![A Mata Atl‚ntica na costa do EspÌrito Santo em Linhares.](http://revistapesquisa.fapesp.br/wp-content/uploads/2016/06/Mata_Mata-Atlântica_Espirito_Santo.jpg)
Leonardo Merçon/Instituto Últimos Refúgios A Mata Atl‚ntica na costa do EspÌrito Santo em Linhares.Leonardo Merçon/Instituto Últimos Refúgios
The Brazilian continental shelf, now submerged along the coastline, may have contained an extensive area of Atlantic Forest some 21,000 years ago—a period known as the Last Glacial Maximum. The idea has been posited by biologists Yuri Leite and Leonora Costa, husband-and-wife professors at the Federal University of Espírito Santo (UFES). Their hypothesis contradicts the long-accepted view that cold temperatures forced the Atlantic Forest to contract and confine small populations of plants and animals to isolated forest fragments, or refugia. “The continental shelf appears on Google Maps, but no one thinks of it as part of the continent,” Leite explains. The idea, which he and colleagues posited in a paper published in the journal PNAS in January 2016, was based on the knowledge that the sea level dropped 120 meters during the Ice Age. It was therefore dubbed the Atlantis Forest hypothesis, in reference to the mythical continent swallowed up by the sea.
The hypothesis emerged from a set of collaborative studies initiated by the Mammalogy and Biogeography Laboratory at the Federal University of Espírito Santo (UFES), which is headed by Leite and Costa. Working with ecological models to infer past conditions, Carolina Loss, a postdoctoral researcher at the laboratory, had the idea of looking at the configuration of the continent during the Ice Age, when the sea level dropped and the coastline advanced hundreds of kilometers eastward, exposing 270 square kilometers of the continental shelf—the equivalent of three times the size of Portugal. At the same time, in a project in collaboration with biologist Renata Pardini of the University of São Paulo (USP) and other colleagues, the group was seeking to evaluate how small mammals responded to the fragmentation of the Atlantic Forest. These situations in which the available habitat is constricted and isolated into distant segments are presumed to cause population reduction and consequent loss of genetic variety. But this was not seen in the demographic models analyzed by Portuguese biologist Rita Rocha, another postdoctoral researcher at UFES. There was no genetic signature of population reduction, and all such scenarios were rejected by the computer models.
“I put the two things together and decided to test out a scenario in which the Atlantic Forest expanded,” Leite says. The model showed this to be the most plausible explanation for the genetic diversity found in strands of DNA from five species of small mammals typical of that type of forest. The analyses indicated that the species moved northward and to lower altitudes, where the temperature was higher, in line with what other studies had indicated previously. The surprises came in seeing that the area suitable for these animals was less fragmented during the Last Glacial Maximum compared to the present and the pre-glacial period, and that the species distribution moved along the then-exposed continental shelf. These findings conflict with the refuge theory—the principal explanation for the formation of the biodiversity in the Brazilian forests, and the Amazon in particular (see Pesquisa FAPESP issue nº 208).
Palynological studies carried out more than a decade ago by biologist Aline Freitas, now a postdoctoral researcher at the University of Murcia, Spain, under the supervision of botanist José Carrión, corroborate the notion that the Atlantic Forest occupied the coastline, and expanded outward onto the continental shelf during the Last Glacial Maximum. Freitas, who did her graduate research at the Federal University of Rio de Janeiro (UFRJ) under the guidance of biologist Marcelo Carvalho, has been analyzing evidence taken from the sea floor in the Campos Basin, in the northern part of that state. The samples of fossilized pollen drawn from this material indicate that the area was a mosaic of trees, shrubs, tree ferns and herbaceous plants typical of the Atlantic Forest and the associated restinga coastal woodland vegetation. “All indications are that the vegetation of this region seems not to have undergone major changes during the glacial and interglacial periods, but rather adapted in accordance with the relative changes in sea level,” she notes, pointing out that her data do not as yet allow for attribution under the Atlantis Forest hypothesis.
![Potential for expansion Wide along Brazil’s south-central coast, the continental shelf (shown in light blue) narrows from Bahia northward](http://revistapesquisa.fapesp.br/wp-content/uploads/2016/06/Mata_Mapa-Litoral-300x239.jpg)
Google Maps
Potential for expansion: wide along Brazil’s south-central coast, the continental shelf (shown in light blue) narrows from Bahia northwardGoogle MapsPatchwork quilt
The researchers have not discarded the idea that there were isolated patches of forest in locations that enabled them to withstand glaciation and, like Noah’s ark, sustain a collection of animals and plants that evolved separately, thus giving rise to the present-day diversity. But the history of the area was likely more complex than that. “The main thing is the topography as a whole,” Leite explains. The terrain in the regions more to the north, where the mammals included in the study remained during glaciation, is less irregular and, partly for this reason, probably permitted a more continuous distribution. The same holds true for the continental shelf, which has flatter topography.
