Where Did They Come From? The Origins of South American Fauna

By Lee E. Harding

Where Did They Come From? The Origins of South American Fauna offers a fascinating journey into the origins of South American flora and fauna. Exploring life on the continent before and after the breakup of Gondwana, it delves into how creatures arrived in South America, be it through drifting across oceans or traversing land bridges. From birds and reptiles to mammals and fish, this book provides a comprehensive compendium of biological diversity, discussing their origins and evolutionary paths. Readers will gain insights into the mechanisms of animal dispersal, evolution, and the impact of the Great Biotic Interchange. The book also lists references for further exploration of the subject.

The book is structured into five parts:
Building South America: Covers tectonic movements, climate changes, and breaching isolation.
Shaping South America: Explores the landforms and diverse biomes across the continent.
Vertebrates within South America: Discusses unique amphibians, reptiles, fish, mammals, and birds that evolved on the continent.
Vertebrates immigrating to South America: Examines exotic reptiles, birds, and mammals that found their way to the continent. The author also lists the families of almost all genera of South American animals, while giving knowledge of their origins.
Recent Arrivals - the Great Biotic Interchange: Explores the significant interchange of various species that occurred later.

Ideal for students, biologists, and anyone curious about the natural world, this book is a captivating read that uncovers the incredible history of South American fauna and its evolutionary tapestry.

• Language: English
• Publisher: Bentham Science Publishers
• Release date: Feb 21, 2001
• ISBN: 9789815124873

Book preview

Part 1 - Building South America

INTRODUCTION

The Neotropical zone boasts seven of the world’s 25 Biodiversity Hotspots: Mesoamerica, Caribbean, Choc´o/Dari´en/western Ecuador, Tropical Andes, Brazil’s Cerrado, Atlantic Forest and Central Chile (Myers et al. 2000 cited by Bagley & Johnson, 2014). This distinctiveness results in part from the continent’s high level of endemism.

As will be seen in Chapter 1, when Gondwana broke up ~110 million years ago (MYA), for about 56 million years, South America remained connected to a land mass that would later become New Zealand, Australia, Antarctica, and adjacent lands, collectively termed Australasia. This created a single biotic realm where, for another 54 million years, plants and animals evolved separately from former Gondwana lands and separately from the rest of South America.

The vast majority of South America’s biota are endemic: they occur nowhere else on Earth. Many lineages of extant plants and animals evolved completely within the continent from ancestors that were within South America or the southern continents when they separated from Africa and began drifting west, the rift creating the Atlantic Ocean. This is vicariance: isolation by physical events such as tectonic plate movement or sea level rise.

Yet other lineages, like New World monkeys, arrived during the vast reach of time from then until about 3.5 million years ago when South America finally connected to North America via the Isthmus of Panama. This is dispersal: organisms moving from their place or origin to colonize new lands. These, too, became endemic, evolving into forms unique to the continent. Their source and route have long bedeviled naturalists. Only recently has the DNA revolution, combined with fossil study, revealed their distant ancestors—some in Africa, some in Australasia, and some even in Europe. How did they get there? By what route?

These questions are complicated by the fact—relatively recently realized—that the continent is not now as—or where—it was. The similarity of rocks, minerals, and fossil animals on the coasts of Africa and South America had perplexed geologists and biogeographers for centuries, and the dovetailed shape of these continents seemed too perfect for coincidence. Yet no one had any idea of how this could be until Wegener (1912) proposed the idea of continental drift. But it was slow to gain acceptance and, when I entered high school in 1960, continental drift was not mentioned in our geology textbooks. It needed new advances in seismology, global magnetism, bathymetry, and marine geology. Plate tectonics was proposed as the mechanism of drift in a series of papers from 1965, when I started university, to 1968, culminating in a paper by W. Jason Morgan (Morgan, 1968).

This was only the beginning of understanding how South America came to be. Sea levels rose and fell with global climate changes, driving ocean waters alternately inland and back out. Tectonic forces drove the Nazca Plate against the South American plate, crumpling it to form the Andes, the highest mountain range outside of the Himalayas. This changed regional climates as prevailing easterly winds hit the mountains and rose, cooling, and dropped their loads of rain. The high runoff to the east formed the great rivers—the Orinoco, Amazon and Paraná—and these deposited sediments eroded from the mountains into deep layers in lowlands, filling basins and extending coastlines from Venezuela to Patagonia.

