First North American fossil monkey and early Miocene tropical biotic interchange

Abstract

New World monkeys (platyrrhines) are a diverse part of modern tropical ecosystems in North and South America, yet their early evolutionary history in the tropics is largely unknown. Molecular divergence estimates suggest that primates arrived in tropical Central America, the southern-most extent of the North American landmass, with several dispersals from South America starting with the emergence of the Isthmus of Panama 3-4 million years ago (Ma). The complete absence of primate fossils from Central America has, however, limited our understanding of their history in the New World. Here we present the first description of a fossil monkey recovered from the North American landmass, the oldest known crown platyrrhine, from a precisely dated 20.9-Ma layer in the Las Cascadas Formation in the Panama Canal Basin, Panama. This discovery suggests that family-level diversification of extant New World monkeys occurred in the tropics, with new divergence estimates for Cebidae between 22 and 25 Ma, and provides the oldest fossil evidence for mammalian interchange between South and North America. The timing is consistent with recent tectonic reconstructions of a relatively narrow Central American Seaway in the early Miocene epoch, coincident with over-water dispersals inferred for many other groups of animals and plants. Discovery of an early Miocene primate in Panama provides evidence for a circum-Caribbean tropical distribution of New World monkeys by this time, with ocean barriers not wholly restricting their northward movements, requiring a complex set of ecological factors to explain their absence in well-sampled similarly aged localities at higher latitudes of North America.

Full text

00 MONTH 2016 | VOL 000 | NATURE | 1
LETTER doi:10.1038/nature17415
First North American fossil monkey and early
Miocene tropical biotic interchange
Jonathan I. Bloch1, Emily D. Woodruff1,2, Aaron R. Wood1,3, Aldo F. Rincon1,4, Arianna R. Harrington1,2,5, Gary S. Morgan6,
David A. Foster4, Camilo Montes7, Carlos A. Jaramillo8, Nathan A. Jud1, Douglas S. Jones1 & Bruce J. MacFadden1
New World monkeys (platyrrhines) are a diverse part of modern
tropical ecosystems in North and South America, yet their early
evolutionary history in the tropics is largely unknown. Molecular
divergence estimates suggest that primates arrived in tropical
Central America, the southern-most extent of the North American
landmass, with several dispersals from South America starting
with the emergence of the Isthmus of Panama 3–4 million years
ago (Ma)1. The complete absence of primate fossils from Central
America has, however, limited our understanding of their history
in the New World. Here we present the first description of a fossil
monkey recovered from the North American landmass, the oldest
known crown platyrrhine, from a precisely dated 20.9-Ma layer in
the Las Cascadas Formation in the Panama Canal Basin, Panama.
This discovery suggests that family-level diversification of extant
New World monkeys occurred in the tropics, with new divergence
estimates for Cebidae between 22 and 25 Ma, and provides the
oldest fossil evidence for mammalian interchange between South
and North America. The timing is consistent with recent tectonic
reconstructions
2,3
of a relatively narrow Central American Seaway
in the early Miocene epoch, coincident with over-water dispersals
inferred for many other groups of animals and plants
4
. Discovery
of an early Miocene primate in Panama provides evidence for a
circum-Caribbean tropical distribution of New World monkeys by
this time, with ocean barriers not wholly restricting their northward
movements, requiring a complex set of ecological factors to explain
their absence in well-sampled similarly aged localities at higher
latitudes of North America.
The Miocene epoch (23.8–5.3 Ma) is marked by substantial climatic
and ecological changes that had profound effects on terrestrial mam
-
mal communities in the New World tropics
5
. Fossils from the tropical
lowlands of Central America are rare owing to a lack of relevant rock
exposures; however, an important exception can be found in Panama
where, since 2009, expansion of the Panama Canal has exposed fossil-
bearing rocks of early Miocene age. The lower Miocene Las Cascadas
Formation (Fig. 1) represents the oldest fossiliferous continental
sequence exposed along the Panama Canal area
3
and includes a diverse
fossil mammal assemblage6–9 that, while compositionally different
from that of the younger overlying Centenario Fauna
10
, similarly has
almost entirely North American affinities despite its proximity to South
America, consistent with a peninsular connection to North America10,11.
Palaeontological fieldwork in 2012–2013 resulted in the discovery of
seven isolated fossil primate teeth that shed new light on the early
evolution of crown New World monkeys in the tropics and pro-
vide insight into the timing and dynamics of the earliest stages of the
Great American Biotic Interchange (GABI) between North and South
America12.
Primates Linnaeus, 1758
Anthropoidea Mivart, 1864
Platyrrhini Geoffroy, 1812
Cebidae Bonaparte, 1831
Panamacebus transitus gen. et sp. nov.
Et y molog y. Generic name combines ‘Panama’ with ‘Cebus’, root taxon
for Cebidae. Specific name ‘transit’ (Latin, crossing) refers to its implied
early Miocene dispersal between South and North America.
Holotype. UF 280128, left upper first molar (M1; Fig. 2a, b).
Referred material. L eft upper second molar (M2; UF 281001; Fig. 2a, b),
partial left lower first incisor (I1; UF 280130), right lower second
incisor (I2; UF 267048), right lower canine (C1; UF 280131), possibly
associated left lower second (P2; UF 280127) and fourth (P4; UF
280129) premolars (Fig. 2e–g).
Local i ty. Lirio Norte (site key YPA-024 in the Florida Museum of
Natural History Vertebrate Paleontology Collection), Panama Canal
area, Panama, Central America (Extended Data Fig. 1).
