‘The scattered islands of the southern, cold ocean have clear affinities with one another . . . They are the basis for my Insulantarctica province.’
Udvardy, 1987, p. 190
Analyses of the distributional patterns of weevils (Coleoptera: Curculionoidea) from several Subantarctic islands, namely, Campbell, Auckland, Snares, Antipodes, Chatham, Falklands, Tierra del Fuego, Tristan da Cunha, Inaccessible, Nightingale, Gough, Marion, Prince Edward, Crozet, Kerguelen, and Heard, as well as South America and New Zealand, were carried out in order to determine their historical relationships, and to test the validity of Udvardy's (1987) Insulantarctica province. Three parsimony analyses of endemicity (PAE) considering (a) only species, (b) only supraspecific taxa, and (c) species and supraspecific taxa together, were undertaken. The following four groups emerged from the analyses: (1) New Zealand with the Snares, Auckland, Campbell, and Chatham Is., where New Zealand is the sister area to the Chatham Is., and the Auckland Is. are the sister area to Campbell I.; (2) South America with the Falkland Is. and Tierra del Fuego, where South America and Tierra del Fuego together are the sister area to the Falkland Is.; (3) Tristan da Cunha-Gough group, with the islands following the sequence Gough, Tristan da Cunha, Inaccessible, and Nightingale Is.; and (4) Kerguelen, Heard, Crozet, Marion, and Prince Edward Is., with Kerguelen and Heard Is. being sister areas, and Marion and Prince Edward Is. together being the sister area to the Crozet Is. It is concluded that the weevil fauna does not support the existence of an Insulantarctica province; the similarities among the different Subantarctic islands are due more to similar environmental conditions rather than to a common history.