This is the backdrop of the creation of South America as we know it today.

REFERENCES

Tectonic Plate Movements

Lee E. Harding

INTRODUCTION

In the 1700s, with specimens of exotic animals and plants flooding into European capitols from their far-flung colonies, natural historians rushed to publish catalogues and pictorial folios. The foundations of natural history were laid down by luminaries including Linnæus, Buffon, Cuvier, Audebert, Lacépèd and Daubenton. South America was among the sources of animals sent by early naturalists (see Chapter 5 Biological Explorations). Already, naturalists knew from fossils in earlier and later strata that plants and animals somehow transformed or evolved from older forms into related, newer forms. Theories of evolution were proposed. A related question was how related lineages travelled from one continent to another. In the 1800s, catalogues of fossil plants and animals were published (e.g., Brongniart, 1828; Lydekker, 1885; Pidgeon, 1830). Charles Lyell’s Principles of geology (1830-1833) brought these questions into focus. Charles Darwin (see Charles Darwin, Chapter 5) took Lyell’s volume 1 with him on the Beagle and received volume 2 by mail while moored at Montevideo. Lyell argued against Lamarkian evolution (Lamark, 1809) and inspired Darwin to seek a better explanation. Naturalists began defining biological realms characterized by related forms of life that contrasted with other realms. The two-volume work by Alfred Russell Wallace—who spent four years collecting in the Amazon and co-published with Darwin, the theory of natural selection (Wallace, 1855, 1858)—The Geographical Distribution of Animals (Wallace, 1876), which included analysis of fossil distribution in relation to living forms, refined and extended earlier works.

The idea of continental drift was put forward in the 1910s and elaborated in the 1920s, but not widely accepted by geologists until the 1950s; the hypothesis has been replaced by plate tectonics, which explains the drift (reviewed by Briggs, 1987).

Prior to continental drift and plate tectonics, biologists thought that odd, disjunct distributions with related plants and animals in far distant continents and islands must be explained by dispersal: somehow, they flew, swam or drifted across vast expanses of ocean. Now they had a better explanation: vicariance. Vicariant distributions are explained by animals and plants getting on board a continent, or

a piece of one, and riding it across. This surely explains how some ancient lineages came to be in both Africa and South America with last common ancestors diverging around the time that the latter continent broke away from Gondwana in the mid-Cretaceous, about 110 MYA (Fig. 1.1).

Fig. (1.1))

Position of Gondwanan elements 94 MYA. Christofer R. Scotese, Paleomap Project, www.scotese.com, accessed 8 October 2022, used with permission.

As the movements of the plates and their positions at various times since then, as well as ancient climate reconstructions, became refined, vicariance also explained somewhat more recent lineages shared among the southern continents: Australasia, Antarctica and South America. These were connected until around the start of the Eocene, 54 MYA (Fig. 1.2). South America remained connected to Antarctica for another 20 million years after that, part of which time the globe became much warmer than today, giving Antarctica a semi-tropical environment.

When the idea of continental drift of tectonic plates finally became accepted, scientists jumped on it to explain the disjunct distributions of many South American plants and animals: as continents drifted apart, the biota on them became isolated from those left behind and evolved independently (e.g., Croizat, 1962; Croizat et al., 1974). This resulted in remnants of an ancestral biota that underwent geographical fragmentation followed by allopatric speciation (vicariance) (Rosen, 1975). This model of speciation is opposed to natural dispersal: organisms moving (or their seeds being spread, in the case of plants) to new territories, while the land stays in one place. For the last few decades, taxonomists and paleoecologists have debated whether vicariance or dispersal was the dominant model in creating the strange distributions of South American plants and animals, with vicariance gaining more leeway until a recent resurgence of dispersal. The recognition that many taxa could only have achieved their disjunct distributions by trans-oceanic dispersal heralded a counter-revolution in biogeography (de Queiroz, 2005) forcing many biogeographers to rethink some long-held hypotheses.