Age and horizon. Primate fossils were collected from a single horizon
of sub-aerially exposed ash fall deposits in the upper part of the Las
Cascadas Formation (Fig. 1a). Magmatic zircons from a non-subaerially
exposed andesitic tuff, ~ 0.25 m below the primate-bearing horizon
(Extended Data Fig. 2), yield an age of 20.93 ± 0.17 Ma (2σ) (Fig. 1b and
Supplementary Table 1), interpreted to be the eruption age of the tuff and
a close approximation of the absolute age of the primate-bearing horizon.
The mammalian assemblage (Supplementary Table 6) recovered from
this horizon is consistent with what is known from the late Arikareean
Ar4 faunal zone (22.8–19.05 Ma; refs 10, 13) North American Land
Mammal Age (NALMA) at higher latitudes13,14 (Extended Data Fig. 3).
See also Supplementary Information (results section).
Diagnosis. Medium-sized cebid platyrrhine (~ 2.7 kg) that differs
from all other cebines in having lower incisors (I1, I2) that are higher
crowned with an incomplete lingual cingulum, and a first upper
molar (M1) that is considerably larger than the second upper molar
(M2). Further differs from most cebines: (except Aotus) in having
an I1 cross-sectional area only somewhat smaller than that of I2;
(except Cebus) in having somewhat inflated lower premolars (P2, P4),
and a P4 talonid and trigonid of similar length; and (except Saimiri
and Neosaimiri) in having M1,2 with a short and strong anterior
cingulum. Further differs: from Cebus and Acrecebus in having M1,2
with a mesiobuccally oriented prehypocrista, lacking a metaconule, a
postprotocrista that is continuous with a hypometacrista that reaches
to the base of the metacone, a sharp and distinct hypometacrista,
and lack of lingual inflation of the protocone; from Cebus in having
a P4 with a smaller hypoconid; and from Saimiri in lacking a
pericone on M1,2. Differs from all callitrichines in having a discrete
1Florida Museum of Natural History, University of Florida, Gainesville, Florida 32611-7800, USA. 2Department of Biology, University of Florida, Gainesville, Florida 32611-7800, USA. 3Department
of Geological and Atmospheric Sciences, Iowa State University, Ames, Iowa 50011-1027, USA. 4Department of Geological Sciences, University of Florida, Gainesville, Florida 32611-7800, USA.
5Department of Evolutionary Anthropology, Duke University, Durham, North Carolina 27708-9976, USA. 6New Mexico Museum of Natural History and Science, Albuquerque, New Mexico 87104,
USA. 7Geociencias, Universidad de los Andes, Calle 1A # 18A-10, Edificio IP, Bogotá DC 111711, Colombia. 8Smithsonian Tropical Research Institute, Box 0843-03092, Balboa, Ancon, Republic of
Panama.
© 2016 Macmillan Publishers Limited. All rights reserved

2 | NATURE | VOL 000 | 00 MONTH 2016
Letter
reSeArCH
entoconid on P4, and M1,2 with a large hypocone and lacking a
buccal cingulum; from callitrichines (but also Aotus) in having an
M2 with a buccally expanded paracone; and from callitrichines
(but also Saimiri and Neosaimiri) in having M1,2 with a well-developed
prehypocrista that extends to the postprotocrista and encloses the talon
lingually. For additional description and metrics see Extended Data
Figs 4–7 and Supplementary Information (results section).
Panamacebus is similar to other platyrrhines in having a single-rooted
P4 rather than double-rooted as in catarrhines. Among cebids, the
dental morphology is most similar to that of fossil cebines, including
middle Miocene Neosaimiri from C olombia (Fig. 2 and Extended Data
Figs 4–7). For additional comparisons see Supplementary Information
(results section) and Supplementary Figs 2 and 3.
Until recently, the oldest evidence for the arrival of primates in
South America from the Old World was from the late Oligocene
epoch (26 Ma) of Bolivia15. The recent discovery of anthropoids from
eastern Peru
16
suggests an earlier arrival into tropical South America,
perhaps during the late Eocene epoch (~ 37–34 Ma), although the date
is uncertain
17
. Dispersals from Africa or Asia through Antarctica or
North America have been suggested
18
, but ‘rafting’ on floating islands
from Africa across the Atlantic is considered the most likely mechanism
for their arrival into South America19. Results from phylogenetic
analyses of new morphological data coded into a previously published
matrix1 vary with the use of different molecular constraints1,20, but
both support Panamacebus within crown Platyrrhini, specifically
within Cebidae (Fig. 3; Supplementary Information (results
section) and Supplementary Figs 4–10), suggesting a northward
early Miocene dispersal across the Central American Seaway (CAS)
from South to North America, perhaps associated with the onset of
extensive development of terrestrial landscapes in Central America
as a consequence of the initial collision with South America2,3 (Fig. 4
and Extended Data Fig. 8).
Using our most resolved tree (Fig. 3a), the minimum age for a split
between Callitrichinae and Cebinae can be calibrated with the radio-
isotopic date (20.77–21.90 Ma, posterior probability (PP): 1.0). The
results of our divergence dating analysis (Extended Data Figs 9, 10 and
Supplementary Information (results section)) using this date as a prior
(Supplementary Table 2), considerably narrow previously reported
range estimates (16.07–23.5 Ma; ref. 20; and 15.66–24.03 Ma; ref. 21).
