Phylogeny and classification of Ochlerotatus and allied taxa (Diptera: Culicidae: Aedini) based on morphological data from all life stages

Transcription

1 Zoological Journal of the Linnean Society, 2008, 153, With 2 figures Phylogeny and classification of Ochlerotatus and allied taxa (Diptera: Culicidae: Aedini) based on morphological data from 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 , USA 2 Department of Entomology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK Received 2 March 2007; accepted for publication 11 July 2007 The phylogenetic relationships and generic assignments of Ochlerotatus and related taxa of uncertain taxonomic position in the classification of Aedini previously proposed by the authors in 2004 and 2006 are explored using 297 characters from eggs, fourth-instar larvae, pupae, adults and immature habitat coded for 158 exemplar species. The ingroup comprises 54 species and the outgroup includes four non-aedine species and 100 aedine species, 21 of which were previously classified as incertae sedis. Data are analysed in a total-evidence approach using implied weighting. The analysis produced 158 most parsimonious cladograms. The strict consensus tree (SCT) corroborates the monophyly of the 30 generic-level taxa recognized previously that are included in the analysis. Overall, the results show remarkable congruence with those obtained previously despite differences in the taxa and morphological characters analysed in this and the two previous studies. All species of Ochlerotatus s.s., subgenus Ochlerotatus sensu auctorum, Geoskusea, Levua, Pseudoskusea and Rhinoskusea included in the analysis fall within a single clade that is treated as genus Ochlerotatus; thus, the last four taxa are restored to their previous subgeneric rank within this genus. Nine additional subgenera, of which four are new, are proposed for monophyletic clades of Ochlerotatus species based on the strength of character support and application of the principle of equivalent rank. Acartomyia stat. nov., Culicelsa stat. nov., Gilesia stat. nov., Protoculex stat. nov. and Chrysoconops stat. nov. are resurrected from synonymy with Ochlerotatus; and Empihals subgen. nov. (type species: Culex vigilax Skuse), Pholeomyia subgen. nov. (type species: Aedes calcariae Marks), Buvirilia subgen. nov. (type species: Aedes edgari Stone & Rosen) and Sallumia subgen. nov. (type species: Aedes hortator Dyar & Knab) are described as new. The sister group of Ochlerotatus includes a number of species that were previously regarded as incertae sedis in Oc. (Finlaya) and Oc. (Protomacleaya). Based on previous observations, refined relationships and new character support, three additional genera are recognized for species previously included in Finlaya, i.e. Danielsia stat. nov. (type species: Danielsia albotaeniata Leicester), Luius gen. nov. (type species: Aedes fengi Edwards) and Hopkinsius gen. nov. (type species: Aedes ingrami Edwards). Additionally, Alloeomyia subgen. nov. (type species: Culex pseudotaeniatus Giles) and Yamada subgen. nov. (type species: Aedes seoulensis Yamada) are introduced as subgenera of Collessius and Hopkinsius, respectively. As is usual with generic-level groups of Aedini, the newly recognized genera and subgenera are polythetic taxa that are diagnosed by unique combinations of characters. The analysis corroborates the previous observation that Oc. (Protomacleaya) is a polyphyletic assemblage of species The Linnean Society of London, Zoological Journal of the Linnean Society, 2008, 153, ADDITIONAL KEYWORDS: cladistics mosquitoes taxonomy systematics. INTRODUCTION Genus Ochlerotatus, described by Lynch Arribalzaga (1891), was formally downgraded to a subgenus of *Corresponding author. John.Reinert@ars.usda.gov Aedes by Edwards (1917) and remained so until it was returned to generic rank by Reinert (2000a) based on differential characters of the male and female genitalia and supplemental features of the fourth-instar larvae and pupae. As traditionally recognized, Ochlerotatus (as a subgenus of Aedes) comprised a group of 29

2 30 J. F. REINERT ET AL. about 180 species. When Ochlerotatus was reinstated to generic rank, it included about 500 species divided between 21 subgenera that were transferred to it from genus Aedes. In our previous studies (Reinert, Harbach & Kitching, 2004, 2006), we examined the higher-level relationships within tribe Aedini (sensu Belkin, 1962), and Ochlerotatus subgenus Finlaya (sensu auctorum) and allied taxa. As a result of taxonomic changes proposed by Reinert et al. (2004, 2006), genus Ochlerotatus now includes some 276 species, most of which are unassigned to one or other of two currently recognized subgenera, Ochlerotatus and Rusticoidus. Ochlerotatus (as a subgenus of Aedes) has not been reviewed on a worldwide basis since Edwards (1932), but a number of changes have been proposed subsequently to Edwards species groups by various authors based on faunal treatments. The definitions of species groups and the inclusion of species in these groups of Ochlerotatus differ somewhat among authors and are based primarily on morphological features of adult females and genital characters of males. Belkin (1962: ) stated that Ochlerotatus (as a subgenus of Aedes) is a heterogeneous complex of several different types which appear to have in common some rather basic structures in the male genitalia.... He further stated It seems probable that the subgenus as now understood is a polyphyletic assemblage and that it will be split up into several subgenera when the numerous species are revised. The groups recognized by Edwards (1932: ) are based in general on very superficial characters, and several of them include completely unrelated forms. Belkin (1962: 388). With one exception (i.e. Rey et al., 2001), Reinert et al. (2004, 2006) reviewed the results of molecular (DNA) studies of aedine taxa published up to the end of 2005 (i.e. Wesson, Porter & Collins, 1992; Besansky & Fahey, 1997; Kumar, Black & Rai, 1998; Behbahani et al., 2005; Cook et al., 2005), all of which were limited by taxon sampling. Rey et al. (2001) used a 763-bp segment of the mitochondrial cytochrome c oxidase I (COI) gene to examine the phylogenetic relationships of 14 species (12 endemic to France) traditionally placed in genus Aedes. Overall, the relationships inferred from their analyses support the generic status of Aedes, Ochlerotatus and Stegomyia proposed by Reinert (2000a) and Reinert et al. (2004). More