thumb|upright=1.3|Examples of migrant species in both Americas. Olive green silhouettes denote North American species with South American ancestors; blue silhouettes denote South American species with North American ancestors.
The Great American Biotic Interchange (GABI), also known as the Great American Interchange and the Great American Faunal Interchange, was an important late Cenozoic paleozoogeographic biotic interchange event in which land and freshwater fauna migrated from North America to South America via Central America and vice versa, as the volcanic Isthmus of Panama rose up from the sea floor, forming a land bridge between the previously separated continents. Although earlier dispersals had occurred, probably over water, the migration accelerated dramatically about 2.7 million years (Ma) ago during the Piacenzian age. It resulted from the joining of the Neotropic (roughly South American) and Nearctic (roughly North American) biogeographic realms definitively to form the Americas. The interchange is visible from observation of both biostratigraphy and nature (neontology). Its most dramatic effect is on the zoogeography of mammals, but it also gave an opportunity for reptiles, amphibians, arthropods, weak-flying or flightless birds, and even freshwater fish to migrate. Coastal and marine biota were affected in the opposite manner; the formation of the Central American Isthmus caused what has been termed the Great American Schism, with significant diversification and extinction occurring as a result of the isolation of the Caribbean from the Pacific. Wallace had spent five years exploring and collecting specimens in the Amazon basin. Others who made significant contributions to understanding the event in the century that followed include Florentino Ameghino, W. D. Matthew, W. B. Scott, Bryan Patterson, George Gaylord Simpson and S. David Webb. The Pliocene timing of the formation of the connection between North and South America was discussed in 1910 by Henry Fairfield Osborn.
Analogous interchanges occurred earlier in the Cenozoic, when the formerly isolated land masses of India and Africa made contact with Eurasia about 56 and 30 Ma ago, respectively.
Before the interchange
Isolation of South America
thumb|†[[Thylacosmilus, a sparassodont]]
After the late Mesozoic breakup of Gondwana, South America spent most of the Cenozoic era as an island continent whose "splendid isolation" allowed its fauna to evolve into many forms found nowhere else on Earth, most of which are now extinct. Its endemic mammals initially consisted primarily of metatherians (marsupials and sparassodonts), xenarthrans, and a diverse group of native ungulates known as the Meridiungulata: notoungulates (the "southern ungulates"), litopterns, astrapotheres, pyrotheres and xenungulates. A few non-therian mammals – monotremes, gondwanatheres, meridiolestidans and possibly cimolodont multituberculates – were also present in the Paleocene; while none of these diversified significantly and most lineages did not survive long, forms like Necrolestes and Patagonia remained as recently as the Miocene.
thumb|upright=0.8|left|The [[monito del monte, Dromiciops gliroides, South America's only australidelphian marsupial]]
thumb|†[[Astrapotherium magnum]]
Marsupials appear to have traveled via Gondwanan land connections from South America through Antarctica to Australia in the late Cretaceous or early Tertiary. One living South American marsupial, the monito del monte, has been shown to be more closely related to Australian marsupials than to other South American marsupials (Ameridelphia); however, it is the most basal australidelphian, meaning that this superorder arose in South America and then dispersed to Australia after the monito del monte split off. Paleognath birds (ratites and South American tinamous) may have made a similar migration around the same time to Australia and New Zealand. Other taxa that may have dispersed by the same route (if not by flying or oceanic dispersal) are parrots, chelid turtles, and the extinct meiolaniid turtles.
Marsupials remaining in South America included didelphimorphs (opossums), paucituberculatans (shrew opossums) and microbiotheres (monitos del monte). Larger predatory relatives of these also existed, such as the borhyaenids and the saber-toothed Thylacosmilus; these were sparassodont metatherians, which are no longer considered to be true marsupials. As the large carnivorous metatherians declined, and before the arrival of most types of carnivorans, predatory opossums such as Thylophorops temporarily attained larger size (about 7 kg).