Our inferred divergence of Cebidae from Atelidae (21.84–24.93 Ma,
PP: 0.99) also greatly reduces the interval estimated in a previous study
(18.14–26.11 Ma; ref. 20). Our divergence estimates for other primate
clades are generally congruent with previous studies (Supplementary
Information (results section)). Panamacebus does not seem to share a
ab cd
eh
f
g
i
j
Figure 2 | Comparison of Panamacebus with middle Miocene cebid
Neosaimiri fieldsi from La Venta, Colombia. a, b, Occlusal (a) and
lingual (b) views of P. transitus left M1 (UF 280128: http://dx.doi.
org/10.17602/M2/M8531) and M2 (UF 281001: http://dx.doi.org/10.17602/
M2/M8550). c, d, Occlusal (c) and lingual (d) views of N. fieldsi right M1
(IGM-KU 89008, mirrored: http://dx.doi.org/10.17602/M2/M8541) and
M2 (IGM-KU 89018, mirrored: http://dx.doi.org/10.17602/M2/M8543).
e–g, Occlusal (e), lingual (f) and buccal (g) views of P. transitus partial
left I1 (UF 280130: http://dx.doi.org/10.17602/M2/M8400), right I2 (UF
267048, mirrored: http://dx.doi.org/10.17602/M2/M8395), right C1 (UF
280131, mirrored: http://dx.doi.org/10.17602/M2/M8401), left P2 (UF
280127: http://dx.doi.org/10.17602/M2/M8397) and left P4 (UF 280129:
http://dx.doi.org/10.17602/M2/M8399). h–j, Occlusal (h), lingual (i) and
buccal (j) views of N. fieldsi left dentary with I2–M2 (UCMP 39205: http://
dx.doi.org/10.17602/M2/M1879, http://dx.doi.org/10.17602/M2/M1880).
Images generated from microCT scan data (see Methods and links
associated with specimen numbers). Scale bars:1 mm (a–d); 5 mm (e–j).
Bas
Obispo
Fm.
Las Cascadas Formation Culebra
Formation
Cucaracha
Formation
Pedro
Miguel
Fm.
Centenario
Fauna (~19 Ma)
Lirio Norte
local fauna
0
5
10
15
Epoch
Miocene
Ma
16
18
20
22
24
26
28
30
Ar1
Ar2
Ar3
Ar4
Biochronol.
Oligocene
Arikareean Hemingf.
He2
He1
25.37 ± 0.13 Ma (b)
19.29 ± 0.4 Ma (a)
*
a
* 20.93 ± 0.17 Ma
18
19
20
21
22
23
24
25
206Pb/238U age (Ma)
.
Sandstones
Conglomerates
Ash falls and lapilli tuffs
Purple–grey mudstones
Grey and black mudstones
Agglomerates
Mean = * 20.93 ± 0.17 Ma
MSWD = 0.85, probability = 0.73
Fossil
vertebrates
b
Figure 1 | Stratigraphy of the primate-bearing locality (YPA-024) in
central Panama. a, Measured stratigraphic section (in metres) in the
Las Cascadas Formation showing the positions of the dated rock sample
(asterisk) and of the Panamacebus fossils as a silhouette of a monkey (right),
correlated to a schematic stratigraphic position of the Lirio Norte Local
Fauna and the Centenario Fauna with previously published radiometric
dates (a, ref. 3; b, ref. 28) indicated (centre), and North American land
mammal faunal zonation (left). Biochronol., Biochronology; Fm.,
formation; Hemingf., Hemingfordian North American Land Mammal Age;
MSWD, mean square of weighted deviates. b, Results from U–Pb isotopic
analyses plotted as a weighted mean of the 206Pb/238U ages from 37 zircons
(see Methods and Supplementary Information).
© 2016 Macmillan Publishers Limited. All rights reserved

00 MONTH 2016 | VOL 000 | NATURE | 3
Letter reSeArCH
close relationship with fossil primates from the Quaternary period of
Cuba, Jamaica and Hispaniola (Fig. 3a, b), although they probably also
dispersed from tropical South America into the Greater Antilles by the
early Miocene1,22. This result suggests two unrelated migrations from
South America to both tropical North America (a crown platyrrhine)
and the Caribbean islands (probable stem platyrrhines) during the
Oligocene/early Miocene epochs.
While attribution of fossils to crown versus stem platyrrhines has
been the subject of substantial ongoing debate23, our results support
an exclusively tropical crown platyrrhine radiation with only the
more distantly related (stem) platyrrhines distributed into the higher
latitudes of South America
1
. However, if the alternative view proves
to be correct, that primates of similar age in Patagonia are also crown
platyrrhines (the long-lineage hypothesis)23, Panamacebus would
still be at least equal in age to the oldest disputed crown platyrrhine.
Coupled with fossil evidence for primates in the lower Miocene of
the Greater Antilles24, this new record in Central America shows that
primates were dispersing northward out of South America with a previ-
ously unrecognized early Miocene circum-Caribbean distribution. The
primate record is consistent with new tectonic and palaeogeographic
reconstructions2,3 of a relatively narrow CAS in the early Miocene
and corresponding dispersals inferred on the basis of molecular and
fossil data for many terrestrial organisms, including amphibians,
reptiles, freshwater fishes, insects and plants4. In this context, it is
perhaps surprising to not also find fossil evidence of caviomorph
rodents and sloths in the early Miocene of Panama since they are found
in a slightly younger locality that includes evidence of a primate in the
Greater Antilles
24
. Absence of primates from the overlying Centenario
Fauna
10
, which is ~ 2 million years (Myr) younger than the Lirio Norte
Local Fauna, might suggest a single waif dispersal event of a short-lived
population. Alternatively, with increased collecting efforts the fossil
record may yet provide evidence for other earlier mammal dispersals
out of South America, and/or primates having a longer history in trop-
ical North America than is currently known.
