservation of the osteoderm precludes any taxonomic hypothesis, and more material is needed.
As mentioned above, the only species in which the skull is preserved corresponds to U. antiquum (=Neuryurus antiquus), a species recognized by Ameghino (1888) from the Monte Hermoso Formation that requires an urgent taxonomic revision.In addition to associated osteoderms, this taxon includes the dorsal carapace, a relatively well preserved skull and a left hemimandible.However, an examination of the referred materials (Ameghino 1889, Plate LXII, Figs. 6 and 7) indicates two different types of osteoderms that can not belong to the same species: one of them shows a typical Doedicurinae affi nity (see Zurita et al. 2014) while the other shows a more similar condition observed in the "Urotherium pattern".This situation suggests that the material could include more than one specimen.With regard to the skull referred by Ameghino (1889, Plate LXIII), although preliminary, a comparison with a Pliocene Doedicurinae recently described in Zurita et al. (2014), suggests that U. antiquum does not show a Doedicurinae morphology.

REFERRED SPECIES
Lomaphorus imperfectus (Gervais and Ameghino, 1880) (*type species) L. elevatus (Nodot, 1857) L. compressus Ameghino, 1882 L. elegans (Burmeister, 1871) L. cingulatus Ameghino, 1889 L. chapalmalensis Ameghino, 1908 TAXONOMIC AND ANATOMIC REMARKS Lomaphorus is a genus created by Ameghino (1889) to originally include six species, of which one, Lomaphorus chapalmalensis Ameghino, 1908, is from the Chapadmalal Formation.From a morphological point of view, Ameghino (1889: 819), as was common at the time, characterized this genus on the basis of the exposed surface of the osteoderms of the dorsal carapace and some morphological details of the caudal tube.Among these morphological details, the exposed surface of the osteoderms presents a "rosette" pattern, although less noticeable than in "Hoplophorus" (i.e.Neosclerocalyptus Paula Couto), and in some cases the osteoderms are not fi rmly united among them, a characteristic also observed in Urotherium.In addition, it is possible to observe some radiating foramina surrounding this central figure, replacing the peripheral figures present in Eosclerocalyptus and Neosclerocalyptus.The general morphology of the caudal tube resembles that of Neosclerocalyptus, but the central fi gures of each osteoderm lack peripheral fi gures (Ameghino 1895: 327).Remarkably, Ameghino (1889: 819) noted that the species of Lomaphorus were smaller than those of "Hoplophorus" (=Neosclerocalyptus).This is an important observation, because several of the characteristics postulated by Ameghino agree with those described for juvenile specimens of Neosclerocalyptus (Zurita et al. 2005; e.g.GCF 83) and Eosclerocalyptus (Zurita 2007b; e.g.MFCA 676) (Fig. 4d), and even for other clades such as Glyptodontinae (Zurita et al. 2011c).In fact, the assignment of this particular morphology to juvenile specimens was first mentioned by Lydekker (1895: 21), supporting his idea that the caudal tube of Lomaphorus corresponded in fact to "Sclerocalyptus" (i.e.Neosclerocalyptus).This taxonomic and ontogenetic interpretation was rejected by Ameghino (1895).However, it seems probable that Ameghino himself had doubts about the validity of Lomaphorus when he stated that: "In my opinion, this is the name that should continue to be used, even if it were proven that Lomaphorus is not a distinct genus" (Ameghino 1895: 525).
In this scenario, Lomaphorus chapalmalensis was created by Ameghino (1908) on the basis of a distal fragment of the caudal tube (MACN Pv 5806) (Fig. 3c).As observed in U. simplex, the main characteristics correspond to a particular morphology of the exposed surface of the osteoderms.This is mainly rugose with a variable number of small foramina.In dorsal view, in some osteoderms it is possible to observe an elevated central fi gure, which is surrounded by a large number of small radiating foramina.This morphology completely agrees with that observed for several specimens of Eosclerocalyptus exhumed from the El Polvorín and Chapadmalal Formations (e.g.Mam-63-4) (Fig. 4c).Actually, this morphology was also observed in some juvenile specimens of the Pleistocene and Pliocene genera Neosclerocalyptus (e.g.GCF 83) and Eosclerocalyptus (e.g.MFCA 676) (Fig. 4d,d´).Our observations, including those of complete specimens of Neosclerocalyptus and Eosclerocalyptus (e.g.Xen 30; MMP 5303), show that some areas of the dorsal carapace show a typical "Lomaphorus" ornamentation pattern whereas others preserve the real "rosette" ornamentation pattern (Figs. 3 a,a´a´´ and 5d).
