The Clovis culture is a prehistoricPaleo-Indian culture, named for distinct stone tools found in close association with Pleistocene fauna at Blackwater Locality No. 1 near Clovis, New Mexico, in the 1920s and 1930s. The Clovis culture appears around 11,500–11,000 uncal RCYBP (uncalibratedradiocarbon yearsbefore present), at the end of the last glacial period, and is characterized by the manufacture of "Clovis points" and distinctive bone and ivory tools. Archaeologists' most precise determinations at present suggest that this radiocarbon age is equal to roughly 13,200 to 12,900 calendar years ago. Clovis people are considered to be the ancestors of most of the indigenous cultures of the Americas.
The only human burial that has been directly associated with tools from the Clovis culture included the remains of an infant boy named Anzick-1. Researchers from the United States and Europe conducted paleogenetic research on Anzick-1's ancient nuclear, mitochondrial, and Y-chromosome DNA. The results of these analyses reveal that Anzick-1 is closely related to modern Native American populations, which lends support to the Beringia hypothesis for the peopling of the Americas.
The Clovis culture was replaced by several more localized regional cultures from the time of the Younger Dryas cold climate period onward. Post-Clovis cultures include the Folsom tradition, Gainey, Suwannee-Simpson, Plainview-Goshen, Cumberland, and Redstone. Each of these is commonly thought to derive directly from Clovis, in some cases apparently differing only in the length of the fluting on their projectile points. Although this is generally held to be the result of normal cultural change through time, numerous other reasons have been suggested to be the driving force for the observed changes in the archaeological record, such as the Younger Dryas post-glacial climate change which exhibited numerous faunal extinctions.
After the discovery of several Clovis sites in eastern North America in the 1930s, the Clovis people came to be regarded as the first human inhabitants who created a widespread culture in the New World. However, this theory has been challenged, in the opinion of many archaeologists, by several archaeological discoveries, including sites such as Cactus Hill in Virginia, Paisley Caves in the Summer Lake Basin of Oregon, the Topper site in Allendale County, South Carolina, Meadowcroft Rockshelter in Pennsylvania, the Friedkin site in Texas, Cueva Fell in Chile and, especially, Monte Verde, also in Chile. The claim to the oldest human archaeological site known in the Americas belongs to the Pedra Furada human remains and hearths, a site in Brazil that precedes the Clovis culture and the other sites already mentioned by 19,000 to 30,000 years. This discovery has become an issue of contention between North American archaeologists and their South American and European counterparts, who disagree on whether it is conclusively proven to be older.
A hallmark of the toolkit associated with the Clovis culture is the distinctively shaped, fluted stone spear point, known as the Clovis point. The Clovis point is bifacial and typically fluted on both sides. Archaeologists do not agree on whether the widespread presence of these artifacts indicates the proliferation of a single people, or the adoption of a superior technology by diverse population groups.
The culture was originally named for a small number of artifacts found between 1932 and 1936 at Blackwater Locality No. 1, an archaeological site between the towns of Clovis and Portales, New Mexico. These finds were deemed especially important due to their direct association with mammoth sp. and the extinct Bison antiquus. The in situ finds of 1936 and 1937 included most of four stone Clovis points, two long bone points with impact damage, stone blades, a portion of a Clovis blade core, and several cutting tools made on stone flakes. Clovis sites have since been identified throughout much, but not all, of the contiguous United States, as well as Mexico and Central America, and even into northern South America.
It is generally accepted that Clovis people hunted mammoths, as Clovis points have repeatedly been found in sites containing mammoth remains. However, mammoth was only a small part of the Clovis diet; extinct bison, mastodon, gomphotheres, sloths, tapir, camelops, horse and a host of smaller animals have also been found at Clovis sites where they were killed and eaten. In total, more than 125 species of plants and animals are known to have been used by Clovis people in the portion of the Western Hemisphere they inhabited.
The oldest Clovis site in North America is believed to be El Fin del Mundo in northwestern Sonora, Mexico, discovered during a 2007 survey. It features occupation dating around 13,390 calibrated years BP. In 2011, remains of Gomphothere were found; the evidence suggests that humans did in fact kill two of them here. There's also the Aubrey site in Denton County, Texas, which produced a radiocarbon date that is almost identical.
Disappearance of Clovis
Further information: Younger Dryas
The most commonly held perspective on the end of the Clovis culture is that a decline in the availability of megafauna, combined with an overall increase in a less mobile population, led to local differentiation of lithic and cultural traditions across the Americas. After this time, Clovis-style fluted points were replaced by other fluted-point traditions (such as the Folsom culture) with an essentially uninterrupted sequence across North and Central America. An effectively continuous cultural adaptation proceeds from the Clovis period through the ensuing Middle and Late Paleoindian periods.
Whether the Clovis culture drove the mammoth, and other species, to extinction via overhunting – the so-called Pleistocene overkill hypothesis – is still an open, and controversial, question. It has also been hypothesized that the Clovis culture saw its decline in the wake of the Younger Dryas cold phase. This 'cold shock', lasting roughly 1500 years, affected many parts of the world, including North America. This appears to have been triggered by a vast amount of meltwater – possibly from Lake Agassiz – emptying into the North Atlantic, disrupting the thermohaline circulation.
The Younger Dryas impact hypothesis or Clovis comet hypothesis originally proposed that a large air burst or earth impact of a comet or comets from outer space initiated the Younger Dryas cold period about 12,900 BPcalibrated (10,900 14C uncalibrated) years ago. The hypothesis has been largely contradicted by research showing that most of the conclusions cannot be repeated by other scientists, and criticized because of misinterpretation of data and the lack of confirmatory evidence.
Main articles: Burnet Cave and Dent Site
A cowboy and former slave, George McJunkin, found an Ancient Bison (Bison antiquus, an extinct relative of the American bison) skeleton in 1908 after a flash flood. The site was first excavated in 1926, near Folsom, New Mexico, under the direction of Harold Cook and Jesse Figgins. On August 29, 1927, they found the first in situFolsom point with the extinct B. antiquus bones. This confirmation of a human presence in the Americas during the Pleistocene inspired many people to start looking for evidence of early humans.
In 1929, 19-year-old Ridgely Whiteman, who had been closely following the excavations in nearby Folsom in the newspaper, discovered the Clovis site near the Blackwater Draw in eastern New Mexico. Despite several earlier Paleoindian discoveries, the best documented evidence of the Clovis complex was collected and excavated between 1932 and 1937 near Clovis, New Mexico, by a crew under the direction of Edgar Billings Howard until 1935 and later by John Cotter from the Academy of Natural Sciences/University of Pennsylvania. Howard's crew left their excavation in Burnet Cave, New Mexico (the first truly professionally excavated Clovis site) in August, 1932, and visited Whiteman and his Blackwater Draw site. By November, Howard was back at Blackwater Draw to investigate additional finds from a construction project.
The American Journal of Archaeology (January–March, 1932 V36 #1) in its "Archaeological Notes" mentions E. B. Howard's work in Burnet Cave, including the discovery of extinct fauna and a "Folsom type" point four feet below a Basketmaker burial. This brief mention of the Clovis point found in place predates any work at the Dent Site in Colorado. Reference is made to a slightly earlier article on Burnet Cave in The University Museum Bulletin of November, 1931.
The first report of professional work at the Blackwater Draw Clovis site is in the November 25, 1932, issue of Science News. The publications on Burnet Cave and Blackwater Draw directly contradict statements by several authors (for example see Haynes 2002:56 The Early Settlement of North America) that Dent, Colorado was the first excavated Clovis site. The Dent Site, in Weld County, Colorado, was simply a fossil mammoth excavation in 1932. The first Dent Clovis point was found November 5, 1932 and the in situ point was found July 7, 1933. The in situ Clovis point from Burnet Cave was excavated in late August, 1931 (and reported early in 1932). E. B. Howard brought the Burnet Cave point to the 3rd Pecos Conference, September 1931, and showed it around to several archaeologists interested in early humans (see Woodbury 1983).
Also, in 1968, in Montana, a Clovis burial site was found where the remains of a two-year-old child were studied. These remains have been named as Anzick-1 and recently, in 2014, have been used in scientific research.
Main article: Anzick-1
Available genetic data shows that the Clovis people are the direct ancestors of roughly 80% of all living Native American populations in North and South America, with the remainder descended from ancestors who entered in later waves of migration. As reported in February 2014, DNA from the 12,600-year-old remains of Anzick boy, found in Montana, has affirmed this connection to the peoples of the Americas. In addition, this DNA analysis affirmed genetic connections back to ancestral peoples of northeast Asia. This adds weight to the theory that peoples migrated across a land bridge from Siberia to North America.
Clovis First/Single origin hypothesis
Main article: Settlement of the Americas
Known as "Clovis First," the predominant hypothesis among archaeologists in the latter half of the 20th century had been that the people associated with the Clovis culture were the first inhabitants of the Americas. The primary support for this was that no solid evidence of pre-Clovis human habitation had been found. According to the standard accepted theory, the Clovis people crossed the Beringia land bridge over the Bering Strait from Siberia to Alaska during the period of lowered sea levels during the ice age, then made their way southward through an ice-free corridor east of the Rocky Mountains in present-day Western Canada as the glaciers retreated.