Prior to this discovery, the oldest fossil evidence for mammalian
dispersal from South to North America is 8.5–9 Ma mylodontid and
megalonychid sloths that, along with later immigrants associated with
different pulses of the GABI, would have necessarily had to move
through the tropics, but also quickly dispersed to higher latitudes. By
contrast, New World monkeys clearly crossed into the tropical lowlands
of Central America at least once by 21 Ma, but there is no record of
them in localities of similar age at higher northern latitudes. This is
especially notable in the Gulf Coastal Plain, where most other mammals
Cebus
Brachyteles
Chiropotes
Saguinus
Mazzonicebus†
Mohanamico†
Homunculus†
Callimico
Tremacebus†
Callicebus
Ateles
Antillothrix†
Branisella†
Leontopithecus
Nuciruptor†
Saimiri
Alouatta
Cacajao
Aotus dindensis†
Proteropithecia†
Acrecebus†
Aotus
Soriacebus†
Xenothrix†
Lagonimico†
Dolichocebus†
Callithrix
Cebuella
Panamacebus†*
Lagothrix
Paralouatta†
Cebupithecia†
Neosaimiri†
Pithecia
Stirtonia†
Chilecebus†
Carlocebus†
Aotus dindensis†
Callithrix
Mazzonicebus†
Branisella†
Chiropotes
Alouatta
Dolichocebus†
Panamacebus†*
Cebupithecia†
Xenothrix†
Tremacebus†
Aotus
Ateles
Cebus
Homunculus†
Cebuella
Saimiri
Callimico
Paralouatta†
Pithecia
Neosaimiri†
Chilecebus†
Acrecebus†
Mohanamico†
Lagonimico†
Soriacebus†
Brachyteles
Saguinus
Nuciruptor†
Callicebus
Antillothrix†
Carlocebus†
Cacajao
Proteropithecia†
Leontopithecus
Stirtonia†
Lagothrix
a
b
Callitrichinae
Cebinae
Atelidae
Pitheciidae
Atelidae
Pitheciidae
Cebinae
Figure 3 | Results from phylogenetic analyses showing Panamacebus within crown Platyrrhini. a, Strict consensus of two most parsimonious trees
analysed using molecular constraint 1 (ref. 1). b, Strict consensus of four most parsimonious trees analysed using molecular constraint 2 (ref. 20). Circles
at nodes correspond to bootstrap support values: orange, greater than 85; blue, 50–85; black, less than 50. Daggers indicate fossil (extinct) taxa. Asterisk
indicates P. transitus.
Caribbean Plate
Culebra and
Cucaracha Fms
250 km
15
30
0
7590
600 km
105
Gulf of Mexico
Pacic
Ocean
Atlantic
Ocean
Caribbean Sea
CAS
South
America
Las Cascadas Fm.
Reconstruction
blocks
Farallon Plate
Figure 4 | Palaeogeographic reconstruction showing hypothetical
dispersal route of Panamacebus across the CAS in the early Miocene.
Yellow and ochre colours indicate subaerial environments, blue colours
indicate marine environments (dark, coastal and platform; light, abyssal).
Criteria used to arrive at this reconstruction include regional tectonic
reconstructions, local and regional palaeomagnetic data, and regional
strain markers and piercing points (see Extended Data Fig. 8, Methods,
and Supplementary Methods). Fm., formation; Fms, formations.
© 2016 Macmillan Publishers Limited. All rights reserved

4 | NATURE | VOL 000 | 00 MONTH 2016
Letter
reSeArCH
share close affinities with those found in Panama
6–9
. Absence of early
Miocene New World monkeys at higher northern latitudes could be
explained by the limited extent of suitable tropical forest habitats, much
like today. However, their distribution in South America, including very
high latitudes of Patagonia, at localities of very similar age
25
, introduces
a potential paradox in which primates would be limited to tropical
forests in North but not South America26. A possible resolution is
found in the taxonomic composition and historical biogeography of
the forests themselves. Early Miocene forests of tropical South America
have a shared Gondwanan history with those at higher southern
latitudes, including Patagonia, and southern Central America (Costa
Rica and Panama), which are dominated by South American-derived
tropical rainforest taxa
27
. Northern tropical Central American forests,
however, have predominantly Laurasian affinities in the early Miocene
(Supplementary Tables 3 and 7 and Supplementary Information (results
section)). Thus, dispersal of New World monkeys further northward in
the early Miocene was probably limited more by their niche conserva-
tism and a boundary between forests with very different evolutionary
histories than by differences in climate or the existence of major geo-
graphical barriers. The New World tropics may have acted as a holding
pen for some tropical mammals for at least 12 Myr before xenarthrans
first appeared at higher latitudes of North America
12
, suggesting that
the distribution of South American-derived tropical forests played an
important part in the early dynamics of the GABI.
Online Content Methods, along with any additional Extended Data display items and
Source Data, are available in the online version of the paper; references unique to
these sections appear only in the online paper.
Received 4 November 2015; accepted 09 February 2016.
Published online 20 April 2016.
1. Kay, R. F. Biogeography in deep time—what do phylogenetics, geology, and
paleoclimate tell us about early platyrrhine evolution? Mol. Phylogenet. Evol.
82, 358–374 (2015).
2. Farris, D. W. et al. Fracturing of the Panamanian Isthmus during initial collision
with South America. Geology 39, 1007–1010 (2011).
3. Montes, C. et al. Evidence for middle Eocene and younger land emergence in
central Panama: implications for Isthmus closure. Geol. Soc. Am. Bull. 124,
780–799 (2012).
4. Bacon, C. D. et al. Biological evidence supports an early and complex
emergence of the Isthmus of Panama. Proc. Natl Acad. Sci. USA 112,
6110–6115 (2015).