But both processes have been in operation during the last 100 million years of South American biological history. Dispersal is discussed below, following a review of plate tectonics.

SOUTH AMERICA SPLITS FROM AFRICA

During the Jurassic, all the continents were aggregated in one giant land mass, Pangaea. Later, about 130–180 MYA (Cox, 2000), Pangaea began to break up and at first, created two supercontinents, Laurasia (with what would become North America, Europe, and Asia) and Gondwana. These were separated by the Tethys Sea, but there were evidently ways for animals to cross it, because some taxa that originated in Laurasia got to Gondwana (or some of its parts as it splits up) and vice versa. Gondwana had broken up to form South America and Africa sufficiently to prevent most dispersal of 120 MYA (Nishihara et al., 2009). By about 110 MYA, Antarctica, Australia, New Zealand, New Guinea, Madagascar and India had also broken away from Pangaea.

The break-up was not uniform or symmetrical. As the continents pulled apart, the lithosphere (rocks comprising the mantle of the Earth) was stretched, with accompanying volcanism and rift valley formation (Pitman III et al., 1995). The new continents weren’t simply pulled apart, but also rotated with respect to one another. Not until new oceanic lithosphere was created from magma welling up into the rift and spreading out on either side of it were the new continents fully formed.

Nor were these processes steady and continuous through time; the starts and stops left magnetic lineations and fracture zones that help geologists identify the sequence and plot the relative positions of the plates and the continents that ride on them (Pitman III et al., 1995). Africa and South America began rifting at the north and south ends earlier than in the middle, near the equator. By 137–131 MYA, South America was probably completely separate from North America and the South Atlantic was opening between South America and Africa south of the Niger Delta, although they were still connected north of there.

Fig. (1.2))

Tectonic plate positions during the Eocene-Oligocene. Scotese 2014, used with permission.

Uncertainties remain about the separation of Gondwana from Laurasia: distribution of animal fossils suggests that many taxa were able to disperse across the Tethys Sea even well after the geologists say that they were separate.

Since there were also climate variations from north to south, creating diverse habitats, some of which would have been barriers to dispersal, taxa in some regions were isolated from their conspecifics earlier than those in other areas, giving a temporally variable dimension to vicariant speciation. Hence, Patagonian and southern African biota had a longer time to evolve unique forms than their respective tropical taxa.

An unresolved issue is whether there was a connection between North America and South America in the latest Cretaceous, which ended 66 MYA. In general, South America remained isolated as the two continents developed different fauna (Kielan-Jaworowska et al., 2007):

"During the Cretaceous the terrestrial vertebrate faunas of North and South America evolved in isolation, resulting in the development of different adaptative types on the two continents. While the North American mammalian faunas were dominated by multituberculates and therians, in South America the dominant mammals were Dryolestoidea (eupantotherians), accompanied by Gondwanatheria, ‘triconodontans’, ‘symmetrodontans’.."

Nevertheless, similarities of late Cretaceous mammal fossil fauna on the two continents suggest some interchange. For example, primitive marsupials predominate and placentals are a small component of both (see Marsupials, below). Also, multituberculates (see below), common in North America, were not thought, until recently, to have invaded South America. Recent findings show they did, but Antarctica is also a possible source.

However, it could not have been much of a connection because many common Cretaceous North American taxa never reached South America, and vice-versa, until the Great Biotic Interchange, or a little before. At any rate, from the Palaeocene (66–55 MYA) onward, the very high level of endemism among mammals on both continents indicates that faunal interchange was very limited (Gingerich, 1985:124). Assuming there was a late Cretaceous connection since Laurasia split from Gondwana, after which North America drifted away from Eurasia while Gondwana remained partially intact, dispersal was likely both ways. Conversely, low sea levels and high global temperatures in the early Eocene (see below) permitted Eurasia–North America interchange while South America remained isolated. Hence, the cosmopolitan nature of mammalian faunas in the Northern Hemisphere (reviewed by Gingerich, 1985), to the exclusion of the Southern Hemisphere.