This hypothesis came to be challenged by studies suggesting a pre-Clovis human occupation of the Americas. In 2011, following the excavation of an occupation site at Buttermilk Creek, Texas, a prominent group of scientists claimed to have definitely established the existence "of an occupation older than Clovis."
According to researchers Michael Waters and Thomas Stafford of Texas A&M University, new radiocarbon dates place Clovis remains from the continental United States in a shorter time window beginning 450 years later than the previously accepted threshold (13,200 to 12,900 BP).
Recently the scientific consensus has changed to acknowledge the presence of pre-Clovis cultures in the Americas, ending the "Clovis first" consensus.
The results of a multiple-author study by Danish, Canadian and American scientists published in Nature in February 2016 revealed that "the first Americans, whether Clovis or earlier groups in unglaciated North America before 12.6 cal. kyr BP", are "unlikely" to "have travelled to North America from Siberia via the Bering land bridge "via a corridor that opened up between the melting ice sheets in what is now Alberta and B.C. about 13,000 years ago" as many anthropologists have argued for decades. The lead author, Mikkel Pedersen – a PhD student from University of Copenhagen – explained, "The ice-free corridor was long considered the principal entry route for the first Americans ... Our results reveal that it simply opened up too late for that to have been possible." The scientists argued that by 10,000 years ago, the ice-free corridor in what is now Alberta and B.C "was gradually taken over by a boreal forest dominated by spruce and pine trees" and that "Clovis people likely came from the south, not the north, perhaps following wild animals such as bison."
Alternatives to Clovis-first
Evidence of human habitation before Clovis
Further information: § Other sites
Archaeological sites that predate Clovis that are well documented include the following:
- Pedra Furada, Piauí, Brazil (10,500–12,000 yr BP; possibly >50,000 yr BP but this is disputed)
- Topper, South Carolina, US (16,000–20,000 yr BP; possibly 50,000 yr BP but this is disputed)
- Meadowcroft Rockshelter, Pennsylvania, US (16,000 yr BP)
- Buttermilk Creek Complex, Salado, Texas, US (15,500 14C yr BP)
- Cactus Hill, Virginia, US (15,070 14C yr BP)
- Monte Verde, Chile (18,500 to 14,80014C yr BP)
- Saltville (archaeological site), Virginia, US (14,510 14C yr BP)
- Taima-Taima, Venezuela (14,000 yr BP)
- Connley Caves, Oregon, US (13,000 yr BP)
- Page-Ladson prehistory site, Florida, US (12,425 ± 32 14C yr BP [15,405–14,146 cal yr BP])
- Lapa do Boquete, Brazil (12,070 ±170 14C yr BP)
- Paisley Caves, Oregon, US (14,300 cal yr BP)
- Tanana Valley, Alaska, US (13,000–14,000 cal yr BP)
- El Abra, Colombia (12,460 ±140 14C yr BP)
- Nenana Valley, Alaska, US (12,000 yr BP)
- Tibitó, Colombia (11,740 ±110 14C yr BP)
- Tagua-Tagua, Chile (11,380 ±380 14C yr BP)
Predecessors of the Clovis people may have migrated south along the North American coastlines, although there are arguments for many migrations along several different routes. Radiocarbon dating of the Monte Verde site in Chile place Clovis-like culture there as early as 18,500 to 14,500 years ago. Remains found at the Channel Islands of California place coastal Paleoindians there 12,500 years ago. This suggests that the Paleoindian migration could have spread more quickly along the Pacific coastline, proceeding south, and that populations that settled along that route could have then begun migrations eastward into the continent.
The Pedra Furada sites in Brazil include a collection of rock shelters, which were used for thousands of years by diverse human populations. The first excavations yielded artifacts with C14 dates of 48,000 to 32,000 years BP. Repeated analysis has confirmed this dating, carrying the range of dates up to 60,000 BP. The best-analyzed archaeological levels are dated between 32,160 ± 1000 years BP and 17,000 ± 400 BP.
In 2004, worked stone tools were found at Topper in South Carolina that have been dated by radiocarbon techniques possibly to 50,000 years ago. But, there is significant scholarly dispute regarding these dates. Scholars agree that evidence of humans at the Topper Site date back to 22,900 cal yr BP.
A more substantiated claim is that of Paisley Caves, Oregon, where rigorous carbon-14 and genetic testing appears to indicate that humans related to modern Native Americans were present in the caves over 1000 14C years before the earliest evidence of Clovis. Traces and tools made by another people, the "Western Stemmed" tradition, were documented.
A study published in Science presents strong evidence that humans occupied sites in Monte Verde, Chile, at the tip of South America, as early as 13,000 years ago. If this is true, then humans must have entered North America long before the Clovis Culture – perhaps 16,000 years ago.
The Tlapacoya site in Mexico is located along the base of a volcanic (remnant) hill on the shore of the former Lake Chalco. Seventeen excavations along the base of Tlapacoya Hill between 1956 and 1973 uncovered piles of disarticulated bones of bear and deer that appeared to have been butchered, plus 2,500 flakes and blades presumably from the butchering activities, plus one non-fluted spear point. All were found in the same stratum containing three circular hearths filled with charcoal and ash. Bones of many other animal species were also present, including horses and migratory waterfowl. Two uncalibrated radiocarbon dates on carbon from the hearths came in at approximately 24,000 and 22,000 years ago. At another location a prismatic micro-blade of obsidian was found in association with a tree trunk radiocarbon dated (uncalibrated) at approximately 24,000 years ago. This obsidian blade has recently been hydration dated by Joaquín García-Bárcena to 22,000 years ago. The hydration results were published in a seminal article that deals with the evidence for pre-Clovis habitation of Mexico.
Coastal migration route
Studies of the mitochondrial DNA of First Nations/Native Americans published in 2007 suggest that the people of the New World may have diverged genetically from Siberians as early as 20,000 years ago, far earlier than the standard theory suggests. According to one alternative theory, the Pacific coast of North America may have been free of ice, allowing the first peoples in North America to come down this route prior to the formation of the ice-free corridor in the continental interior. No evidence has yet been found to support this hypothesis except that genetic analysis of coastal marine life indicates diverse fauna persisting in refugia throughout the Pleistocene ice ages along the coasts of Alaska and British Columbia; these refugia include common food sources of coastal aboriginal peoples, suggesting that a migration along the coastline was feasible at the time. Some early sites on the coast, for example Namu, British Columbia, exhibit maritime focus on foods from an early point with substantial cultural continuity.
In February 2014, researchers reported on their DNA analysis of the remains of Anzick boy (referred to as Anzick-1) of Montana, the oldest skeleton found in the Americas and dated to 12,600 years ago. They found the mtDNA to be D4h3a, "one of the rare lineages associated with Native Americans." This was the same as the mtDNA associated with current coastal populations in North and South America. The study team suggest that finding this genetic evidence so far inland shows that "current distribution of genetic markers are not necessarily indicative of the movement or distribution of peoples in the past." The Y haplotype was found to be Q-L54*(xM3). Further testing found that Anzick-1 was most closely related to Native American populations (see below).
Main article: Solutrean hypothesis
The controversial Solutrean hypothesis proposed in 1999 by Smithsonian archaeologist Dennis Stanford and colleague Bruce Bradley (Stanford and Bradley 2002), suggests that the Clovis people could have inherited technology from the Solutrean people who lived in southern Europe 21,000–15,000 years ago, and who created the first Stone Age artwork in present-day southern France. The link is suggested by the similarity in technology between the projectile points of the Solutreans and those found at Clovis (and pre-Clovis) sites. Its proponents point to tools found at various pre-Clovis sites in eastern North America (particularly in the Chesapeake Bay region) as progenitors of Clovis-style tools. The model envisions these people making the crossing in small watercraft via the edge of the pack ice in the North Atlantic Ocean that then extended to the Atlantic coast of France, using skills similar to those of the modern Inuit people, making landfall somewhere around the then-exposed Grand Banks of the North American continental shelf.
In a 2008 study of the relevant paleoceanographic data, Kieran Westley and Justin Dix concluded that "it is clear from the paleoceanographic and paleo-environmental data that the Last Glacial Maximum (LGM) North Atlantic does not fit the descriptions provided by the proponents of the Solutrean Atlantic Hypothesis. Although ice use and sea mammal hunting may have been important in other contexts, in this instance, the conditions militate against an ice-edge-following, maritime-adapted European population reaching the Americas."
University of New Mexico anthropologist Lawrence G. Straus, a primary critic of the Solutrean hypothesis, points to the theoretical difficulty of the ocean crossing, a lack of Solutrean-specific features in pre-Clovis artifacts, as well as the lack of art (such as that found at Lascaux in France) among the Clovis people, as major deficiencies in the Solutrean hypothesis. The 3,000 to 5,000 radiocarbon year gap between the Solutrean period of France and Spain and the Clovis of the New World also makes such a connection problematic. In response, Bradley and Stanford contend that it was "a very specific subset of the Solutrean who formed the parent group that adapted to a maritime environment and eventually made it across the north Atlantic ice-front to colonize the east coast of the Americas" and that this group may not have shared all Solutrean cultural traits.
