Journal of Biogeography (J. Biogeogr.

) (2016) 43, 1645–1655

ORIGINAL Geographical distribution patterns of

ARTICLE Carcharocles megalodon over time
reveal clues about extinction
mechanisms
Catalina Pimiento1,2,3* Bruce J. MacFadden2, Christopher F. Clements4,
Sara Varela5,6, Carlos Jaramillo3, Jorge Velez-Juarbe7 and Brian R. Silliman8

1
Paleontological Institute and Museum, ABSTRACT
University of Zurich, CH-8006 Zurich,
Aim Given its catastrophic consequences, the extinction of apex predators has
Switzerland, 2Florida Museum of Natural
History, University of Florida, Gainesville, FL
long been of interest to modern ecology. Despite major declines, no present-
32611, USA, 3Smithsonian Tropical Research day species of marine apex predator has yet become extinct. Because of their
Institute, Balboa, Panama, 4Department of vulnerability, understanding the mechanisms leading to their extinction in the
Evolutionary Biology and Environmental past could provide insight into the natural factors that interact with human
Studies, University of Zurich, Zurich CH-8057, threats to drive their loss. We studied the geographical distribution patterns of
Switzerland, 5Departamento de Ciencias de la the extinct macro-predatory shark Carcharocles megalodon in order to elucidate
Vida, Edificio de Ciencias, Campus Externo, its pathway to extinction.
Universidad de Alcal
a, 28805 Alcal
a de
Location World-wide from the Miocene to the Pliocene (c. 23–2.6 Ma).
Henares Madrid, Spain, 6Museum f€ ur
Naturkunde, Leibniz Institute for Evolution Methods A meta-analysis of C. megalodon occurrence records was performed
and Biodiversity Science, 10115 Berlin, using the Paleobiology Database as a platform. The data were binned into geo-
Germany, 7Department of Mammalogy, logical time slices, and the circular home range around each data point was
Natural History Museum of Los Angeles mapped in reconstructions made in GPlates. We then quantitatively assessed
County, Los Angeles, CA 90007, USA, the species’ geographical range and global abundance over time, and the rela-
8
Division of Marine Science and Conservation, tionship between distribution and climate.
Duke University, Beaufort, NC 28516, USA
Results The pathway to extinction of C. megalodon probably started in the
late Miocene with a decrease in its global abundance. This decrease was then
followed by a decline in its geographical range during the Pliocene. Although
the extinction of C. megalodon has been attributed to climate change, we found
no evidence of direct effects of global temperature. Instead, we found that the
collapse in geographical distribution coincided mainly with a drop in the diver-
sity of filter-feeding whales and the appearance of new competitors (large
predatory whales and the great white shark).
Main conclusions This research represents the first study of the distributional
trends of an extinct, cosmopolitan apex predator in deep-time. Our results sug-
gest that biotic factors, and not direct temperature limitations, were probably
the primary drivers of the extinction of the largest marine apex predators that
*Correspondence: Catalina Pimiento, ever lived.
Paleontological Institute and Museum,
University of Zurich, Karl Schmid-Strasse 4, Keywords
CH-8006 Zurich, Switzerland.
E-mail: catalina.pimientohernandez@pim.
apex predator, area coverage, extinction drivers, Miocene, Pliocene, sharks,
uzh.ch species occupancy

(Myers et al., 2007; Estes et al., 2011; DeLong et al., 2015).

INTRODUCTION
For example, it has been shown that the removal of sharks
Apex predators are ecologically important because they considerably degrades marine ecosystems (Myers et al., 2007;
stabilize the populations of their prey (Estes et al., 2011). Ferretti et al., 2010). This is a concern because large sharks
Accordingly, their local elimination or global extinction can are declining significantly throughout the world’s oceans
trigger large shifts in ecosystem structure and function (Dulvy et al., 2014).

ª 2016 John Wiley & Sons Ltd http://wileyonlinelibrary.com/journal/jbi 1645

doi:10.1111/jbi.12754
C. Pimiento et al.

