<<

Current Biology Magazine

Correspondence PILOSA

Dasypus kappleri The phylogenetic * septemcinctus * Dasypus hybridus affi nities of the extinct * Dasypus novemcinctus FG Dasypus yepesi glyptodonts Long-nosed * Dasypus sabanicola * Dasypus novemcinctus US (Dasypodinae) 0.50 / - / - * Frédéric Delsuc1,*, Gillian C. Gibb1,2, Dasypus pilosus Melanie Kuch3, Guillaume Billet4, Euphractus sexcinctus Hairy, Six-banded & Armadillos 1 5 * Chaetophractus villosus Lionel Hautier , John Southon , 0.99 / 1.0 / 88 () Jean-Marie Rouillard6, Zaedyus pichiy * Chaetophractus vellerosus 7 * Juan Carlos Fernicola , 35 ± 3 Mya sp. (Glyptodontinae) 8 9 [42-30] Sergio F. Vizcaíno , Ross D.E. MacPhee , * Chlamyphorus truncatus Fairy Armadillos and Hendrik N. Poinar3,* 0.97 / 1.0 / 69 Calyptophractus retusus (Chlamyphorinae) Priodontes maximus 0.93 / 1.0 / 73 * Tolypeutes tricinctus PP / PP / BP Among the fossils of hitherto unknown CAT PART PART * Tolypeutes matacus Giant, tatouay Three-banded that Darwin collected in South * * & Naked-tailed Armadillos America between 1832 and 1833 during Cabassous chacoensis () Cabassous centralis * the Beagle expedition [1] were examples * Cabassous unicinctus of the large, heavily armored later known as glyptodonts. Ever since, E M L E M L E L E M L glyptodonts have fascinated evolutionary Paleo. Eocene Oligocene Pli. P. biologists because of their remarkable 66 62 59 56 48 38 34 28 23 16 12 5 3 0 Million years ago skeletal adaptations and seemingly isolated phylogenetic position even Figure 1. Phylogenetic position of glyptodonts. within their natural group, the cingulate Phylogeny and molecular timescale of extant armadillos including the extinct glyptodont Doedi- xenarthrans (armadillos and their allies curus sp. (in red). Bayesian chronogram was obtained using a rate-autocorrelated log-normal [2]). In possessing a carapace comprised relaxed molecular clock model using PhyloBayes under the CAT-GTR-G mixture model with a of fused , the glyptodonts birth death prior on the diversifi cation process, and six soft calibration constraints. Mean diver- gence dates and associated 95% credibility intervals are represented as node bars. Plain black were clearly related to other cingulates, node bars indicated calibration constraints. The main geological periods follow Geological Time but their precise phylogenetic position Scale of the Geological Society of America (E = Early, M = Middle, L = Late; Paleo. = Paleocene, as suggested by morphology remains Pli. = , P. = Pleistocene). Statistical support values obtained from three different phy- unresolved [3,4]. To provide a molecular logenetic reconstruction methods (PPCAT: Bayesian Posterior Probability under the CAT-GTR+G mixture model; PP : Bayesian PP under the best partition model; BP : Maximum likelihood perspective on this issue, we designed PART PART sequence-capture baits using in silico Bootstrap Percentage under the best partition model) are indicated with stars corresponding to nodes with PP > 0.95 and BP > 90. The full chronogram and phylogram are provided in Figure S2. reconstructed ancestral sequences and successfully assembled the complete recent proposal that, despite numerous a recently assembled dataset mitochondrial genome of Doedicurus differences in body size and carapace encompassing all modern xenarthran sp., one of the largest glyptodonts. structure, glyptodonts do not constitute a species [5], we reconstructed, in silico, a Our phylogenetic reconstructions sister-group to armadillos, as traditionally set of ancestral mitogenomic sequences, establish that glyptodonts are in fact assumed [2], but are instead nested which permitted the synthesis of a deeply nested within the within them [4,6]. This hypothesis is, set of target capture RNA baits. Baits crown-group, representing a distinct however, based on a restricted set constructed in this way may allow for subfamily (Glyptodontinae) within family of cranio-dental characters. Here, a more specifi c sequence capture Chlamyphoridae [5]. Molecular dating we put this proposition to the test by of phylogenetically distant ancient suggests that glyptodonts diverged analyzing the mitochondrial genome of a specimens than baits based solely no earlier than around 35 million years specimen of the late surviving glyptodont on available modern sequences. This ago, in good agreement with their fossil Doedicurus. One of the largest members permitted the reconstruction of a nearly record. Our results highlight the derived of its , with an estimated body mass complete mitochondrial genome of nature of the glyptodont morphotype, of ~1.5 tons [7], Doedicurus exhibited Doedicurus at 76x coverage. Illumina one aspect of which is a spectacular numerous distinctive characters, reads mapping to the newly assembled increase in body size until their famously including a club-shaped, Doedicurus mitogenome were 45 base at the end of the last . armored tail adorned with spikes, pairs on average and displayed C-to-T Although the phylogenetic unity of presumably used in intraspecifi c combat. damage patterns at both 3’ and 5’ order Cingulata has never been seriously Using ancient DNA (aDNA) extraction ends, characteristic of authentic aDNA. questioned, how its three constituent techniques, we recovered endogenous We have ruled out the possibility of groups (armadillos, glyptodonts, and DNA from a carapace fragment the inadvertent enrichment of nuclear pampatheres) are related to one another (MACN Pv 6744) dated to 12,015 ± 50 copies of mitochondrial origin (NUMTs) has been diffi cult to resolve in fi ne detail. 14C radiocarbon years before present by performing additional phylogenetic Of special interest in this regard is the (Supplemental information). Utilizing controls (Supplemental information).