Genetic evidence of east/west dichotomy
Main article: Genetic history of indigenous peoples of the Americas
Mitochondrial DNA analysis in 2014 has found that members of some native North American tribes have a maternal ancestry (called haplogroup X) linked to the maternal ancestors of some present-day individuals in western Asia and Europe, albeit distantly. This has also provided some support for pre-Clovis models. More specifically, a variant of mitochondrial DNA called X2a found in many Native Americans has been traced to western Eurasia, while not being found in eastern Eurasia.
Mitochondrial DNA analysis of Anzick-1 concluded that the boy belonged to what is known as haplogroup or lineage D4h3a. This finding is important because the D4h3a line is considered to be a lineage "founder", belonging to the first people to reach the Americas. Although rare in most of today's Native Americans in the US and Canada, D4h3a genes are more common among native peoples of South America, far from the site in Montana where Anzick-1 was buried. This suggests a greater genetic complexity among Native Americans than previously thought, including an early divergence in the genetic lineage 13,000 years ago. One theory suggests that after crossing into North America from Siberia, a group of the first Americans, with the lineage D4h3a, moved south along the Pacific coast and, over thousands of years, into Central and South America, while others may have moved inland, east of the Rocky Mountains. The apparent early divergence between North American and Central plus South American populations may or may not be associated with post-divergence gene flow from a more basal population into North America; however, analysis of published DNA sequences for 19 Siberian populations does not favor the latter scenario.
Spearheads and DNA found at the Paisley Caves site in Oregon suggest that North America was colonized by more than one culture, and that the Clovis culture was not the first. There is evidence to suggest an east/west dichotomy, with the Clovis culture located to the east.
But in 2014, the autosomal DNA of a 12,500+-year-old infant from Montana was sequenced. The DNA was taken from a skeleton referred to as Anzick-1, found in close association with several Clovis artifacts. Comparisons showed strong affinities with DNA from Siberian sites, and virtually ruled out any close affinity with European sources (the "Solutrean hypothesis"). The DNA also showed strong affinities with all existing Native American populations, which indicated that all of them derive from an ancient population that lived in or near Siberia, the Upper Palaeolithic Mal'ta population. Mal'ta belonged to Y-DNA haplogroup R and mitochrondrial macrohaplogroup U.
The data indicate that Anzick-1 is from a population directly ancestral to present South American and Central American Native American populations. This rules out hypotheses which posit that invasions subsequent to the Clovis culture overwhelmed or assimilated previous migrants into the Americas. Anzick-1 is less closely related to present North American Native American populations (including a Yaqui genetic sample), suggesting that the North American populations are basal to Anzick-1 and Central and South American populations. The apparent early divergence between North American and Central plus South American populations may or may not be associated with post-divergence gene flow from a more basal population into North America; however, analysis of published DNA sequences for 19 Siberian populations did not favor the latter scenario. The boy Anzick-1, which is 12,6 thousand years old on the territory of Montana, belonged to Y-haplogroup Q-L54(xM3). Q-L54 is by far the largest haplogroup among Native Americans.
In approximate reverse chronological order:
- Pedra Furada, Serra da Capivara National Park, in the state of Piauí, Brazil. Site with evidence of non-Clovis human remains, a rock painting rupestre art drawings from at least 12,000–6,000 BP. Hearth samples C-14 dates of 48–32,000 BP were reported in a Nature article (Guidon and Delibrias 1986). New hearth samples with ABOX dates of 54,000 BP were reported in the Quaternary Science Reviews. Paleoindian components found here, have been challenged by American researchers such as Meltzer, Adovasio, and Dillehay.
- The Monte Verde site in Chile, was occupied from 14,800 years BP, with bones and other finds dating on average 12,500 yrs BP. The earliest finds at the site were from between 32,840 and 33,900 years BP, but are controversial.
- Lagoa Santa, Minas Gerais, Brazil, is erroneously asserted to be Clovis age or even possibly Pre-Clovis in age. The recent discussion of this site (specifically Lapa Vermelha IV) and the Luzia skull, reportedly 11,500 years old by Neves and Hubb, makes it clear that this date is a chronological date in years Before Present and not a raw radiocarbon date in eastern Brazil. Clovis sites mostly date between 11,500 and 11,000 radiocarbon years which means 13,000 years before present at a minimum. "Luzia" is at least 1,000 years younger than Clovis and Lapa Vermelha IV should not be considered a Pre-Clovis site.
- Cueva del Milodón, in Patagonian Chile dates at least as early as 10,500 BP. This is a site found particularly early in the New World hunt for Early Man, circa 1896, and needs additional basic research, but 10,500 B.P. would be 1,500 years younger than Clovis, or if the dating is 10,500 RCYBP, it would still be roughly 500–700 years younger than Clovis. In either case this should not be considered a Pre-Clovis site.
- Cueva Fell and Pali Aike Crater sites in Patagonia, with hearths, stone tools and other elements of human habitation dating to at least as early as 11,000 BP.
- The Big Eddy Site in southwestern Missouri contains several claimed pre-Clovis artifacts or geofacts. In situ artifacts have been found in this well-stratified site in association with charcoal. Five different samples have been AMS dated to between 11,300 and 12,675 BP (Before Present).
- Taima Taima, Venezuela has cultural material very similar to Monte Verde II, dating to 12,000 years BP. Recovered artifacts of the El Jobo complex in direct association with the butchered remains of a juvenile mastodon. Radiocarbon dates on associated wood twigs indicate a minimum age of 13,000 years before the present for the mastodon kill, a dating significantly older than that of the Clovis complex in North America.
- A cut mastodon tusk found at Page-Ladson, Jefferson County, Florida on the Aucilla River has been dated to 12,300 years BP near a few in situ artifacts of similar age.
- The Schaefer and Hebior mammoth sites in Kenosha County, Wisconsin indicate exploitation of this animal by humans. The Schaefer Mammoth site has over 13 highly purified collagen AMS dates and 17 dates on associated wood, dating it to 12,300–12,500 radiocarbon years before the present. Hebior has two AMS dates in the same range. Both animals show conclusive butchering marks and associated non-diagnostic tools.
- A site in Walker, Minnesota with stone tools, alleged to be from 13,000 to 15,000 years old based on surrounding geology, was discovered in 2006. However, further examination suggests that the site does not represent a human occupation.
- In a 2011 article in Science, Waters et al. 2011 describe an assemblage of 15,528 lithic artifacts from the Debra L. Friedkin site west of Salado, Texas. These artifacts (including 56 tools, 2,268 macrodebitage and 13,204 microdebitage) define the Buttermilk Creek Complex formation, which stratigraphically underlies a Clovis assemblage. While carbon dating could not be used to directly date the artifacts, 49 samples from the 20 cm Buttermilk floodplain sedimentary clay layer in which the artifacts were embedded were dated using optically stimulated luminescence (OSL). Eighteen OSL ages, ranging from 14,000 to 17,500 ka were obtained from this layer. The authors report "the most conservative estimate" of the age of the Buttermilk clays range from 13,200 to 15,500 ka, based on the minimum age represented by each of the 18 OSL ages.
- Human coprolites have been found in Paisley Caves in Oregon, carbon dated at 14,300 years ago. Genetic analysis revealed that the coprolites contained mtDNA haplogroups A2 and B2, two of the five major Native American mtDNA haplogroups.
- The Mud Lake site, in Kenosha County, Wisconsin consists of the foreleg of a juvenile mammoth recovered in the 1930s. Over 100 stone tool butchering marks are found on the bones. Several purified collagen AMS dates show the animal to be 13,450 RCYBP with a range of plus or minus 1,500 RCYBP variance.
- Meadowcroft Rockshelter in southwestern Pennsylvania, excavated 1973–78, with evidence of occupancy dating back from 16,000 to 19,000 years ago.
- Cactus Hill in southern Virginia, with artifacts such as unfluted bifacial stone tools with dates ranging from c. 15,000 to 17,000 years ago.
- Sixty-eight stone and bone tools discovered in an orchard in East Wenatchee, Washington in 1987, excavated in 1988 and 1990. Five of the Clovis points are on display at the Wenatchee Valley Museum & Cultural Center.
- ^"Clovis Culture" entry in the Concise Oxford Dictionary of Archaeology: "9,500–9,000 BC".
- ^ abWaters, Michael; et al. (2007). "Redefining the Age of Clovis: Implications for the Peopling of the Americas". Science (315): 1122–1126. doi:10.1126/Science.1127166.
- ^David R. Starbuck, The Archeology of New Hampshire: Exploring 10,000 Years in the Granite State, UPNE, 2006, p. 25.
- ^ abSharon Bigley, "Ancient native boy's genome reignites debate over first Americans", Reuters, 12 February 2014
- ^Owsley, Douglas W; Hunt, David (May 2001). "Clovis and early Archaic crania from the Anzick site (24PA506), Park County, Montana". Plains Anthropologist.
- ^New Rdiocarbon Dates for the Anzick Clovis Burial by Juliet E. Morrow and Stuart J.Fiedel. In Paleoindian Archaeology, edited by J.E.Morrow and C.G.Gnecco. University Press of Florida, Gainesville.