Given their importance and vulnerability, understanding carried out, but only for higher taxonomic levels (e.g. the
the extinction of apex predators is of interest to modern generic level in Finnegan et al., 2015). Because C. megalodon
ecology. Despite major shifts in baseline abundances (Harnik is a relatively long-lived species with a widely distributed fos-
et al., 2012; McCauley et al., 2015), so far no species of sil record, it provides the opportunity to study the geograph-
modern marine apex predator has become extinct. However, ical distribution patterns of a predator at the species level in
because apex predators are vulnerable to extinction today, deep-time.
understanding the mechanisms that led to their extinction in We assessed the pathway to extinction of C. megalodon by
the past could provide insight into the natural factors that studying its geographical distribution over geological time.
interact with human threats to drive their loss. Analysis of the results provided with insights into the mech-
Carcharocles megalodon (megalodon) is the largest marine anisms of extinction in light of the hypotheses that have
apex predator to have ever lived. It reached a maximum been proposed (see above), which we categorized into two
length of 18 m and ranged world-wide during the Miocene groups, i.e. biotic and climatic. Accordingly, we assessed
and Pliocene (Gottfried et al., 1996; Purdy, 1996; Pimiento C. megalodon’s geographical distribution trends over time
& Balk, 2015). Recent studies suggest that it became extinct and how they related to biotic and climatic events. The dis-
c. 2.6 Ma, at the Pliocene/Pleistocene boundary (Pimiento & tributional patterns of a cosmopolitan apex predator species
Clements, 2014). Bite marks on fossil bones and associated have never been studied before over a geological time-scale.
fauna suggest that this species preyed on marine mammals Hence, this research provides a deep-time perspective on the
(Spizuco et al., 1993; Morgan, 1994; Gottfried et al., 1996; extinction of apex predators. In a broader context, our study
Purdy, 1996; Aguilera & De Aguilera, 2004; Aguilera et al., has the potential to inform conservation efforts to address
2008). Based on its trophic level, cosmopolitan distribution the current global extinction crisis of many species of large
and the impact that modern sharks have on prey populations sharks.
and ecosystem structure (Myers et al., 2007; Estes et al.,
2011), it can be inferred that the extinction of C. megalodon
MATERIALS AND METHODS
had a major impact on ancient marine communities
(Pimiento & Clements, 2014).
Data collection
It has been hypothesized that the extinction of C. megalodon
was caused by: (1) a decline in diversity of filter-feeding whales All published records of C. megalodon were collected using
and pinnipeds; (2) competition with large predatory whales the Paleobiology Database (PaleoBioDB, https://paleo
(e.g. odontocetes, killer whales); (3) climate change (mainly biodb.org; parameters: species name = Carcharocles megalodon,
cooling); or (4) a shift in the distribution of large marine last search date June 2013). The PaleoBioDB takes all
mammal prey species towards higher latitudes (Morgan, 1994; synonyms into consideration when searching for a taxon. For
Allmon et al., 1996; Gottfried et al., 1996; Allmon, 2001; a discussion of our taxonomic assignment of the species
Lindberg & Pyenson, 2006). The first two hypotheses imply megalodon to the genus Carcharocles, see Pimiento et al.
that biotic factors such as competition and prey availability (2010), Ehret et al. (2012), Pimiento et al. (2013a,b) and
were major extinction drivers, whereas the second two Pimiento & Balk (2015).
hypotheses imply that C. megalodon was physiologically We searched for additional C. megalodon reports using the
constrained to warmer temperatures. These hypotheses remain ISI Web of Science (http://webofknowledge.com), GeoRef
untested because the extinction pathway of this apex predator (http://geoscienceworld.org), Google Scholar (http://scholar.-
has not been studied. google.com) and Shark-References (http://shark-references.-
Geographical distribution correlates strongly with extinc- com) with the search terms: Megalodon OR Carcharodon
tion susceptibility in a wide variety of modern and fossil taxa AND megalodon OR Carcharocles AND megalodon. Any new
(Jablonski, 1986; Jablonski & Flessa, 1986; McKinney, 1997; data were catalogued as occurrences in PaleoBioDB (Archive
Purvis et al., 2000; Kiessling & Aberhan, 2007; Payne & Fin- #20: http://goo.gl/PpIh0G). In total, the data archive com-
negan, 2007; Brook et al., 2008; Davidson et al., 2009, 2012). prised 284 records consisting of 82 pre-existing records and
Geographical distribution consists of two closely related vari- 124 new records.
ables: geographical range and abundance (Kiessling & Aber- Each of our data points represented a fossil collection (i.e.
han, 2007; Lyons & Smith, 2010). Therefore, studying a set of fossil occurrences that were co-located geographically
species’ geographical range and abundance patterns over geo- and temporally) and were mostly represented by isolated
logical time could further our understanding of extinction teeth. Each collection was treated as an occurrence and was
pathways (e.g. Payne & Finnegan, 2007). linked to one or more supporting references. Although C.
The fossil record of extinct terrestrial predators (e.g. dino- megalodon teeth are typically large and easily recovered
saurs) is particularly scarce (Wang & Dodson, 2006; Barrett from their respective faunas, we also collected records of
et al., 2009; O’Gorman & Hone, 2013), making it difficult to C. megalodon absences (i.e. all other records from the
assess their distributional patterns, and hence pathways of Miocene and Pliocene that reported shark assemblages but
extinction of a single species. Research into the geographical not C. megalodon). We used the absences in our analysis of
distribution of marine predators and extinction risk has been occupancy (see below).

1646 Journal of Biogeography 43, 1645–1655

ª 2016 John Wiley & Sons Ltd
 Megalodon’s pathway to extinction

Miocene to the Pliocene (23.03–2.57 Ma), following the

Data filtering
geological time-scale of Gradstein et al. (2012). We did not
We performed a standardized evaluation following the strati- subdivide Pliocene into early and late in order to maintain a
graphic assessment process described in Pimiento & Cle- relatively equal time span for each bin (i.e. early Miocene,
ments (2014) to determine which occurrence records to middle Miocene, late Miocene and Pliocene). We then
include in our analyses. When necessary, and taking into constructed C. megalodon occurrence maps for each time
consideration the diagnostic characteristics of the dentition bin following the methods of Jaramillo et al. (2013) and the
of the species (e.g. Pimiento et al., 2013a), we performed a tectonic reconstructions provided by GPlates (Boyden et al.,
taxonomic assessment to verify that the specimens reported 2011).
were indeed C. megalodon. This was done by examining the We placed the occurrences on the maps by creating a cir-
morphology of the fossil specimens. To do so, we either vis- cular home range area around each point using ‘gBuffer’ in
ited the museum collections housing the specimens, or we the R package rgeos. Each circular area had a diameter of
examined available published images. This was particularly 1000 km2, based on the maximum longitudinal movement
useful when assessing records from the early Miocene, a per- reached by the great white shark (Carcharodon carcharias), as
iod in which Carcharocles chubutensis, an ancestor of C. reported in Domeier & Nasby-Lucas (2008). Given the differ-
megalodon, occurs (Applegate & Espinosa-Arrubarrena, 1996; ences in body size between these two species (maximum total
Ward & Bonavia, 2001; Pimiento et al., 2013b; Pimiento & length = c. 6 m in C. carcharias versus c. 18 m in C. mega-
Balk, 2015). We distinguished the two species by the absence lodon; Fergusson et al., 2009; Pimiento & Balk, 2015), our
of lateral cusplets in the teeth of adults of C. megalodon proxy of the home range of C. megalodon represents a con-
(Pimiento et al., 2013b). As it is not possible to assess onto- servative estimate. Despite the differences in size (and possi-
geny based merely on specimen examination, we referred bly physiology), C. carcharias is widely considered to be a
those specimens showing a clear absence of lateral cusplets to modern ecological analogue of C. megalodon. This analogy is
C. megalodon, which were therefore only adults. This had the based on similarities in trophic level (and hence ecological
potential to reduce our sample size, but in turn it provided a function), overall tooth and vertebral centrum morphology
robust taxonomic assessment for this time period (see Dis- and taxonomic proximity (Applegate & Espinosa-Arrubar-
cussion). We did not include any Pleistocene records in our rena, 1996; Purdy et al., 2001; Pimiento et al., 2010; Ehret
analyses because: (1) Pleistocene records of C. megalodon et al., 2012; Pimiento & Balk, 2015), and it is therefore a
have been regarded as unconfirmed or unreliable (Applegate useful model for studying the extinct species (e.g. Gottfried
& Espinosa-Arrubarrena, 1996), and (2) a recent analyses et al., 1996; Wroe et al., 2008; Pimiento et al., 2010;
concluded that C. megalodon was probably extinct by this Pimiento & Balk, 2015).
time period (Pimiento & Clements, 2014). In order to qualitatively assess C. megalodon’s pathway
As a result of our age evaluation process, we selected the to extinction, we calculated the global area coverage, occu-
206 records that were deemed to be reliable (see Table S1 in pancy range and various latitude metrics per time bin. In
Appendix S1 in the Supporting Information) and disregarded addition, we built an ecological niche model with a subset
the records that did not meet the evaluation criteria (28 of our samples. The global area coverage was the total
records; see Table S2). Additional information can be found area (km2) of the home range areas per time bin (calcu-
in Appendix S2, where we describe the rationale behind the lated using ‘gArea’ from the rgeos package), which was
evaluation of each record. More detailed information and used as a proxy for geographical range (Kiessling & Aber-
supporting references can be accessed by searching for the han, 2007). We only considered marine areas (i.e. we did
PaleoBioDB number for each record on the PaleoBioDB not include any area projected to be on land), and over-
website: https://paleobiodb.org. lapping home areas were combined and only included in
the area calculation once. Furthermore, given the differ-
ences in sample size for each time period, we standardized
Analyses
our bins, bootstrapping the data 1000 times and resam-
All analyses were performed in R (R Core Team, 2015). To pling the bins based on the lowest sample size (i.e. the
visualize all global occurrences of C. megalodon, a map with early Miocene had 23 geographically unique records,
the fossil records was constructed following the methods whereas the middle Miocene had 75, the late Miocene had
described in Jaramillo et al. (2013). We included all known 65 and the Pliocene had 43).
records, distinguishing between those with adequate strati- Occupancy was determined as the ratio of the number of
graphic and taxonomic information, and those without such localities where a species was found to the total number of
information (see Appendix S2). localities sampled, i.e. occurrences/(occurrences + absences)
Using only the reliable records (see Table S1), we resam- and was used as a proxy of global abundance (Lyons &
pled the age of each occurrence 1000 times, bootstrapping Smith, 2010). This metric standardizes the occurrence of
the timing of each record from a uniform distribution each time bin in relation to the total number of samples
between its upper and lower age. We then binned the available. For latitude, we calculated the total range, maxi-
resampled data into geological periods, from the early mum, minimum, mean and modal latitude per time bin.

