Mastodon Temporal range: Late Miocene – early Holocene 8–0.011 Ma PreꞒ Ꞓ O S D C P T J K Pg N (Possible earliest record of up to ~10 Ma)
Mounted M. americanum skeleton ("Warren mastodon"), American Museum of Natural History
Scientific classification
Domain:	Eukaryota
Kingdom:	Animalia
Phylum:	Chordata
Class:	Mammalia
Order:	Proboscidea
Family:	†Mammutidae
Genus:	†Mammut Blumenbach, 1799
Type species
†Elephas americanus (= †Mammut americanum) Kerr, 1792
Other species
†M. matthewi Osborn, 1921 †M. vexillarius Matthew, 1930 †M. raki Frick, 1933 †M. nevadanum Stock, 1936 †M. cosoensis Schultz, 1937 †M? furlongi Shotwell & Russell, 1963 †M. pacificum Dooley et al., 2019 Species pending reassessment †M. borsoni Hays, 1834 †M. obliquelophus Mucha, 1980 †M. lufugense Zhang, 1982 †M. zhupengensis Zhang et. al., 1991
Synonyms
Genus synonymy Harpagmotherium Fischer von Waldheim, 1808 Mastotherium Fischer von Waldheim, 1814 Mastodon Cuvier, 1817 Tetracaulodon Godman, 1830 Missourium Koch, 1840 Leviathan Koch, 1841 Pliomastodon Osborn, 1926 Synonyms of M. americanum Elephas americanus Kerr, 1792 Mammut ohioticum Blumenbach, 1799 Elephas macrocephalus Camper, 1802 Harpagmotherium canadense Fischer de Waldheim, 1808 Elephas mastodontus Barton, 1810 Mastotherium megalodon Fischer de Waldheim, 1814 Tapirus mastodontoides Harlan, 1825 Tetracaulodon mastodontoideum Godman, 1830 Mastodon ohioticum Eichwald, 1832 Mastodon cuvieri Hays, 1834 Mastodon jeffersoni Hays, 1834 Tetracaulodon collinsii Hays, 1834 Tetracaulodon godmani Hays, 1834 Tetracaulodon tapyroides Hays, 1834 Elephas ohioticus de Blainville, 1839–1864 Missourium kochii Koch, 1840 Leviathan missourii Koch, 1840 Tetracaulodon osagii Koch, 1841 Tetracaulodon kochii Koch, 1841 Tetracaulodon bucklandii Grant, 1842 Missourium theristocaulodon Koch, 1843 Mastodon rugatum Koch, 1845 Elephas rupertianus Richardson, 1854 Trilophodon ohioticus Falconer, 1868 Mammut progenium Hay, 1914 Mastodon americanus plicatus Osborn, 1926 Mammut oregonense Hay, 1926 Mastodon moodiei Barbour, 1931 Mastodon americanus alaskensis Frick, 1933 Mastodon acutidens Osborn, 1936 Synonyms of M. matthewi Mastodon matthewi Osborn, 1921 Pliomastodon sellardsi Simpson, 1930 Pliomastodon adamsi Hibbard, 1944 Synonyms of M. vexillarius Pliomastodon vexillarius Matthew, 1930 Synonyms of M. raki Mastodon raki Frick, 1933 Synonyms of M. nevadanum Pliomastodon nevadanus Stock, 1936 Synonyms of M. cosoensis Pliomastodon cosoensis Schultz, 1937 Synonyms of "M." borsoni Mastodon vellavus Aymard, 1847 Mastodon vialleti Aymard, 1847 Mastodon buffonis Pomel, 1848 Mastodon affinis Pomel, 1859 Zygolophodon borsoni Osborn, 1926 Mastodon pavlowi Osborn, 1936 Mammut shansiense Chow & Chang, 1961 Synonyms of "M." obliquelophus M. praetypicum? Schlesinger, 1917

A mastodon (mastós 'breast' + odoús 'tooth') is a member of the genus Mammut (German for "mammoth"), which, strictly defined, was endemic to North America and lived from the late Miocene to the early Holocene. Mastodons belong to the order Proboscidea, the same order as elephants and mammoths (which belong to the family Elephantidae). Mammut is the type genus of the extinct family Mammutidae, which diverged from the ancestors of modern elephants at least 27-25 million years ago, during the Oligocene.