recently, Shepard, Andreadis & Vossbrinck (2006) showed that species of Aedes (four species) and Ochlerotatus (15 species) from the north-eastern United States fall into two separate and distinct clades based on phylogenetic analyses of 18S rrna genes. Thus, the elevation of Ochlerotatus to generic level by Reinert (2000a) based on morphological data is corroborated by mitochondrial and rrna gene sequence data. Shortly after the present paper was written, Cywinska, Hunter & Hebert (2006) and Kumar et al. (2007) reported on the genetic divergence of 37 Canadian and 63 Indian mosquito species, respectively, based on short fragments of the COI region of mitochondrial DNA ( barcodes ) used to provide species identification. The study of Cywinska et al. (2006) included 24 species of Aedes (genera Aedes and Ochlerotatus of Reinert et al., 2004, 2006) and that of Kumar et al. (2007) included 27 aedine species of genera Aedes, Armigeres, Collessius, Diceromyia, Fredwardsius, Heizmannia, Lorrainea, Phagomyia, Rhinoskusea, Stegomyia and Verrallina (all sensu Reinert et al., 2004, 2006). Neighbour-joining (NJ) trees were used to cluster haplotypes, and the distinct clusters were interpreted as species. Neighbour-joining is a phenetic method, not a phylogenetic method, and the observed clusters are based on overall similarity, not synapomorphy; hence, the relationships among the species portrayed in the NJ trees of Cywinska et al. (2006) and Kumar et al. (2007) are not relevant to the conclusions drawn from the cladistic analyses of Reinert et al. (2004, 2006). Belkin (1962: 61) believed that all stages must be studied to reconstruct phyletic lines and that in different lines evidence of relationships cannot always be found readily in the same stage. Reinert (2002b: 27) also indicated that All life stages of the included species, including their bionomics, should be evaluated to establish a natural classification. In the present study, we analyse morphological characters of females, female genitalia, males, male genitalia, pupae, fourth-instar larvae and eggs, as well as habitats of immature stages, of 54 species of Ochlerotatus representing taxa worldwide to build on our previous studies (Reinert et al., 2004, 2006), with the aim of reconstructing the phylogeny and classification of Ochlerotatus and allied taxa. Several species of Oc. (Finlaya) of uncertain placement (i.e. incertae sedis) are also included. To avoid confusion, species of uncertain generic and subgeneric placement, and genericlevel names used in the previous sense that include these species, are enclosed within single quotation marks to distinguish them from formally recognized monophyletic taxa. MATERIAL AND METHODS Study taxa include the 158 species listed in Appendix 1 (see also for authorship and geographical distribution). The ingroup includes 54 species, which represent established species groups of Ochlerotatus, groups that are recognized as being closely related to Ochlerotatus and groups previously included in the genus. The outgroup includes 104 species: 100 aedine species, 21 of which we previously classified as incer-

3 PHYLOGENY AND CLASSIFICATION OF OCHLEROTATUS 31 tae sedis (Reinert et al., 2006), and the four nonaedine species that we used in our previous analyses (Reinert et al., 2004, 2006), i.e. Culex quinquefasciatus, Culiseta inornata, Mansonia titillans and Orthopodomyia signifera. Taxa examined include type species of most genus-level names. Unfortunately, specimens of several important Australasian and Neotropical species were unavailable or had several missing life stages. The data (Appendix 2) comprise 297 characters from eggs (four), fourth-instar larvae (91), pupae (38), females (94), males (14), female genitalia (26), male genitalia (29) and habitat of immature stages (one). Some autapomorphic characters are included because they were present in more than one generic-level taxon in one or both of our previous studies (Reinert et al., 2004, 2006). Individually reared, pin-mounted adults with associated slide-mounted fourth-instar larval and pupal exuviae were studied when available. A phase contrast or differential interference contrast microscope with 400 magnification is needed to observe the very slender distal parts of many setae of larvae and pupae for the measurement of total length. Pinned adults were examined with a binocular stereomicroscope utilizing cold white light delivered by a fibre-optic illumination system. An adjustable examination stage with biaxial rotation capability (see fig in Russell et al., 1963) allowed observation and illumination of specimens at any angle. In general, character data were derived from three to six specimens (range 1 20) of each life stage and structure for each species. However, we coded missing data for a few species from literature sources [i.e. characteristics of most eggs; pupae of Oc. (Fin.) seoulensis, Oc. (Och.) andersoni, Oc. (Och.) flavifrons, Oc. (Och.) nivalis, Oc. (Och.) ratcliffei, Oc. (Och.) theobaldi and Oc. (Och.) trivittatus; and larvae of Oc. (Och.) andersoni, Oc. (Och.) flavifrons and Oc. (Och.) nivalis] or from collection data sheets. Pupae are unknown for Da. echinus, Gi. pulchriventer and Oc. (Fin.) alboannulatus. The pupal description of Oc. (Och.) zammitii is incomplete and is based on two dark, slide-mounted pupae with developing adults visible inside. The remaining missing data are denoted by a? in the data set (Appendix 2). States of continuous characters were determined either by clear gaps in the observed counts or measurements (e.g. characters 34 and 69), or by reference to observed intraspecific variation (e.g. character 47). Characters that could not be scored due to absence of homologous structures ( dependent characters ) are indicated by a dash ( ), e.g. absence of setae 6-S and 9-S on the larval siphon of Mansonia. All multistate characters were treated as unordered. Polymorphic characters are explicitly coded as exhibiting only those states observed. Specimens from the following collections were examined during the study: National Museum of Natural History, Smithsonian Institution, Washington, DC; The Natural History Museum, London, UK; Department of Zoology and Entomology, University of Queensland, Brisbane, Australia; Australian National Insect Collection, CSIRO Division of Entomology, Canberra, Australia; Florida State Collection of Arthropods, Division of Plant Industry, Gainesville, Florida; Laboratoire/Cellule Entomologie, EID Mediterranée, Montpellier, France; Bohart