- ^ abcdefghijklRasmussen, M.; Anzick, S. L.; Waters, M. R.; Skoglund, P.; DeGiorgio, M.; Stafford, T. W.; Rasmussen, S.; Moltke, I.; Albrechtsen, A.; Doyle, S. M.; Poznik, G. D.; Gudmundsdottir, V.; Yadav, R.; Malaspinas, A. S.; White, S. S.; Allentoft, M. E.; Cornejo, O. E.; Tambets, K.; Eriksson, A.; Heintzman, P. D.; Karmin, M.; Korneliussen, T. S.; Meltzer, D. J.; Pierre, T. L.; Stenderup, J.; Saag, L.; Warmuth, V. M.; Lopes, M. C.; Malhi, R. S.; Brunak, S. R.; Sicheritz-Ponten, T.; Barnes, I.; Collins, M.; Orlando, L.; Balloux, F.; Manica, A.; Gupta, R.; Metspalu, M.; Bustamante, C. D.; Jakobsson, M.; Nielsen, R.; Willerslev, E. (2014-02-13). "The genome of a Late Pleistocene human from a Clovis burial site in western Montana". Nature. 506 (7487): 225–229. Bibcode:2014Natur.506..225R. doi:10.1038/nature13025. PMC 4878442. PMID 24522598.
- ^ abRaff, J. A.; Bolnick, D. A. (2014-02-13). "Palaeogenomics: Genetic roots of the first Americans". Nature. 506 (7487): 162–163. Bibcode:2014Natur.506..162R. doi:10.1038/506162a. PMID 24522593.
- ^ abHaynes, Gary (2002). The Early Settlement of North America: The Clovis Era. New York: Cambridge University Press. p. 52. ISBN 0-521-52463-6.
- ^ abcWaters, Michael R.; Forman, Steven L.; Jennings, Thomas A.; Nordt, Lee C.; Driese, Steven G.; Feinberg, Joshua M.; Keene, Joshua L.; Halligan, Jessi; Lindquist, Anna; Pierson, James; Hallmark, Charles T.; Collins, Michael B.; Wiederhold, James E. (2011). "The Buttermilk Creek Complex and the Origins of Clovis at the Debra L. Friedkin Site, Texas"(PDF). Science. 331 (25): 1599–1603. Bibcode:2011Sci...331.1599W. doi:10.1126/science.1201855. PMID 21436451. Retrieved 2011-03-24.
- ^"Monte Verde Under Fire". Archaeology.org. 1999-10-18. Retrieved 2011-02-06.
- ^"Pedra Furada, Brazil: Paleoindians, Paintings, and Paradoxes. An interview". Athena Review. 3 (2 - Peopling of the Americas). 2002. Retrieved 2011-03-27.
- ^Bellos, Alex (14 February 2000). "Brazilian Findings Spark Archeological Debate". Cabrillo.edu. Retrieved 2011-03-27 – via The Guardian.
- ^Meltzer, David (24 June 1995). "Stones of Contention". New Scientist. Retrieved 2011-03-27.
- ^ abMann, Charles C. (November 2013). "The Clovis Point and the Discovery of America's First Culture". Smithsonian.com. The Smithsonian Institution. Retrieved 20 January 2015.
- ^Pearson, Georges; Ream, Joshua (2005). "Clovis on the Caribbean Coast of Venezuela". Current Research in the Pleistocene. 22: 28–31. ISSN 8755-898X.
- ^Haynes, G. (2009). "Estimates of Clovis-Era Megafaunal Populations and Their Extinction Risks". American Megafaunal Extinctions at the End of the Pleistocene. Vertebrate Paleobiology and Paleoanthropology. Springer Science + Business Media B.V. pp. 39–53.
- ^Surovell, T. A.; Waguespack, N. M. (2008-11-15). "How many elephant kills are 14?: Clovis mammoth and mastodon kills in context". Quaternary International. 191 (1): 82–97. Bibcode:2008QuInt.191...82S. doi:10.1016/j.quaint.2007.12.001.
- ^Sanchez, G.; Holliday, V. T.; Gaines, E. P.; Arroyo-Cabrales, J.; Martínez-Tagüeñ, N.; Kowler, A.; Lange, T.; Hodgins, G. W. L.; Mentzer, S. M.; Sanchez-Morales, I. (2014-07-14). "Human (Clovis)–gomphothere (Cuvieronius sp.) association ∼13,390 calibrated yBP in Sonora, Mexico". PNAS. 111: 10972–10977. Bibcode:2014PNAS..11110972S. doi:10.1073/pnas.1404546111. PMC 4121807. PMID 25024193.
- ^Ferring, C. Reid (2001) The Archaeology and Paleoecology of the Aubrey Clovis Site (41DN479) Denton County, Texas. Center for Environmental Archaeology, Department of Geography, University of North Texas, Denton
- ^"Southeastern Prehistory: Paleoindian Period". National Park Service. Archived from the original on 2008-05-08. Retrieved 2008-04-28.
- ^Lepper, Bradley T. (1999). "Pleistocene Peoples of Midcontinental North America". In Bonnichsen, Robson; Turnmire, Karen. Ice Age People of North America. Corvallis: Oregon State University Press. pp. 362–394.
- ^Barnosky, A. D.; et al. (1 October 2004). "Assessing the Causes of Late Pleistocene Extinctions on the Continents". Science. 306: 70–5. Bibcode:2004Sci...306...70B. doi:10.1126/science.1101476. PMID 15459379. Retrieved 20 January 2015.
- ^Anderson, D. G.; et al. (December 2009). "Human Population Decline in North America during the Younger Dryas". American Geophysical Union. Retrieved 20 January 2015.
- ^Teller, James T. (16 November 2012). "Importance of freshwater injections into the Arctic Ocean in triggering the Younger Dryas cooling"(PDF). PNAS. 109: 19880–1. Bibcode:2012PNAS..10919880T. doi:10.1073/pnas.1218344109. PMC 3523840. PMID
The incisors of proboscideans (tusks and tushes) are one of the most important feature in conservation, ecology and evolutionary history of these mammals. Although the absence of upper incisors is rare in proboscideans (occurring only in deinotheres), the independent losses of lower incisors are recognized for most of its lineages (dibelodont condition). The presence of lower incisors in the Pan-American gomphothere Cuvieronius hyodon was reported a few times in literature, but it was neglected in systematic studies. We analyzed several specimens of Cuvieronius hyodon from the Americas and recognized that immature individuals had lower incisors during very early post-natal developmental stages. Subsequently, these are lost and lower incisors alveoli close during later developmental stages, before maturity. Moreover, for the first time in a formal cladistic analysis of non-amebelodontine trilophodont gomphotheres, Rhynchotherium and Cuvieronius were recovered as sister-taxa. Among several non-ambiguous synapomorphies, the presence of lower incisors diagnoses this clade. We recognize that the presence of lower incisors in Cuvieronius and Rhynchotherium is an unique case of taxic atavism among the Elephantimorpha, since these structures are lost at the origin of the ingroup. The rediscovery of the lower incisors in Cuvieronius hyodon, their ontogenetic interpretation and the inclusion of this feature in a revised phylogenetic analysis of trilophodont gomphotheres brought a better understanding for the evolutionary history of these proboscideans.
The Pan-American proboscidean Cuvieronius hyodon (Fischer de Waldheim, 1814) is the only gomphothere known throughout the New World from the early to late Pleistocene [1–3]. Traditionally, it is stated that South American gomphotheres (Notiomastodon platensis and Cuvieronius hyodon, sensu [3,4]) lack lower defenses (both incisors, the deciduous—di1 or tushes—and the permanents—i1 or tusks) , such as observed in mammoths and extant elephants . Accordingly, the dental formula of C. hyodon consists of one pair of upper tushes (deciduous incisors), one pair of upper tusks (evergrowing permanent incisors), three pairs of lower and upper deciduous premolars (first bilophodont—dp2/dP2—and the other two trilophodonts—dp3/dP3and dp4/dP4) and three pairs of upper and lower permanent molars (first two trilophodonts—m1/M1 and m2/M2 and last one tetra or pentalophodonts—m3/M3 ). The horizontal replacement of the deciduous premolars and permanent molars, together with the progressive wear on these teeth occlusal surface generate a wearing pattern that can be associated with developmental stages of the proboscidean (Table 1 [7–9]).
Developmental stages and age classes for immature individuals of Cuvieronius hyodon based on lower tooth eruption, wear stage and estimated age in years (modified from [7,8]).
However, some authors reported the presence of lower incisors in two juvenile specimens of C. hyodon from Tarija, Bolivia [10–12]. Also, a mandible of another juvenile of C. hyodon with a pair of lower incisor alveoli was recognized from Costa Rica . Additionally, the lower incisors of C. hyodon were considered as vestigial lower tushes , because these teeth are present only in a few immature individuals, whereas in most juvenile mandibles there are no tushes nor detectable traces of lower incisor alveoli. No further detailed morphological study on the lower incisors of Cuvieronius hyodon has been made since those reports, although this taxon has been widely studied in the last decades [1,3,14–21].