M. americanum, known as an "American mastodon" or simply "mastodon," had a long and complex paleontological history spanning all the way back to 1705 when the first fossils were uncovered from Claverack, New York in the American colonies. Naturalists struggled to explain the affinities of the proboscidean because of its uniquely shaped molars, which have no modern analogues in terms of large mammals. The American mastodon caught the attention of not only European researchers but also influential Americans before and after the American Revolution. American historians of the 21st century, such as Paul Semonin and Keith Stewart Thomson, have argued that findings and displays of its fossils had helped to bolster American nationalism and contributed to a greater understanding of extinctions. It is known by many skeletons which are now typically on display in American museums, a trend first started by Charles Willson Peale in 1804.

Taxonomically, M. americanum was first recognized as a distinct species by Robert Kerr in 1792 then classified to its own genus Mammut by Johann Friedrich Blumenbach in 1799, thus making it amongst the first fossil mammal genera to be erected with undisputed taxonomic authority. However, the original genus name was overshadowed by "Mastodon" in the 19th century, which was established formally by the French naturalist Georges Cuvier in 1817. The genus became problematic as it served as a wastebasket taxon for over 20 fossil proboscidean species whose dentitions more closely resembled that of M. americanum than those of elephantids or deinotheres up to the early 20th century. After major revisions of proboscidean taxonomy, the genus Mammut today is defined as the valid genus name and includes 7 definite species, 1 of questionable affinities, and 4 other species from Eurasia that are pending reassessments to other genera.

As a member of the Mammutidae, it is defined by its molars which have a zygodont morphology (where the cusps of the teeth are merged into ridges) which have remained evolutionarily conservative (little changed) throughout the evolution of the family. In comparison to its likely ancestor Zygolophodon, Mammut is evolutionarily derived with a reduction to loss of the lower tusks, shortening of the mandibular symphysis (the frontmost part of the lower jaw), and particularly long and upward curving upper tusks. As a result of its conservative evolutionary history, Mammut consistently occupied specializations to browsing on leaves, fruits, and woody plants. This allowed mastodons to niche partition with other members of Proboscidea in North America, like gomphotheres and the Columbian mammoth, who had shifted to mixed feeding or grazing by the late Neogene-Quaternary. Mammut as a result was a highly successful proboscidean that had achieved maximum diversity in the Pliocene then is known from abundant fossil evidence in the Late Pleistocene. It is thought to have had behaviors not much different from elephants and mammoths, living in herds, displaying strong degrees of sexual dimorphism, and entering phases of aggression under musth in the case of males.

Archeological evidence reveals that mastodons for at least a few thousand years coexisted with Paleoindians, who were the first humans to have dispersed to North America and at least sometimes hunted them based on kill sites found in both pre-Clovis and Clovis cultural phases.

Mastodons disappeared along with many other North American animals, including most of its largest animals (megafauna), as part of the Late Pleistocene extinctions around the end of the Late Pleistocene-early Holocene, the causes typically being attributed to Clovis culture hunting, severe climatic phases like the Younger Dryas, or some combination of the two. The American mastodon had its last recorded occurrence in the earliest Holocene around 11,000 years ago, which is considerably later than other North American megafauna species. Today, the American mastodon is one of the most well-known fossil species in both academic research and public perception, the result of its inclusion in American popular culture.

Research history

Earliest finds

In a letter dating to 1713, Edward Hyde, 3rd Earl of Clarendon (known also as Lord Cornbury) from New York reported to the Royal Society learned society of Great Britain that in 1705, a large-sized tooth was found near the side of the Hudson River by a Dutch country-fellow and was sold to New York General Assembly member Van Bruggen for a gill of rum, and Bruggen eventually gave it to Cornbury. He then stated that he sent Johannis Abeel, a recorder of Albany, New York to dig near the original site of the tooth to find more bones.^[1][2]

Abeel reported in a later that he went to the town of Claverack, New York where the original bones were found. American historian Paul Semonin said that the accounts written by Cornbury and Abeel match up with that written by in the July 30, 1705 entry in The Boston News-Letter.^[3] The account reported skeletal evidence of an antediluvian (or biblical) "giant" uncovered from Claverack. The femur and one of the teeth both dissolved before they could be further observed, however.^[4][1]