Museum, University of California, Davis, California; National Institute of Malariology, Parasitology and Entomology, Hanoi, Vietnam; Florida Medical Entomology Laboratory, University of Florida, Vero Beach, Florida; Connecticut Agricultural Experiment Station, New Haven, Connecticut; Canadian National Collection of Insects, Arachnids and Nematodes, Ottawa, Ontario, Canada; and the collection of the first author. Abbreviations used for generic-level taxa follow Reinert (2001a) and Reinert et al. (2006). New generic-level taxa proposed herein are described in Appendix 3. Changes to the generic and subgeneric placement of affected species are indicated in Appendix 4. Due to the unavailability of specimens for study and inadequate published descriptions, a number of species could not be assigned to generic-level taxa with certainty and remain in the category of incertae sedis. Abbreviations for newly recognized genera are listed in Appendix 5. See Reinert et al. (2004, 2006) for the classification of other aedine taxa. CHARACTER DESCRIPTIONS Character numbers used in Reinert et al. (2006) are included in parentheses following current character numbers. Adult characters were derived from females unless otherwise noted. Males of many species often have fewer setae and scales than females, and would skew the coding of some character states. We have reinterpreted the homologies of some structures from those used in some published articles and additional explanations are given below in the character descriptions. Setal branching characters are coded to reflect observed intraspecific variation. Hence, some such characters are coded in the form of (0) single; (1) branched whereas others are coded in the form of (0) single or two-branched; (1) > 3 branches (or similar form) to reflect the actual development of the particular seta in each species. Anatomical nomenclature and chaetotaxy follow Harbach & Knight (1980, 1982) except for terms proposed by Reinert (1990, 1999, 2000b). Many

4 32 J. F. REINERT ET AL. structures of characters used in the current study are illustrated in these references. However, references to character states illustrated in other publications are provided in the explanations of the characters listed below. Explanation of characters and references to illustrations in our previous papers (Reinert et al., 2004, 2006) should be consulted as only information on new and modified characters is provided herein. Some previously used characters, however, include supplemental information about ingroup taxa. Also, some previously used multistate characters were divided into two or more characters to simplify coding regimes. EGGS Reinert (2005) noted that the egg stage has been described and/or illustrated for only approximately 16% of the currently recognized species of Aedini, and provided a list of those species. The four egg characters below are coded as missing, in part or in total, for approximately half of the taxa included in this study. 1(1). Deposition: (0) laid singly; (1) laid in a raft; (2) laid in a mass. Numerous authors have indicated that the eggs of species of Aedes and Ochlerotatus (as genus Aedes and its previously recognized subgenera) are laid singly. Egg deposition, however, is unknown for numerous species of Aedini. 2(2). Shape: (0) anterior end flattened, posterior end rounded; (1) both ends tapered; (2) anterior end elongate and narrow. 3(new). Width, midlength: (0) somewhat expanded; (1) greatly expanded, more or less diamond-shaped in lateral outline. Most aedine eggs exhibit state (0) (see examples in Ross & Horsfall, 1965: fig. 188, St. aegypti; fig. 189, Oc. (Pro.) triseriatus; fig. 204, Oc. dorsalis). Examples of eggs greatly expanded at midlength often appear more or less diamond-shaped in lateral outline (state 1) and include Oc. (Rus.) rusticus (see Marshall, 1938: fig. 57h; Service, Duzak & Linley, 1997: figs 13, 14a), Ps. (Pso.) ciliata and Ps. (Pso.) howardii (see Horsfall, Voorhees & Cupp, 1970: figs 32A, 33A) and Mu. alternans (see Linley, Geary & Russell, 1991: fig. 12). 4(3). Outer chorion: (0) pattern absent or weakly developed; (1) pattern well developed; (2) with spiny appearance, each cell with elongate anteriorly inclined tubercle. LARVAE (FOURTH-INSTARS) 5(4). Labiogula: (0) short, length < width; (1) moderate to long, length width. 6(5). Antenna, length: (0) short, 0.40 median length of dorsal apotome (DAp); (1) moderate to long, 0.42 median length of DAp. 7(6, modified). Antenna, spicules: (0) absent; (1) present. Character 6 of Reinert et al. (2006) included two states, few and numerous, that are now represented by state (1). 8(7). Seta 1-A, length: (0) short, 3.0 times antennal width at point of attachment; (1) longer, 3.1 times antennal width at point of attachment. 9(8). Seta 1-A, development: (0) single or 2-branched; (1) 3 branches. 10(9). Setae 2,3-A, position: (0) insertion apical or nearly apical; (1) insertion noticeably subapical. 11(10). Seta 1-C, development: (0) spiniform; (1) single, thinner, distal part attenuate; (2) forked or branched, proximal part stout. 12(11, modified). Seta 4-C, position: (0) insertion anterior to seta 6-C; (1) insertion at same level as seta 6-C; (2) insertion posterior to seta 6-C. Character 11 of Reinert et al. (2006) was modified by separating state (2) from state (1). The positions of setae 4 7-C are determined on head capsules that are horizontal with the thorax and abdomen, normally exuviae. This also applies to characters 16 and (12). Seta 4-C, length: (0) short, 0.19 median length of DAp; (1) moderate, median length of DAp; (2) long, median length of DAp; (3) very long, 0.90 median length of DAp. The lengths of setae 4 7-C are compared with the median length of the dorsal apotome, preferably in larval exuviae. 14(13). Seta 5-C, length: (0) short, 0.19 median length of DAp; (1) moderate, median length of DAp; (2) long, median length of DAp; (3) very long, 0.90 median length of DAp. 15(14). Seta 5-C, development: (0) single; (1) branched. 16(15). Seta 6-C, position: (0) insertion anterior to seta 7-C; (1) insertion at same level or posterior to seta 7-C. 17(16). Seta 6-C, length: (0) short, 0.19 median length of DAp; (1) moderate, median length of DAp; (2) long, median length of DAp; (3) very long, 0.90 median length of DAp. 18(new). Seta 6-C, development: (0) single; (1) branched. See Wood, Dang & Ellis (1979) for illustrations of state (0) (pl. 47, Oc. sollicitans) and state (1) (pl. 19, Ae. cinereus).