In the mid-Cretaceous (~110 MYA), Australia, New Zealand and Antarctica formed the eastern part of Gondwana and were farthest south, while South America formed the western part and was farther north. In this vast area, with different climatic regions, it should not be surprising that different fauna and flora had evolved in different parts of the supercontinent. As they broke up, Africa drifted north, South America more or less maintained its latitudinal position while moving west and Australia-New Zealand-Antarctica drifted south while maintaining—barely—its connection to the southern tip of South America. Weather was mild (and became milder still in the Paleogene, 66–54 MYA; see below), leaving Antarctica unglaciated and forested (see below). This resulted in vicariant distributions of many higher-order (orders, families) Gondwana taxa, which were free to disperse throughout the region. That they did so is revealed by the marine, freshwater and terrestrial species common to Australia-New Zealand-Antarctica and South America, especially the southern tip, which maintained a similar climate, during successive epochs (reviewed by Briggs, 1987).

A long-held assumption was that South America was completely isolated from other continents since then, until the Isthmus of Panama formed at the beginning of the Pleistocene, 3.5 to 2.5 MYA (see Part 5, the Great Biotic Interchange). However, new geophysical data and recent analysis (e.g., Lawver et al., 2009) have refined this view to include a series of islands or transatlantic ridges that could have served as stepping-stones between Africa and South America, between North America and South America and between Antarctica and South America,

greatly reducing the distance over water that plants and animals would have had to traverse. This issue is discussed further below.

The maps in Fig. (1.2) show several areas of shallow water that may have been dry land during periods of low sea level and therefore could have provided routes to island-hop across the Atlantic, and between North and South America, without crossing whole oceans. The dropping sea level (see below), combined with tectonic uplift in some areas, exposed large areas of the southern South Atlantic, from the Malvina’s plains to the Walvis Ridge off Namibia, as late as about 28 MYA (Fleagle, 2013). This is shown by terrestrial sediments in many now-submerged parts of the South Atlantic (Browning et al., 2008; Fischer et al., 1999; Rau et al., 1987; Tissot et al., 1980). Land also emerged as island groups, plateaux or ridges in the mid- and North Atlantic. There was a dry plain extending east from Tierra del Fuego, encompassing the Islas Malvinas (Falkland Islands) and extending from there far out into the South Atlantic. It may even have reached the Walvis Ridge, or at least provided a short sea passage with island stepping stones along the way.

The Antarctic Peninsula, Thurston Island (= West Antarctica, WANT), Patagonia and Magallanes and are, in fact, small plates or crustal blocks that stayed more or less together throughout. Reguero & Goin (2021) note that,

"In the Late Cretaceous the continental breakup and assembly of West Gondwana influenced the geographic context in three major areas: the Antarctic Peninsula crustal block and southern SAM (Magallanes Region and Patagonia), within which the development of a particular Weddellian biota took place. ..the Antarctic Peninsula crustal block [has] been part of the southern South American continent since ~80 Ma forming a Weddellian area."

The Arctic Peninsula and Thurston Island crustal blocks maintained a lingering, or intermittent connection to Patagonia and Magallanes Region areas until the Early Paleogene, about 26 MYA (Reguero & Goin, 2021).

Ezcurra and Agnolín (2012) further proposed a new biogeographical model for late Mesozoic terrestrial ecosystems in which Europe and Gondwana were connected and possessed a common Eurogondwanan fauna during the earliest Cretaceous. Subsequently, the European territories severed from Africa and then connected to Asiamerica resulting in a faunal interchange, followed by a brief reconnection to Africa during the Campanian (83.6–72.1 MYA) to Maastrichtian (72.1–66.0 MYA) ages. They assert that this model explains the presence of Gondwanan taxa in Laurasia and the absence of Laurasian forms in the southern continents (Africa, South America, Antarctica and Australia) during the Cretaceous. If true, it would explain the trans-Atlantic disjunct distributions of several extant plant and animal groups between North America and Europe/Africa, but not between them and South America.

