Osteology and relationships of the southern freshwater lower euteleostean fishes

Temperate lands of the Southern Hemisphere have freshwater f ish faunas that are dominated by a series of lower euteleostean genera that seem to fall into three clusters: 1) Retropinna, Stokellia and Prototroctes, forming a family Retropinnidae; 2) Aplochi- ton, Lovettia, Galaxias, Neochanna, Brachygalaxias, Paragalaxias and Galaxiella in- cluded in either a single family Galaxiidae or split between Galaxiidae and Aplochito- nidae (the latter including Lovettia); and 3) Lepidogalaxias. Although the generic classif ication seems relatively well-settled, the familial allocations and phylogenetic relationships among these three groups of genera have been much discussed in recent decades, dating back into the 1960s. There is wide agreement that Retropinna, Stokel- lia, and Prototroctes are closely related to each other; that the other genera apart from Lepidogalaxias are also closely related to each other and well separated from the ret- ropinnids; but what Lepidogalaxias is related to has been a subject of considerable debate and highly divergent phylogenetic hypotheses for many years. Our study pre- sents extensive mor phological information, mostly osteological, that seeks to clarify all of these relationships. Using the Northern Hemisphere Osmeridae as the principal out- group, our data support the historical relationships discussed above, for all but Lepido- galaxias, though we also suggest that Aplochiton and Lovettia are quite deeply sepa- rated from the galaxiid genera. Our conclusions relating to Lepidogalaxias are highly equivocal, given the large numbers of autapomor phic characters in the analysis and the lack of convincing synapomor phies shared by that genus and other southern lower euteleosteans, we conclude that Lepidogalaxias is deeply separated from the other southern genera. Incor porating additional molecular information, we conclude that Le- pidogalaxias is possibly the sister group of all other euteleostean f ishes, a conclusion that also emerges more explicitly from other, recent molecular studies of this highly controversial taxon. It has so many distinctive mor phological characters that it seems diff icult, if not impossible, to identify its relationships from osteology alone, though the extent of autapomor phy may result in part from our use of Osmeridae as the prin- cipal outgroup. Further studies comparing Lepidogalaxias with other outgroups, or per- haps the use of other types of mor phological characters, may reveal more of its rela- tionships.