5 PHYLOGENY AND CLASSIFICATION OF OCHLEROTATUS 33 19(17). Seta 7-C, position: (0) insertion anterior to seta 5-C; (1) insertion at approximately same level as seta 5-C; (2) insertion posterior to seta 5-C. 20(18). Seta 7-C, length: (0) short, 0.19 median length of DAp; (1) moderate, median length of DAp; (2) long, median length of DAp. 21(new). Seta 7-C, development: (0) single (rarely 2-branched); (1) 3 branches. See Belkin (1962) for examples of state (0) (fig. 277, Mu. alternans) and state (1) (fig. 269, Oc. (Och.) vigilax). 22(19). Seta 12-C, position: (0) insertion mesal to or posterior in line with seta 13-C; (1) insertion lateral to seta 13-C. 23(20). Seta 13-C, development: (0) single; (1) branched. 24(21). Seta 14-C, development: (0) single; (1) branched. 25(22). Seta 19-C: (0) absent; (1) present. 26(23). Ventromedian cervical sclerite: (0) absent; (1) present. 27(24). Setae 1 3-P: (0) not attached to a common setal support plate; (1) two or three of these setae attached to a common setal support plate. 28(25). Seta 1-P, length: (0) length of seta 2-P; (1) > length of seta 2-P. 29(new). Seta 4-P, length: (0) length of seta 3-P; (1) > length of seta 3-P. See Belkin (1962) for examples of state (0) (figs 218, 234, Ge. longiforceps and Pm. argyronotum) and state (1) (figs 274, 316, Le. geoskusea and Sc. albolineata). 30(new). Seta 4-P, development: (0) single; (1) branched. Seta 4-P is single in a number of taxa, e.g. Oc. (Och.) excrucians and Ta. savoryi (see Tanaka, Mizusaka & Saugstad, 1979: figs 82, 99), whereas in other taxa it is branched (see Tanaka et al., 1979: figs 79, 102, Oc. (Och.) vigilax and Oc. (Fin.) seoulensis). 31(26). Seta 5-P, length: (0) length of seta 6-P; (1) > length of seta 6-P. 32(27). Seta 5-P, development: (0) single; (1) branched. 33(new). Seta 7-P development: (0) single; (1) branched. Belkin (1962) illustrated state (0), e.g. Ha. australis and Mu. alternans (figs 211, 277), and state (1), e.g. Oc. (Och.) edgeri and Oc. (Och.) mcdonaldi (figs 267 and 271). 34(28). Seta 8-P, length: (0) 1.2 times length of seta 4-P; (1) 1.8 times length of seta 4-P. 35(29). Seta 8-P, development: (0) single or occasionally 2-branched; (1) multiple-branched (occasionally 3-branched). 36(30). Seta 1-M, length: (0) 2.5 times length of seta 2-M; (1) 3.5 times length of seta 2-M. 37(31). Seta 1-M, development: (0) single, rarely 2-branched; (1) 3 branches. 38(32). Seta 4-M, development: (0) single; (1) branched. 39(33). Seta 7-M, length: (0) < length of seta 5-M; (1) length of seta 5-M. 40(new). Seta 1-T, development: (0) slender to slightly thickened; (1) very stout. Most species of Aedini have seta 1-T slender to slightly thickened. Seta 1-T is distinctly stout and inserted on a tubercle in some species, e.g. Co. hatorii (see Tanaka et al., 1979: fig. 98) and Co. elsiae (see Barraud, 1934: fig. 39c). 41(new). Seta 1-T, insertion: (0) on surface of integument; (1) on tubercle or plate. Most aedine taxa have seta 1-T inserted on the surface of the integument. 42(34). Seta 2-T, development: (0) single; (1) branched. 43(35). Seta 4-T, development: (0) single; (1) 2 branches, not stellate; (2) 3 branches, stellate. 44(new). Seta 6-T, development: (0) single; (1) branched. Many taxa of Aedini have seta 6-T single (see Arnell & Nielsen, 1972: figs 9, 13, Ja. (Jar.) varipalpus and Ja. (Jar.) sierrensis), whereas some taxa have seta 6-T branched (see Reinert, 1999: fig. 14, Za. longipalpis). 45(36). Seta 3-I, development: (0) single; (1) branched. 46(37). Seta 7-I, length: (0) < 0.45 length of seta 6-I; (1) 0.55 length of seta 6-I. 47(38). Seta 7-I, development: (0) single to 3-branched; (1) 4 branches. 48(39). Seta 12-I: (0) absent; (1) present. 49(40). Seta 1-II, development: (0) single; (1) 2 branches, not stellate; (2) 3 branches, stellate. 50(41). Seta 2-II, development: (0) single; (1) branched. 51(new). Seta 3-II, development: (0) single; (1) branched. See Zavortink (1972) for illustrations of seta 3-II single (figs 32, 34, Ko. purpureipes and Ab. papago) and seta 3-II branched (figs 18, 28, Oc. (Pro.) kompi and Oc. (Pro.) knabi). 52(42). Seta 5-II, development: (0) single; (1) 2 branches, not stellate; (2) 3 branches, stellate.

6 34 J. F. REINERT ET AL. 53(43). Seta 6-II, length: (0) < length of seta 6-III; (1) length of seta 6-III. 54(44). Seta 6-II, development: (0) single; (1) branched. 55(45). Seta 7-II, development: (0) similar to seta 7-I; (1) different than seta 7-I. 56(46). Seta 8-II, development: (0) single; (1) branched. 57(47). Seta 6-III, development: (0) single; (1) branched. 58(48). Seta 3-V, length: (0) 1.55 times length of seta 5-V; (1) 1.90 times length of seta 5-V. 59(49). Seta 1-VII, length: (0) short, 0.42 dorsal length of segment X; (1) moderately long, dorsal length of segment X; (2) long to very long, 0.94 dorsal length of segment X. Segment X length is the middorsal length measured along a straight line parallel to the longitudinal axis from the posterior margin of the saddle (minus marginal spicules) anteriorly to the first annulation in the membrane cephalad of the saddle. Fourth-instar larvae, not exuviae, should be used for measurements of segment X. 60(50). Seta 2-VII, position: (0) insertion near seta 1-VII; (1) insertion far anterior to seta 1-VII. 61(51). Seta 2-VII, development: (0) single; (1) branched. 62(52). Seta 3-VII, position: (0) insertion anterior to seta 1-VII; (1) insertion at same level or posterior to seta 1-VII. 63(53). Seta 3-VII, development: (0) single; (1) branched. 64(54). Seta 3-VII, length: (0) short, 0.42 dorsal length of segment X; (1) moderately long, dorsal length of segment X; (2) long to very long, 0.94 dorsal length of segment X. See information under character 59 for dorsal length of segment X. 65(55). Seta 10-VII, development: (0) single; (1) branched. 66(56). Seta 12-VII, position: (0) insertion anterior to seta 13-VII; (1) insertion at approximately same level as seta 13-VII; (2) insertion posterior to seta 13-VII. 67(57). Seta 12-VII, development: (0) single; (1) branched. 68(58). Setae 1,2-VIII: (0) not inserted on common setal support plate; (1) both inserted on common setal support plate. 69(59). Seta 1-VIII, length: (0) short, 0.38 length of seta 2-VIII; (1) moderately long, length of seta 2-VIII; (2) long, 1.10 length of seta 2-VIII. 70(new). Seta 1-VIII, development: (0) single; (1) branched. Seta 1-VIII is normally branched in Aedini, e.g. Oc. dorsalis (see Tanaka et al., 1979: fig. 98) and Oc. (Pro.) terrens (see Schick, 1970: fig. 16). Seta 1-VIII is single in a few taxa, e.g. Co. hatorii (see Tanaka et al., 1979: fig. 98) and Oc. (Och.) spilotus (see Dobrotworsky, 1965: fig. 47). 71(60). Seta 2-VIII, development: (0) single; (1) branched. 72(61). Seta 4-VIII, development: (0) single; (1) branched. 73(62). Comb: (0) few to several scales in one or two irregular rows; (1) numerous scales in a patch. 74(63). Comb plate: (0) absent; (1) present. 75(64). Siphon, acus: (0) absent; (1) present. 76(65, in part). Pecten: (0) absent; (1) present. 77(65, in part). Pecten: (0) spines evenly spaced; (1) distal one or more spines more widely spaced. Species without a pecten, e.g. Ma. titillans and Or. signifera, are coded ( ). 78(66). Seta 1-S, development: (0) one seta at base of siphon; (1) one seta some distance from base of siphon; (2) two or more setae some distance from base of siphon. 79(new). Seta 1a-S, length: (0) short, 0.40 width of siphon; (1) moderately long, width of siphon; (2) long, width of siphon; (3) very long, 2.17 width of siphon. The length of seta 1a-S is compared with the width of the siphon of fourthinstar larval exuviae at the point of insertion. 80(new). Seta 1a-S, development: (0) single; (1) branched. Seta 1a-S is single in a few taxa, e.g. Ps. (Pso.) ciliata (see Belkin, Heinemann & Page, 1970: fig. 56) and Ps. (Pso.) howardii, whereas it is branched in most Aedini (see Belkin et al., 1970: fig. 69, Oc. (Och.) taeniorhynchus). 81(new). Seta 1a-S, insertion: (0) within pecten; (1) distal to pecten. See Zavortink (1972) for taxa with seta 1a-S inserted within the pecten (figs 47, 50, Gc. (Gec.) atropalpus and Gc. (Gec.) epactius) and seta 1a-S inserted distal to the pecten (fig. 9, Oc. (Pro.) triseriatus). Taxa without a pecten, e.g. Ma. titillans and Or. signifera, are coded ( ). 82(67). Seta 6-S, development: (0) short; (1) long. Seta 6-S is short if its length is less than the distal width of the siphon (see Belkin, 1962: fig. 320, St. aegypti) and long if the length is noticeably greater than the distal width of the siphon (see Belkin, 1962: fig. 245, Fl. franclemonti).