Journal of Biogeography 43, 1645–1655 1647

ª 2016 John Wiley & Sons Ltd
C. Pimiento et al.

In order to explore further the relationship between geo- Caribbean and western Atlantic (Antigua and Barbuda,
graphical distribution and climate, we carried out two analy- Barbados, Cuba, Grenada, Jamaica and Puerto Rico), and
ses. For the first, we created time bins using the methods Europe (Austria, Belgium, Cyprus, Czech Republic, Den-
described above, but for every 2 Myr and latitudinal bins mark, France, Germany, Greece, Italy, Malta, the Nether-
every 15°. We then calculated the occupancy range per time lands, Turkey, Poland, Portugal, Slovakia and Spain)
and latitudinal bin, and generated a raster image (using (Fig. 1). Fossil occurrences were also located in Asia (Japan,
‘geom_raster’ from the ggplot2 package). We used occupancy India and Myanmar), Africa (Angola, Libya, Nigeria and
rather than occurrence to take into account the difference in South Africa) and Oceania (Australia, Fiji and New Zealand).
total number of sites (= samples) per bin. For the second Reports with inadequate or insufficient stratigraphical, geo-
analysis, we developed an ecological niche model using a graphical or taxonomic information included occurrences
subsample from the Pliocene. Our subsample consisted of from Austria, Colombia, Germany, India, Italy, Japan, Malta,
the 18 occurrences (marked with an asterisk in Table S1) Mexico, UK, USA (California and South Carolina), South
that overlapped with the climatic global circulation model Africa, Spain and Venezuela. Areas with a lack of occurrences
CCSM4 (community climate system model; Peter et al., included Brazil, the Pacific coast of northern Central Amer-
2011) and represented a solid estimate of the distribution of ica, the northern part of the Indian Ocean, the Arctic Ocean
the species because it covered its entire latitudinal range. We and the Southern Ocean (Fig. 1).
built the model for the Pliocene using the climatic envelope Our data-quality screening process and time-binning
‘Bioclim’ (Busby, 1991), then projected the model into the analysis revealed that the oldest reliable records of C. mega-
Last Glacial Maximum, Holocene and Recent to test whether lodon were early Miocene in age (see Appendix S2). For
the climatic changes during the glacial–interglacial periods this time period, fossils of the species were mostly dis-
would have been related to changes in the distribution of the tributed in the Northern Hemisphere, with the majority of
species. For a detailed methodology see Appendix S2. the fossil occurrences being in the Caribbean, western
Atlantic and Mediterranean Sea, and a few in the central
Paratethys (Austria), eastern Atlantic (Portugal), Indian
RESULTS
Ocean (India) and USA (California). Southern Hemisphere
Carcharocles megalodon was distributed globally, with occur- records only included a few occurrences in Australia
rences in the Pacific, Atlantic and Indian oceans. The species (Fig. 2d; see Appendix S2). For the middle Miocene,
had a geographical range from 55.28° N to 43.99° S and records of C. megalodon were present in the major ocean
from 178.32° E to 122.35° W and it was found mostly in the basins of the world, occurring widely in the Neotropics
Americas (Argentina, Chile, Colombia, Costa Rica, Ecuador, (the Caribbean and central Eastern Pacific), the western
Mexico, Panama, Peru, USA, Venezuela and Uruguay), the Atlantic (New Jersey and Maryland) and Europe (eastern

60

30

−30

−60

−140 −100 −60 −20 20 60 100 140 180

Figure 1 Carcharocles megalodon fossil occurrences. Blue dots represent fossil occurrences of the species, yellow dots represent dubious
occurrences (i.e. occurrences without adequate associated stratigraphic, taxonomic or geographical information).

