Gnaphalieae in New Zealand comprises an estimated 70 to 80 species in ten genera. Most species, and three of the genera, are endemic. All genera were earlier included, at least in part, in Gnaphalium or Helichrysum with the exceptions of Rachelia, which is newly described, and Craspedia, which belongs to a quite different alliance. Interspecific hybrids are not infrequent in the New Zealand flora, but reports of intergeneric hybrids are rare except in Gnaphalieae, for which a surprising number of anecdotal reports exist, many of them supported by herbarium material. We report here on the first rigorous testing of these hybrid hypotheses. Putative intergeneric hybrids were located in the wild. Four alternative hypotheses were tested for each specimen: that it was (a) within the range of variation of an existing species, (b) a mutant form of an existing species, (d) a hybrid, but of different parentage to that postulated. Specimens shown to fit none of these hypotheses were analyzed for intermediacy between the putative parent species, proximity to one or both parents, and fertility. Artificial resynthesis was attempted. Results of the analyses of two putative hybrids are presented. The existence of such putative intergeneric hybrid links led us to hypothesize that most New Zealand species evolved very recently within New Zealand rather than arriving separately by long distance dispersal from several sources. These would include the species of Anaphalioides, Leucogenes, Rachelia, Raoulia, those currently retained in Ewartia and Helichrysum, and possibly Euchiton, but not Ozothamnus or Pseudognaphalium. Indepedent support for this hypothesis has come from a recent analysis of ITS sequences which showed the group of all hybridising genera except Euchiton to be monophyletic.

Key words: Compositae, Gnaphalieae, intergeneric hybrids, New Zealand