7 PHYLOGENY AND CLASSIFICATION OF OCHLEROTATUS 35 83(68). Seta 8-S, development: (0) short; (1) long. The length of seta 8-S is determined by comparison with the distal width of the siphon. See the annotation under character (69). Seta 9-S, development: (0) short, slender, nearly straight or slightly curved; (1) stouter, hooklike. 85(70). Sclerotization of segment X (saddle): (0) incomplete ventrally; (1) completely encircles segment X. 86(71). Saddle, acus: (0) absent; (1) present. 87(72). Saddle, moderate to well-developed spicules on posterior margin: (0) absent; (1) present. 88(73). Seta 1-X, position: (0) insertion on saddle; (1) insertion ventral to saddle. 89(new). Seta 2-X, length: (0) long; (1) short to moderately long. Seta 2-X exhibits state (0) if it is slightly shorter to slightly longer than seta 3-X (see Tyson, 1970: fig. 17, Mu. (Muc.) laniger); it exhibits state (1) if it is 0.70 length of seta 3-X (see Wood et al., 1979: pl. 30, Oc. (Och.) hexodontus). 90(new). Seta 2-X, development: (0) single or 2-branched (rarely 3-branched); (1) > 5 branches (rarely 4-branched). In state (0) seta 2-X is normally single or 2-branched but rarely 3-branched (see Tyson, 1970: fig. 17, Mu. (Muc.) laniger), whereas in state (1) it has numerous branches (see Wood et al., 1979: pl. 30, Oc. (Och.) hexodontus). 91(74). Seta 3-X, development: (0) single; (1) branched. 92(75). Sclerotization supporting seta 4-X (ventral bush): (0) absent; (1) grid with only transverse bars; (2) grid with both lateral and transverse bars; (3) boss. 93(76). Precratal/preboss setae (i.e. two or more setae anterior to grid or boss): (0) absent; (1) present. 94(77). Seta 4a-X of ventral brush, length: (0) short; (1) long. The length of the most caudal seta (4a-X) of the ventral brush is determined in comparison with the length of the next cephalad seta (4b-X). Seta 4a-X is short if it is less than 0.25 of the length of seta 4b-X (see Zavortink, 1972: figs 32, 37, Ko. purpureipes and Az. ramirezi). Most aedine species have seta 4a-X as long as or only slightly shorter than seta 4b-X (see Zavortink, 1972: fig. 47, Gc. (Gec.) epactius). 95(78, in part). Seta 4a-X of ventral brush, development: (0) single, slender, simple; (1) single or forked distally, stout, plumose; (2) branched, short stem; (3) branched, moderate to long stem. Belkin (1962) illustrated several states of seta 4a-X, e.g. state (0) (fig. 350, St. tongae), state (2) (fig. 269, Oc. (Och.) vigilax) and state (3) (fig. 237, Fl. bougainvillensis). See Reinert (1993: fig. 8, Mo. pecuniosa) for state (1). PUPAE 96(79). Cephalothorax, with clear, unpigmented spots: (0) absent; (1) present. 97(80, in part). Trumpet, tracheoid area: (0) absent; (1) present. 98(80, in part). Trumpet, tracheoid area: (0) at base, weakly developed; (1) distal to base, well developed. Species without a tracheoid area, e.g. Op. (Opi.) fuscus, Ps. (Gra.) jamaicensis and Ps. (Pso.) howardii, are coded ( ). 99(81). Seta 1-CT, development: (0) similar to seta 3-CT; (1) very strongly developed, considerably longer than seta 3-CT; (2) weakly developed, considerably shorter than seta 3-CT. 100(82). Seta 5-CT, length: (0) 1.2 times length of seta 4-CT; (1) > 1.3 times length of seta 4-CT. 101(83). Seta 7-CT, length: (0) length of seta 6-CT; (1) times length of seta 6-CT; (2) 6.0 times length of seta 6-CT. 102(84). Seta 11-CT, development: (0) single; (1) branched. 103(85). Seta 13-CT: (0) absent; (1) present. 104(86). Seta 3-I, length: (0) < length of seta 6-I; (1) length of seta 6-I. 105(87). Seta 3-I, development: (0) single, rarely split apically; (1) branched. 106(88). Seta 6-I, length: (0) length of seta 7-I; (1) > length of seta 7-I. 107(new). Seta 1-II, development: (0) 3 branches (rarely 4-branched); (1) 5 branches. 108(89). Seta 2-II, position: (0) insertion mesal to or anterior in line with seta 1-II; (1) insertion lateral to seta 1-II. 109(90). Seta 2-II, length: (0) < length of seta 1-II; (1) length of seta 1-II. 110(91). Seta 3-II, position: (0) insertion mesal to or posterior in line with seta 2-II; (1) insertion lateral to seta 2-II. 111(new). Seta 3-II, development: (0) single; (1) branched. Belkin (1962) illustrated state (0) (fig. 230, Ra. albilabris), and state (1) (fig. 278, Mu. painei). 112(92). Seta 3-II, length: (0) length of seta 6-II; (1) > length of seta 6-II. 113(new). Seta 3-II: (0) slender; (1) stout. The thickness of seta 3-II is determined by comparison with seta 1-II. See Belkin (1962) for illustrations of state (0) (fig. 210, Ha. australis) and state (1) (fig. 236, Fl. bougainvillensis).