In the remaining species of the genus, particularly in L. cingulatus Ameghino 1889, the single osteoderm from the dorsal carapace (Fig. 3d), which characterizes the species and comes from the "Mesopotamian" (late Miocene-early Pliocene) of the surroundings of Paraná, Entre Ríos province, is currently lost.However, there is a calcotype (MACN A-592).According to Castellanos (1940: 262) and Scillato-Yané et al. (2013: 125), this osteoderm does not correspond to the genus Lomaphorus (Trachycalyptus?).In fact, the characteristics postulated by Ameghino (1889: 821) agree with those of the remaining species of Lomaphorus, and the scarcity of the material precludes a precise taxonomic assignation, indicating that this species must be labeled as nomen dubium.
Lomaphorus corallinus is another species recognized and characterized by Rovereto (1914: 103, Plate VII, fi g. 1, 1a) on the basis of some associated osteoderms of the dorsal carapace (MACN Pv 8331) (Fig. 3e).The holotype comes from "Arau-canian" levels (late Miocene-Pliocene) of northwestern Argentina (Valle de Santa María, Catamarca province).Later, this species was transferred by Castellanos (1932) to the genus Lomaphorops, although this was rejected by Cabrera (1944) due to the scarcity of the holotype and the lack of characteristics.Castellanos (1948) later referred to this species new material, particularly numerous osteoderms of the dorsal carapace and a caudal tube.The holotype corresponds to osteoderms of the antero-lateral region of the dorsal carapace, and their exposed surface is almost identical to that observed in glyptodonts with a "rosette" ornamentation pattern.In fact, Castellanos himself (1948: 8) indicated that some osteoderms were very similar to those of Trachycalyptus, whereas others were similar to those of Eosclerocalyptus and Neosclerocalyptus.Therefore, no diagnostic characteristics are visible, and this species must be considered as nomen dubium (sensu Mones 1986b).In addition, the original description given by Rovereto (1914) completely agrees with the morphology of the exposed surface of the osteoderms of the antero-lateral region of the dorsal carapace of glyptodonts with a "rosette" ornamentation pattern (e.g.Eosclerocalyptus).
The Pleistocene species of Lomaphorus correspond to L. imperfectus (Gervais and Ameghino, 1880, the type species of the genus), L. compressus Ameghino, 1882, L. elegans Burmeister, and L. gracilis Nodot, 1857.Lomaphorus has been mentioned as a valid taxon in the early Pleistocene (Ensenadan Age) and middle Pleistocene (Bonaerian Age) of the Pampean region of Argentina (Ameghino 1889, Scillato-Yané et al. 1995, Carlini and Scillato-Yané 1999, Cione and Tonni 2005, Soibelzon et al. 2010, Luna and Giraudo 2012, Cruz et al. 2012).L. imperfectus is mainly known on the basis of osteoderms of the dorsal carapace and caudal tubes.Even if the holotype is not known with certainty, some material was illustrated by Ameghino (1889, Plate 58, fi g. 1, 2, 3 and Plate 60, fi g. 6) and Lydekker (1895, Plate 3, fi g. 1) (Fig. 3f).Our comparisons with specimens of Neosclerocalyptus pseudornatus and N. ornatus including both structures (e.g.MMP 4600, MACN Pv 7075; Fig. 4d,e) show a remarkable morphological similarity (see Zurita 2007a).L. compressus is only known by osteoderms of the dorsal carapace (see Ameghino, 1889, Plate 69, fi g. 18).As in the previous case, the morphology of the exposed surface of the osteoderms is also present in several specimens of Neosclerocalyptus (e.g.MFCA 758).The poor characterization of this species becomes evident when Ameghino (1895) stated that: "Sclerocalyptus scrobiculatus should disappear because the armor attributed to this species belongs to Lomaphorus compressus…" In turn, the morphology of the osteoderms of L. elevatus clearly resembles that of juvenile specimens of Neosclerocalyptus, as pointed out by Luna and Giraudo (2012) and Oliva et al. (2013).Supporting this interpretation, none of the remains assigned to Lomaphorus included the skull.Genus Trachycalyptus Ameghino in Rovereto (1914) STRATIGRAPHIC AND GEOGRAPHIC DISTRIBUTION Late Miocene-Pliocene of Argentina.

