COOCCURRENCE OF AUSTRALIAN LAND BIRDS - DIAMONDS ASSEMBLY RULES REVISITED

Using null model simulations, we tested for non-random patterns of loc al co-occurrence in 28 congeneric guilds of the Australian avifauna. A t the scale of 1 degrees latitude-longitude blocks, species in most gu ilds co-occurred more often than expected by chance. However, coexiste nce was significantly less than expected for six of the 28 guilds. In four of these guilds (Climacteris, Cinclosoma, Manorina, and Psophodes ), the species were segregated by habitat use and/or geographic range. The remaining two cases were complicated by uncertainty in taxonomy ( Malurus) and unreliable field records (Corvus). We also examined distr ibution patterns in five Australian guilds that are analogs of avian g uilds designated by Diamond for the Bismarck Archipelago. For two of t he five guilds (Pachycephala and Zosterops), co-occurrence in Australi a was less than expected, mirroring an insular pattern of ''checkerboa rd distributions'' in the Bismarck Archipelago. For the remaining thre e guilds (Ptilonopus, Myzomela, and Lonchura), co-occurrence was signi ficantly greater than expected. Overall, our results suggest that comp etitive-based assembly rules are not important in determining species coexistence within most congeneric guilds of the Australian avifauna, at least at the large spatial scale of our analysis.