5. MacFadden, B. J. Extinct mammalian biodiversity of the ancient New World
tropics. Trends Ecol. Evol. 21, 157–165 (2006).
6. Rincon, A. F., Bloch, J. I., Suarez, C., MacFadden, B. J. & Jaramillo, C. A. New
oridatragulines (Mammalia, Camelidae) from the early Miocene Las Cascadas
Formation, Panama. J. Vertebr. Paleontol. 32, 456–475 (2012).
7. Rincon, A. F., Bloch, J. I., Macfadden, B. J. & Jaramillo, C. A. First Central
American record of Anthracotheriidae (Mammalia, Bothriodontinae) from the
early Miocene of Panama. J. Vertebr. Paleontol. 33, 421–433 (2013).
8. Rincon, A. F., Bloch, J. I., Macfadden, B. J. & Jaramillo, C. A. New early Miocene
protoceratids (Mammalia, Artiodactyla) from Panama. J. Verteb. Paleontol.
http://dx.doi.org/10.1080/02724634.2015.970688 (2015).
9. Wood, A. R. & Ridgwell, N. M. The rst Central American chalicothere
(Mammalia, Perissodactyla) and the paleobiogeographic implications for
small-bodied schizotheriines. J. Vertebr. Paleontol. 35, e923893 (2015).
10. MacFadden, B. J. et al. Temporal calibration and biochronology of the
Centenario Fauna, early Miocene of Panama. J. Geol. 122, 113–135
(2014).
11. Kirby, M. X., Jones, D. S. & MacFadden, B. J. Lower Miocene stratigraphy along
the Panama Canal and its bearing on the Central American Peninsula. PLoS
One 3, e2791 (2008).
12. Woodburne, M. O. The Great American Biotic Interchange: dispersals,
tectonics, climate, sea level and holding pens. J. Mamm. Evol. 17, 245–264
(2010).
13. Albright, L. B., III et al. Revised chronostratigraphy and biostratigraphy of the
John Day Formation (Turtle Cove and Kimberly Members), Oregon, with
implications for updated calibration of the Arikareean North American land
mammal age. J. Geol. 116, 211–237 (2008).
14. Tedford, R. H. et al. in Late Cretaceous and Cenozoic Mammals of North
America: Biostratigraphy and Geochronology 169–231 (Columbia Univ. Press,
2004).
15. MacFadden, B. J. Chronology of Cenozoic primate localities in South America.
J. Hum. Evol. 19, 7–21 (1990).
16. Bond, M. et al. Eocene primates of South America and the African origins of
New World monkeys. Nature 520, 538–541 (2015).
17. Kay, R. F. New World monkey origins. Science 347, 1068–1069 (2015).
18. Jameson Kiesling, N. M., Yi, S. V., Xu, K., Gianluca Sperone, F. & Wildman,
D. E. The tempo and mode of New World monkey evolution and biogeography
in the context of phylogenomic analysis. Mol. Phylogenet. Evol. 82, 386–399
(2015).
19. Houle, A. The origin of platyrrhines: an evaluation of the Antarctic scenario and
the oating island model. Am. J. Phys. Anthropol. 109, 541–559 (1999).
20. Springer, M. S. et al. Macroevolutionary dynamics and historical biogeography
of primate diversication inferred from a species supermatrix. PLoS ONE 7,
e49521 (2012).
21. Perelman, P. et al. A molecular phylogeny of living primates. PLoS Genet. 7,
e1001342 (2011).
22. Cooke, S. B., Rosenberger, A. L. & Turvey, S. An extinct monkey from Haiti and
the origins of the Greater Antillean primates. Proc. Natl Acad. Sci. USA 108,
2699–2704 (2011).
23. Rosenberger, A. L. Platyrrhines, PAUP, parallelism, and the Long Lineage
Hypothesis: a reply to Kay et al. (2008). J. Hum. Evol. 59, 214–217 (2010).
24. MacPhee, R. D., Iturralde-Vinent, M. A. & Ganey, E. S. Domo de Zaza, an early
Miocene vertebrate locality in south-central Cuba, with notes on the tectonic
evolution of Puerto Rico and the Mona Passage. Am. Mus. Novit. 3394, 1–42
(2003).
25. Dunn, R. E. et al. A new chronology for middle Eocene–early Miocene
South American Land Mammal Ages. Geol. Soc. Am. Bull. 125, 539–555
(2013).
26. Rosenberger, A., Tejedor, M., Cooke, S. & Pekar, S. in South American Primates
Developments in Primatology: Progress and Prospects (eds Garber, P. A., Estrada,
A., Bicca-Marques, J. C., Heymann, E. W. & Strier, K. B.) Ch. 4, 69–113
(Springer, 2009).
27. Jaramillo, C. et al. Palynological Record of the Last 20 Million Years in Panama.
Ch. 8, 134–251 (Missouri Botanical Garden Press, 2014).
28. Rooney, T. O., Franceschi, P. & Hall, C. M. Water-saturated magmas in the
Panama Canal region: a precursor to adakite-like magma generation? Contrib.
Mineral. Petrol. 161, 373–388 (2011).
Supplementary Information is available in the online version of the paper.