1648 Journal of Biogeography 43, 1645–1655

ª 2016 John Wiley & Sons Ltd
 Megalodon’s pathway to extinction

(a) Geographical analyses indicated that during the early and

middle Miocene C. megalodon presented its narrowest cover-
age (early Miocene median = 20.05 million km2, middle
50

Miocene median = 19.22 million km2; calculated from 1000

randomizations). During the late Miocene it reached its
0

broadest coverage (median = 24.94 million km2), to then

decline in the Pliocene (median = 21.91 million km2)
50

(Fig. 3, blue line). Analyses of relative abundance showed

Pliocene
that the minimum occupancy range occurred during the
(b) early Miocene (25%), and the maximum during the middle
Miocene (47%). During the late Miocene and Pliocene, the
occupancy of C. megalodon decreased (36% and 29% respec-
50

tively) (Fig. 3, green line). Similarly, latitudinal range estima-

tions showed maximum values during the middle and late
0

Miocene, a tropical modal value in the late Miocene, and

mostly tropical mean values overall (Table 1).
50

We found no latitudinal gradient, and no evidence of cor-

Late Miocene
relation between occupancy values per latitude and tempera-
(c) ture trends over time (Fig. 4). Major events in global climate
during the Miocene and Pliocene, such as the high tempera-
ture values reached during the Mid-Miocene Climatic Opti-
50

mum (MMCO; 17–15 Ma) and the subsequent cooling as a

result of the re-establishment of a major ice sheet on Antarc-
0

tica (c. 10 Ma) (Zachos et al., 2001), did not correspond

with either an increase or decrease in C. megalodon latitudi-
50

nal abundance (Fig. 4). Finally, our ecological niche models

Middle Miocene
based on Pliocene records indicated that C. megalodon
(d) inhabited oceans with a mean annual temperature from 12
50
0
50

Early Miocene

−150 −100 −50 0 50 100 150 180

Figure 2 Carcharocles megalodon buffer circles area over

geological time. (a) Pliocene, (b) late Miocene, (c) middle
Miocene, (d) early Miocene.

Atlantic and Mediterranean–central Paratethys). In the Figure 3 Carcharocles megalodon geographical distribution over
Indo-Pacific Ocean, C. megalodon occurred in Australia and time. Blue = global area coverage (a proxy of geographical
India, and in the western Pacific it occurred in Japan range). The box plot shows the interquartile range of the 1000
(Fig. 2c). For the late Miocene, records of C. megalodon randomizations. The line connects the median value of each
were distributed along the coasts of the Americas (from period. Green = species occupancy (a proxy of global
abundance).
California to Chile in the Pacific, and from Florida to
Argentina in the Atlantic) and Europe, with some occur-
rences in southern Africa, New Zealand and Japan
(Fig. 2b). For the Pliocene, the fossil record of C. mega- Table 1 Carcharocles megalodon latitudinal metrics over time.
lodon had a scattered distribution along the Americas, with Range Maximum Minimum Mean Mode
occurrences in Europe, Australia, New Zealand and Angola
(Fig. 2a). Taken together, the temporal maps revealed that Pliocene 87.60° 44.90° N 42.70° S 14.74° 28.00°
Late Miocene 102.50° 55.50° N 47.00° S 13.3° 9.10°
the Atlantic Ocean (including the Caribbean), California
Middle Miocene 102.60° 53.10° N 49.50° S 29.56° 34.50°
and Australia were persistently inhabited by this species Early Miocene 98.20° 48.70° N 49.50° S 19.95° 38.70°
throughout its duration.

Journal of Biogeography 43, 1645–1655 1649

ª 2016 John Wiley & Sons Ltd
C. Pimiento et al.

to 27 °C (maximum 18–33 °C, minimum = 1–24 °C; see (a)

Table S3). Our projections showed that the global areas cli-
matically suitable for C. megalodon would have not been
affected significantly by the temperature changes during the
Pleistocene, Holocene and Recent. For instance, there was a
decrease of only 2% in the predicted area during the Last
Glacial Maximum, and an increase of 8% and 9% in the pre-
dicted areas during the Holocene and Recent respectively
(Fig. 5).
(b)

DISCUSSION
Our analyses of fossil occurrences indicate that C. megalodon
was a cosmopolitan species with a widespread distribution
that covered the major ocean basins in the tropical, subtropi-
cal and temperate regions (Fig. 1; see Table S3). Analyses of
occurrences over time showed that C. megalodon originated
during the early Miocene (Fig. 3), rather than the middle
Miocene as previously assumed (Applegate & Espinosa-Arru- (c)
barrena, 1996; Purdy et al., 2001; Pimiento et al., 2010,
2013a,b; Pimiento & Balk, 2015). Furthermore, it has been
widely accepted that C. chubutensis (or its synonym Car-
charocles subauriculatus) preceded C. megalodon and occurred
in the early Miocene (Applegate & Espinosa-Arrubarrena,
1996; Pimiento et al., 2013b). However, our data evaluation
process (see Appendix S2) and time-binning analysis indi-
cated that C. megalodon also occurred in the early Miocene.
These early Miocene records were carefully examined follow-
(d)
ing our evaluation process (see Appendix S2) to corroborate
that they were indeed C. megalodon, and not C. chubutensis
(see Materials and Methods). Among these records, those
stratigraphically constrained to the stage level reported C.
megalodon in the Burdigalian (20.43–15.97 Ma). Given that
both C. megalodon and C. chubutensis are chronospecies
within the megatooth lineage (Ward & Bonavia, 2001;
Pimiento & Balk, 2015), it is possible that C. chubutensis
Temperature trends (ºC)

60 4
Latitudinal bins

45
30
15 Figure 5 Potential environmentally suitable areas for
2
0 Carcharocles megalodon in the Pliocene, projected to the Last
-15 Glacial Maximum (LGM), Holocene and Recent time periods.
-30 0
-45 Black dots indicate the fossil occurrences. Suitability ranges from
-60 0 (not suitable) to 1 (optimum).

22.6 20.6 18.6 16.6 14.6 12.6 10.6 8.6 6.6 4.6 2.6
occurred during the Aquitanian and C. megalodon originated
Time bins
in the Burdigalian. More studies are needed to establish the
Occupancy origination and extinction times of all chronospecies within
0.50

0.25
0.75
1.00

values the megatooth lineage.

Carcharocles megalodon reached its maximum geographical
Figure 4 Species occupancy over time and latitude. Blue range value (area coverage) in the late Miocene (Fig. 3, blue
squares = occupancy values per latitude bin. Grey area = cells line). However, at the same time its global abundance started
with no data. Gold line = temperature trends from Zachos et al. decreasing (Fig. 3, green line). Therefore, C. megalodon’s
(2001). maximum geographical range coincides with a decline in its