Metatherians and a few xenarthran armadillos, such as Macroeuphractus, were the only South American mammals to specialize as carnivores; their relative inefficiency created openings for nonmammalian predators to play more prominent roles than usual (similar to the situation in Australia). Sparassodonts and giant opossums shared the ecological niches for large predators with fearsome flightless "terror birds" (phorusrhacids), whose closest living relatives are the seriemas. North America also had large terrestrial predatory birds during the early Cenozoic (the related bathornithids), but they died out before the GABI in the Early Miocene, about 20 million years ago. Through the skies over late Miocene South America (6 Ma ago) soared one of the largest flying birds known, Argentavis, a teratorn that had a wing span of 6 m or more, and which may have subsisted in part on the leftovers of Thylacosmilus kills. Terrestrial sebecid (metasuchian) crocodyliforms with ziphodont teeth were also present at least through the middle Miocene and maybe to the Miocene-Pliocene boundary. Some of South America's aquatic crocodilians, such as Gryposuchus, Mourasuchus and Purussaurus, reached monstrous sizes, with lengths up to 12 m (comparable to the largest Mesozoic crocodyliforms). They shared their habitat with one of the largest turtles of all time, the 3.3 m (11 ft) Stupendemys.
thumb|The [[giant anteater, Myrmecophaga tridactyla, the largest living descendant of South America's early Cenozoic mammalian fauna]]
Xenarthrans are a curious group of mammals that developed morphological adaptations for specialized diets very early in their history. In addition to those extant today (armadillos, anteaters, and tree sloths), a great diversity of larger types was present, including pampatheres, the ankylosaur-like glyptodonts, predatory euphractines, various ground sloths, some of which reached the size of elephants (e.g. Megatherium), and even semiaquatic to aquatic marine sloths.
thumb|left|Life restoration of [[extinction|†Macrauchenia, an ungulate belonging to the extinct South American native ungulate order Litopterna]]
The notoungulates and litopterns had many strange forms, such as Macrauchenia, a camel-like litoptern with a small proboscis. They also produced a number of familiar-looking body types that represent examples of parallel or convergent evolution: one-toed Thoatherium had legs like those of a horse, Pachyrukhos resembled a rabbit, Homalodotherium was a semibipedal, clawed browser like a chalicothere, and horned Trigodon looked like a rhinoceros. Both groups started evolving in the Lower Paleocene, possibly from condylarth stock, diversified, dwindled before the great interchange, and went extinct at the end of the Pleistocene. The pyrotheres and astrapotheres were also strange, but were less diverse and disappeared earlier, well before the interchange.
The North American fauna was a typical boreoeutherian one, supplemented with Afrotherian proboscids.
Pre-interchange oceanic dispersals
thumb|[[Capybara, Hydrochoerus hydrochaeris]]
thumb|[[Emperor tamarin, Saguinus imperator]]
African origin
The invasions of South America started about 40 Ma ago (middle Eocene), when caviomorph rodents arrived in South America. Their subsequent vigorous diversification displaced some of South America's small marsupials and gave rise to – among others – capybaras, chinchillas, viscachas, and New World porcupines. The independent development of spines by New and Old World porcupines is another example of parallel evolution. This invasion most likely came from Africa. The crossing from West Africa to the northeast corner of Brazil was much shorter then, due to continental drift, and may have been aided by island hopping (e.g. via St. Paul's Rocks, if they were an inhabitable island at the time) and westward oceanic currents. Crossings of the ocean were accomplished when at least one fertilised female (more commonly a group of animals) accidentally floated over on driftwood or mangrove rafts. Hutias (Capromyidae) would subsequently colonize the West Indies as far as the Bahamas, Over time, some caviomorph rodents evolved into larger forms that competed with some of the native South American ungulates, which may have contributed to the gradual loss of diversity suffered by the latter after the early Oligocene.
Later (by 36 Ma ago), primates followed, again from Africa in a fashion similar to that of the rodents.
Fossil evidence presented in 2020 indicates a second lineage of African monkeys also rafted to and at least briefly colonized South America. Ucayalipithecus remains dating from the Early Oligocene of Amazonian Peru are, by morphological analysis, deeply nested within the family Parapithecidae of the Afro-Arabian radiation of parapithecoid simians, with dental features markedly different from those of platyrrhines. The Old World members of this group are thought to have become extinct by the Late Oligocene. Qatrania wingi of lower Oligocene Fayum deposits is considered the closest known relative of Ucayalipithecus.
