SOUTH OF THE SAHARA (AFROTROPICAL REGION)

Transcription

1 OF THE PUBLICATIONS AFRICAN INSTITUTE FOR MEDICAL RESEARCH SOUTH No. 55 SUPPLEMENT TO THE ANOPHELINAE OF A SOUTH OF THE SAHARA AFRICA of Biological Sciences School of Sussex University Johannesburg South Africa (AFROTROPICAL REGION) by M. T. GILLIES Brighton England and MAUREEN COETZEE of Medical Entomology Department African Institute for Medical Research South THE SOUTH AFRICAN INSTITUTE FOR MEDICAL RESEARCH P O BOX IO38 JOHANNESBURG x987

2 ISBN o 62o o32x 3 and printed by Typeset South African Institute for Medical Research The (C) M T Gillies and M Coetzee, x987

3 TABLE OF CONTENTS Page Introduction Malaria transmission 2 Classification 2 Taxonomic changes 2 nomenclature 3 Morphological 7 Distribution List of species of Afrotropical 9 to the females 12 Key to the fourth stage larvae 36 Key to the pupae 59 Key account of species 64 Supplementary References cited 127 Additional uncited references 139

4 house-spraying, has died down. Cataloguing the anopheline fauna of many countries in the by is now well advanced, that of Ethiopia and Angola having been the most recent to be added. region both as to differences between member species and between populations of the same complex, and two, the recognition of complexes of sibling species in other groups. In both these areas species, role of cytogenetics guided by evolutionary theory has been paramount. The details of the banding sequences brought about by inversion arrangements have no place in a work chromosome this sort. But the importance of the genetic insights provided by such studies into the behaviour of of and their relevance to the control of disease cannot be over-emphasised. vectors supplement is meant to be used in close conjunction with the original volume. In it we give This accounts of the small number of newly described species together with such new synoptic on the biology and distribution of the fauna as has been published since the appearance information the second edition of 'De Meillon'. Coupled with this we provide new, illustrated keys to the of and larvae. The search for species complexes, recognisable by their karyotypes, has onlyjust adults and already it has revealed that these complexes are a common feature not only of the begun fauna of Africa but of other regions of the world as well. These groups are known as anopheline or 'cryptic' species. It is worth emphasising that the two disciplines of classical taxonomy 'sibling' evolutionary genetics are ideally complementary and mutually supporting. Both are essential. and the accuracies of identification provided by the classical method are less than was assumed Clearly be the case by earlier workers. to work was made possible by the award of a Wellcome fellowship to one of us (MTG) by the This Trust in London, to whom we express our sincere appreciation. We are most grateful to Wellcome C. D. Ramsdale for his valuable help in the compilation ofthe illustrated keys, the figures being Mr contributed by him. We are also indebted to Prof H. E. Paterson, Dr B. A. Harrison, Dr C. entirely Green, Mr R. H. Hunt and Dr G. B. White for helpful discussions and comments on the A. and to Dr P. S. Cranston for providing access to the collections at the British Museum manuscript History). (Natural acknowledge permission from Plexus Publishing Inc. to reproduce figs. 66 and 77a from We and Knight's Taxonomist's Glossary of Mosquito Anatomy and Blackwell Scientific Harbach to reproduce figs. 5 and 6 from Lambert and Coetzee 982), fig. 3 from Coetzee 983) Publications fig. 2 from Coetzee (i984). and we would like to express our feelings of warm respect towards Dr Botha de Meillon, still Lastly, in partial retirement, for his encouragement and good wishes in undertaking this task. only INTRODUCTION In the 8 years or so since the publication of the 'Anophelinae of Africa South of the Sahara', in understanding this group of medically important insects has been rather uneven. The progress wave of interest in the bionomics ofvectors, spurred on by the prospects ofmalaria eradication great Two areas of study, however, have been vigorously pursued. One, the analysis of the gambiae

