A BAYESIAN PERSPECTIVE ON THE STREPSIPTERA PROBLEM

摘要: parasitize at least seven orders and 33 families of insects (Kathirithamby 2000). The phylogenetic placement Strepsiptera with respect to other the ‘Strepsiptera problem’ has proven be remarkably difficult resolve. They have been variously placed sister Coleoptera (Hendlirsch 1903, Boerner 1904, Shipley 1904), within coleopteran suborder Polyphaga as Rhipiphoridae (Crowson 1960), or Diptera (Whiting & Wheeler 1994, Whiting et al. 1997). Even Endopterygota questioned (Kristensen 1991). Many morphological characters relating one insect group another are interpret. Strepsipteran individuals small possess a unique combination that either specialized reduced. For example, single characterposteriormoterism -places strepsipterans coleopterans. A number suggest close relationship 1998). Unfortunately, quality especially quantity applicable determining likely remain limited. Molecular data hold promise resolving Strepsiptera. However, first molecular studies phylogeny ambiguous. only sequence data, date, from 18S 28S ribosomal genes (Carmean Crespi 1995, Phylogenetic analysis these using different methods produce quite placements for Parsimony trees having (a called Halteria by Maximum likelihood analyses, on hand, do not strongly place any particular insects. Interestingly, problem catalyzed interest in properties resulted an increased understanding potential biases affect methods. Specifically, directly motivated differences between likelihood-based parsimony In this paper, I review some strengths limitations maximum inference. used infer strepsipteran is important interpreting degree support various hypotheses. will also outline advances Bayesian inference describe how method can measure + dipteran monophyly changed light data.