REFERRED SPECIES
T. chapadmalensis Ameghino in Rovereto (1914) TAXONOMIC AND ANATOMIC REMARKS Trachycalyptus is another poorly characterized genus, originally recognized by Ameghino (1908) as a nomen nudum and later illustrated by Rovereto (1914 Plate 27, fig.2).T. chapadmalensis is a species that has never been formally described.The holotype (MACN Pv 5823) comprises seven associated osteoderms of the lateral region of the dorsal carapace, possibly from the Chapadmalal Formation (Fig. 3g).Reexamination of the material has shown that the osteoderms only preserved ALFREDO E. ZURITA et al.
the outer zone (see Krmpotic et al. 2009).The genus was characterized by Castellanos in several opportunities (e.g.Castellanos 1940Castellanos : 263, 1948: 11): 11).As in the above-mentioned genera, the main diagnostic characteristic completely agrees with that of glyptodonts possessing a "rosette" pattern at the level of the dorsal carapace, and the morphology is almost identical to that observed in the specimens referred to Lomaphorus and Urotherium.Subfamily Glyptodontinae Gray, 1869 Genus Paraglyptodon Castellanos, 1932 STRATIGRAPHIC AND GEOGRAPHIC DISTRIBUTION Pliocene of the Pampean and Northwestern regions of Argentina (Ameghino 1908, Castellanos 1953, Reguero et al. 2007, Reguero and Candela 2011, Zurita et al. 2013).

REFERRED SPECIES
P. chapadmalensis (Ameghino, 1908) P. dubius (Ameghino, 1908) P. uquiensis Castellanos, 1953P. paranensis Castellanos, 1953 TAXONOMIC AND ANATOMIC REMARKS Two species of the subfamily Glyptodontinae were recognized as originally coming from the Chapadmalal Formation: P. dubius Castellanos and P. chapadmalensis (Ameghino in Rovereto).The latter species was originally recognized by Ameghino (1908: 426) as belonging to the Pleistocene genus Glyptodon Owen (G.chapadmalensis) but neither illustrated nor descriptions were available in that contribution.Some years later, Rovereto (1914: Plate 27) illustrated the holotype and classified it as "Sclerocalyptus" chapadmalensis (Fig. 3H).In turn, Castellanos (1932) recognized the genus Paraglyptodon to include this species (P.chapadmalensis).Some years later Castellanos (1953) included two articulated osteoderms (MACN Pv 6285), previously illustrated by Rovereto (1914, Plate 29) with the same stratigraphic and geographic provenance as P. chapadmalensis, as P. dubius.As pointed by Oliva et al. (2010) both holotypes (P.chapadmalensis and P. dubius) are morphologically almost identical and undoubtedly correspond to the subfamily Glyptodontinae, taking into account the morphology of their exposed surface (see Zurita et al. 2013).However, a new revision of the holotype of P. chapadmalensis and P. dubius demonstrates that it is not possible to observe clear diagnostic characteristics.This is mainly because the 16 associated osteoderms forming the holotype of P. chapadmalensis (Fig. 3h) correspond to the lateral region of the dorsal carapace, which retains a conservative morphology in most glyptodonts (see Zurita et al. 2011bZurita et al. , 2013)).Our observations suggest that the exposed surface of the osteoderms presents the same taphonomic alteration observed in other glyptodonts with "rosette" ornamentation pattern.In addition to this, it seems possible that the osteoderms do not come from Pliocene sediments but from upper Pleistocene units (MT per.obs., see also Reig 1958).This possibility is congruent with the original observations of Ameghino (1908), who identifi ed the osteoderms as belonging to Glyptodon.In fact, comparisons with juvenile specimens of Glyptodon suggest that the holotype represents an early ontogenetic stage of Glyptodon (Zurita et al. 2011c) (Fig. 4g).This is an important observation because P. chapadmalensis characterizes the upper Chapadmalalan Age/Stage (see Cione and Tonni 2005).According to this, the only valid species of the genus corresponds to P. uquiensis Castellanos, a well characterized species that shows several synapomorphies with Glyptodon (Zurita et al. 2013).In turn, Oliva et al. (2010) demonstrated the status of nomen dubium of P. paranensis.