8 36 J. F. REINERT ET AL. 114(93). Seta 5-II, position: (0) insertion lateral to or posterior in line with seta 4-II; (1) insertion mesal to seta 4-II. 115(94). Seta 5-II, length: (0) length of seta 3-II; (1) > length of seta 3-II. 116(95). Seta 6-II, length: (0) length of seta 7-II; (1) > length of seta 7-II. 117(new). Seta 3-III, length: (0) length of seta 5-III; (1) > length of seta 5-III. Seta 3-III is noticeably longer than seta 5-III in most Aedini (see Belkin, 1962: fig. 207, Op. (Opi.) fuscus), however, seta 3-III is shorter than seta 5-III is some taxa (see Belkin, 1962: fig. 210, Ha. australis). 118(96). Seta 3-III, development: (0) single; (1) branched. 119(97). Seta 6-III, development: (0) single; (1) branched. 120(98). Seta 5-V, length: (0) < medial length of tergum VI; (1) medial length of tergum VI. 121(99). Seta 2-VI, position: (0) insertion mesal to or anterior in line with seta 1-VI; (1) insertion lateral to seta 1-VI. 122(100). Seta 3-VI, position: (0) insertion mesal to or anterior in line with seta 1-VI; (1) insertion lateral to seta 1-VI. 123(101). Seta 6-VII, position: (0) insertion anterior to seta 9-VII; (1) insertion posterior to or mesal at same level as seta 9-VII. 124(102). Seta 9-VII, length: (0) length of seta 6-VII; (1) > length of seta 6-VII. 125(new). Seta 9-VIII, position: (0) insertion on posterolateral corner; (1) insertion slightly anterior or mesal to posterolateral corner; (2) insertion noticeably anterior to posterolateral corner. Seta 9-VIII is inserted on the posterolateral corner of tergum VIII in many aedine taxa (see Zavortink, 1972: fig. 46, Gc. (Gec.) epactius). See Berlin (1969: fig. 43, Hw. sexlineata) for state (1). A few taxa have seta 9-VIII inserted noticeably anterior to the posterolateral corner of tergum VIII (see Reinert, 1999: fig. 11, Za. longipalpis). 126(103). Seta 9-VIII, development: (0) single or 2-branched; (1) 3 branches. 127(104). Paddle midrib: (0) weakly developed, terminating well before apex of paddle; (1) well developed, extending to or near apex of paddle. 128(105). Paddle, fringe of hair-like spicules: (0) absent; (1) present. 129(106). Paddle, apical margin: (0) without emargination; (1) with moderately deep emargination. Species with state (0) have the apical margin sharply to broadly rounded or flat. 130(108, in part). Seta 1-Pa: (0) absent; (1) present. 131(107, modified). Seta 1-Pa, length: (0) 0.33 length of paddle; (1) length of paddle; (2) 0.80 length of paddle. Species without seta 1-Pa, e.g. Ma. titillans, are coded ( ). 132(108, in part). Seta 1-Pa, development: (0) single (rarely 2-branched); (1) 3 branches. Species without seta 1-Pa are coded ( ). 133(109). Seta 2-Pa: (0) absent; (1) present. ADULTS (FEMALES EXCEPT WHERE OTHERWISE NOTED) 134(110, in part). Erect scales of head: (0) absent; (1) present. 135(110, in part). Erect scales of head: (0) restricted to occiput; (1) on both occiput and vertex. Tanaka et al. (1979) illustrated state (0) (fig. 234, St. aegypti) and state (1) (fig. 215, Oc. (Och.) sticticus). Species without erect scales on the head, e.g. Op. (Opi.) fuscus, are coded ( ). 136(111). Decumbent scales of vertex: (0) all broad; (1) all narrow; (2) both broad and narrow. 137(112). Ocular line, width: (0) narrow; (1) broad. 138(113). Ocular scales: (0) all narrow; (1) all broad; (2) both narrow and broad. See Reinert et al. (2004, 2006) for quantification of the character states. 139(114). Eyes, immediately above antennal pedicels: (0) contiguous; (1) narrowly to moderately separated; (2) broadly to very broadly separated. The distance separating the eyes is determined on the area immediately above the antennal pedicels and is measured in numbers of eye facets (diameter). Species with eyes that touch or are separated by less than or equal to one eye facet are considered contiguous and are scored (0), e.g. Ph. assamensis, those separated by 2 4 (rarely 5) facets are scored (1), e.g. Da. geniculata, and those separated by 6 or more facets are scored (2), e.g. Op. (Opi.) fuscus. Scales protruding from the median, anterior area of the vertex or upper part of the interocular space may obscure this area, and in some cases, e.g. Howardina, the scales must be removed to determine the distance between the eyes. Gutsevich (1974a, b, 1975a, b) used the diameter of eye facets compared with the distance between the eyes as a means of distinguishing several genera and subgenera, although he used a different criterion than the number of facets used here.