Current Biology 26, R141–R156, February 22, 2016 ©2016 Elsevier Ltd All rights reserved R155 Current Biology Magazine

By comparing our ancient mitogenome the global congruence observed with 2. Hoffstetter, R. (1958). . In Traité de paléontologie, Vol. 2, no. 6., P. Piveteau, ed. to those of living xenarthrans (Figure previous nuclear-based phylogenies (Masson et Cie: Paris), pp. 535–636. 1), we were able to confi dently place as well as molecular dating analyses 3. Engelmann, G.F. (1985). The phylogeny of Doedicurus within armadillos as the provides convincing evidence for the the Xenarthra. In The evolution and ecology of armadillos, sloths, and vermilinguas. G.G. sister-group of a clade composed of proposed xenarthran evolutionary Montgomery, ed. (Smithsonian Institution Press: Chlamyphorinae (fairy armadillos) and history [5]. On this evidence, glyptodonts Washington DC), pp. 51–64. 4. Gaudin, T.J., and Wible, J.R. (2006). The phylogeny Tolypeutinae (three-banded, naked-tailed (Glyptodontinae) comprised a of living and extinct armadillos (Mammalia, and giant armadillos; Supplemental distinct, Late Paleogene lineage of Xenarthra, Cingulata): a craniodental analysis. In Amniote paleobiology: perspectives on the information). This clearly contradicts the chlamyphorid armadillos [5]. Such a evolution of mammals, birds and reptiles, M.T. old view that glyptodonts must have radical repositioning of glyptodonts within Carrano, T.J. Gaudin, R.W. Blob, and J.R. Wible, diverged from other cingulates at a very the armadillo crown group has major eds. (University of Chicago Press: Chicago), pp. 153–198. early point in their phylogenetic history, consequences for interpreting aspects 5. Gibb, G.C., Condamine, F.L., Kuch, M., Enk, J., on the grounds that, for example, they of cingulate evolution. For example, the Moraes-Barros, N., Superina, M., Poinar, H.N., and Delsuc, F. (in press). Shotgun mitogenomics possessed such features as a completely dome-shaped, tightly-fused carapace provides a reference phylogenetic framework and fused carapace lacking movable bands of glyptodonts has long been thought to timescale for living xenarthrans. Mol. Biol. Evol. http://dx.doi.org/10.1093/molbev/msv250. [3]. Our results are more compatible be fundamentally different from that of 6. Billet, G., Hautier, L., De Muizon, C., and Valentin, but still incongruent with recent armadillos and pampatheres, in which the X. (2011). Oldest cingulate skulls provide morphological cladistic analyses [4,6] carapace consists of articulated sections. congruence between morphological and molecular scenarios of armadillo evolution. Proc. Biol. Sci. that position glyptodonts within a more Our results imply that the unarticulated 278, 2791–2797. inclusive but nevertheless paraphyletic carapace is in fact a derived feature, 7. Vizcaíno, S.F., Cassini, G.H., Toledo, N., and Bargo, M.S. (2012). On the evolution of large size Euphractinae. which in turn provides an explanation for in mammalian herbivores of Cenozoic faunas To examine the consequences of the apparent presence of movable bands of southern South America. In Bones, clones this novel phylogenetic placement, in some Miocene glyptodonts [3]. and biomes: an 80-million year history of recent Neotropical mammals, B. Patterson and L. Costa, we incorporated Doedicurus into Glyptodonts were a group of eds. (University of Chicago Press: Chicago), pp. the morphological character matrix ambulatory specialized herbivores that 76–101. 