476 CHAPTER 18(Populations or species that have been left behind in this way may be called relicts).
In this case, dispersal expanded the range, and extinction of populations in inter-
vening habitat caused the vicariant disjunction of populations.Historical Explanations of Geographic
DistributionsBiogeographers have used a variety of guidelines for inferring the histories of dis-
tributions. Some of these guidelines are well founded. For example, the distribu-
tion of a taxon cannot be explained by an event that occurred before the taxon
originated: a genus that originated in the Miocene, for example, cannot have
achieved its distribution by continental movements that occurred in the Cre-
taceous. Some other guidelines are more debatable. Some authors in the past
assumed that a taxon originated in the region where it is presently most diverse.
But this need not be so; as we have seen, wild species of the horse family today
are found only in Africa and Asia, even though the fossil record shows that the
family originated in North America. Changes in environment can radically alter
a taxon’s distribution. Species of beetles that were present in England during the
Pleistocene became extinct there (probably due to glacial changes in climate) and
are restricted today to various remote parts of Asia and Africa [5].
Several sources of evidence cast light on the historical causes of geographic dis-
tributions. The fossil record can show that a taxon proliferated in one area before
appearing in another, and geological data may describe the appearance or dis-
appearance of barriers [24]. For example, fossil armadillos are limited to South
America throughout the early Cenozoic and are found in North American deposits
only from the Pliocene and Pleistocene, after the Isthmus of Panama was formed.
We may infer that armadillos dispersed into North America from South America.
Paleontological data must be interpreted cautiously, because a taxon may be much
older, and have inhabited a region longer, than a sparse fossil record shows.
Phylogenetic methods are the foundation of most modern studies of histori-
cal biogeography. Inferring ancestral distributions from a phylogeny is much like
inferring ancestral character states (see Fig ure 2.16), although biologists are con-
tinuing to develop phylogenetic methods for determining the roles of dispersal,
vicariance, and extinction in the history of distributions [35]. The following sec-
tions include several examples of biogeographic inferences made from phylogenies.Vicariance
Changes in climate and in the configuration of land and sea have separated popu-
lations that became different species. We have already mentioned the emergence
of the Isthmus of Panama, which fully closed about 3 Mya and separated Carib-
bean and western Pacific populations of marine species [30].
The breakup of Pangaea first into Laurasia and Gondwana, and later into the
modern land masses (see F i g u r e 17.19), has long seemed to be a wonderful expla-
nation for many disjunct distributions, especially among pieces of Gondwana in
the Southern Hemisphere. Cichlid fishes, whose spectacular adaptive radiation
was introduced in Chapter 9, are limited to fresh water in Madagascar, Africa, and
tropical America. Araucaria pines are native to South America, Australia, and Nor-
folk Island and New Caledonia in the southern Pacific. Biologists hypothesized
that in each such case, the living species are descended from ancestors that were
distributed across Gondwana and became isolated on the several land masses
they now occupy, after Gondwana started splitting apart. But this scenario can
apply only to clades that are older than the split between the land masses they now18_EVOL4E_CH18.indd 476 3/22/17 1:39 PM