Big Bone Lick

In 1739, a French military expedition under the command of Charles III Le Moyne (known also as "Longueil") explored the locality of "Big Bone Lick" (located in what is now the US state of Kentucky) and gathered fossil bones and teeth there.^[5] The French naturalist Louis Jean-Marie Daubenton examined the fossil collection brought by Longueuil and compared it with specimens of extant elephants and Siberian mammoths in 1762. Daubenton said that the bones were discovered by Native Americans (probably Abenaki hunter-warriors). He came to the conclusion that the femur and tusk belonged to an elephant while the molars (or cheek teeth) came from a separate giant hippopotamus.^[6][7][8]

In Shawnee tradition, the proboscideans roamed in herds and were hunted by giants, who both eventually died out. The accounts told by the Shawnee individuals in 1762 are the oldest known documented interpretations of the "Ohio" fossils, although the traditions may have had been told for generations.^[9][10]

In 1767, Peter Collinson credited Irish trader George Croghan for having sent him and Benjamin Franklin fossil evidence of the mysterious proboscideans, using them for his studies. He concluded that the peculiar grinders (the molars) were built for herbivorous diets of branches of trees and shrubs as well as other vegetation, a view later followed by Franklin.^[11][12]

In 1768, Scottish anatomist William Hunter recorded that he and his brother John Hunter observed that the teeth were not like those of modern elephants. He determined that the "grinders" from Ohio were of a carnivorous animal but believed that the tusks belonged to the same animal. After examining fossils from Franklin and Lord Shelburne, Hunter was convinced that the "pseudo-elephant", or "animal incognitum" (shortened as "incognitum"), was an animal species separate from elephants that might have also been the same as the proboscideans found in Siberia. He concluded his article with the opinion that although regrettable to philosophers, humanity should be thankful to heaven that the animal, if truly carnivorous, was extinct.^[13]

Early American observations

In 1785, Reverend Robert Annan wrote an account recalling an event in which workers discovered bones in his farm near the Hudson River in New York in fall of 1780. The workers found four molars in addition to another that was broken and thrown away. They also uncovered bones, including vertebrae that broke shortly thereafter. Annan expressed his confusion at what the animal could be but speculated based on its "grinders" that it was carnivorous in diet. He speculated also that it was probably extinct due to some catastrophe within the globe.^[14]

American statesman Thomas Jefferson stated his thoughts on Notes on the State of Virginia (published by 1785) that the fossil proboscideans may have been carnivorous, still exist in the northern parts of North America, and are related to mammoths whose remains were found in Siberia. Jefferson referenced the theory of American social degeneracy by Georges-Louis Leclerc, Comte de Buffon, countering it by using extant and extinct animal measurements, including those of "mammoths," as proof that North America faunas were not "degenerative" in size.^[15] Semonin pointed out that social degeneracy was an offensive concept to Anglo-American naturalists and that the American proboscidean fossils were used as political tools to inspire American nationalism and counter against the theory of American degeneracy.^[16][17]

In 1799, laborers recovered a thighbone while digging a marl pit at John Masten's farm in Newburgh, New York, and subsequent excavations were observed by a crowd of over a hundred people.^[18] American painter and exhibitionist Charles Willson Peale visited the locality in 1801, where he first sketched the fossils then purchased excavation privileges and full ownership of the fossils from Masten and borrowed a loan from the American Philosophical Society (APS) in Philadelphia, Pennsylvania. In addition to the first skeleton, the second was excavated using a mill-like device to drain a 12 ft (3.7 m) deep marl pit. Peale assembled a complete skeleton in his Philadelphia Museum in 1804, and its exhibit was open first to invited members of the American Philosophical Society on December 24 then to the general public on December 25 for an exhibit admission fee in addition to the general admission fee.^[19]

The special exhibition attracted thousands of visitors, and the skeleton became a US national symbol.^[20] Charles Peale's son Rembrandt Peale took the skeleton to Europe used to promote the fossil proboscidean and have it used as support for Jefferson's final rebuttals against Buffon's arguments for supposed inferiority of American faunas. Author Keith Stewart Thomson argued that the promotion of the "mastodon" skeleton made it a symbol of the strength of American nationalism and that "mammoth" as a term became associated with gigantism. Decades later, the museum bankrupted, and the first skeleton's specimens were sold to some German spectators in around 1848, who eventually sold it to Hessisches Landesmuseum Darmstadt in Germany where it is now displayed. The second skeleton's specimens landed eventually at the American Museum of Natural History.^[21]