As a result, all subsequent phylogenetic analyses including this Pan-American gomphothere have neglected the observations of those authors [12–14], by neither considering Cuvieronius hyodon as having lower incisors, nor rating this taxon as polymorphic for this particular feature [1,5,6,21,22]. Here, we recognize ontogenetic and evolutionary patterns for the lower incisors, deciduous premolars and molars, and lower jaw of Cuvieronius hyodon. In addition, we present a phylogenetic hypothesis to non-amebelodontine trilophodont gomphotheres with emphasis in the position of C. hyodon. Moreover, we discuss the significance of the lower incisors features to the evolution of non-amebelodontine trilophodont gomphotheres and also to Elephantimorpha broadly speaking (sensu ).
Results and Discussion
The lower incisors of Cuvieronius hyodon
Among 46 individuals of C. hyodon analyzed, we identified seventeen individuals (eleven juveniles and six adults) with lower incisor alveoli (approximately 37% of the sample analyzed, see Table 2, Fig 1A and 1D). Two mandibles with lower incisors preserved the lower incisors in loci [12, 24] (Fig 1B). Unfortunately, these specimens were damaged and the lower incisors were lost, probably during the transfer from the Museo de Historia Natural de Bolivia (La Paz) to the Museo Nacional de Paleontología y Arqueología de Tarija (Fig 1C). A small fragment of the right incisor was still preserved in mandible MNPA-V 005867 (Fig 2A) when one of the present authors visited the Tarija Museum in 2003.
Analyzed mandibles of Cuvieronius hyodon, their developmental stage, the presence/absence of lower incisors, and shape of mandibular symphysis.
Juvenile mandibles of Cuvieronius hyodon.
Lower tusks of Cuvieronius, Rhynchotherium and Gomphotherium.
The lower incisors of Cuvieronius are small, well-preserved and containing enamel ; and two mandible specimens were figured (MNPA-V 005867 and MNPA-V 005888) with a pair of elongated, thin, parallel with convergent tips incisors (Figs 1B and 2A). Besides the morphological description of the lower incisors , a picture of specimen MNPA-V 005867 including a small lower incisor is the only direct record of this tooth (Fig 2A). Unlike the upper permanent tusks, the lower incisor of C. hyodon is elongated and thin, with 11 mm at its widest diameter, straight and not twisted (see Figs 1B and 2A). According with the shape of the lower alveoli, we infer that the lower incisor probably had an oval/rounded cross section (Fig 1A–1C). The presence of enamel in the lower incisor of Cuvieronius, is dubious: these teeth were described as clearly having enamel , however, no trace of enamel was noted by the only present author that visited the Tarija Museum before the incisor get lost (MPF). The proximal end of the incisor (pulp cavity) possessed and open pulp, although of limited size, which indicates that these incisors were evergrowing.
In fact, newly born and immature individuals of Gomphotheriidae are rare, especially those with complete deciduous dentition. Rhynchotherium is known only from a few specimens ( Fig 2B), whereas Cuvieronius has a more significant fossil record from Bolivia and Mexico [11,12,26,27]. One of the youngest records of immature Cuvieronius is a newborn skull specimen from Lake Chapala, Jalisco, Mexico, which presents the upper tushes in loci . The upper tushes are quite similar to the upper tushes of Gomphotherium angustidens , since both are small, conical, with pointed and slightly wrinkled enamel collar.
However, concerning immature individuals, Gomphotherium angustidens has a more complete fossil record than Cuvieronius and Rhynchotherium, which includes several specimens of tushes (deciduous incisors or ‘‘milk tusks”), tusks (the evergrowing permanent incisors), deciduous premolars, permanent premolars and molars [9,25,28–30]. Rhynchotherium is known only from one immature specimen (F:A.M. 18216) with deciduous premolars in use and erupted lower incisors. The lower incisors are slightly upcurved, coniform, not twisted; with a small lateral enamel band and oval cross section (see Fig 2A and 2B). Later in ontogeny, the lower incisors of Rhynchotherium become slightly upcurved and may lose their lateral enamel band. The presence of enamel bands on lower incisors of Rhynchotherium specimens is usually not easily observed . However, the immature lower incisors of Rhynchotherium (Fig 2B) present this enamel structure, which generally is lost with usage and/or during ontogenetic development.
The lower tushes of Gomphotherium angustidens are represented by the di1  and are small, straight, with slightly inflated apices, and piriform cross section (Fig 2C ). The apex of each tush is covered by an enamel cap, which is flattened, wrinkled at the tip and presents well-defined borders (“collar”, according to , Fig 2C). On the other hand, the immature lower tusks (i1) have pointed apices, with bean-shaped cross sections and a uniform enamel cap, which has no collar or well-defined borders . As the individual grows, the enamel cap wears away, the tusks become longer and dorsoventrally flattened, and develop a piriform cross section [25,28] (see Fig 2D).
The lower tushes of Gomphotherium angustidens are very particular and only few immature individuals have these teeth preserved in loci . The tissues responsible for the attachment of the lower incisors in the alveoli are very soft, fragile and they could deteriorate easily, thus causing the lower incisors to detach from the alveoli before the final burial of the mandible . Therefore, loss of the lower incisors before fossilization is more likely to be a common pattern among gomphotheres, and the presence of open alveoli is the only evidence of the existence of these teeth.
Growth and ontogeny of lower incisors in Cuvieronius hyodon
The developmental stages recognized for Cuvieronius hyodon in the sample analyzed here were J1, J3, J4, J5, Y1, Y2, Y3 and Adult. We did not recognized individuals in the developmental stages of Fetus, J2 and Y4, because none of those specimens presented the deciduous molars with the combination of wear pattern of these developmental stages (Table 1). There are 21 juvenile specimens between J1 and Y3, and 25 Adult specimens (Table 2). It is important to note that among juveniles between developmental stages J1 to J5 in which it was possible to observe the symphysial region, the frequency of lower incisor alveoli is 100% (i.e., all J1 to J5 specimens presented open lower incisor alveoli). Concerning the specimens on developmental stages Y1 to Y3, five specimens presented closing lower incisor alveoli (38.5% of Y1 to Y3 juveniles) and seven presented no trace of alveoli.
Among the Adults with preserved symphysis (24 individuals), only six specimens presented closing alveoli (25% of Adults), whereas 24 presented no trace of alveoli (Table 2). Therefore, we observed that until the developmental stage J5, all studied individuals of Cuvieronius carried lower incisors before death (except for two with broken symphysial region), suggesting their lower tusks were still being used before death, and the subsequent loss of these incisors was due taphonomic processes . In the eleven individuals from Y1 to Adult which lower incisors alveoli, these structures were partially closing (i.e., filled with spongy tissue, shallow and usually reduced in size—Fig 1D and 1E).
It is possible to infer that, in the ontogenetic development of C. hyodon, the lower tusks were still being used until the developmental stage J5. These were subsequently lost around the developmental stage Y2, and the alveoli started to heal/close from this stage onwards. No individuals of stage Y1 present the symphysial region bearing lower incisor alveoli, thus, it was not possible to precisely infer the status of the lower incisors and lower incisors alveoli in this age group.
In other non-amebelodontine trilophodont gomphotheres with lower incisors, the youngest specimen of Rhynchotherium (F:A.M. 18216) has erupted lower incisors and with worn tips at the stage Y1, whereas the lower incisors of most immature Cuvieronius are probably lost at this developmental stage. At the stage Y3, the lower tusks of Rhynchotherium are longer, slightly upcurved and with divergent tips, following the downturned curvature of the mandibular symphysis. In Gomphotherium angustidens, the lower tushes are probably formed/erupted before the stage J1 stage, when both dp2 and dp3 become erupt . The lower tushes are replaced by the lower tusks probably at the stage J2, when the dp2 and dp3 are erupted and slightly worn. As soon as they erupt, the lower tusks also become worn rapidly, and the small enamel cap at their tips wears away [25,28,32,33]. Cuvieronius individuals at J1 developmental stage already present a large open alveoli, which indicates the presence of completely erupted lower incisors. Probably, the lower incisors develop and erupt at the Fetus and J1 developmental stages and there is no evidence of replacement of the lower incisors until the stage J5. In addition, the absence of isolated lower incisors of Cuvieronius in the fossil record could be also an evidence of the no replacement of the lower incisor in Cuvieronius.
In addition, the brevirostrine non-amebelodontine trilophodont gomphotheres do not develop second generation lower teeth, as Gomphotherium had true permanent pre-molars and lower tusks [9,25,28]. The dentition of Notiomastodon, Sinomastodon and Stegomastodon comprises the upper pair of tushes, the upper and lower deciduous pre-molars and permanent molars, which derived from primary lamina. The upper tusks (I2) are the only teeth originated on the secondary lamina, which replaces the upper tushes in immature individuals (around J3 developmental stage). Furthermore, Rhynchotherium and Cuvieronius present the lower incisors, which are not replaced during lifespan. Probably, except for the presence of the I2, the secondary lamina was lost in the brevirostrine non-amebelodontine trilophodont gomphotheres. Thus, considering the eruption of lower incisor at Fetus/J1 developmental stage and the absence of replacement by the permanent lower incisors in Cuvieronius, we believe that is more parsimonious to consider its lower incisors as tushes (di1), developed from the primary lamina.