Remarkably, the descendants of those few bedraggled "waifs" that crawled ashore from their rafts of African flotsam in the Eocene now constitute more than twice as many of South America's species as the descendants of all the mammals previously resident on the continent (372 caviomorph and monkey species versus 136 marsupial and xenarthran species).
Many of South America's bats may have arrived from Africa during roughly the same period, possibly with the aid of intervening islands, although by flying rather than floating. Noctilionoid bats ancestral to those in the neotropical families Furipteridae, Mormoopidae, Noctilionidae, Phyllostomidae, and Thyropteridae are thought to have reached South America from Africa in the Eocene,
thumb|left|[[Red-footed tortoise, Chelonoidis carbonaria]]
Tortoises also arrived in South America in the Oligocene. They were long thought to have come from North America, but a recent comparative genetic analysis concludes that the South American genus Chelonoidis (formerly part of Geochelone) is actually most closely related to African hingeback tortoises. Tortoises are aided in oceanic dispersal by their ability to float with their heads up, and to survive up to six months without food or water. and Galápagos Islands (the Galápagos tortoise). A number of clades of American geckos seem to have rafted over from Africa during both the Paleogene and Neogene. Skinks of the related genera Mabuya and Trachylepis apparently dispersed across the Atlantic from Africa to South America and Fernando de Noronha, respectively, during the last 9 Ma. Surprisingly, South America's burrowing amphisbaenians and blind snakes also appear to have rafted from Africa, as does the hoatzin, a weak-flying bird of South American rainforests. Gavialids and crocodiles are believed to have radiated across the Americas from Africa via oceanic dispersal, although an Australia/Asian origin has also been considered.
thumb|†[[Megalonyx|Megalonyx wheatleyi]]
South American origin
Caimans are believed to have repeatedly colonized North America as early as the Eocene, with more derived caimans (such as Centenariosuchus and Purussaurus) and podocnemidid turtles appearing by the early Miocene. Sebecid crocodylomorphs may have island hopped into the Caribbean during the Eocene-Oligocene, going extinct around 4.5Ma ago.
Megalonychid and mylodontid ground sloths island-hopped to North America by 9 Ma ago. In contrast, megatheriid and nothrotheriid ground sloths did not migrate north until the formation of the isthmus. Sloths first appear in Florida after a major sea level lowstand at the terminus of the Miocene. Terror birds may have also island-hopped to North America as early as 5 Ma ago.
The Caribbean Islands were populated primarily by species from South America, due to the prevailing direction of oceanic currents, rather than to a competition between North and South American forms. Except in the case of Jamaica, oryzomyine rodents of North American origin were able to enter the region only after invading South America.
North American origin
The earliest traditionally recognized mammalian arrival from North America was a procyonid that island-hopped from Central America before the Isthmus of Panama land bridge formed, around 7.3 Ma ago. This was South America's first eutherian carnivore. South American procyonids then diversified into forms now extinct (e.g. the "dog-coati" Cyonasua, which evolved into the bear-like Chapalmalania). However, all extant procyonid genera appear to have originated in North America. The first South American procyonids may have contributed to the extinction of sebecid crocodilians by eating their eggs, but this view has not been universally viewed as plausible.) South American native ungulates did poorly, with only a handful of genera withstanding the northern onslaught. (Several of the largest forms, macraucheniids and toxodontids, have long been recognized to have survived to the end of the Pleistocene. Recent fossil finds indicate that one species of the horse-like proterotheriid litopterns did, as well. The notoungulate mesotheriids and hegetotheriids also managed to hold on at least part way through the Pleistocene.) South America's small marsupials, though, survived in large numbers, while the primitive-looking xenarthrans proved to be surprisingly competitive and became the most successful invaders of North America. The African immigrants, the caviomorph rodents and platyrrhine monkeys, were less impacted by the interchange than most of South America's 'old-timers', although the caviomorphs suffered a significant loss of diversity, Eurasia was connected in turn to Africa, which contributed further to the species that made their way to North America. South America, though, was connected only to Antarctica and Australia, two much smaller and less hospitable continents, and only in the early Cenozoic. Moreover, this land connection does not seem to have carried much traffic (apparently no mammals other than marsupials and perhaps a few monotremes ever migrated by this route), particularly in the direction of South America. This means that Northern Hemisphere species arose within a land area roughly six times greater than was available to South American species. North American species were thus products of a larger and more competitive arena,