Biologist Henrique Batalha-Filho of the Federal University of Bahia (UFBA), a coauthor of the article in PNAS, has not abandoned the refuge theory. On the same day on which the collaborative study with his UFES colleagues was published, another of his papers was published online by the Journal of Zoological Systematics and Evolutionary Research. In that study, Batalha infers that the Star-throated Antwren (Rhopias gularis) followed the pattern predicted by the refuge theory for the Atlantic Forest, in accordance with the model proposed in 2008 by Brazilian biologist Ana Carolina Carnaval from the City University of New York, along with Australian biologist Craig Moritz of Australian National University. Batalha says that, in the southern part of their distribution, these birds are found at sea level. To the north, in the state of Bahia, they exist only at elevations above 600 meters, which restricts their distribution considerably.
He sees no contradiction between the two studies. “In my view, the Atlantic Forest appears to be a mosaic of histories, and each species responds differently to the situation to which it is subjected,” he suggests. He started to think about this during his doctoral studies (see Pesquisa FAPESP issue nº 210), when he saw signs that bird species underwent different histories during the glacial period. Although he has observed in many cases that forest-dependent animals appear to have followed the classic pattern in which their distribution is restricted to refugia, others have remained stable. Such is the case for another bird he studied, the White-browed Warbler (Myiothlypis leucoblepharar), which is specialized for higher, and therefore colder, elevations, though its habitat does not appear to have been restricted during glaciation. “I began to think that ecologies played an important role in this history.”
![At some points along the coast of Espírito Santo, the Atlantic Forest comes close to the water’s edge, as seen in the area around Linhares, near the Doce River](http://revistapesquisa.fapesp.br/wp-content/uploads/2016/06/Mata_Mata-Atlântica_Espirito_Santo.jpg)
Leonardo Merçon/Instituto Últimos Refúgios At some points along the coast of Espírito Santo, the Atlantic Forest comes close to the water’s edge, as seen in the area around Linhares, near the Doce RiverLeonardo Merçon/Instituto Últimos Refúgios
The new perspective also enables a new interpretation for the division that many researchers have observed in the region around the Doce River, in terms of the population genetics of several types of animals. “Since my Ph.D., I’ve been closely studying the northern and southern Atlantic Forest, looking at the rivers,” Costa says. “I came to realize that rivers are not barriers in the deepest history of the Atlantic Forest.” Other features appear to have been more important in the region. To the south of the Doce River, the mountains come quite close to the coast, while in the north they are farther away. The continental shelf narrows somewhat to the south of that river, while farther to the north, starting at the Abrolhos Archipelago, it becomes very narrow. All of this topography should be part of the analyses as we go forward.
“The article adds fuel to the fire and introduces yet another hypothesis to be tested,” says biologist Maria Tereza Thomé of São Paulo State University (Unesp) in Rio Claro, who was not involved in the research. The new perspective may cause her to look at her own findings through new eyes, in view of the fact that she arrived at results consistent with the Atlantis Forest hypothesis in her studies of frogs of the group Rhinella crucifer, which found no substantial demographic fluctuations. In the southern part of the forest, however, she infers that the populations have remained stable. In a 2014 paper published in Molecular Ecology, Thomé and colleagues suggest a need to identify barriers to animal movement that are invisible today. The continental shelf hypothesis fits right in with it. “For my animals, it makes complete sense,” she says.
Thomé stresses the importance of the study being well thought-out, conducted only by Brazilian researchers and published in a well-reputed journal, simply because it is a good idea in need of wider exposure. “In our field, we suffer from a shortage of hypotheses; now it is imperative that we all embrace this one,” she comments. But it will still need to be tested to become established.
“New hypotheses are important for enriching the discussion,” says Fabio Raposo do Amaral, a biologist at the Diadema campus of the Federal University of São Paulo (Unifesp). In his opinion, future studies will need more species and a broader range of genetic material. “I wonder if that data set has the statistical power to explore the events in question, with the necessary degree of precision to separate periods that differ by only a few thousand years—precision that probably only genomic data can provide,” he comments. Batalha was involved in testing historical scenarios for how populations were sustained, based on coalescent theory, which infers DNA modifications by looking backwards in time from the present. He concurs that “these analyses call for more markers.” Even if researchers see a reliable sign, he agrees that the use of several strands of DNA would strengthen the hypothesis raised in the PNAS paper. “We may never know what happened in the Atlantic Forest, but we’re adding more pieces to the puzzle,” he says.
Scientific articles
BATALHA-FILHO, H. & MIYAKI, C. Late Pleistocene divergence and postglacial expansion in the Brazilian Atlantic Forest: multilocus phylogeography of Rhopias gularis (Aves: Passeriformes). Journal of Zoological Systematics and Evolutionary Research. On-line. January 11, 2016.
FREITAS, A. G. de et al. Pollen grains in quaternary sediments from the Campos Basin, state of Rio de Janeiro, Brazil: Core BU-91-GL-05. Acta Botanica Brasilica. V. 27, No. 4, p. 761-72. October/December 2013.
LEITE, Y. L. R. et al. Neotropical forest expansion during the last glacial period challenges refuge hypothesis. PNAS. V. 113, No. 4, p. 1008-13. January 26, 2016.
THOMÉ, M. T. C. et al. Barriers, rather than refugia, underlie the origin of diversity in toads endemic to the Brazilian Atlantic Forest. Molecular Ecology. V. 23, No. 24, p. 6152-64. November 24, 2014.