Usually, in mammals, the secondary generation (permanent) teeth develop only after the first generation (deciduous) teeth [34,35]. In the case of brevirostrine non-amebelodontine trilophodont gomphotheres, we believe that the secondary lamina was suppressed and the incisors of Cuvieronius are originated in the primary lamina, with no replacement. In fact, Cuvieronius retained in the mandible a pair of deciduous incisors, the deciduous premolars and molars, all originated in first generation. This "replacement pattern” (a modified “monophyodont" dentition, since there is no replacement) is similar to the one found in Murid rodents, which have eliminated primary and secondary generations of canines and premolars, and retained only a pair of deciduous incisors and the molars, with no dental replacement at all . Since there is also no record of isolated incisors recognized as Rhynchotherium, and the fossil record does not present evidence of incisors replacement, probably the same inference can be made to this taxon.
The lower incisors of C. hyodon were considered as vestigial because they are present only in a few specimens of Tarija (Bolivia) . However, the definition of a vestigial structure  is: 1) it occurs regularly in all members of a population; 2) it is present in the parents and recent ancestors; and, 3) it either occurs transiently during development or may persist into maturity (depending on the degree of evolutionary suppression). In this way, the presence of lower tushes in Cuvieronius hyodon is not vestigial. The argument presented by that author  better fits the evolutionary process of spontaneous atavism, in which rare atavistic anomalies occur in individuals of natural populations . Nevertheless, we identified the presence of incisors in other populations of C. hyodon from Central and North America, and that persistent lower tushes are characteristic of the developmental stages J1 to J5, which correspond to the juvenile age class. Yet, the lower tushes of C. hyodon fails in two of the four criteria created to identify atavism –the persistence of a feature in adults and the presence in only one or a few individuals within a population. For this reason, we agree with the vestigial presence of lower tushes in C. hyodon, but not for the same arguments provided by this author .
Phylogenetic position of Cuvieronius hyodon
As previously mentioned, most cladistics analysis of non-amebelodontine trilophodont gomphotheres have not considered the presence of lower tusks in Cuvieronius [1,5,6,21,22]. Regardless of the particular taxonomic assumptions for the South American endemic gomphothere (Notiomastodon in , Stegomastodon platensis and S. waringi in , Haplomastodon chimborazi and “S.”platensis in , “S.” waringi and “S.” platensis in , and Notiomastodon platensis in ), all those previous phylogenetic hypotheses recovered Cuvieronius as its sister taxon.
To better understand the “dance” of tusks in the non-amebelodontine trilophodont gomphotheres, we conducted a phylogenetic revision of this group, with especial emphasis on the addition of new characters related to lower incisors. Our phylogenetic analysis resulted in one most parsimonious tree (32 steps; CI = 0.781, RI = 0.8) and recovered Cuvieronius hyodon and Rhynchotherium falconeri as sister-taxa (Clade A in Fig 3). This clade is defined by seven unambiguous synapomorphies: the presence of an enamel band in the upper tusks (character 1, state 2); twisted upper tusks (character 3, state 1); presence of lower tushes (character 4, state 0); presence of the incisors fossa (character 11, state 0); ventral torsion of mandible symphysis ≥35° (character 13, state 1); oval/circular cross section of lower tushes (character 16, state 1); and upper tusks alveoli greatly diverging distally (character 19, state 2).
Phylogenetic position of Cuvieronius hyodon.
We are not the first to suggest the sister-taxa relationship between C. hyodon and R. falconeri. An almost complete skull and mandible of Rhynchotherium from Arizona (USA) was described  and the author suggested Rhynchotherium as a precursor of Cuvieronius. Likewise, a close relationship was stated between both taxa , based on the spiral enamel bands on the upper tusks and brevirostrine mandibles, although the authors did not perform a formal cladistic analysis to test this hypothesis. Additionally, these authors indicated that the presence of lower incisors and downturned symphysis distinguish R. falconeri from C. hyodon. We disagree with their interpretation of about both features, because C. hyodon also presents lower incisors (in juvenile individuals from J1 to J5 developmental stages) and the mandibular symphysis in C. hyodon is as much downturned as it is in R. falconeri (35° or more). However, we identified differences between C. hyodon and R. falconeri in this study; such as R. falconeri having a longer symphysis than C. hyodon (an exceptional condition among brevirostrine gomphotheres), which accommodates robust lower tushes. The lower tushes of R. falconeri are slightly upcurved and with divergent tips (Fig 2B), whereas C. hyodon had slender, straight and parallel to convergent incisors (Figs 1B and 2A). In addition, both lower tushes have open pulp cavity, an evidence of an evergrowing condition, however, Cuvieronius has a delayed loss of the lower tushes (J5 developmental stage), compared to Gomphotherium (J1 developmental stage), while Rhynchotherium retained the evergrowing lower tushes during all lifespan.
Here, we provide the first phylogenetic hypothesis that take into consideration the presence of lower tushes in Cuvieronius hyodon. Furthermore, this is the first analysis in which Cuvieronius and Rhynchotherium were found as a natural group (Clade A, Fig 3), supported by the presence of the lower tushes and six other unambiguous synapomorphies. Additionally, if we consider the presence of lower tushes in Cuvieronius in previously published phylogenies [1,5,6,21,22] this feature is inferred to have occurred three times, independently, within the non-amebelodontine trilophodont gomphotheres (in Gomphotherium, Rhynchotherium and Cuvieronius). In the present study, the presence of lower tushes occurs only two times, independently, in Gomphotherium and in Clade A (Fig 3). It is important to highlight that, probably, the ancestor of the Clade A had lower tushes, at 6 Mya (Fig 3), and this feature may be a case of evolutionary reversal, because the lower tushes were lost (apomorphic/derived condition) at the base of the Clade C (Fig 3).
Another synapomorphy of Clade A related to the incisors is the presence of a pair of twisted upper tusks (Fig 4), which is also an exclusive and unique condition within the order Proboscidea [6,29]. The twisted upper tusks of Rhynchotherium falconeri and Cuvieronius hyodon (I2) are usually long, straight to slightly upcurved in lateral view, and symmetric (they mirror each other, in right-hand torsion, Fig 4). The growth of naturally twisted structures is usually interpreted as a associated response to environmental stresses. Besides, such structures are mechanically advantageous, enhancing their impact resistance and fracture toughness. In addition, in a study on narwhal tusks, the spiral mode of growth of the tooth ensures overall straightness of the incisor, even if it grows irregularly . In this way, the same process was probably positively selected in Clade A, as a response to a demand for long and resistant upper tusks. It is important to highlight that the upper tusks of Rhynchotherium and Cuvieronius are not spiraled to the same extent (the enamel band in the former seems to present more turns around the ivory than in the latter), and therefore Rhynchotherium probably presents upper tusks that are more resistant to mechanical stress than Cuvieronius. However, as we did not evaluate a large sample so far, the level of spiralization in the enamel band of upper tusks could also be individually or ontogenetically variable in both taxa.
Upper tusks morphotypes of Clade B.
The “dance” of tusks in Cuvieronius hyodon and its closest partners
An intense and fast aridification, along with cooling climatic changes occurred during the middle-late Miocene, Early Barstovian to Early Clarendonian (E2, Fig 3; ) in North America. Especially in the Southwest United States and the Great Plains, this drastic climatic shift also promoted an environmental change from Woodland to Savanna , which also possibly led to the loss of lower tusks in the Clade C (Fig 3). At some point between the early Clarendonian and the middle-late Hemphillian, the North American Savanna was the scenario for the origin and diversification of Clade B (Fig 3). This clade includes all New World non-amebelodontine trilophodont gomphotheres (Stegomastodon, Notiomastodon, Cuvieronius and Rhynchotherium) and the Asian Sinomastodon. Clade B is diagnosed by four unambiguous synapomorphies: the upper tusks upturned, absence of lower incisors, brevirostrine mandible and upper tusk alveoli slightly diverging. The brevirostrine condition of mandible occurred independently of (and later than) the loss of lower incisors (Fig 3). The association of longirostrine and tuskless mandibles is very rare in proboscideans, and only Choerolophodon (family Choerolophodontidae, sensu ), a recently described new genus of Mammutidae , as well as the non-amebelodontine trilophodont gomphotheres Gnathabelodon and Eubelodon show these two conditions simultaneously. Moreover, the emergence of brevirostry in Clade B appears to be related to a more open environment, the Savanna (Fig 3), which is in accordance with previous studies .
The presence of lower tushes is an unambiguous synapomorphy of Clade A, and it can be explained by the evolutionary process of taxic atavism [38,39,45,46]. This evolutionary process refers to the reappearance of a character state typical of a remote ancestor (in this case, the presence of lower tushes, the symplesiomorphic state in Gomphotherium) in a later deriving lineage (the common ancestor of Clade A). Taxic atavism is very rare in vertebrates [38,46]. Accordingly, the retrieval of the lower tushes in the ancestor of the Clade A probably occurred at 6 Mya (middle to late Hemphillian, Fig 3) at the Great Plains of North America  and it is a unique case within Elephantimorpha (sensu , Fig 3). The second peak of warmth and humidity (P2) at the E3 climatic event affected the Great Plains , and this climatic event is associated herein to the reacquisition of lower tushes in Clade A (Fig 3). However, after this warming period, the environmental humidity and temperature levels rapidly and drastically decreased again, allowing for the domination of C4 grasses in the Great Plains grasslands .