9 PHYLOGENY AND CLASSIFICATION OF OCHLEROTATUS (115, in part). Interocular space, scales: (0) absent; (1) present. In Geoskusea, Levua and Rhinoskusea, the interocular space is reduced to a very small triangle and is apparently without scales. In these subgenera, the eyes are contiguous for nearly the entire mesal length, resulting in a very small interocular space. Most taxa have interocular scales. 141(115, in part). Interocular space, scales: (0) all narrow; (1) all broad; (2) both narrow and broad. Many taxa have narrow scales on this area. Broad scales on the interocular space occur in other taxa, e.g. Abraedes, Downsiomyia, Gymnometopa, Haemagogus, Kenknightia, Kompia, Stegomyia and Zavortinkius. Opifex (Opi.) fuscus has only a few scattered small broad scales on this area. Both narrow and broad scales occur on the interocular space of Mc. (Cha.) wattensis. Species without interocular scales are coded ( ). See the annotation under character (116). Interocular space, setae: (0) absent; (1) present. 143(new). Interocular space, setae: (0) 5; (1) 6. Many aedine taxa have fewer than five setae on the interocular space, e.g. Ab. papago and Fl. kochi, whereas other taxa have six or more, often numerous, setae on this area, e.g. Oc. dorsalis. Taxa without setae on the interocular space, e.g. Gy. mediovittata and Ko. purpureipes, are coded ( ). 144(117, in part) Antennal pedicel, vestiture on mesal surface: (0) absent; (1) present. 145(117, in part). Antennal pedicel, development of vestiture on mesal surface: (0) few to numerous scattered scales (not overlapping to only slightly overlapping and not silvery) and/or setae; (1) patch of broad, overlapping, silvery scales. Species without vestiture on the mesal surface of the antennal pedicel, e.g. Vansomerenis, are coded ( ). 146(118). Antennal pedicel, scales on lateral surface: (0) absent; (1) present. 147(119). Apical flagellomeres (males): (0) both apical flagellomeres disproportionately long compared with other flagellomeres; (1) these flagellomeres not disproportionately long in comparison with others. 148(120). Antenna, development of flagellar whorls (males): (0) few short setae, dispersed more or less around flagellomeres; (1) several moderately long to long setae, directed more or less dorsally and ventrally; (2) numerous long setae, normally directed dorsally and ventrally. 149(121). Maxillary palpus, pale scales: (0) absent; (1) present. 150(122). Maxillary palpomeres (males): (0) five, palpomeres 2 and 3 fused/ankylosed; (1) three, palpomere 4 absent or vestigial. 151(new). Maxillary palpomeres, position of palpomeres 4 and/or 5 in relationship to 3 (males): (0) down-turned; (1) up-turned; (2) nearly straight. State (0) normally includes both palpomeres 4 and 5 that usually bear numerous long setae on the ventrolateral margins of at least palpomere 4 (see Tanaka et al., 1979: fig. 213, Oc. (Och.) excrucians), but in a few taxa only palpomere 5 is down-turned. Up-turned palpomeres 4 and 5 occur in a few taxa, e.g. Gymnometopa, Psorophora, Stegomyia (see Tanaka et al., 1979: fig. 234, St. aegypti) and the non-aedine outgroup species Cx. quinquefasciatus. Palpomeres 4 and 5 are nearly straight with palpomere 3 in some taxa, e.g. Bruceharrisonius, Geoskusea (see Tanaka et al., 1979: fig. 239, Ge. baisasi), Isoaedes, Rhinoskusea and Scutomyia. Species without palpomeres 4 and 5, e.g. Aedes, Levua and Rhinoskusea, are coded ( ). 152(new). Maxillary palpus, palpomere 3 length (males): (0) short, 0.14 length of proboscis; (1) long, 0.21 length of proboscis. Palpomere 3 is short in some Aedini, e.g. Aedes, Levua and Rhinoskusea (see Reinert, 1976: fig. 2, Rh. longirostris) and long in many taxa, e.g. Aztecaedes, Kompia and Ochlerotatus (see Tanaka et al., 1979: figs 212, 213, Oc. dorsalis and Oc. (Och.) excrucians). 153(new). Maxillary palpus, palpomere 5 length (males): (0) short, 0.55 length of palpomere 4; (1) long, 0.67 length of palpomere 4. Only a few taxa included in the present study have a short palpomere 5, e.g. Ia. cavaticus, Mc. (Mac.) tremula and the non-aedine outgroup species Or. signifera. Most species have a long palpomere 5. Species without palpomeres 4 and/or 5, e.g. Aedes, Levua and Rhinoskusea, are scored ( ). 154(123). Maxillary palpus, length (males): (0) short, 0.25 proboscis length; (1) moderate, proboscis length; (2) long, 0.84 proboscis length. 155(124, modified). Maxillary palpus, development of setae on palpomeres 3 (distally) and 4 (males): (0) absent or few, short to moderately long; (1) moderate to numerous, long. Numerous long setae are present on the ventral surface of palpomere 4 and distally on the ventral surface of palpomere 3 in a number of aedine taxa, e.g. Mucidus (see Tyson, 1970: fig. 6, Mu. quasiferinus), Ochlerotatus and Ae. (Aedimorphus) (see Tanaka et al., 1979: figs 213, 215, 237, 238, Oc. (Och.) excrucians, Oc. (Och.) sticticus, Ae. (Adm.) alboscutellatus and Ae. (Adm.) vexans vexans), whereas some taxa possess moderate numbers of long setae on these palpomeres, all

10 38 J. F. REINERT ET AL. are scored (1). In other taxa, setae are absent or only a few short to moderately long ones are present (state 0), see Tanaka et al. (1979): figs 234, 239, 243, St. aegypti, Ge. baisasi and Ae. yamadai, and Edwards (1941): fig. 68, Er. quinquevittatus. Taxa without palpomere 4, e.g. Aedes, Opifex and Rhinoskusea, are coded ( ). 156(new). Proboscis, length: (0) < length of forefemur; (1) length of forefemur. The length of the proboscis is greater than, or occasionally equal to, the length of the forefemur in most aedine taxa, e.g. Bruceharrisonius, Downsiomyia, Haemagogus (see Arnell, 1973: fig. 38, Hg. splendens) and Ochlerotatus (most species), but the proboscis is shorter than the forefemur in a few taxa, e.g. Fl. kochi, Mu. quadripunctis and Za. longipalpis. 157(125). Proboscis, pale scales: (0) absent; (1) present. 158(new). Proboscis, pale-scaled band near midlength: (0) absent; (1) present. Most aedine taxa do not have a pale-scaled band near the middle of the proboscis, e.g. Ae. esoensis (see Tanaka et al., 1979: fig. 242), Ge. baisasi, Hg. (Hag.) equinus, Oc. (Och.) rubrithorax (see Russell, 1996: 81) and St. aegypti. Other taxa have a white-scaled band, e.g. Fl. kochi (see Russell, 1996: 61), Ra. notoscripta, Oc. (Och.) mitchellae and Oc. (Och.) taeniorhynchus (see Carpenter & LaCasse, 1955: pl. 86). 159(126). Antepronota: (0) approximated; (1) widely separated. 160(146). Antepronotum, scales: (0) absent; (1) present. Most species of Aedini have scales on the antepronotum, but a few taxa do not, e.g. Geoskusea, Levua and Rhinoskusea. 161(new). Antepronotum, scales: (0) all narrow; (1) all broad; (2) both narrow and broad; (3) broad and erect twisted. Erect, twisted scales occur in species of Mu. (Mucidus). Also, see the annotations under characters 166, 184 and (127). Anterior acrostichal setae: (0) absent; (1) present. 163(128). Posterior acrostichal setae: (0) absent; (1) present. 