1650 Journal of Biogeography 43, 1645–1655

ª 2016 John Wiley & Sons Ltd
 Megalodon’s pathway to extinction

global abundance. This discrepancy indicates that the exten- resources (e.g. smaller cetaceans such as small whales and
sive geographical range of C. megalodon in the late Miocene dolphins).
could have been a response to a global population decline. In the Pliocene, the relative abundance of C. megalodon
This result also suggests that geographical range alone can- not only continued to decrease, but its geographical range
not be considered an indicator of extinction vulnerability, also declined (Fig. 3). At the end of the late Miocene and
as a widespread range can be a response to population during part of the Zanclean (early Pliocene), there was a
decline. temporary increase in the diversity of cetaceans (Morgan,
1994). However, during the rest of the Pliocene, mysticete
diversity declined again (Allmon et al., 1996; Allmon, 2001;
Biotic mechanisms
Marx & Uhen, 2010). Lindberg & Pyenson (2006) noted the
The pathway to extinction of C. megalodon started in the late permanent attenuation of mysticete diversity coincident with
Miocene, with a drop in its global abundance (Fig. 3). Dur- the last records of C. megalodon during the Pliocene.
ing this time, a number of biotic events occurred that could Although odontocetes also declined during the first half of
have been related to this trend. A massive diversity crash of the Pliocene, they had a slight recovery at the end of the
cetaceans took place around this time (Marx & Uhen, 2010: Pliocene and throughout the Pleistocene (Marx & Uhen,
fig. 1). Whether or not C. megalodon preyed upon marine 2010). In addition, the appearance and diversification of
mammals remains an open question; however, cetacean fos- other potential competitors for C. megalodon also occurred
sils [e.g. mysticetes (filter-feeding whales) and odontocetes during the Pliocene, including large delphinids, for example,
(toothed whales)] are frequently found together with C. the killer whale (Orcinus) (Lindberg & Pyenson, 2006), and
megalodon teeth, and C. megalodon bite marks have been the modern great white shark, C. carcharias (Ehret et al.,
identified in some cetacean material, leading to the hypothe- 2012).
sis of a predator–prey relationship (Spizuco et al., 1993; Previous studies have found that macro-evolutionary
Morgan, 1994; Allmon et al., 1996; Gottfried et al., 1996; trends in large sharks are linked to the origin and diversifica-
Purdy, 1996; Allmon, 2001; Purdy et al., 2001; Aguilera & tion of marine mammals (Adnet & Martin, 2007). Because
De Aguilera, 2004; Aguilera et al., 2008; Antunes et al., the predator–prey and competitive interactions of C. mega-
2015). This hypothesis is further supported by the significant lodon have not been rigorously studied yet, we cannot know
increase in mysticete body size after the extinction of for certain whether C. megalodon fed directly upon filter-
C. megalodon (Pimiento & Clements, 2014). The coincidental feeding whales, or competed directly with large odontocetes
timing of the decline in global abundance of C. megalodon and the great white shark (e.g. Morgan, 1994). However, the
and the drop in diversity of mysticetes and odontocetes dur- evidence from the marine fossil record suggests that biotic
ing the late Miocene suggests that the evolutionary history of factors in the late Miocene (a drop in potential diversity of
cetaceans played a role in the extinction of C. megalodon. cetaceans and the appearance of new competitors) and Plio-
Another biotic factor that could have been connected with cene (a persistent diversity drop in mysticetes and the
the decline in global abundance of C. megalodon during the appearance and diversification of new apex predators) coin-
late Miocene is the appearance of new potential competitors. cide with the continued and prominent decrease in geo-
During the middle Miocene, raptorial sperm whales appeared graphical distribution of C. megalodon prior to its extinction
and persisted through the late Miocene (Lambert et al., at the Pliocene/Pleistocene boundary. Future studies are
2010). It has been proposed that these pan-physeteroids needed to clarify the identity of species interacting with
occupied feeding niches similar to that of extant killer whale, C. megalodon.
Orcinus orca. If so, some of the largest of these raptorial apex
predators probably fed mostly on smaller baleen whales and
Climatic mechanisms
other marine mammals, and therefore competed with C.
megalodon for available food resources (Lambert et al., 2010, It has been suggested that the extinction of C. megalodon
2014). Furthermore, Carcharodon hubbelli, the ancestor of was associated with climate change (Gottfried et al., 1996).
the modern great white shark (C. carcharias), first appeared Various climatic events can be associated with the distribu-
during the late Miocene (Ehret et al., 2012). Evidence of tional trends of this species over time. For instance, the max-
competition between C. megalodon and C. hubbelli has not imum global abundance of C. megalodon during the middle
been reported, but the co-occurrence of both has been Miocene (Fig. 3) overlaps with the exceptionally warm cli-
recorded in the Pisco Formation of Peru (Ehret et al., 2012). mates of MMCO. Moreover, its subsequent drop in global
Moreover, based on tooth morphology (a good proxy of abundance during the late Miocene coincides with the rela-
diet), the great white shark and C. megalodon are considered tively cooler climates that occurred thereafter. During the
to be ecological analogues (Applegate & Espinosa-Arrubar- Pliocene, the continuing decline of C. megalodon corresponds
rena, 1996; Purdy et al., 2001; Pimiento et al., 2010; Ehret with the well-known oscillation of warm and cold tempera-
et al., 2012; Pimiento & Balk, 2015). It therefore can be tures (Fig. 4; Zachos et al., 2001). Thus, it is possible that C.
inferred that, when they co-occurred, these two closely megalodon benefited from warmer climates and was nega-
related sharks were competing for some of the same tively affected during colder (or transitional) periods, as

Journal of Biogeography 43, 1645–1655 1651

ª 2016 John Wiley & Sons Ltd
C. Pimiento et al.