Another standard paradigm for much of the last half-century—that Madagascar was isolated by deep, wide sea expanses for nearly the whole of the Jurassic to the present—is also in doubt. Ali and Aitchison (2008) showed that, at various times from the Jurassic to the mid-Cretaceous or later, minor physiographic features (principally ocean islands) could have provided causeways and/or stepping-stone dispersal pathways between Africa, Madagascar and India. They suggest a likely island chain between Africa and Madagascar helped the ancestors of five land-mammal orders plus bats to cross the > 400-km-wide Mozambique Channel at various times during the Cenozoic.

Related to this idea is the Kerguelen Plateau, a group of volcanic islands formed during the breakup of the Gondwana about 110 MYA. Its sedimentary rocks show that the bulk of the plateau was above the sea for much of this time, sinking about 20 million years ago, making it a submerged continent. It may once have been continuous with Australia and/or India and would have been covered by coniferous forest during the Cretaceous. Most of the plateau is now about 600 m below sea-level (from http://sciencythoughts.blogspot.ca/2012/11/volcanic- activity-on-heard-island.html, December 22, 2014).

Ali and Aitchison (2008:145) suggest that:

"it is possible that the Kerguelen Plateau connected India and Australia–Antarctica in the mid-Cretaceous (approximately 115–90 Ma). Later, the Seychelles–Mascarene Plateau and nearby elevated sea-floor areas could have allowed faunas to pass between southern India and Madagascar in the Late Cretaceous, from around 85–65 Ma, with an early Cenozoic extension to this path forming as a result of the Reunion hot-spot trace islands growing on the ocean floor to the SSW of India. ..[however] direct terrestrial migration routes that have been proposed for the last 15 m.y. of the Cretaceous (Asia to India; Antarctica to Madagascar and/or India) can probably be dismissed because the marine barriers, likely varying from > 1000 up to 2500 km, were simply too wide."

The Madagascar connection is relevant to South American biodiversity because of shared taxa that prove dispersal from one to the other. Examples are the extinct elephant bird (a ratite), a snake member of the Boidae, and an iguanid lizard (see Part 3).

Likewise, a currently submerged land mass around New Zealand is large enough to quality as a continent and has been given a name as such: Zealandia (Luyendyk, 1995). It had, at one time, Araucaria/Podocarp forests like Antarctica and Patagonia and could have served as a centre of endemism and dispersal origin. Indeed, at least one mammal, a non-therian of an unknown Gondwanan ghost lineage lived there along with a sphenodontid, a crocodilian, geckos, skinks, bats, at least 24 bird species, until the Miocene (Worthy et al., 2006).

These new analyses have made it clear that monkeys, ostriches and flamingos may not have had to raft across the whole Atlantic Ocean to disperse between South America and Africa, and have clarified the connectivity among the southern territories.

ISTHMUS OF PANAMA CONNECTS THE CONTINENTS CONTINENTS CONTINENTS

Bagley and Johnson (2014) summarize the connection between North and South America as follows: By 60–50 MYA (Palaeocene), Central America was an island archipelago and dispersal into the region could only have occurred over a ~400 km ocean gap to the north and a ~400–1,500 km ocean gap or gaps to the south, based on plate reconstructions. Volcanism intensified 50–38 MYA, raising the Central Cordillera and creating a phylogeographic divide between Pacific and Atlantic coasts of Central America. Uplifting of tectonic plates in the Miocene extended the land area, although it is unclear whether this left a peninsula projecting from the north and a long ocean gap to Colombia, or multiple islands with smaller ocean gaps between them. Uplift, volcanism and erosional sedimentation in marine basins continued until the isthmus became fully closed 3.5–3.1 MYA, although intermittent openings associated with episodic sea-level rise continued until 1.8 MYA.

The creation of the Isthmus of Panama caused the Great Biotic Interchange, a pre-eminent event in South American biogeography. Many animals from North America invaded South America, and vice-versa (see Part 5). Many books and hundreds of scholarly articles have been written about it (e.g., Flynn & Wyss, 1998; Marshall, 1988; Stehli & Webb, 1988; Webb, 2006). The timing has for many years been held to be 3.5–3.1 MYA, based on the geology described above and the two-way flood of taxa that first appear in the animal fossil record of either continent after that period. For plants, two-way exchange and lineage diversification began earlier, as some plants are better able to disperse than most animals (Cody et al., 2010).