Excavation of a specimen in a golf course in Heath, Ohio, 1989 (left) and a replica of the "Burning Tree mastodon" complete skeleton (right)

Other skeletons of Mammut americanum were excavated within the United States in the first half of the 19th century. One of them was collected by American showman Albert C. Koch in what is today the Mastodon State Historic Site at Missouri in 1839. He hypothesized in 1840 that the proboscidean, which he classified as Missourium, was much larger than an elephant, had horizontal tusks plus trunks, and occupied aquatic habitats.^[22] He acquired additional fossils from a spring on the Pomme de Terre River to assemble a mounted skeleton of the "Missouri Leviathan" and briefly exhibited it at St. Louis. After exhibiting the skeleton throughout Europe, he sold the skeleton to the British Museum of Natural History. Richard Owen then properly reassembled the skeleton, and it today is on display there.^[23][24]

In 1845, another skeleton was excavated from Newburgh by laborers hired by Nathaniel Brewster initially to remove lacustrine deposits to fertilize the neighboring fields. They were observed by a large amount of spectators and uncovered relatively complete fossil evidence of M. americanum.^[25][26] The skeleton was exhibited in New York City and other New England towns then was acquired by John Collins Warren for study.^[27][28] After Warren's death in 1856, the skeleton was sent to Warren's family but was traded to Harvard Medical School for John Warren's skeleton. The "Warren mastodon", under the request of American paleontologist Henry Fairfield Osborn, was purchased by the American financier J. P. Morgan for $30,000 in 1906 and donated to the American Museum of Natural History where it is exhibited today.^[29][26]

Early taxonomic history

In the 1790s, the "American incognitum" was subject to research by multiple taxonomists. Scottish writer Robert Kerr erected the species name Elephas americanus in 1792 based on fossil tusks and "grinders" from the Big Bone Lick locality. He stated that the tusks were similar to elephants while the molars were completely different because they were covered with enamel and had a double row of high conical cusp processes. Kerr was unsure about the taxonomic affinities of the molars and referenced that Thomas Pennant supposed that they belong to an unknown species within the genus Elephas, giving the common name "American elephant."^[30]

German naturalist Johann Friedrich Blumenbach also followed up with more taxonomic descriptions of fossil proboscideans in 1799. The first fossil species, recovered from Germany, was described as belonging to the newly erected species Elephas primigenius? (now known as Mammuthus primigenius). The second was what he considered to be an unknown "colossal land monster of the prehistoric world," considering it to be the "mammoth." He created the genus Mammut and erected the species Mammut ohioticum based on fossil bones dug up from Ohio in North America. He said that the species was distinguished from other animals of the prehistoric world based on the unusual shapes of the large molars. The genus name "Mammut" refers to the German translation for "mammoth."^[31] The naming of the genus Mammut in 1799 makes it the second or third genus to be recognized with taxonomic authority given that Megalonyx had been named the same year.^[32]

French naturalist Georges Cuvier also described known fossil proboscidean species back in 1796, although his account was later published in 1799. He considered that the remains uncovered from Siberia were true "mammoths" that had similar dentitions to extant elephants but had some morphological differences. He mentioned the fossil remains that were brought back by Longueil from Ohio back in 1739 and several researchers from previous decades who noted the unusual molars and thought that they belonged to different animals like hippopotamuses. He followed recognition in the previously established species "Elephas americanus" and argued that the species was different from elephants and mammoths and cannot be found amongst living animals due to extinction from catastrophism.^[33][34]

The proboscidean species was subject to several other species names given by other taxonomists within the earliest 18th century as well as the genus name Harpagmotherium by the Russian naturalist Gotthelf Fischer von Waldheim in 1808.^[24]