previously proposed (Gottfried et al., 1996; Purdy, 1996). If clues regarding the pathway to extinction for the largest
this was the case, then during cold periods higher occupancy shark that ever lived. Further studies of predator–prey inter-
values should be expected towards the tropics compared with actions in relevant ancient marine ecosystems are needed to
temperate regions. Even when we found a higher frequency identify the causes of its extinction.
of occurrences in tropical latitudes during the late Miocene
(modal value; Table 1), our analysis of occupancy range per
ACKNOWLEDGEMENTS
latitude and time did not show such a gradient (Fig. 4).
Moreover, we found that the distribution of C. megalodon This project was funded by the NSF EAR 0418042, PIRE
did not contract during cold periods, nor did it expand dur- 0966884 (OISE, DRL, EAR). We thank the following muse-
ing warm periods (Fig. 4). ums for allowing us to study their specimens: the British
Distributional shifts in marine mammals, i.e. cetaceans Natural History Museum; Museo Argentino de Ciencias Nat-
(Gaskin, 1982) and pinnipeds (Dem
er
e et al., 2003: fig. 3.7), urales ‘Bernardino Rivadavia’; Museo de La Plata; Museo de
to anti-tropical latitudes have been proposed as another Historia Natural de la Universidad de San Marcos, Lima;
potential cause of the extinction of C. megalodon (Gottfried Museo Nacional de Historia Natural de Chile; Florida
et al., 1996). This hypothesis implies that this shark was Museum of Natural History; Natural History Museum of
physiologically unable to follow its prey to higher latitudes. Los Angeles County; San Diego Natural History Museum;
Conversely, our results indicate that C. megalodon occurred University of California Museum of Paleontology; Smithso-
in a wide geographical and temperature range, including nian Institution National Museum of Natural History; Field
anti-tropical latitudes (Figs 2 & 5; Table 1; see Table S3). Museum of Natural History; North Carolina Museum of
Most importantly, our ecological niche models show that Natural Sciences; Mus
eum national d’Histoire naturelle;
global temperature changes (such as the extreme cold climate Royal Belgian Institute of Natural Sciences, Museum f€ur Nat-
of the Last Glacial Maximum) would not have affected the urkunde and Museo de Nacional de Ciencias Naturales de
distribution of C. megalodon. These results are in agreement Madrid. Special thanks go to P. Hietz and O. Rodriguez for
with the physiological capability of large sharks (such as the their assistance accessing the collections at the Museum of
great white shark, an ecological analogue of C. megalodon) to Natural History, Vienna, Austria, and to D. I. Hastie and E.
conserve metabolic heat by maintaining a higher body tem- Fitzgerald for assistance accessing the collections at the
perature than the surrounding water, i.e. mesothermy or Museum Victoria, Australia. Thanks to J. Fischer and K.
regional endothermy (Bernal et al., 2001; Grady et al., 2014), Tsukui-Shockey for their help in translating papers to Eng-
which in turn gives them the capacity to exist in colder lish, to M. Kowalewski for his advice on the analyses, and to
waters (Goldman, 1997; Bonfil et al., 2005) and have a wide- B. Kent, J. Kriwet and an anonymous referee for their con-
spread distribution (Fergusson et al., 2009; Dambach & structive comments, which substantially improved this
R€odder, 2011). Taken together, our results demonstrate that manuscript. Finally, thanks to J. N. Griffin and D. Hone for
the pathway to extinction for C. megalodon was not driven their insight. This is University of Florida Contribution to
directly by temperature limitations. Paleobiology number 695 and Paleobiology Database publi-
cation number 253.

CONCLUDING REMARKS
REFERENCES
The study of the geographical distribution of C. megalodon
over its temporal range reveals that the pathway to extinction Adnet, S. & Martin, R.A. (2007) Increase of body size in six
of this species started in the late Miocene with a decline in gill sharks with change in diet as a possible background of
its global abundance. During the Pliocene both its geographi- their evolution. Historical Biology, 19, 279–289.
cal range and the global abundance showed a final drop Aguilera, O. & De Aguilera, D.R. (2004) Giant-toothed white
(Fig. 3). We interpret the overall decline in the distribution sharks and wide-toothed mako (Lamnidae) from the
of C. megalodon to be a response to prey availability and Venezuela Neogene: their role in the Caribbean, shallow-
other factors that indirectly regulated prey abundance, such water fish assemblage. Caribbean Journal of Science, 40,
as competition from other predators. Although these biotic 368–382.
factors could have been driven by oceanographic changes Aguilera, O.A., Garcia, L. & Cozzuol, M.A. (2008) Giant-
(Allmon et al., 1996; Allmon, 2001; Dem
er
e et al., 2003; toothed white sharks and cetacean trophic interaction
Marx & Uhen, 2010) we found no correlation between tem- from the Pliocene Caribbean Paraguana Formation.
perature and distribution. Furthermore, the coincidental tim- Palaeontologische Zeitschrift, 82, 204–208.
ing between the maximum values for geographical range and Allmon, W.D. (2001) Nutrients, temperature, disturbance,
the initial decrease in global abundance during the late Mio- and evolution: a model for the late Cenozoic marine
cene (Fig. 3) demonstrates that, in cosmopolitan species, record of the western Atlantic. Palaeogeography, Palaeocli-
geographical range alone cannot be considered an indicator matology, Palaeoecology, 166, 9–26.
of extinction vulnerability, as range broadening can be a Allmon, W.D., Emslie, S.D., Jones, D.S. & Morgan, G.S.
long-term response to population decline. This study reveals (1996) Late Neogene oceanographic change along Florida’s