Therefore, due to this fast climatic change, when environment conditions became cooler and dryer, the evolution of lower tushes followed different paths in Rhynchotherium falconeri and Cuvieronius hyodon. The ontogenetic process that explains such structure in Clade A could be the Neotenic Paedomorphism, in which a feature grows in a slower velocity (heterochrony) when compared with the velocities of growth in an ancestor. Rhynchotherium could represent the full process of paedomorphism, since it retained the evergrowing lower tushes (di1) during all lifespan. Cuvieronius, in turn, presented a shorter period of retention of the lower tushes (typical juvenile structure), until an older developmental stage (J5; since Gomphotherium lost the di1 in Fetus/J1 developmental stages). Moreover, the pedomorphic neoteny is more evident in Rhynchotherium because, by definition, it generates adults that resemble or, in this case, bear juvenile features of their ancestors (presence of di1).
The “dance” of tusks in Elephantimorpha
Cuvieronius hyodon was recovered as a derived gomphothere in most previous phylogenetic hypotheses [1,5,6,21,22]. This position is extremely interesting, since Cuvieronius had lower tushes, a condition usually considered as plesiomorphic for Proboscidea . Moreover, the tetrabelodont condition is also considered plesiomorphic to Elephantimorpha. The loss of lower incisors is recorded in most lineages of Elephantimorpha, and only a few (Mammutidae, Amebelodontidae and non-amebelodontine trilophodont gomphotheres) retained lower incisors after the Miocene (Fig 5). Accordingly, we believe that the loss of lower incisors in Elephantimorpha occurred independently, in three occasions during the Miocene (E1, E2 and E3, Fig 5).
Time-calibrated phylogenetic relationship of Elephantimorpha in association to climatic fluctuations from Paleogene to Quaternary.
Therefore, what drove the negative selection of lower incisors in most lineages of Elephantimorpha? This natural group of Proboscidea was distributed throughout all continents from the Late Oligocene to the Pliocene/Pleistocene transition (except in Antarctica, Australia and South America; ) and only two taxa of non-amebelodontine trilophodont gomphotheres were recorded in South America during Pleistocene [3,14]. Here, we infer that due to the wide geographic range of Elephantimorpha during Miocene, the independent negative selection of its lower incisors may have been caused by global scale climatic events. Accordingly, the Miocene epoch is known by its climatic instability  and climatic fluctuations are one of the best candidates to promote such evolutionary selections (Fig 5). The climatic variations in global temperature and humidity throughout the Miocene, the "Zachos Curve” , is compared to a time-calibrated phylogenetic hypothesis for Elephantimorpha [6,23,44,49], most of the rapid cooling/drought events (E1, E2 and E3) coincides with the loss of lower incisors in distinct lineages of Elephantimorpha (Fig 5).
However, how did these climatic events negatively select the lower incisors in most lineages of Elephantimorpha? The highest heat transfer in Proboscidea occurs in the ears, trunk and incisors tips [51,52]. Thus, in a colder climatic condition, a tetrabelodont proboscidean (with four incisors: a lower and an upper pairs) would lose more heat to the environment than a dibelodont proboscidean (with only the upper pair of incisors). In addition, the heat conductivity of proboscidean ivory is low, mainly when compared to isolated dentine and enamel . Thus, incisors with large amount of enamel should also be responsible for a higher heat loss than incisors with a smaller amount of enamel (an enamel tip or small band).
Furthermore, the incisors of proboscideans have several important roles on their biology, including feeding and mating  and probably, because of these functions, the natural selection did not act negatively on both upper and lower incisors and positively preserved only one pair; in the case of Elephantimorpha, the upper one. In addition, we also infer that the loss of the enamel in tusks and the loss of lower incisors may be a response to drastic decrease in global temperature and humidity, as a way to preserve body thermal energy in a cold and dry environment.
The E1 climatic shift, which we infer as the one driving the negative selection of the lower tusks in Elephatimorpha, occurred at the transition Paleogene/Neogene (approximately 23.3 to 21 Mya) and comprised the loss of lower incisors in the most recent ancestor to Choerolophodon and other lineages of Elephantimorpha (Fig 5). Subsequently, Earth experienced a period of gradual increase in global temperatures and humidity, reaching the apex of these conditions around 17 to 15 Mya, a period known as the middle Miocene Climatic Optimum . The Miocene Climatic Optimum was followed by E2 (approximately 15 to 13 Mya), which is characterized by an abrupt decrease in global temperatures and humidity.
The second climatic event is marked by the radiation of dibelodont forms in two lineages of Elephantimorpha: the Mammutidae and the non-amebelodontine trilophodont gomphotheres (green bars in Fig 5), around 13.5 Mya. The loss of lower incisors occurred simultaneously in these two lineages, probably as a response to the colder and dryer climatic conditions after the Miocene Climatic Optimum. The decrease in global temperatures and humidity was less abrupt after the E2, but persisted until 8.5 Mya (Fig 5), when it was interrupted by E3 (the third and last climatic event). Broadly speaking, the E3 was a period of slow global cooling and drought. However, it also comprised two abrupt peaks of warmth and humidity (P1 and P2; Fig 5). No significant event in the evolution of Elephantimorpha occurred during the ascending of P1, but at the decline from P1 (cooling and drought climatic change) occurred the rise of dibelodont forms in tetralophodont gomphotheres (green bar in Fig 5), followed by the extinction of its tetrabelodont forms at the end of P1 (black bar in Fig 5).
The ascending to P2 is marked by the rise of tetrabelodont forms in the non-amebelodontine trilophodont gomphotheres, throughout the recovery of lower tushes in the common ancestor of the Rhynchotherium-Cuvieronius clade (clade A in Fig 5) after the loss of these structures at the origin of Clade C (Fig 4), which occurred in E2 (the yellow bar in Fig 5), probably as a response to the increasing of global temperature and humidity. Several evolutionary events occurred at the decline of P2, as the extinction of the tetrabelodont trilophodont gomphothere Gomphotherium (the black bar in Fig 5), the rise of dibelodont forms (green bars in Fig 5) of Stegodontidae and Elephantidae, and the extinction of the tetralophodont stegodontids and elephantids at the end of P2. After E3, the vertiginous increase of cold and dry climatic conditions was recorded, and all derived lineages of Elephantimorpha became extinct during the Pleistocene or at the Pleistocene/Holocene boundary, except for the dibelodont extant elephants Loxodonta and Elephas (Elephantidae; ).
Among Elephantimorpha, Cuvieronius hyodon is the unique tetrabelodont non-amebelodontine trilophodont gomphothere to reach the latest Pleistocene (INAH uncatalogued specimen, ). Nonetheless, to understand the “dance” of incisors in the non-amebelodontine trilophodont gomphotheres, the presence of lower tushes in Cuvieronius hyodon must be scored in a phylogenetic context.
We confirmed here the presence of vestigial lower tushes in Cuvieronius hyodon, which probably were in use from all juvenile stages (J1–J5) to the first of the young stage (Y1). The presence of lower tushes is one of the seven synapomorphies that diagnoses the well-supported clade including Cuvieronius hyodon and Rhynchotherium falconeri. Moreover, we recognized that the reappearance of lower tushes in this clade is an evolutionary process of taxic atavism in Proboscidea, and a case of pedomorphic neoteny, since adult Rhynchotherium presented these “milk and juvenile” structures.
The loss of enamel in incisors and the loss of lower incisors in all lineages of Elephantimorpha is probably a result of a negative selection related to the decrease of global temperature and humidity, occurred after the Miocene Climatic Optimum (26 to 15 Mya) on. The recovering of the lower tushes by the clade including Cuvieronius hyodon and Rhynchotherium falconeri is an unique case of taxic atavism within the Elephantimorpha, and is possibly related to an increasing apex of temperature and humidity during the Late Miocene (around 6 Mya) at the Great Plains. In addition, the twisted upper tusks of Rhynchotherium and Cuvieronius is also an exclusive synapomorphy of this clade and probably occurred in response to the need of long and resistant upper incisors.
Materials and Methods
We analyzed 46 mandibles of Cuvieronius hyodon, including young and adult individuals from Tarija (Bolivia), Mexico, Costa Rica and United States (Table 1). All mandibles analyzed in this study were previously identified as Cuvieronius hyodon through the association of diagnostic materials, such as upper tusks and/or skulls . The studied specimens are housed in the paleontological collections of Museo de Ciencias Naturales ‘Bernardino Rivadavia’ (MACN-Pv), Argentina; Museo Nacional de Paleontología y Arqueología de Tarija (MNPA or TAR), Bolivia; American Museum of Natural History (AMNH-DVP, F:A.M.), United States of America; Escuela Centroamericana de Geología de la Universidad de Costa Rica (UCR), Costa Rica; Instituto Nacional de Antropología y Historia (INAH) and Centro de Geociencias de la Universidade Nacional Autónoma de México, Campus Juriquilla, Mexico; Swedish Museum of Natural History, Sweden (NMR); and Muséum National d’Histoire Naturelle, France.