Cuvier's taxonomy

In 1806, Cuvier wrote multiple extended research articles on fossil proboscideans of Eurasia and the Americas. He stated that the bones that Buffon previously described from North America were not of elephants but another animal that he referred to as the "mastodonte," or the "animal of Ohio."^[35] He reinforced the idea that the extinct "mastodon" was an animal close in relationship to elephants that differed by jaws with large tubercles. He suggested that "mammoth" and "carnivorous elephant" be discontinued as names for the species and that it receive a new genus name instead. Cuvier said that for "mastodonte," he derived the name's etymology (compound μαστός (mastós, "breast") + ὀδούς (odoús, "tooth") from Ancient Greek to mean "nipple tooth," since he thought that it expressed the characteristic form of the teeth.^[36]

In 1817, the French naturalist officially established the genus name Mastodon, reaffirming that it is extinct and has left no living descendants. He established that it had an overall body form similar to elephants but had molars more similar to hippopotamuses and pigs that did not serve to grind meat. The first species he erected within Mastodon was Mastodon giganteum, giving it the informal name "great mastodon" and writing that that it is designated to the Ohio proboscidean with abundant fossil evidence, equal size but greater proportions to modern elephants, and diamond-shaped points of the molars. The naturalist also created the second species name Mastodon angustidens and gave it the informal name "narrow-toothed mastodon," diagnosing it as having narrower molars, smaller sizes compared to M. giganteum, and range distributions in Europe and South America.^[37] Cuvier also erected several other species of Mastodon originating from other continents in 1824.^[38] Despite Cuvier's genus name being younger than multiple other genus names, Mastodon became the most commonly used genus name for the 19th century.^[39][24]

Taxonomic problems

"Mastodon" was riddled with major taxonomic problems since species now determined as belonging to other proboscidean genera were classified to Mastodon on the basis of similar dentitions to that of "Mastodon giganteum" (= Mammut americanum), effectively making it a wastebasket taxon.^[37][38][40] Various fossil proboscidean species were classified into Mastodon in the 19th century before eventually being reclassified into distinct genera.^[24] In addition to still-valid species names, several synonymous or dubious species names ultimately belonging to different genera were erected within the Americas as well throughout the 19th century.^[41][42][43] Also, many species names erected based on M. americanum remains were erected. As a result, M. americanum has many synonymous names. The issue of synonymous species names were especially apparent in the first half of the 19th century.^[24]

Today, the genera that include species formerly classified into Mastodon include Gomphotherium (G. angustidens, G. pyrenaicum, G. productum, G. libycum, G. subtapiroideum, G. steinheimense),^[44][45][46] Zygolophodon (Z. turicensis, Z. proavus),^[47][48] Cuvieronius (C. hyodon),^[49] Stegodon (S. elephantoides),^[50] Stegolophodon (S. latidens, S. cautleyi),^[51] Anancus (A. avernensis, A. sivalensis, A. perimensis),^[52] Tetralophodon (T. longirostris),^[53] Choerolophodon (C. pentelici),^[54] Stegomastodon (S. mirificus),^[55] Rhynchotherium ("R." euhypodon),^[42] Stenobelodon (S. floridanus),^[56] and Notiomastodon (N. platensis).^[41]

In 1830, American naturalist John Davidson Godman created the genus Tetracaulodon plus its species T. Mastodontoideum based on what he determined to be differences between it and Mastodon based on the skull and dentition.^[57] Both Richard Harlan and William Cooper pointed out that except for the tusks, all other characteristics of the specimens were consistent with M. giganteum. They therefore argued that there was no reason to assume that the tusks were not just individual variations, a view followed also by George William Featherstonhaugh. Isaac Hays comparatively defended Godman's taxon, which led to a bitter debate regarding the validity of the genus amongst American naturalists.^[58]

The validities of both Tetracaulodon and Missourium were rejected by Owen in 1842, although he retained the former name informally.^[59] By 1869, American paleontologist Joseph Leidy determined that Mastodon americanus is the senior species synonym and listed M. giganteum as a junior synonym. He also listed Mammut, Harpagmotherium, Mastotherium, Missourium, and Leviathan as synonyms of Mastodon. He also noted that M. americanum as a species was highly variable in morphology.^[60][61]