1652 Journal of Biogeography 43, 1645–1655

ª 2016 John Wiley & Sons Ltd
 Megalodon’s pathway to extinction

west coast: evidence and mechanisms. The Journal of Geol- D.A. & O’Connor, M.I. (2015) The body size dependence
ogy, 104, 143–162. of trophic cascades. The American Naturalist, 185,
Antunes, M.T., Legoinha, P. & Balbino, A. (2015) Mega- 354–366.
lodont, mako shark and planktic foraminifera from the Dem
er
e, T.A., Berta, A. & Adam, P.J. (2003) Chapter 3: Pin-
continental shelf off Portugal and their age. Geologica Acta, nipedimorph evolutionary biogeography. Bulletin of the
13, 181–190. American Museum of Natural History, 279, 32–76.
Applegate, S.P. & Espinosa-Arrubarrena, L. (1996) The fossil Domeier, M.L. & Nasby-Lucas, N. (2008) Migration patterns
history of Carcharodon and its possible ancestor, Creto- of white sharks Carcharodon carcharias tagged at Guada-
lamna: a study in tooth identification. Great white sharks: lupe Island, Mexico, and identification of an eastern Paci-
the biology of Carcharodon carcharias (ed. by A.P. Klimley fic shared offshore foraging area. Marine Ecology: Progress
and D.G. Ainley), pp. 19–36. Academic Press, San Diego, Series, 370, 221–237.
CA. Dulvy, N.K., Fowler, S.L., Musick, J.A. et al. (2014) Extinc-
Barrett, P.M., McGowan, A.J. & Page, V. (2009) Dinosaur tion risk and conservation of the world’s sharks and rays.
diversity and the rock record. Proceedings of the Royal Soci- Elife, 3, e00590.
ety B: Biological Sciences, 276, 2667–2674. Ehret, D.J., MacFadden, B.J., Jones, D.S., Devries, T.J., Fos-
Bernal, D., Dickson, K.A., Shadwick, R.E. & Graham, J.B. ter, D.A. & Salas-Gismondi, R. (2012) Origin of the white
(2001) Review. Analysis of the evolutionary convergence shark Carcharodon (Lamniformes: Lamnidae) based on
for high performance swimming in lamnid sharks and recalibration of the upper Neogene Pisco Formation of
tunas. Comparative Biochemistry and Physiology Part A: Peru. Palaeontology, 55, 1139–1153.
Molecular & Integrative Physiology, 129, 695–726. Estes, J.A., Terborgh, J., Brashares, J.S. et al. (2011) Trophic
Bonfil, R., Me€y er, M., Scholl, M.C., Johnson, R., O’Brien, S., downgrading of planet earth. Science, 333, 301–306.
Oosthuizen, H., Swanson, S., Kotze, D. & Paterson, M. Fergusson, I., Compagno, L.J.V. & Marks, M. (2009) Car-
(2005) Transoceanic migration, spatial dynamics, and pop- charodon carcharias. The IUCN Red List of Threatened Spe-
ulation linkages of white sharks. Science, 310, 100–103. cies, e.T3855A10133872. Available at: http://dx.doi.org/
Boyden, J.A., M€ uller, R.D., Gurnis, M., Torsvik, T.H., Clark, 10.2305/IUCN.UK.20092.RLTS.T3855A10133872.en
J.A., Turner, M., Ivey-Law, H., Watson, R.J. & Cannon, (Downloaded 18 November 2015).
J.S. (2011) Next-generation plate-tectonic reconstructions Ferretti, F., Worm, B., Britten, G.L., Heithaus, M.R. & Lotze,
using GPlates. Geoinformatics: Cyberinfrastructure for the H.K. (2010) Patterns and ecosystem consequences of shark
Solid Earth Sciences, 9, 5–114. declines in the ocean. Ecology Letters, 13, 1055–1071.
Brook, B.W., Sodhi, N.S. & Bradshaw, C.J.A. (2008) Syn- Finnegan, S., Anderson, S.C., Harnik, P.G., Simpson, C., Tit-
ergies among extinction drivers under global change. tensor, D.P., Byrnes, J.E., Finkel, Z.V., Lindberg, D.R.,
Trends in Ecology and Evolution, 23, 453–460. Liow, L.H., Lockwood, R., Lotze, H.K., McClain, C.R.,
Busby, J.R. (1991) BIOCLIM: a bioclimatic analysis and pre- McGuire, J.L., O’Dea, A. & Pandolfi, J.M. (2015) Paleon-
diction system. Nature conservation: cost effective biological tological baselines for evaluating extinction risk in the
surveys and data analysis (ed. by C.R. Margules and M.P. modern oceans. Science, 348, 567–570.
Austin), pp. 64–68. Plant Protection Quarterly, Canberra, Gaskin, D. (1982) The ecology of whales and dolphins. Heine-
Australia. mann, London.
R Core Team (2015) R: A language and environment for sta- Goldman, K.J. (1997) Regulation of body temperature in the
tistical computing. R Foundation for Statistical Computing, white shark, Carcharodon carcharias. Journal of Compara-
Vienna. tive Physiology B, 167, 423–429.
Dambach, J. & R€ odder, D. (2011) Applications and future Gottfried, M.D., Compagno, L.J.V. & Bowman, S.C. (1996)
challenges in marine species distribution modeling. Aqua- Size and skeletal anatomy of the giant ‘megatooth’
tic Conservation: Marine and Freshwater Ecosystems, 21, shark Carcharodon megalodon. Great white sharks: the
92–100. biology of Carcharodon carcharias (ed. by A.P. Klimley
Davidson, A.D., Hamilton, M.J., Boyer, A.G., Brown, J.H. & and D.G. Ainley), pp. 55–89. Academic Press, San
Ceballos, G. (2009) Multiple ecological pathways to extinc- Diego, CA.
tion in mammals. Proceedings of the National Academy of Gradstein, F.M., Ogg, G. & Schmitz, M. (2012) The geologic
Sciences USA, 106, 10702–10705. time scale 2012. Elsevier, Amsterdam.
Davidson, A.D., Boyer, A.G., Kim, H., Pompa-Mansilla, S., Grady, J.M., Enquist, B.J., Dettweiler-Robinson, E., Wright,
Hamilton, M.J., Costa, D.P., Ceballos, G. & Brown, J.H. N.A. & Smith, F.A. (2014) Evidence for mesothermy in
(2012) Drivers and hotspots of extinction risk in marine dinosaurs. Science, 344, 1268–1272 (Supplemental Materi-
mammals. Proceedings of the National Academy of Sciences als).
USA, 109, 3395–3400. Harnik, P.G., Lotze, H.K., Anderson, S.C., Finkel, Z.V., Fin-
DeLong, J.P., Gilbert, B., Shurin, J.B., Savage, V.M., Barton, negan, S., Lindberg, D.R., Liow, L.H., Lockwood, R.,
B.T., Clements, C.F., Dell, A.I., Greig, H.I., Harley, McClain, C.R., McGuire, J.L., O’Dea, A., Pandolfi, J.M.,
D.D.G., Kratina, P., McCann, K.S., Tunney, T.D., Vasseur, Simpson, C. & Tittensor, D.P. (2012) Extinctions in

Journal of Biogeography 43, 1645–1655 1653

ª 2016 John Wiley & Sons Ltd
C. Pimiento et al.