Several recent studies have proposed—an idea not fully accepted—that the Isthmus of Panama may have closed, or began closing, much earlier, at least by 15–13 MYA (Coates & Stallard, 2013) and possibly by 23 MYA (Bacon et al., 2012; Bacon et al., 2015; Montes et al., 2015). Because so many animals dispersed from North America to South America or vice versa during the Miocene (24–5 MYA) and Pliocene (5–1.8 MYA), as discussed in Part 4, many biologists and palaeontologists have defined the Great Biotic Interchange as dating from these earlier dispersals, rather than the traditional 2.5–3.5 MYA.

For example, Croft (2016) defines two episodes of intercontinental dispersal: a Trans-Atlantic Dispersal Interval (TADI) of around 30–41 MYA when caviomorph rodents and platyhrine primates arrived from Africa (Part 4 gives many more examples of animals that arrived during that time) and the Great American Biotic Interchange (GABI) that began about 9 MYA in the Miocene. This would resolve several mysteries about when several groups of non-volant (non-flying) mammals, snakes, frogs, turtles and other taxa reached South America from North America, and vice verse, before 3.5 MYA. It does not, however, address the question of why only a few taxa crossed until 3.5 MYA, when the two-way trickle became a torrent. The new geographic, geological and biological reconstructions suggest intermittent connections associated with the Panama Block’s movements and sea level changes that facilitated two or more waves of dispersal before full, permanent connection permitted unimpeded flow of plants and animals in the Great Biotic Interchange. A comprehensive review and reanalysis of geological, paleontological, and molecular data suggest completion of the connection between South- and Central America and formation of the Isthmus of Panama around 2.8 MYA. The authors conclude that evidence used to support an older isthmus is inconclusive, and we caution against the uncritical acceptance of an isthmus before the Pliocene (O’Dea et al., 2016). Because much of this book—the whole of Part 4—concerns animals that arrived in South America during its isolation, I hold to the standard definition of the Great Biotic Interchange as beginning when a permanent connection was established, notwithstanding that many creatures crossed earlier.

THE CARIBBEAN

The Caribbean region is complicated. Biologically, it has some strange animals and, in the not so distant past, had many more so. Charles Darwin (1859) said of it,

"I do not deny that there are many and grave difficulties in understanding how several of the inhabitants of the more remote islands, whether still retaining the same specific form or modified since their arrival, could have reached their present homes."

It is also complicated geologically. Considered a tectonic plate, it is more like a collection of fragments with numerous volcanos interjecting land masses at various times.

As Pangea broke apart, the Caribbean Plate remained connected to North America until the Oligocene, and has been moving east since then. Yet, from the mid-Oligocene to the Miocene, the Panama islandm arc may have provided a filter causeway (Pitman III et al., 1995). Filter means a route that some species—the good dispersers—can navigate but others cannot.

This movement resulted in biogeographic separation (vicariance) in the ancestral lineages of two odd mammals: Nesophontes and Solenodon. They are insectivores, but are so distantly related to other families in the Order Soricomorpha (shrews and moles) that both have their own families. Their lineage is so ancient that their crania retain structures that were common in Mesozoic mammals including a few Cretaceous eutherians (see Part 3: Mammals), but are otherwise unknown in placental mammals (Wible, 2008). Their ancestors, apternodontids and possibly geolabidids of North America, became extinct by the late Oligocene, but these two lineages persisted until after Europeans arrived on the islands (MacFadden, 1980). Nesophontes went extinct in the late 19th or early 20th Century; two Solenodon species exist today, but are rare (MacPhee et al., 1999). Mitogenomic phylogeny has confirmed the basal position of solenodons relative to shrews and moles, with Solenodon (Photo 1.1) having diverged about 78 MYA (Brandt et al., 2016).

Photo 1.1)

Solenodon paradoxus, captured in the Reserva Ecologica Ojos Indigenos, Dominican Republis, in 2015. Photographed by Nathan S. Upham. (Mammal Image Library, American Society of Mammalogists, 5 February 2018).