In 1902, American paleontologist Oliver Perry Hay listed Mammut as the prioritized genus name given its status as the oldest genus name, making Mastodon, Tetracaulodon, and Missourium classified as junior synonyms. He also established M. americanum as the type species.^[39] The genus name Mastodon was subsequently abandoned by many American paleontologists in favor of Mammut within the early 20th century.^{[62][63][64][24]} In 1942, American paleontologist George Gaylord Simpson said that for his study, he prioritized the historic plus taxonomically correct name Mammut over Mastodon.^[65] He continued prioritizing Mammut in 1945, stating that people were generally aware of its taxonomic priorities over Mastodon and that people had refused to use it. He stated that he did not want to either but reluctantly set aside his personal preferences to follow taxonomic rules.^[66]

Additional species

In 1921, Osborn created the species name Mastodon matthewi based on distinct molars from the Snake Creek Formation of western Nebraska, naming it in honor of William Diller Matthew. He also erected another species M. merriami from the Thousand Creek Formation in Nevada, which was eventually synonymized with Zygolophodon proavus.^[67][48] Osborn in 1926 followed up for Mastodon matthewi by establishing the genus Pliomastodon for the species based on cranial differences from "Miomastodon" (= Zygolophodon).^[68]

In 1930, Matthew erected a second species for Pliomastodon named P. vexillarius based on fossil material from the locality of Elephant Hill in California, determining that it differs from Mammut by differences in the skull and that the etymology of the species name was made in honor of paleontological contributions by the Standard Oil Company of California.^[69][48]

In 1933, Childs Frick named the species Mastodon raki from the locality of Truth or Consequences, New Mexico based on differences on the heel and M₃ tooth from M. americanus, otherwise having proportions similar to it.^[70][48] In 1936, Chester Stock published the species name Pliomastodon nevadanus based on fossils from the Thousand Creek Beds of northwestern Nevada.^[71] In 1937, John R. Schultz created the species name Pliomastodon? cosoensis, naming it after the Coso Mountains in Inyo County, California where skull fossils were recovered.^[72]

In 1963, J. Arnold Shotwell and Donald E. Russell created another species Mammut (Pliomastodon) furlongi, assigning it to fossils collected from the Juntura Formation of Oregon. The species name was created in honor of Eustace L. Furlong, who made early fossil collections from the western side of the Juntura Basin.^[73]

The genus Pliomastodon was synonymized with Mammut while Miomastodon was synonymized with Zygolophodon by Jeheskel Shoshani and Pascal Tassy in a 1996 appendix,^[74] a view that was followed by other authors in later years.^[75][76][48]

In 2019, Alton C. Dooley Jr. et al. established Mammut pacificus based on fossils collected from the Diamond Valley Lake in Hemet, California. They also stated that M. oregonense is a nomen dubium and that further analysis needs to be done to confirm whether or not M. furlongi belongs to Zygolophodon instead.^[76]

In 2023, Wighart von Koenigswald et al. reviewed the North American species of Zygolophodon and Mammut. They synonymized P. adamsi and P. sellardsi with Mammut matthewi and emended M. nevadanus and M. pacificus to M. nevadanum and M. pacificum, respectively. They also said that they were uncertain of the taxonomic status of M. furlongi, specifically whether or not it was a variant of sexual dimorphism of Z. proavus. Some authors have considered M. nevadanum to be synonymous with M. matthewi while others had retained validity of the species name.^[48][76]

Several mammutid species outside of North America are classified to Mammut (or "Pliomastodon"), namely M. borsoni, M. obliquelophus, M. zhupengensis, and M. lufugense (possibly synonymous with M. obliquelophus).^{[77][78][79][80]} Recent research such as that of von Koenigswald et al. in 2023 warned that the genus Mammut should be carefully used for non-North American species.^[48]

Classification and evolution

Mammut is the type genus of the Mammutidae, the sole family of the elephantimorph clade Mammutida (the other elephantimorph clade is Elephantida). The Mammutidae is characterized by molars with zygodont-form crests, which have remained morphologically conservative throughout the evolutionary history of the family. Mammut is considered to be a derived genus of the family because of strong zygodont development.^[81] As a family of the Elephantimorpha clade, it is only distantly related to the Deinotheriidae due to major differences in dentition and emergence of adult teeth.^[82] The Mammutidae is identified as a monophyletic clade, meaning that it did not leave any derived descendant groups in its evolutionary history.^[83] The monophyly of the Mammutidae makes it differ from the Elephantida, where the Gomphotheriidae is paraphyletic (or ancestral to more derived descendant groups in the cladistic sense) in relation to the derived elephantoid families Stegodontidae and Elephantidae (elephants, mammoths, and relatives).^[84]

