Biogeography of <i>Nothofagus</i> supports the sequence of Gondwana break‐up

<jats:title>Summary</jats:title><jats:p>Swenson, U., Hill, R. S. & McLoughlin, S.: Biogeography of <jats:italic>Nothofagus</jats:italic> supports the sequence of Gondwana break‐up. ‐ Taxon 50: 1025–1041. 2001. ‐ ISSN 0040–0262. The Austral biota reveals many links between Australasia and South America that have challenged biogeographers for many years. <jats:italic>Nothofagus</jats:italic>, the Southern Beech, is probably the classical example. With the general acceptance of continental drift, the break‐up of Gondwana is regarded as primarily responsible for many disjunct patterns expressed in the Southern Hemisphere biota. Vicariance biogeography is the principal tool used to investigate biogeographic patterns of extant plant groups, resulting in areagrams or general area cladograms. These are often at odds with current geological knowledge, and on this basis, alternative hypotheses of area relationships and geological history have, therefore, been suggested. One such areagram was recently advocated by Linder & Crisp (1995) in a biogeographic analysis of <jats:italic>Nothofagus.</jats:italic> Three explanations, often in combination, account for incongruence: long‐distance dispersals, extinctions, and erroneous geological models. All of these parameters ought to be considered in the analysis. Here we report the result of a historical biogeographic analysis of <jats:italic>Nothofagus</jats:italic> where we compare the reconciled trees between a well‐supported <jats:italic>Nothofagus</jats:italic> phylogeny and two geological hypotheses: (1) the current view of Gondwana break‐up, and (2) the areagram by Linder & Crisp. Our analysis makes use of extant and extinct taxa, as well as the assumption of long‐distance dispersals as defined by maximized vicariance. Our results show that <jats:italic>Nothofagus</jats:italic> existed prior to the break‐up of Gondwana and, most importantly, its present distribution supports, and is dependent upon, the traditional break‐up sequence of East Gondwana, compatible with three vicariance events. The areagram, conceived as an alternative geological hypothesis, presents a more parsimonious solution, but fails to explain numerous past distributions in areas such as Antarctica, South America, and Tasmania. We therefore recommend a conservative approach to use (general) areagrams in historical biogeography.</jats:p>