Each mandible was assigned to a developmental stage based on the functional teeth under usage before death and their stages of wearing (modified from [7,8,20]). The developmental stages for Mammut americanum were based on pre-molars and molars eruption and wear stages, from "Fetus" to "Young adult” (total of nine developmental stages, ), while others 23 dental age classes were suggested to Gomphotherium angustidens . However, comparisons across Gomphotheriidae and Mammutidae indicate their series of teeth wear patterns may differ, since G. angustidens presents permanent pre-molars and subsequent teeth usually differ from one to two wear stages in brevirostrine gomphotheres , which does not happen in Mammut. In this way, we adapted the wear stages proposed to Mammut and Gomphotherium to best fit the developmental stages identified in Cuvieronius hyodon (total of eleven developmental stages; see Table 1), by changing the developmental stage "Young adult” to Y3 and proposing the developmental stages Y4 (dp4 in severe wear and m1 in wear stage 2) and Adult (only permanent molars in use).
Moreover, we analyzed the presence of lower incisors and/or lower incisors alveoli in each mandible, with the alveoli being classified into two categories : open alveoli (a pair or a single cylindrical cavity, usually symmetric and posteriorly directed) or closing alveoli (a pair or a single oval depression, usually symmetric, filled with spongy tissue). The morphological features of Cuvieronius hyodon lower incisors were described and compared to those of immature individuals of Rhynchotherium falconeri  and Gomphotherium angustidens [9,28,30].
We performed a phylogenetic analysis in order to elucidate the relationships of Cuvieronius hyodon within the non-amebelodontine trilophodont gomphotheres (sensu ). In this study, we use this term instead of trilophodont gomphotheres to represent the lineage inclusive of Gomphotherium, Rhynchotherium, Gnathabelodon, Eubelodon, Stegomastodon, Cuvieronius, Notiomastodon and Sinomastodon . We propose a new data matrix of twenty-four dental, mandible and cranial characters (data in S1 Appendix). Our dataset consists of characters modified from the literature [1,6,21] as well as new characters, such as the downward deflection of the mandibular symphysis and the difference in height between the mandibular condyle and the coronoid process (data in S1 Appendix). We run the data matrix in the software TNT , using the exact search algorithm “implicit enumeration” and characters with equal weights. The outgroup is represented by Gomphotherium angustidens, and the ingroup includes Eubelodon morrilli, Gnathabelodon thorpei, Rhynchotherium falconeri, Cuvieronius hyodon, Stegomastodon (only North American species, see ), Notiomastodon platensis (= Haplomastodon chimborazi, “Stegomastodon” waringi, “Stegomastodon” platensis) and the Asian taxon Sinomastodon, as it is commonly considered closely related to New World gomphotheres . However, we did not include Amahuacatherium and Megabelodon in the present analysis. Whereas Amahuacatherium has been previously considered to be an invalid genus , Megabelodon is probably part of Gomphotherium , a position which we agree with. We considered all species of Stegomastodon (North American species ) and Sinomastodon  during character-state scoring because the species diversity of these taxa is still under revision.
The authors are grateful to all curators of the collections cited above for allowing the access to the Proboscidea specimens that supported this study, to Dr. Wendy Dirks, Dr. Marco Aurélio Gallo de França and Doctoral Candidate Tiago Simões for the English improvements and suggestions and to Doctoral Candidate Arnaud Schmitt for contribution accessing Cuvieronius specimens.
This work was funded by Conselho Nacional de Desenvolvimento Científico e Tecnológico (140453/2012-01, 201081/2014-8 - DM; 248772/2013-9 - LSA), www.cnpq.br, and Fundação Carlos Chagas de Amparo à Pesquisa do Estado do Rio de Janeiro (E-26/100.246/2014 – DM; 204036-E_25/2014-Jovem Cientista do Nosso Estado - LSA), www.faperj.br. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
All relevant data are within the paper and its Supporting Information files.
1. Ferretti MP, Anatomy of Haplomastodon chimborazi (Mammalia, Proboscidea) from the late Pleistocene of Ecuador and its bearing on the phylogeny and systematics of South American gomphotheres. Geodiversitas. 2010; 32 (4): 663–721.
2. Lucas SG, The palaeobiogeography of South American gomphotheres. J Palaeogeogr. 2013; 2 (1): 19–40.
3. Mothé D, Avilla LS, Mythbusting evolutionary issues on South American Gomphotheriidae (Mammalia: Proboscidea). Quatern Sci Rev. 2015; 110 (2015): 23–35.
4. Mothé D, Avilla LS, Cozzuol MA, Winck GR, Revision of the Quaternary South America lowland gomphotheres (Mammalia: Proboscidea: Gomphotheriidae). Quatern Int. 2012; 276 (27): 2–7.
5. Prado JL, Alberdi MT, A cladistic analysis among trilophodont gomphotheres (Mammalia, Proboscidea) with special attention to the South America genera. Palaeontology. 2008; 51 (4), 903–915.
6. Shoshani J, Tassy P, The Proboscidea: evolution and palaeoecology of elephants and their relatives. 1st ed Oxford: Oxford University Press; 1996.
7. Green JL, Hulbert RC Jr, The deciduous premolars of Mammut americanum (Mammalia, Proboscidea). J Vert Paleo. 2005; 25: 702–715.
8. Mothé D, Avilla LS, Winck GR, Population structure of the gomphothere Stegomastodon waringi (Mammalia: Proboscidea: Gomphotheriidae) from the Pleistocene of Brazil. An Acad Bras Cienc. 2010; 82 (4): 983–996. [PubMed]
9. Tassy P, L’odontologie de Gomphotherium angustidens (Cuvier, 1817) (Proboscidea, Mammalia): données issues du gisement d’En Péjouan (Miocène moyen du Gers, France). Geodiversitas. 2014; 36 (1): 35–115.
10. Nordenskiöld E, Über die Saugetierfossilien des Tarijatals, Södamerika. I. Mastodon andium Cuv. Kungle Svenska Vetenskaps Akademiens Handlingar. 1903; 37: 1–30.
11. Boule M, Thevenin A, Mammifères fossiles de Tarija, 1st ed Paris: Imprimerie nationale, Mission scientifique G. de Crequi-Monfort et F. Senechal de la Grange; 1920.
12. Liendo-Lazarte M, Algunas consideraciones sobre los Mastodontes de ‘Tarija. Mus Nac Tiahuanacu Sec Paleont. 1946; 1–14.
13. Laurito CA, Estudio de un ejemplar de Cuvieronius hyodon hallado en el lecho del Río Nacaome, Ouanacaste, Costa Rica. Rev Geol Amér Central. 1990; 11: 41–50.
14. Ferretti MP, Enamel structure of Cuvieronius hyodon (Proboscidea, Gomphotheriidae) with a discussion on enamel evolution in Elephantoids. J Mammal Evol. 2008; 15: 37–58.
15. Montellano-Ballesteros M, New Cuvieronius finds from the Pleistocene of central Mexico. J Paleontol. 2002; 76: 578–583.
16. Prado JL, Alberdi MT, Sanchez B, Azanza B, Diversity of the Pleistocene Gomphotheres (Gomphotheriidae, Proboscidea) from South America. In: Reumer, J.W.F., De Vos, J., Mol, D. (Eds.), Advances in Mammoth Research, Proceedings of the Second International Mammoth Conference, Rotterdam, Deinsea. 2003; 9: 347–363.
17. Alberdi MT, Prado JL, Salas R, The Pleistocene Gomphotheriidae (Proboscidea) from Peru. N. Jb. Geol. Paleaont. Abh. 2004; 231, 423–452.
18. Lucas SG, Cuvieronius (Mammalia, Proboscidea) from the Pleistocene of Ulloma, Bolivia In: Lucas S.G., Morgan G.S., Spielmann J.A., Prothero D.R. (Eds.), Neogene Mammals, Albuquerque, N Mex Nat Hist Sci Mus Bull. 2008; 44: 399–402.
19. Lucas SG, Taxonomic nomenclature of Cuvieronius and Haplomastodon, proboscideans from the Plio-Pleistocene of the New World. N Mex Nat Hist Sci Mus Bull. 2009; 44: 409–415.
20. Lucas SG, Alvarado GE, Fossil Proboscidea of the upper Cenozoic of Central America: taxonomy, evolutionary and paleobiogeographic significance. Rev Geol Am Cent. 2010; 42: 9–42.
21. Mothé D, Avilla LS, Cozzuol M, The South American Gomphotheres (Mammalia, Proboscidea, Gomphotheriidae): Taxonomy, Phylogeny, and Biogeography. J Mammal Evol. 2013; 20: 23–32.
22. Cozzuol MA, Mothé D, Avilla LS, A critical appraisal of the phylogenetic proposals for the South American Gomphotheriidae (Proboscidea: Mammalia). Quatern Int. 2012; 255: 36–41.
23. Gheerbrant E, Tassy P, L'origine et l'évolution des éléphants. Comptes Rendus Palevol. 2009; 8: 281–294.
24. Hoffstetter R, Les Mammifères Pléistocènes de la République de l’Équateur. In: Mémoires de la Société Géologique de France, Nouvelle Serie. 1952; 31.
25. Frick C, New Remains of Trilophodont-Tetralophodont mastodons. Bull Am Mus Nat Hist. 1926; 59: 505–652.
26. Aguilar RH, Description of a newborn gomphothere skull (Mammalia, Gomphotheriidae) from Lake Chapala, Jalisco, Mexico. New Mex Mus Nat Hist Sc