DISCUSSION AND RESULTS
A large diversity of Cingulata Glyptodontidae has been traditionally proposed as characterizing the Pliocene, especially the Chapadmalalan Age/Stage (late Pliocene).As common during the second half of the 19th century and the fi rst half of the 20th century, these taxa were mainly recognized on the basis of partial dorsal carapace and/or caudal tubes (Soibelzon et al. 2006, Fernicola andPorpino 2012).Since then, this diversity has been regarded as valid in most of the specialized literature (see Pascual et al. 1966, Scillato-Yané et al. 1995, Carlini and Scillato-Yané 1999, Cione and Tonni 1995, 2001, 2005, among others).
In this context, both the reexamination of the holotypes and new findings coming from Chapadmalal and El Polvorín Formations indicate that the real diversity of Glyptodontidae of the Pliocene (and some of the Pleistocene) is much more restricted than previously considered.The particular pattern observed in the exposed surface of the osteoderms of the dorsal carapace and/or caudal tube characterizing several taxa, is due to a taphonomic alteration and/or juvenile specimens (see Zurita et al. 2009Zurita et al. , 2011c)).Although this non-real pattern can be observed in some Pleistocene specimens, this taphonomic process is very common in late Pliocene glyptodonts.To support our hypothesis, it is possible to mention the following: a) none of these taxa include skulls; b) in several taxa, the description given by the authors includes an intriguing combination of characteristics present in well known species (e.g.Neosclerocalyptus and Eosclerocalyptus) with others produced by this taphonomic alteration; c) on several occasions, the authors themselves recognized the notable morphological similarity among taxa ("Lomaphorops", Trachycalyptus and Lomaphorus…….although they present similarities that relate them closely"; Castellanos 1948: 26); d) the notable similarity between Lomaphorus and "Sclerocalyptus" (=Neosclerocalyptus) becomes evident when Ameghino himself transferred some material originally assigned to Lomaphorus to "Sclerocalyptus" (Ameghino 1889) or by the endless debate between Ameghino (1895) and Lydekker (1895) about the taxonomic identity of Lomaphorus and "Sclerocalyptus".
In addition, the strongest argument concerns the new findings from the Chapadmalal and El Polvorín Formations of almost complete specimens of well known species (Eosclerocalyptus and Plohophorus), in which some areas of the dorsal carapace/caudal tubes present a typical Lomaphorus/ Trachycalyptus/Urotherium morphology while others present a clear "rosette" pattern (Fig. 5).
From a biostratigraphic point of view, some species were important.This is especially true for the Glyptodontinae Paraglyptodon chapadmalensis, because it characterizes the upper Chapadmalalan Age/Stage (see Cione andTonni 1995, 2005).In fact, with regard to the Glyptodontinae, our fi eld observations show the almost complete absence of records of this clade in Chapadmalalan levels, a situation that dramatically changes when compared to the Pleistocene, in which Glyptodontinae become the most abundant glyptodonts in South America (Carlini and Zurita 2010).Suggestively, the Glyptodontinae seem to be more frequently recorded in the late Chapadmalalan levels of Uquía Formation (Northwestern Argentina).This could be due to the fact that both regions have had different biogeographical scenarios since late Miocene (see Reguero et al. 2007, Reguero andCandela 2011).
In this context, the Glyptodontidae "Plohophorini" were not formally studied in this contribution, but all the remains recovered from the Chapadmalal and El Polvorín Formations belong to Plohophorus fi guratus, a taxon also present in the Monte Hermoso Formation (Tomassini et al. 2013).It is possible that the remaining species of Plohophorus with the same stratigraphic and geographic provenance and recognized on the basis of caudal tubes (P.cuneiformis Ameghino, P. coronatus Rovereto, and P. sygmaturus Ameghino) (see Castellanos 1939)

CONCLUSIONS
The evidence indicates that the real diversity of Pliocene Glyptodontidae is much more limited than previously supposed, as already proven in some Pleistocene glyptodonts.In fact, most of the species included in the genera Lomaphorus, Lomaphorops, Urotherium, and Trachycalyptus chapadmalensis do not show valid characteristics and must thus be considered as non-valid taxa.From a biostratigraphic perspective, the Glyptodontinae P. chapadmalensis does not show an adequate characterization and must be replaced as a fossil guide in the biozone of the Upper Chapadmalalan Age/Stage.