Acknowledgements We thank M. Silcox, D. Boyer, G. Gunnell, S. Chester,
P. Morse, E. Sargis, D. Steadman, E. Kowalski, Z. Randall, A. Rosenberger,
J. Krigbaum and D. Daegling for advice and discussion, R. Kay and L. Marivaux
for reviews that significantly improved the quality of the manuscript, D. Byerley
and G. Maccracken for finding primate fossils in Panama, M. Droulliard for
assistance with geochronology laboratory and fieldwork, R. Kay, P. Holroyd and
D. Reed for access to comparative specimens, J. Bourque for fossil preparation,
D. Byerley for artwork associated with Fig. 3, D. Boyer for facilitating access
to the Duke University SMIF microCT facility, and D. Boyer, G. Yapuncich and
J. Thostenson for help acquiring and processing microCT scan data (funded in
part by National Science Foundation (NSF) BCS 1304045 to D. Boyer and E. St
Clair, and BCS 0851272 to R. Kay). We thank O. Moskalenko, M. Gitzendanner
and D. Reed for assistance with the high-performance computing resources at
the University of Florida. We thank the Autoridad del Canal de Panama (ACP)
and the Ministerio de Comercio e Industria (MICI) for logistical support and
access to the Panama Canal Zone. Part of this manuscript was written when
J.I.B. was supported as an Edward P. Bass Distinguished Visiting Environmental
Scholar in the Yale Institute for Biospheric Studies (YIBS). The NSF (PIRE project
0966884), Smithsonian Tropical Research Institute Paleobiology Fund, and the
Florida Museum of Natural History funded this research. This is University of
Florida Contribution to Paleobiology 782.
Author Contributions J.I.B., A.R.W., E.D.W. and G.S.M. contributed to project
planning. J.I.B., A.R.W. and A.R.H. contributed to systematic palaeontology
and microCT scans. D.A.F. and A.F.R. contributed to radioisotopic analyses
and stratigraphy. B.J.M., A.F.R., G.S.M., A.R.W. and J.I.B. contributed to
biochronological analysis. E.D.W., J.I.B. and A.R.W. contributed to phylogenetic
analyses. E.D.W. performed divergence dating analyses. C.M. and C.A.J.
contributed to palaeogeographic analysis. N.A.J., J.I.B. and C.A.J. contributed to
the pollen summary. A.R.W., A.F.R., J.I.B., G.S.M., E.D.W. and D.S.J. contributed to
fieldwork. J.I.B., B.J.M., G.S.M., C.A.J. and D.S.J. contributed to financial support.
All authors contributed to manuscript and figure preparation.
Author Information The LSIDs for Panamacebus (genus), urn:lsid:zoobank.
org:act:C33F8967-EE79-47B8-8A98-2C6D2C7557CA, and for Panamacebus
transitus (species), urn:lsid:zoobank.org:act:3E01F3F2-B1F8-433E-B110-
9AD5AAF3DB99, have been deposited in ZooBank. Reprints and permissions
information is available at www.nature.com/reprints. The authors declare no
competing financial interests. Readers are welcome to comment on the online
version of the paper. Correspondence and requests for materials should be
addressed to J.I.B. (jbloch@flmnh.ufl.edu).
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
METHODS
Geochronology. Sample MD11 was collected from an ash layer that forms the
upper-most part of a several-metre thick, welded andesitic tuff. The tuff is exposed
within the conformable section of the Las Cascadas Formation approximately
0.25 m below the mammal fossil-bearing unit (Fig. 1). Zircons were separated
from two 10-kg sample aggregates using conventional magnetic and density con-
centration methods. Clear to light pink, euhedral zircons about 100 μ m in length
were mounted in epoxy and ground to reveal internal surfaces.
U–Pb isotopic analyses were conducted on 20-μ m spots on zircon grains
using a Nu-Plasma multicollector plasma source mass spectrometer coupled to
a New Wave 213-nm laser (LA-MC-ICP-MS). The data were acquired using the
Nu-Instruments Time Resolved Analysis software. Data calibration and drift
were based on multiple analyses of the reference zircon FC-1 (ref. 29) repeated
between every 10 ablations of the unknowns. Concordant U–Pb data were yielded
by 37 zircons (Supplementary Table 1). A weighted mean of the
206
Pb/
238
U ages
gives 20.93 ± 0.17 Ma (2σ) (Fig. 1b), which we interpret to be the eruption age
of the andesitic tuff and the absolute age of the Las Cascadas mammal fossil
horizon (Supplementary Information (results section)). These data are consistent
with a
206
Pb/
238
U age of 19.3 ± 0.4 Ma for a welded tuff unit within the younger
Culebra Formation30.
Three-dimensional data acquisition. Fossil platyrrhine specimens (UF 267048,
28001, 280127-280131; UCMP 38762, 38989, 39205) and casts (IGM-KU 89008,
89011, 89018, 89019, 89021, 89086, 89092, 89104, 90016) were scanned at the
Duke University Shared Materials Instrumentation facility in Durham, North
Carolina, using a Nikon XTH 225 ST MicroCT scanner (Supplementary Table 4).
The resulting scans were reconstructed into tiff stacks and were imported into
Avizo 7 (Visualization Sciences Group) for surface visualization (Fig. 2, Extended
Data Figs 4–7 and Supplementary Figs 2, 3) and manipulation. Data were distrib-
uted and shared using Duke University’s three-dimensional data archive (http://
www.morphosource.org). DOI links to three-dimensional data are provided in
the figure captions.
Comparative morphology. Comparisons were done within the morphological
framework outlined and discussed previously1, with reference to the literature
for fossil taxa as discussed in Supplementary Methods (see also table 1 in ref. 1
and references therein). Additional comparisons were made to specimens and
casts of fossil and extant platyrrhines. Three-dimensional images of relevant fossil
platyrrhines were reconstructed from microCT scans (see earlier).
