Zoological Journal of the Linnean Society, 2004, 142, 289–368. With 7 figures © 2004 The Linnean Society of London, Zoological Journal of the Linnean Society, 2004, 142, 289–368 289 Blackwell Science, LtdOxford, UKZOJZoological Journal of the Linnean Society0024-4082The Lin- nean Society of London, 2004? 2004 1423 289368 Original Article J. F. REINERT ET AL. PHYLOGENY AND CLASSIFICATION OF AEDINI *Corresponding author. E-mail: jreinert@gainesville.usda.ufl.edu Phylogeny and classification of Aedini (Diptera: Culicidae), based on morphological characters of all life stages JOHN F. REINERT 1 *, RALPH E. HARBACH 2 and IAN J. KITCHING 2 1 Center for Medical, Agricultural and Veterinary Entomology, United States Department of Agriculture, Agricultural Research Service, 1600/1700 SW, 23rd Drive, Gainesville, FL 32608–1067, USA 2 Department of Entomology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK Received December 2003; accepted for publication May 2004 Higher-level relationships within Aedini, the largest tribe of Culicidae, are explored using morphological characters of eggs, fourth-instar larvae, pupae, and adult females and males. In total, 172 characters were examined for 119 exemplar species representing the existing 12 genera and 56 subgenera recognized within the tribe. The data for immature and adult stages were analysed separately and in combination using equal (EW) and implied weighting (IW). Since the classification of Aedini is based mainly on adult morphology, we first tested whether adult data alone would support the existing classification. Overall, the results of these analyses did not reflect the generic classifi- cation of the tribe. The tribe as a whole was portrayed as a polyphyletic assemblage of Aedes and Ochlerotatus within which eight (EW) or seven (IW) other genera were embedded. Strict consensus trees (SCTs) derived from analyses of the immature stages data were almost completely unresolved. Combining the adult and immature stages data resulted in fewer most parsimonious cladograms (MPCs) and a more resolved SCT than was found when either of the two data subsets was analysed separately. However, the recovered relationships were still unsatisfactory. Except for the additional recovery of Armigeres as a monophyletic genus, the groups recovered in the EW analysis of the com- bined data were those found in the EW analysis of adult data. The IW analysis of the total data yielded eight MPCs consisting of three sets of two mutually exclusive topologies that occurred in all possible combinations. We carefully studied the different hypotheses of character transformation responsible for each of the alternative patterns of rela- tionship but were unable to select one of the eight MPCs as a preferred cladogram. Overall, the relationships within the SCT of the eight MPCs were a significant improvement over those found by equal weighting. Aedini and all exist- ing genera except Ochlerotatus and Aedes were recovered as monophyletic. Ochlerotatus formed a polyphyletic assemblage basal to Aedes. This group included Haemagogus and Psorophora, and also Opifex in a sister-group rela- tionship with Oc. (Not.) chathamicus. Aedes was polyphyletic relative to seven other genera, Armigeres, Ayurakitia, Eretmapodites, Heizmannia, Udaya, Verrallina and Zeugnomyia. With the exception of Ae. (Aedimorphus), Oc. (Finlaya), Oc. (Ochlerotatus) and Oc. (Protomacleaya), all subgenera with two or more species included in the analysis were recovered as monophyletic. Rather than leave the generic classification of Aedini in its current chaotic state, we decided a reasonable and conservative compromise classification would be to recognize as genera those groups that are ‘weighting independent’, i.e. those that are common to the results of both the EW and IW analyses of the total data. The SCT of these combined analyses resulted in a topology of 29 clades, each comprising between two and nine taxa, and 30 taxa (including Mansonia) in an unresolved basal polytomy. In addition to ten genera (Armigeres, Ayurakitia, Eretmapodites, Haemagogus, Heizmannia, Opifex, Psorophora, Udaya, Verrallina and Zeugnomyia), generic status is proposed for the following: (i) 32 existing subgenera of Aedes and Ochlerotatus, includ- ing nine monobasic subgenera within the basal polytomy, i.e. Ae. (Belkinius), Ae. (Fredwardsius), Ae. (Indusius), Ae. (Isoaedes), Ae. (Leptosomatomyia), Oc. (Abraedes), Oc. (Aztecaedes), Oc. (Gymnometopa) and Oc. (Kompia); (ii) three small subgenera within the basal polytomy that are undoubtedly monophyletic, i.e. Ae. (Huaedes), Ae. (Skusea) and Oc. (Levua), and (iii) another 20 subgenera that fall within the resolved part of the SCT, i.e. Ae. (Aedes), Ae. (Alanstonea), Ae. (Albuginosus), Ae. (Bothaella), Ae. (Christophersiomyia), Ae. (Diceromyia), Ae. (Edwardsaedes), Ae. (Lorrainea), Ae. (Neomelaniconion), Ae. (Paraedes), Ae. (Pseudarmigeres), Ae. (Scutomyia), Ae. (Stegomyia), Oc. (Geoskusea), Oc. (Halaedes), Oc. (Howardina), Oc. (Kenknightia), Oc. (Mucidus), Oc. (Rhinoskusea) and Oc. (Zavortinkius). A clade consisting of Oc. (Fin.) kochi, Oc. (Fin.) poicilius and relatives is raised to generic rank as