Taking into account the real diversity of late Pliocene Glyptodontidae, the scarcity of records of Glyptodontinae draws attention, whereas the Doedicurinae (cf.Eleutherocercus antiquus), Eosclerocalyptus and "Plohophorini" (Plohophorus fi guratus) are the most frequently recorded taxa in the Chapadmalal and El Polvorín Formations (Table I).This situation clearly contrasts with the Pleistocene scenario, in which the "Plohophorini" are completely absent, the Doedicurinae are one of the least recorded glyptodonts, and the Glyptodontinae (Glyptodon and Glyptotherium) are the glyptodonts most frequently recorded (Carlini and Zurita 2010).The evidence suggests that this turnover regarding the Glyptodontidae association started between the older Chapadmalalan and the younger Marplatan (ca.3.3 Ma; see Cione andTonni 2001, Vizcaíno et al. 2004).However, taking into account the results presented in this contribution, an update of the Marplatan Glyptodontidae is much needed.
Finally, a preliminary comparison between the Glyptodontidae association from the Montehermosan levels of the Monte Hermoso Formation (late Miocene?-Pliocene) and Chapadmalalan levels, suggests that it is certainly possible that most of the valid taxa are present in both units, as recently suggested by Zurita et al. (2014).This observation is in agreement (at least in part) with the hypothesis of Tomassini et al. (2013) with regard to the real age of the Monte Hermoso Formation and its chronological relation with the Chapadmalal Formation.

Figure 4 -
Figure 4 -Eosclerocalyptus cf.lineatus (Xen 30).a, dorsal carapace in lateral view; a´, detail of the exposed surface of the osteoderms showing a clear Lomaphorus/Urotherium/Trachycalyptus ornamentation pattern; a´´, detail of the exposed surface of the osteoderms showing a "rosette" ornamentation pattern; b, Eosclerocalyptus sp., dorsal carapace showing the taphonomic alteration of its exposed surface (b´); c, Eosclerocalyptus cf.lineatus (Xen 30), proximal half of caudal tube showing a Lomaphorus/ Urotherium pattern; d, Neosclerocalyptus (GCF 83), caudal tube belonging to a juvenile specimen; d´, detail of the distal portion showing the particular morphology; e, Neosclerocalyptus ornatus (MMP 4600), caudal tube in dorsal and ventral views showing the similitude with the caudal tube assigned to Lomaphorus; f, Glyptodon (PVF 85), associated osteoderms and caudal armor belonging to an early juvenile stage; g, Glyptodon reticulatus (MCA 2017), dorsal carapace belonging to a juvenile specimen showing the morphology of the lateral osteoderms.
Museo de Ciencias Naturales "Carlos ALFREDO E. ZURITA et al.

cf: central fi gure; cs: central sulcus; rs: radial sulci; rf: radiating foramina; pf: peripheral fi gures.
Zárate 2005, Gómez Samus et al. 2014guratus.This is an important point, becauseDeschamps (2005)proposed the Actenomys priscus-P.ALFREDO E. ZURITA et al. cuneiformis zone.An integral taxonomic revision of Plohophorus and allied genera is thus needed.With regard the Glypodontidae Doedicurinae, Zurita et al. (2014) recently suggested the presence of cf.Eleutherocercus antiquus as the only valid species coming from Chapadmalalan levels.From a biostratigraphic point of view, the Glyptodontidae assemblage recognized both for the El Polvorín and the Chapadmalal Formations clearly suggests that both units are, at least partially, synchronous.This interpretation is in agreement with the age estimated for El Polvorín (ca.3.55Ma) and Chapadmalal Formations (ca.4.5-3.2Ma.)(seeZárate2005, Gómez Samus et al. 2014).