Phylogenetic analysis of morphological data. We conducted three phyloge-
netic analyses using maximum parsimony to examine the phylogenetic position
of Panamacebus. The first analysis was conducted using the same parameters
as the original analysis1 with 177 ordered characters and all characters given a
weight other than one (Supplementary Data 1). In the second analysis we used
a different constraint tree derived from a recent molecular supermatrix20,21
(re-analyses of these data used for the constraint tree are described in
Supplementary Methods). The third analysis was performed without enforcing
a constraint tree. Support for the topology of the resulting trees was determined
by bootstrapping (see Supplementary Methods for detailed descriptions of
phylogenetic analyses).
Divergence dating recalibration analyses. Divergence dating analyses were
conducted to determine the effect of a new minimum age for the divergence of
Callitrichinae and Cebinae based on the radioisotopic date obtained for the horizon
where the Panamacebus fossils were discovered. We took a conservative approach
and calibrated the divergence of Cebinae and Callitrichinae as opposed to cal-
ibrating a node within Cebinae as suggested by the placement of Panamacebus
in the phylogeny using the first molecular constraint (Fig. 3a). Using a reduced
molecular data set derived from ref. 20, we conducted two separate Bayesian
Markov chain Monte Carlo (MCMC) analyses (see Supplementary Methods and
Supplementary Data 3). Our analyses employed a lognormal relaxed molecular
clock and a general time reversible (GTR) model with a gamma distribution, using
four rate categories and estimated base frequencies. We set 15 calibration points
(Supplementary Table 2 and Supplementary Data 4); 14 nodes were calibrated as
described previously
20
and a 15th node was calibrated to the date corresponding
to the age of Panamacebus (20.93 ± 0.17 Ma). A random starting tree was used
and analyses were conducted with an MCMC chain length of 200,000,000 states
sampled every 10,000 states (for detailed description of divergence dating analyses
see Supplementary Methods).
Palaeogeographic reconstruction. Figure 4 and Extended Data Fig. 8 show a
reconstruction built using fault-bound tectonic blocks restored to a late Oligocene–
early Miocene palinspastic palaeogeographic position. Criteria used to arrive at
this reconstruction include: (1) regional tectonic reconstructions; (2) local and
regional palaeomagnetic data; and (3) regional strain markers and piercing points.
Published stratigraphic columns with age constraints spanning the late Oligocene–
early Miocene were placed in their corresponding palaeogeographic position. See
Supplementary Methods for detailed methods and references.
Palaeobotanical analysis of Miocene forests. We compiled pollen and spore
occurrence data from nine different formations of Oligocene to middle Miocene age
(Supplementary Table 3). The samples come from Florida, Puerto Rico, southern
Mexico, Costa Rica and Panama; the number of samples per formation ranges from
2 to 52 (Supplementary Table 7). We assigned the plant families and genera to
coarse biogeographic categories on the basis of modern species distributions and
fossil evidence. The categories are Gondwana–Amazonian, Gondwana– northern
Andean, Gondwana–southern Andean, Laurasian, or unassigned. Next, we
calculated the percentage contribution of the biogeographic regions to the species
(morphotype) richness in each fossil flora. Taxa that were not assigned to a
biogeographic region were excluded from calculations of biogeographic affinity
as described previously27. See Supplementary Methods for detailed references and
discussion.
29. Black, L. P. et al. The application of SHRIMP to Phanerozoic geochronology; a
critical appraisal of four zircon standards. Chem. Geol. 200, 171–188 (2003).
30. Montes, C. et al. Arc-continent collision and orocline formation: closing of the
Central American seaway. J. Geophys. Res. 117, B4 (2012).
© 2016 Macmillan Publishers Limited. All rights reserved

Letter
reSeArCH
Extended Data Figure 1 | Map of the southern part of Panama Canal (Gaillard Cut). Black circles mark the position of the Lirio Norte Local Fauna
locality (YPA024) and other specific terrestrial vertebrate collecting sites (Centenario Fauna) in the area (modified from ref. 10).
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
Extended Data Figure 2 | Photograph of the northern wall (south-facing) in the Lirio Norte Local Fauna locality (YPA024). Dated rock sample
MD11 was collected from an ash layer that forms the upper-most part of a several-metre thick, welded andesitic tuff. The tuff is exposed within the
conformable section of the Las Cascadas Formation approximately 0.25 m below, and in close proximity to, the mammal fossil-bearing unit that includes
the Panamacebus fossils.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter
reSeArCH
Extended Data Figure 3 | Generic-level biostratigraphy of selected taxa from the early Miocene (~21 Ma) Las Cascadas Formation, Panama.
Individual biochronologies were interpreted from temporal ranges known from higher-latitudes. The calibration of the NALMA boundaries is as
described previously13,14.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
Extended Data Figure 4 | Detailed upper molar comparisons
of Panamacebus to Neosaimiri. a–d, Left to right, occlusal,
distal, lingual, mesial and buccal views of the left M1 (UF 280128:
http://dx.doi.org/10.17602/M2/M8531) (a) and the left M2 (UF 281001:
http://dx.doi.org/10.17602/M2/M8550) (d) of P. transitus compared with
M1 and M2 of N. fieldsi (b, c, e–g): right M1 (IGM-KU 89008:
http://dx.doi.org/10.17602/M2/M8541) (b), right M1 (IGM 89019:
http://dx.doi.org/10.17602/M2/M8544) (c), right M2
(IGM-KU 89018: http://dx.doi.org/10.17602/M2/M8543) (e),
right M2 (IGM-KU 89104: http://dx.doi.org/10.17602/M2/M8548) (f),
and right M2 (IGM-KU 89011: http://dx.doi.org/10.17602/M2/M8542)
(g). Images were generated from microCT scan data (see Methods and
links associated with specimen numbers). Right-sided teeth were flipped
to facilitate comparison with left-sided teeth. Scalebar, 1 mm.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter
reSeArCH
Extended Data Figure 5 | Detailed upper molar comparisons of
Panamacebus to Neosaimiri and Cebus. a–g, Left to right, occlusal, distal,
lingual, mesial and buccal views of the left M1 (UF 280128: http://dx.doi.