This tectonic activity has been associated with volcanoes in Central America and along the eastern edge of the Caribbean, where they created a string of islands in the Lesser and Greater Antilles. It has long been proposed that this was the route that some animals traversed from South America to the Caribbean and thence to North America by island-hopping before the closure of the Isthmus of Panama. One proposal was that, as the Caribbean Plate moved east, its leading edge pushed up land, the rupture also causing a series of volcanoes. Sea level dropped during the mid-Cenozoic (see below) and combined with uplift and volcanoes to create ridges along which plants and animals could disperse northward. This idea is termed the Greater Antilles and Aves Ridge (GAARlandia) hypothesis (Escalona & Mann, 2011; Iturralde-Vinent & MacPhee, 1999).

Some animal distribution fits this model. For example, phylogenetic investigation of 10 species of the toad genus Peltophryne in Caribbean islands shows that a common ancestor reached the islands about 33 MYA, when the purported land bridge is thought to have existed (Alonso et al., 2011). Also, freshwater Cichlid fishes colonized the Greater Antilles during the time of the proposed land bridges within a narrow window of the middle Oligocene (Říčan et al., 2013).

Although the GAARlandia hypothesis works for some taxa, including certain frogs, fish and reptiles cited by the original authors, Iturralde-Vinent and MacPhee (1999), it does not for others. There is no evidence that the xenarthrans (see below), hutias (caviomorph rodents; see below) and primates (see below) arrived during the time that the Aves Ridge-Lesser Antilles chain might have been connected in the mid-Oligocene. Rather, the DNA timing of divergence and fossils suggest that these taxa arrived later, during the Pliocene, and several authors remain convinced that they could not have walked; they had to swim or drift on floating rafts of vegetation as waif dispersals (Ali, 2012; Fabre et al., 2014; Hedges, 2006; Poux et al., 2006).

Support for waif dispersal is supported by ocean–atmosphere modeling for the Cenozoic that shows that current flowed from northern South America to the Greater Antilles (Huber & Caballero, 2003).

Against the GAARlandia hypothesis, Ali (2012) asserts that, considering the massive two-way flow of taxa across the Isthmus of Panama during the Great Biotic Interchange,

"..the key question is, why do we not observe in both extant taxa and the fossil record [in the Caribbean] a broad range of orders and families? Instead we see a restricted high-order taxonomic composition, plus at the lower levels several broad adaptive radiations due to species exploiting a wide range of unoccupied niches (e.g. ground sloths, capromyid rodents, eleutherodactyline frogs, and anoline and sphaerodactyline lizards)."

New fossils, particularly those from Panama associated with the digging of a new Panama Canal, are illuminating the biogeography of this complicated region. Primate fossils dating to 21 MYA show that a founding colony had crossed the Central American Seaway, also called the Atrato Seaway, by at least that date (Ford, 2006). Obviously, any animals or plants that reached Central America could have dispersed to North America and colonized the Caribbean from the north.

SOUTH AMERICA-ANTARCTICA-AUSTRALIA

Throughout the Eocene, Antarctica was ice-free, wetter and warmer than today and still connected, or nearly so, to South America (summarized by Tavares et al., 2006).

After the breakup of Gondwana, South America remained connected to Antarctica by a long, narrow causeway, the Weddellian Isthmus that extended from the Antarctic Peninsula. This permitted free flow of terrestrial vertebrates between the continents. As Antarctica pulled away and sea levels rose 100 to 200 m above current levels, at the end of the Palaeocene, about 55 MYA (see below), the isthmus broke into a series of islands (Reguero et al., 2014). This was the start of Simpson’s splendid isolation (Simpson, 1980); but the line of islands became not an absolute barrier, but rather a filter that allowed good or lucky dispersers to cross.

However, the rising seas also split East Antarctica from West Antarctica by a seaway connecting the Ross Sea and the Weddell Sea. This drove endemism among animals isolated in West Antarctica and the Antarctic Peninsula. Reguero et al. (2002) reported that:

"..the Seymour Island La Meseta Fauna .. contains at least 10 mammal taxa, predominantly tiny marsupials (mostly endemic and new taxa). The endemism of these marsupials suggests the existence of some form of isolating barrier (climatic and/or geographic) during the Eocene. Faunal similarity between the La Meseta Fauna and the