Highly derived oribatid mites (Brachypylina) seem to fit most predictions of evolutionary theory regarding the ecological, geographical and taxonomic distribution of parthenogenesis. Earlier derivative groups generally do not. We suggest that the ancestors of large, completely parthenogenetic families (for example, Brachychthoniidae, Lohmanniidae, Camisiidae, Trhypochthoniidae, Malaconothridae, Nanhermanniidae) were themselves parthenogenetic, and that 'speciation' and radiation occurred in the absence of sexual reproduction. Further, it is speculated that automixy (meiotic thelytoky) was the process involved -even though some extant species may prove to be secondarily apomictic, just as apomicts evolve from sexually reproducing lineages. If mechanisms for maintaining heterozygosity are effective, automixy can provide all the advantages of parthenogenesis, plus the DNArepair advantage of meiosis. Genotypes can be diversified by the formation of distinct clones, thereby providing the raw material for successful radiation. The disadvantage of such a system would be its slow rate of change due to the absence of an allele-recruitment mechanism. Much of the above remains speculative since available data are meagre. Future work should include surveys of thelytokous mechanisms in oribatid mites, the genetic characterization of populations of a variety of parthenogenetic and sexual species, and the refinement of hypotheses on phylogenetic relationships of parthenogenetic taxa• Invited paper. R. Schuster et al., The Acari

No abstract