org/10.17602/M2/M8531) (a) and the left M2 (UF 281001: http://dx.doi.
org/10.17602/M2/M8550) (c) of P. transitus compared with M1 and M2
of N. fieldsi (b, d): right M1 (IGM 89019: http://dx.doi.org/10.17602/M2/
M8544) (b), and right M2 (IGM-KU 89104: http://dx.doi.org/10.17602/
M2/M8548) (d); and the left M1 and M2 of Cebus capucinus (USNM
291128; http://dx.doi.org/10.17602/M2/M8627) in occlusal (e), lingual
(f) and buccal views (g). Images were generated from microCT scan data
(see Methods and links associated with specimen numbers). Right-sided
teeth were flipped to facilitate comparison with left-sided teeth. Scale
bar,1 mm.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
Extended Data Figure 6 | Detailed comparisons of Panamacebus
lower teeth to those of Neosaimiri and Stir tonia. a–k, Left to right,
occlusal, distal, lingual, mesial and buccal views of the partial left I1
(UF 280130: http://dx.doi.org/10.17602/M2/M8400) (a), right I2 (UF
267048, mirrored: http://dx.doi.org/10.17602/M2/M8395) (d), left P2
(UF 280127: http://dx.doi.org/10.17602/M2/M8397) (g), left P4
(UF 280129: http://dx.doi.org/10.17602/M2/M8399) (i) and right
C1 (UF 280131, mirrored: http://dx.doi.org/10.17602/M2/M8401)
(k) of P. transitus compared with the right I1 (IGM-KU 89086:
http://dx.doi.org/10.17602/M2/M8546) (b), right I2 (IGM-KU 89092:
http://dx.doi.org/10.17602/M2/M8547) (e), left I2 (UCMP 39205: http://
dx.doi.org/10.17602/M2/M1880) (f), right P2 (IGM-KU 90016:
http://dx.doi.org/10.17602/M2/M8549) (h) and right C1 (IGM-KU 89021:
http://dx.doi.org/10.17602/M2/M8432) (j) of N. fieldsi and the right
I2 (UCMP 38989: http://dx.doi.org/10.17602/M2/M1799) of Stirtonia
tatacoensis (c). Images were generated from microCT scan data
(see Methods and links associated with specimen numbers). Right-sided
teeth were flipped to facilitate comparison with left-sided teeth. Scale
bar,5 mm.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter
reSeArCH
Extended Data Figure 7 | Detailed comparisons of Panamacebus lower
teeth to those of Cebus. a–i, Occlusal (a), lingual (c) and buccal (e) views
of P. transitus partial left I1 (UF 280130: http://dx.doi.org/10.17602/M2/
M8400), right I2 (UF 267048, mirrored: http://dx.doi.org/10.17602/M2/
M8395), right C1 (UF 280131, mirrored: http://dx.doi.org/10.17602/M2/
M8401), left P2 (UF 280127: http://dx.doi.org/10.17602/M2/M8397) and
left P4 (UF 280129: http://dx.doi.org/10.17602/M2/M8399); occlusal (b),
lingual (d) and buccal (f) views of C. capucinus right P2–P4 (USNM
291128: http://dx.doi.org/10.17602/M2/M8629); and occlusal (g), lingual
(h) and buccal (i) views of C. capucinus left dentary with I1–M3 (USNM
291236: http://dx.doi.org/10.17602/M2/M8622). Right-sided teeth
were mirrored to facilitate comparison with left-sided teeth. Images
were generated from microCT scan data (see Methods). Scale bars,
5mm.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
Extended Data Figure 8 | Detailed palaeogeographic reconstruction
of the Panama Canal Basin region during the late Oligocene–early
Miocene, showing the location of key geological formations, faults
and tectonic blocks. Sedimentary environments were extrapolated from
published stratigraphic sections that were placed over the palinspastic
reconstruction in the following locations: (1) Pacific sections that
include conglomerate and sandy strata; (2) upper Magdalena Basin;
(3) Amaga Formation coal-bearing and sandy-conglomeratic; (4) Choco;
(5) easternmost Panama; 96) western Panama; (7) Panama; (8) Canal basin;
(9) northwestern Colombia; (10) Sierra Nevada Santa Marta, unnamed
sandy and conglomeratic strata; (11) Guajira Peninsula; (12) Falcon;
(13) Falcon/Lara; (14) middle Magdalena Basin; (15) Floresta Massif;
(16) axial Cordillera Oriental; (17) foothills; (18) southern middle
Magdalena Basin. See Supplementary Methods for references and detailed
discussion.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter
reSeArCH
Extended Data Figure 9 | Maximum clade credibility tree summarizing the concatenated trees from the divergence dating analysis using the birth–
death model showing the Platyrrhini. Ninety-five per cent highest posterior density intervals are shown on key nodes and the corresponding branches
are labelled with posterior probability values. For clarity, only the Platyrrhini are shown here.
© 2016 Macmillan Publishers Limited. All rights reserved

Letter reSeArCH
Extended Data Figure 10 | Maximum clade credibility tree summarizing the concatenated trees from divergence dating the analysis using the Yule
model showing the Platyrrhini. Ninety-five per cent highest posterior density intervals are shown on key nodes and the corresponding branches are
labelled with posterior probability values. For clarity, only the Platyrrhini are shown here.
© 2016 Macmillan Publishers Limited. All rights reserved