164(129). Anterior dorsocentral setae: (0) absent; (1) present. 165(130). Posterior dorsocentral setae: (0) absent; (1) present. 166(new). Scutum, scales: (0) all narrow; (1) all broad; (2) both narrow and broad; (3) narrow, broad and erect twisted. The scutum of most aedine taxa is covered with narrow scales. Broad scutal scaling occurs in some Aedini, e.g. Hg. (Hag.) splendens (see Arnell, 1973: fig. 38). Taxa with broad scales on the anterior promontory and/or supraalar area (anterior to the base of the wing), and narrow scales elsewhere on the scutum, are scored as state (2). Numerous erect, twisted scales in addition to narrow and broad scales on much of the scutum are found in species of Mu. (Mucidus) (see Tyson, 1970: figs 1 3 and 5). 167(new). Scutum, scale colour: (0) all dark; (1) both pale and dark; (2) all pale. Geoskusea baisasi (see Tanaka et al., 1979: fig. 239) is an example of a species with a dark-scaled scutum. The scutum of many aedine taxa has both pale and dark scales (see Arnell, 1976: fig. 35, Oc. scapularis). Ochlerotatus (Och.) caspius has a contrasting pattern of golden and white scales, and is scored (2). 168(131). Acrostichal and dorsocentral areas, anterior covered with pale scales: (0) absent; (1) present. 169(new). Acrostichal and dorsocentral areas, anterior with variable arrangement of pale and darker scales: (0) absent; (1) present. A variable arrangement or random mixture of pale and dark scales on the anterior part of the acrostichal and dorsocentral areas is found in some aedine taxa, e.g. Oc. (Och.) aculeatus, Oc. (Och.) cantans and Oc. (Och) edgari. 170(132). Anterior acrostichal area, with narrow to moderate, pale-scaled stripe: (0) absent; (1) present. 171(133). Posterior acrostichal area, with narrow to moderate, pale-scaled stripe: (0) absent; (1) present. 172(134). Anterior dorsocentral area, with narrow, pale-scaled stripe: (0) absent; (1) present. 173(135). Posterior dorsocentral area, with narrow, pale-scaled stripe: (0) absent; (1) present. 174(136). Scutal fossa, scales: (0) all dark; (1) contrasting pale scales in large patch; (2) contrasting lines or small patches of pale scales on lateral and/or mesal and/or posterior margins. Some species have a variable arrangement of pale and darker scales, and are scored (2). 175(new). Scutal fossal scales: (0) sparse; (1) dense. Most aedine taxa have uniformly dense narrow or broad scales covering the scutal fossa. The scutal fossa has bare areas or sparse, normally narrow scales in a few taxa, e.g. Oc. (Och.) bimaculatus, Oc. fulvus pallens, Ps. (Pso.) ciliata and Ps. (Pso.) howardii.

11 PHYLOGENY AND CLASSIFICATION OF OCHLEROTATUS (137). Prescutellar area, median and/or posterior parts: (0) bare; (1) with scales and/or setae. 177(138). Prescutellar setae: (0) absent; (1) present. 178(new). Prescutellar setae: (0) 5 per side; (1) 6 per side. Taxa exhibiting state (1) normally have 10 or more setae on each side of the prescutellar space, but some have as few as six setae per side. Species without prescutellar setae, e.g. Hg. (Haemagogus), are coded ( ). 179(139). Prescutellar area, pale scales on outer margin mesal to setae: (0) absent; (1) present. 180(140, in part). Antealar area, anterior part, scales: (0) absent; (1) present. 181(140, in part). Antealar area, anterior part, scales: (0) dark; (1) pale. Species without scales on the anterior part of the antealar area, e.g. Co. banksi and Co. pseudotaeniatus, are coded ( ). 182(141). Supraalar area, scales: (0) all dark; (1) with longitudinal patch or stripe of pale scales; (2) with transverse patch of pale scales. 183(142, modified). Scutellum, scales on midlobe: (0) all narrow; (1) all broad; (2) both narrow and broad; (3) narrow, broad and erect twisted. State (3) occurs in species of Mu. (Mucidus). 184(143, modified). Scutellum, scales on lateral lobes: (0) all narrow; (1) all broad; (2) both narrow and broad; (3) narrow, broad and erect twisted. See the annotations under characters 161 and (144). Paratergal scales: (0) absent; (1) present. 186(145). Parascutellar scales: (0) absent; (1) present. 187(147). Postpronotal scales: (0) absent; (1) present. 188(new). Postpronotal scales: (0) all narrow; (1) all moderately broad to broad; (2) both narrow and moderately broad to broad; (3) narrow, broad and erect twisted. Aedes esoensis is an example of a species that exhibits state (0) (see Tanaka et al., 1979: fig. 242). A postpronotum with broad scales is illustrated in Tanaka et al. (1979: fig. 229, Ph. watasei). Taxa with state (2) normally have narrow scales dorsally and a patch of broad scales ventrally on the postpronotum (see Wood et al., 1979: pls 31, 34, Oc. (Och.) impiger and Oc. (Och.) intrudens). State (3) is characteristic of subgenus Mucidus of Mucidus. Also, see the annotation under character (148). Prespiracular setae: (0) absent; (1) present. 190(149). Postspiracular setae: (0) absent; (1) present. 191(150). Postspiracular scales: (0) absent; (1) present. 192(151). Hypostigmal scales: (0) absent; (1) present. 193(152, in part). Subspiracular area, vestiture: (0) absent; (1) present. 194(152, in part). Subspiracular area, vestiture: (0) scales; (1) scales and setae. Species without vestiture on the subspiracular area, e.g. Ae. cinereus, Do. nivea, Ke. dissimilis and Mu. quadripunctis, are coded ( ). 195(153). Upper proepisternal setae: (0) few, 1 4; (1) moderate number, 5 19; (2) numerous, (154). Upper proepisternum, scales: (0) absent; (1) present. 197(155). Lower proepisternum, scales: (0) absent; (1) present. 198(156). Upper mesokatepisternal setae: (0) absent; (1) present. 199(new). Mesokatepisternum, scales: (0) in one large patch; (1) in two patches; (2) in three patches. See Wood et al. (1979) for examples of the mesokatepisternum with a single large patch of scales (pl. 15, Oc. (Och.) campestris), two scale-patches (pl. 16, Oc. (Och.) canadensis) and three scale-patches (pl. 34, Oc. (Och.) intrudens). 200(157). Prealar setae: (0) 20; (1) (158). Upper prealar area, scales: (0) absent; (1) present. Scales on the upper prealar area (prealar knob) are usually intermixed with the prealar setae but some taxa only have several scales on the ventral side of the knob that may be continuous with the scale patch on the lower prealar area. 202(159). Lower prealar area, scales: (0) absent; (1) present. 203(160). Mesepimeron, scales: (0) absent; (1) present. 204(new). Mesepimeron, scales: (0) in one patch; (1) in two patches. Taxa without scales on the mesepimeron are coded ( ). 205(161). Lower anterior mesepimeral setae: (0) absent; (1) present. 206(162). Mesepimeron, fine setae on ventral area: (0) absent; (1) present. 207(163, in part). Metameron, vestiture: (0) absent; (1) present. 208(163, in part). Metameron, development of vestiture: (0) with scales; (1) with setae. Taxa without vestiture on the metameron are coded ( ). 209(164, in part). Alula, marginal scales: (0) absent; (1) present.