Print version ISSN 0001-3765
An. Acad. Bras. Ciênc. vol.74 no.2 Rio de Janeiro June 2002
LEONARDO S. AVILLA 1*, CARLOS R.A. CANDEIRO 2, PAULO A. BUCKUP 1 AND LÍLIAN P. BERGQVIST 2
1 Departamento de Vertebrados, Museu Nacional / UFRJ, RJ.
2 Instituto de Geociências, Departamento de Geologia, UFRJ, RJ.
Presented by ALEXANDER W.A. KELLNER
Reconstruction of biogeographic events associated with the breakup of Gondwana supercontinent has attracted attention of biogeographers for over a century, but progress has been hampered by the scarcity of phylogenetic data. The widespread use of cladistic methodology in phylogenetic studies during the last decades, however, has made it possible to compile enough phylogenetic data to test biogeographic hypotheses using parsimony methods. We have analyzed seven phylogenetic data sets using Brooks Parsimony Analysis (BPA) in order to produce a comprehensive hypothesis of the historic relationships among Gondwanan sub-units.
Our data set includes phylogenies of Dipnoi (lungfishes), "peirosaurids'' (extinct Crocodylomorpha), Neoceratosauria (theropod dinosaurs), "Titanosauria'' (sauropod dinosaurs), Ratites (ostriches and relatives), Nataloidea (bats), and Gondwanatheria (group of extinct mammals) obtained from the literature. Presumptive areas of endemism within the former supercontinent were delimited according to (putative) distribution of terminal taxa. To produce a matrix of areas versus taxa, ancestral taxa specified by the phylogenies were coded according to the presence of their descendants in each presumptive area of endemism. A single most parsimonious area cladogram was selected using global parsimony analysis (consistency index = 0.77; retention index = 0.83). The following set of area relationships was specified: (Mongolia ((New Zealand, Australia), (Africa (Brazil (Argentina (Madagascar, India)))))). According to this hypothesis, the first vicariant event associated with the break up of the Gondwanan supercontinent was the separation of Australia and New Zealand from the remaining landmasses (Upper Jurassic-Early Cretaceous). As expected, the formation of the South Atlantic Ocean and the Strait of Mozambique subsequently lead to early isolation of Africa from the surrounding continental plates (Albian-Aptian). Surprisingly, however, the South American continent does not appear as a monophyletic entity. The Andean-Patagonean portion of the continent is more closely related to India and Madagascar than to the (currently) tropical portion. Such relationship may be explained by a lasting connection of the southern tip of South America to Antarctica and the Indian subcontinent (which was adjacent to Antarctica during most of the Cretaceous). Although Antarctica is not represented in any of the available phylogenies, we predict that further fossil discoveries in that continent will reveal taxa closely related to species in Argentina, India and Madagascar. ( December 20, 2001 ).