Although the separation of the Mammutida and Elephantida is strongly supported based on morphological differences, their origins within the late Paleogene remain uncertain. One hypothesis asserts that the Elephantimorpha is monophyletic if the primitive Elephantiformes genus Phiomia was truly ancestral to both the Elephantida and Mammutida. An alternate hypothesis suggests that the Elephantimorpha is diphyletic because Phiomia is ancestral to gomphotheres while Palaeomastodon is ancestral to mammutids.^[82] The earliest undisputed mammutid genus Losodokodon is recorded in Kenya, Africa and firmly establishes the earliest presence of mammutids in the late Oligocene (~27-24 Ma). The Mammutidae, like other Paleogene proboscideans, was therefore an endemic radiation within the continent akin to other endemic mammals like arsinoitheres, hyracoids, and catarrhine primates plus non-endemics such as anthracotheres and hyaenodonts.^[85]

In the early Neogene phase of evolution, Eozygodon made an appearance in the earliest Miocene (~23-20 Ma) of Africa after Losodokodon. Eozygodon was subsequently succeeded by Zygolophodon by the early Miocene, and the latter dispersed into Eurasia by around 19-18 million years ago, and into North America by the middle Miocene. The dispersal of mammutids between Africa and Eurasia may have occurred multiple times. The Mammutidae eventually went extinct in Africa prior to the late Miocene.^[86][87][77]

Mammut as currently defined sensu lato (in a broad sense) is most likely polyphyletic (comprising several unrelated groups). This is because the inclusion of Eurasian mammutid species into Mammut implies that they share a common origin with North American Mammut, but this relationship has been doubted. As a result, these Eurasian species may belong to either other existing mammutid genera or entirely new genera. "Mammut" borsoni, the last Eurasian mammutid, became extinct during the earliest Pleistocene, around 2.5-2 million years ago.^[77]

The oldest evidence of mammutids in North America is of a fragmentary molar of Zygolophodon sp. from Massacre Lake, Nevada, dating to 16.5-16.4 Ma (during the Hemingfordian stage of the North American land mammal ages (NALMA)). The only definitively defined species of Zygolophodon from North America is Z. proavus, which occurs in the Barstovian and Clarendonian stages. M? furlongi from the Black Butte in Oregon also dates back to the Clarendonian stage, but the affinities of the species remains unclear. If it truly is a species of Mammut, then its earliest temporal range is recorded at about 10 Ma. The earliest undisputed appearance of Mammut is of M. nevadanum from Thousand Creek Beds, dating back to the early Hemphillian, or 8.0-7.1 Ma. Historically, North American paleontologists considered that North American Zygolophodon evolved into Mammut in an endemic fashion while European workers generally thought that Mammut was a Eurasian immigrant that replaced North American Zygolophodon during the Miocene or Pliocene. Current evidence supports an endemic origin of North American Mammut from Zygolophodon without later migration because of the gradual appearance of Mammut morphologies and a lack of solid evidence that Mammut sensu stricto (in a strict sense) ever dispersed outside of North America.^[48]

M. matthewi is recorded from the late Hemphillian to early Blancan stages. Mammutid specimens of the Hemphillian and Blancan had typically previously been assigned to M. matthewi, but this is seemingly the result of overreliance on stratigraphic positions to define taxa. M. vexillarius, M. raki, and M. cosoensis are definitively recorded from the Blancan, and M. raki specifically is thought to not be synonymous with M. pacificum.^[48] M. americanum (known popularly as an "American mastodon" or simply "mastodon") is also stratigraphically recorded first from the early Blancan of the Ringold Formation, Washington. The age of the formation where the mammutid specimen was found dates to about 3.75 Ma. It is also known from multiple other Blancan sites such as Fish Springs Flat in Nevada.^[48][88][89] From the Irvingtonian to the Rancholabrean (from around 1.6 million to 11,000 years ago), only M. americanum and the newly appearing M. pacificum are recorded, the former having an exceptional level of diversity based on abundant skeletal evidences from the late Pleistocene that is unusual for the typical mammutid fossil record.^[76][48]

The following cladogram defines the phylogeny of certain proboscideans, a majority known from endocasts, including M. americanum:^[90]