8. Simpson, G.G. (1948). The beginning of the age of of Billet et al. [6], but were unable to reached giant size bracketed between mammals in South America. Part 1. Introduction. identify any exclusive synapomorphies two extant of armadillos that do Systematics: Marsupialia, Edentata, Condylarthra, Litopterna, and Notioprogonia. Bull. Am. Mus. Nat. justifying grouping of the former with not share either of these characteristics. Hist. 91, 1–227. Chlamyphorinae + Tolypeutinae. In our Based on our new phylogenetic 9. Woodburne, M.O., Goin, F.J., Bond, M., Carlini, study, only two characters, pertaining to framework, we performed a statistical A.A., Gelfo, J.N., López, G.M., Iglesias, A., and Zimicz, A.N. (2014). Paleogene land the shape and position of the mandibular reconstruction of ancestral body masses. faunas of South America; a response to global coronoid process, might qualify as According to our analysis, the mean climatic changes and indigenous fl oral diversity. J. Mammal. Evol. 21, 1–73. potential synapomorphies, but only under ancestral body mass estimate of the last 10. Bergqvist, L.P., Abrantes, E.A., and Dos Santos the assumption that both have reverted common ancestor of Glytodontinae + Avilla, L. (2004). The Xenarthra (Mammalia) of to ancestral states in three-banded Chlamyphorinae + Tolypeutinae was a São José de Itaboraí basin (Upper Paleocene, Itaboraian), Rio de Janeiro, Brazil. Geodiversitas armadillos. This analysis nevertheless mere 6 kg (95% credibility interval: 1–19 26, 323–337. revealed other morphological similarities kg), implying a spectacular increase between glyptodonts and fairy armadillos in glyptodont body mass during the 1Institut des Sciences de l’Evolution, UMR 5554, (Supplemental information). Neogene (Supplemental Information). CNRS, IRD, EPHE, Université de Montpellier, Montpellier, France. 2Ecology Group, Institute We estimate that glyptodonts diverged This inference is in line with the fossil of Agriculture and Environment, Massey from Chlamyphorinae + Tolypeutinae record, which indicates that glyptodonts University, Palmerston North, New Zealand. 35 ± 3 million years ago, close to the evolved from medium-sized forms in the 3McMaster Ancient DNA Centre, Department Eocene–Oligocene transition (Figure 1). Miocene (e.g., , of Anthropology, Biology and Biochemistry, This molecular estimate is compatible ~80 kg) to become true in McMaster University, Hamilton, Canada. 4 with the age of the oldest and widely the Pleistocene (e.g., clavipes, Sorbonne Universités, CR2P, UMR 7207, CNRS, Muséum National d’Histoire naturelle, accepted glyptodont remains (Mustersan ~2,000 kg) before disappearing with most Université Paris 06, Paris, France. 5Keck- Glyptatelus osteoderms [8], ca. 36–38 other South American large mammals Carbon Cycle AMS facility, Department of million years old [9]). Tarsal bones from some 10,000 years ago [7]. Earth System Science, University of California, the Early Eocene locality of Itaboraí Irvine, CA, USA. 6MYcroarray, Ann Arbor, MI (Brazil), currently dated to more than SUPPLEMENTAL INFORMATION 48105, United States / Chemical Engineering 50 Myr [9]), have been interpreted as Department, University of Michigan, Ann Arbor, MI, USA. 7Consejo Nacional de Investigaciones glyptodont, but the elements in question Supplemental Information including results, ac- knowledgements, experimental procedures and Científi cas y Técnicas, Sección Paleontología are better interpreted as belonging de Vertebrados, Museo Argentino de Ciencias two fi gures can be found with this article online to indeterminate dasypodoids [10]. Naturales Bernardino Rivadavia, Buenos Aires, at http://dx.doi.org/10.1016/j.cub.2016.01.039. According to our results, they might Argentina. 8División Paleontología Vertebrados, belong to stem cingulates that evolved Facultad de Ciencias Naturales y Museo, REFERENCES before basal divergences occurred within Universidad Nacional de La Plata, CONICET, La Plata, Argentina. 9Division of Vertebrate Zoology/ the armadillo crown group, an event we 1. Fernicola, J.C., Vizcaíno, S.F., and De Iuliis, G. Mammalogy, American Museum of Natural date to ca. 45 million years ago (Figure 1). (2009). The fossil mammals collected by Charles History, New York, NY, USA. Darwin in South America during his travels on While our results are based strictly board the HMS Beagle. Revist. Asoc. Geol. Argent. *E-mail: [email protected] (F.D.), on the comparison of mitogenomes, 64, 147–159. [email protected] (H.N.P.)

R156 Current Biology 26, R141–R156, February 22, 2016 ©2016 Elsevier Ltd All rights reserved