ancient and modern seas. Trends in Ecology and Evolution, extinction. Proceedings of the National Academy of Sciences
27, 608–617. USA, 104, 10506–10511.
Jablonski, D. (1986) Background and mass extinctions. The Peter, R.G., Danabasoglu, G., Donner, L.J., Holland, M.M.,
alternation of macroevolutionary regimes. Science, 231, Hunke, E.C., Jayne, S.R., Lawrence, D.M., Neale, R.B.,
129–133. Rasch, P.J., Vertenstein, M., Worley, P.H., Yang, Z.L. &
Jablonski, D. & Flessa, K.W. (1986) The taxonomic structure Zhang, M. (2011) The community climate system model
of shallow-water marine faunas: implications for Phanero- version 4. Journal of Climate, 24, 4973–4991.
zoic extinctions. Malacologia, 27, 43–66. Pimiento, C. & Balk, M. (2015) Body size trends of the
Jaramillo, C., Zavada, M., Ortiz, J., Pardo, A. & Ochoa, D. extinct giant shark Carcharocles megalodon: a deep-time
(2013) The biogeography of the araucarian dispersed pol- perspective on marine apex predators. Paleobiology, 41,
len Cyclusphaera. International Journal of Plant Sciences, 479–490.
174, 489–498. Pimiento, C. & Clements, C.F. (2014) When did Carcharocles
Kiessling, W. & Aberhan, M. (2007) Geographical distribu- megalodon become extinct? A new analysis of the fossil
tion and extinction risk: lessons from Triassic-Jurassic record. PLoS ONE, 9, e0117877.
marine benthic organisms. Journal of Biogeography, 34, Pimiento, C., Ehret, D.J., MacFadden, B.J. & Hubbell, G.
1473–1489. (2010) Ancient nursery area for the extinct giant shark
Lambert, O., Bianucci, G., Post, K., Muizon, C., Salas-Gis- Megalodon from the Miocene of Panama. PLoS ONE, 5,
mondi, R., Urbina, M. & Reumer, J. (2010) The giant bite e10552.
of a new raptorial sperm whale from the Miocene epoch Pimiento, C., Gonz
alez-Barba, G., Ehret, D.J., Hendy, A.J.W.,
of Peru. Nature, 466, 105–108. MacFadden, B.J. & Jaramillo, C. (2013a) Sharks and rays
Lambert, O., Bianucci, G. & Beatty, B.L. (2014) Bony out- (Chondrichthyes, Elasmobranchii) from the Late Miocene
growths on the jaws of an extinct sperm whale support Gatun Formation of Panama. Journal of Paleontology, 87,
macroraptorial feeding in several stem physeteroids. 755–774.
Naturwissenschaften, 101, 517–521. Pimiento, C., Gonzalez-Barba, G., Hendy, A.J.W., Jaramillo,
Lindberg, D.R. & Pyenson, N.D. (2006) Evolutionary pat- C., MacFadden, B.J., Montes, C., Suarez, S.C. & Shippritt,
terns in Cetacea. Fishing up prey size through deep time. M. (2013b) Early Miocene chondrichthyans from the
Whales, whaling and ocean ecosystems (ed. by J.A. Estes, Culebra Formation, Panama: a window into marine verte-
D.P. DeMaster, D.F. Doak, T.M. Williams and R.L. Brow- brate faunas before closure of the Central American
nell), pp. 67–81. University of California Press, Los Ange- Seaway. Journal of South American Earth Sciences, 42,
les, CA. 159–170.
Lyons, S.K. & Smith, F.A. (2010) Using macroecological Purdy, R. (1996) Paleoecology of fossil white sharks. Great
approach to study geographic range, abundance and body white sharks: the biology of Carcharodon carcharias (ed. by
size in the fossil record. Quantitative methods for studying A.P. Klimley and D.G. Ainley), pp. 67–78. Academic Press,
paleobiology (ed. by J. Alroy and G. Hunt), pp. 117–141. San Diego, CA.
Paleontological Society Papers. Paleontological Society, Purdy, R.W., Schneider, V.P., Applegate, S.P., McLellan, J.H.,
New Haven, CT. Meyer, R.L. & Slaughter, B.H. (2001) The Neogene sharks,
Marx, F.G. & Uhen, M.D. (2010) Climate, critters, and ceta- rays, and bony fishes from Lee Creek Mine, Aurora, North
ceans: Cenozoic drivers of the evolution of modern whales. Carolina. Geology and paleontology of the Lee Creek
Science, 327, 993–996. Mine, North Carolina. Smithsonian Contributions to Paleo-
McCauley, D.J., Pinsky, M.L., Palumbi, S.R., Estes, J.A., biology, 90, 71–202.
Joyce, F.H. & Warner, R.R. (2015) Marine defaunation: Purvis, A., Gittleman, J.L., Cowlishaw, G. & Mace, G.M.
animal loss in the global ocean. Science, 347, 1255641. (2000) Predicting extinction risk in declining species. Pro-
McKinney, M.L. (1997) Extinction vulnerability and selectiv- ceedings of the Royal Society B: Biological Sciences, 267,
ity: combining ecological and paleontological views. 1947–1952.
Annual Review of Ecology and Systematics, 28, 495–516. Spizuco, M.P., Kinkel, M.D., Morgan, G.S. & Allmon, W.D.
Morgan, G.S. (1994) Whither the giant white shark? Ameri- (1993) Taphonomy of a shark-scavenged baleen whale
can Paleontologist, 2, 1–2. from the Late Pliocene ‘Pinecrest Beds’ of Florida. Geolog-
Myers, R.A., Baum, J.K., Shepherd, T.D., Powers, S.P. & ical Society of America, Southeastern Section. Abstracts w/
Peterson, C.H. (2007) Cascading effects of the loss of Program, 25, A71.
apex predatory sharks from a coastal ocean. Science, 315, Wang, S.C. & Dodson, P. (2006) Estimating the diversity of
1846–1850. dinosaurs. Proceedings of the National Academy of Sciences
O’Gorman, E.J. & Hone, D.W.E. (2013) Body size distribu- USA, 103, 13601–13605.
tion of the dinosaurs. PLoS ONE, 7, e51925. Ward, D. & Bonavia, C. (2001) Additions to, and a review
Payne, J.L. & Finnegan, S. (2007) The effect of geographic of, the Miocene shark and ray fauna of Malta. The Central
range on extinction risk during background and mass Mediterranean Naturalist, 3, 131–146.

1654 Journal of Biogeography 43, 1645–1655

ª 2016 John Wiley & Sons Ltd
 Megalodon’s pathway to extinction

Wroe, S., Huber, D.R., Lowry, M., McHenry, C., Moreno, Appendix S2 Additional methodological information.
K., Clausen, P., Ferrara, T.L., Cunningham, E., Dean,
M.N. & Summers, A.P. (2008) Three-dimensional com- BIOSKETCH
puter analysis of white shark jaw mechanics: how hard can
a great white bite? Journal of Zoology, 276, 336–342. Catalina Pimiento is a postdoctoral researcher interested in
Zachos, J., Pagani, M., Sloan, L., Thomas, E. & Billups, K. the macroecology of apex predators. She studies long-term
(2001) Trends, rhythms, and aberrations in global climate diversity dynamics, environmental constraints, biogeography,
65 Ma to present. Science, 292, 686–693. body size and the extinction of marine mega-fauna.
Author contributions: C.P., B.M. and B.S. conceived the
SUPPORTING INFORMATION ideas; C.P., C.C., S.V. and C.J. designed the analyses; C.P.
collected the data; C.P., C.C. S.V. and J.V.J. analysed the
Additional Supporting Information may be found in the data; C.P. led the writing.
online version of this article:
Appendix S1 Occurrences of Carcharocles megalodon
Editor: Lawrence Heaney
included (Table S1) and excluded (Table S2) in the analyses.

Journal of Biogeography 43, 1645–1655 1655

ª 2016 John Wiley & Sons Ltd