5 though attempted on only a modest scale, the failure of house-spraying campaigns to break Even completely in many parts of Africa is now a well established fact. It appears that the transmission of arabiensis as a vector exacerbates the problem. A major factor in the lack of success of presence control methods in the northern savanna regions was established by the work of conventional (Molineaux and Gramiccia, 1980), who showed that a significant proportion ofpopulations Coluzzi both gambiae s.s. and arabiensis was resting outside. This behaviour was, at least partially, under of control and, as a consequence, contact of this section of the population with insecticide was genetic and its longevity normal. Hence, in this zone of Africa the interruption of transmission by slight was shown to be unattainable at the present time, even when the operation was house-spraying with exemplary thoroughness. conducted the days when malaria eradication was the orthodox view it was assumed that transmission In the main, endophilic vectors, principally the gambiae complex andfunestus, would ultimately be by by house-spraying. This raised the question of whether low level transmission might still controlled maintained by secondary or 'incidental' vectors, on the grounds that they were largely exophilic be would have little contact with insecticide-treated surfaces. A large number of dissections were and out and some dozen or so species, listed by Gillies and De Meillon (1968), were shown to be carried infected with sporozoites at a very low frequency. With the abandonment of eradication naturally question now appears academic and the importance of such minor vectors is for the most part this It is also now recognised that the detection of very low grade transmission is almost negligible. difficult by entomological means alone, and identification of the vector in these impossibly some members of the gambiae complex. Nevertheless, sporadic malaria infections have continued of occur and localised outbreaks have occasionally resulted (Smith et al. 1977, De Meillon et al. to Suspicion fell on a member of the funestus group (referred to by tle authors as aruni? and 1977). here as vaneedeni) and onflavicosta, both of which could be caught biting man outdoors in described numbers. However, these suspicions were not supported by catches in outdoor resting sites, some precipitin tests showed that most of them had fed on bovids. The problem, therefore, remains since unresolved. sum up, the overall vectorial situation in tropical Africa continues to be dominated by To Vectors listed by Gillies and De Meillon remains.unclear. In certain instances, such Incidental pharoensis and squamosus, the position is further complicated: by the demonstration that these as are in reality species complexes. 'species' CLASSIFICATION (1980) has pointed out that Reid and Knight (1961) and Reid (1968), in classifying the Harrison used the term 'section' for a higher category of taxon than Gillies and De Meillon subgenus who employed it for subdivisions of certain series in subgenus Cellia. This latter usage is (1968), here, since it seems to us that the possibilities of confusion are slight and that these retained are helpful for those primarily concerned with the Afrotropical fauna. subdivisions 121 taxa listed below for the Afi'otropical Region include 114 full species, of which 3 are The from cytogenetic studies to represent complexes of as yet unnamed sibling species, and 7 known Following the practice adopted in the.2nd edition we are not dealing with the fauna of subspecies. since this has been comprehensively covered by Grjebine (1966). Madagascar, CHANGES TAXONOMIC the publication of the 2nd edition of the Anophelinae South of the Sahara in 1968 one further Since of the fauna of the Afrotropical Region (formerly Ethiopian Zoogeographical Region) has listing MALARIA TRANSMISSION circumstances is rarely possible. place where this sort of situation exists is in the northern Transvaal, where house-spraying One many years with DDT has led to the apparent elimination offunestus s.s. and the extreme rarity over s.s., arabiensis andfunestus. In those areas where they occur in numbers, melas, merus, bwambae, gambiae and man-biting populations of nili are also important local vectors. The status of the dozen moucheti or so

6 published (White, 1980). This list, as part of the Catalogue of the Diptera of the Afrotropical been also covered the fauna of the Malagasy Region, which is excluded from this Supplement. Region and the elevation of wellcomei subsp, erepens to specific status. multicolor, omitted vaneedeni, the Catalogue by White, is retained on the list of Afrotropical species. from NOMENCLATURE MORPHOLOGICAL their exhaustive compilation, the Taxonomists' Glossary, Harbach and Knight (1980, 1982) In a unified system of nomenclature of the anatomical terms for morphological characters presented in mosquito taxonomy. This glossary had its origins in Belkin's (1962) system of used but has been extended to embrace the anatomy of the whole insect. Until chaetotaxonomy, except in the case of pupal chaetotaxy, Belkin's system has been relatively little used by recently, Meillon (1968) used the old, descriptive names for larval hairs in preference to the, at that time De numerical system. Since this present work is intended to be used in conjunction with the 1968 new, we are continuing to use the old terms in keys and descriptions of new species, while giving volume, new terms in parentheses in every case. By this means it is hoped to ensure, albeit a little the hand a simple, arbitrary system in which the veins are given serial numbers, on the other a one in which an attempt is made to establish homologies with the generalised insect plan. The system has been widely used in the past by mosquito taxonomists, not least because its simplicity former it suitable for the identification ofspecimens by field workers. In recent years, following Belkin made a more rigorous approach has been advocated with the intention ofbringingculicid terminology (1962), into line with that employed in other groups of Diptera. However, complete stability of has yet to be achieved, perhaps because of difficulty in getting agreement on all the nomenclature involved. In view of this, and in view ofthe fact that the present Supplement is intended homologies 1. WING (marshallii group), illustrating nomenclature of markings and veins; nomenclature employed by Harbach and Fig. (1980) is given in parentheses after each name. Pale markings: a humeral (prehumeral pale spot), a' pre-sector Knight spot), c' pale interruption of third main dark area of 1st vein, d subapical pale spot (preapical pale spot), apical spot pale pale spot),ffringe spots. Dark markings: A, B, C, D, the four main dark areas of costa and st vein (presector, median, (apical sector or stem ofsecond vein (radial sector), 2.1 and 2.2 upper and lower branches of Rs vein (R2 and R3). Other veins similarly numbered. second adopted here since the publication of the Catalogue, include the addition of 5 species, Changes lounibosi, letabensis, hughi and bwambae, the description of 2 new species, ethiopicus and namibiensis, those working on in Africa and not all workers are familiar with it. Moreover, Gillies and that the work will be accessible to those familiar with either of the two systems. clumsily, venation. Two approaches to the naming of veins have been evident over the years; on the Wing for use by field workers, we have decided to continue to use the numerical system, as shown in fig. 1. a a' b b' e d D e (humaral pale spot), b sector spot (presector pale spot), b' accessory sector spot (sector pale spot), sub-costal spot (subcostal preapical, apical dark spots). Veins:

7 Table I we show the different systems currently employed. In adopting the numerical system, the In entomologist will note that he is not required to assign different names to different sections applied Costa Subcosta Vein 2. 3 Vein 4 Vein 4.1 Vein 4.2 Vein Vein Costa Subcosta (proximally) Radius (distally) R Radial sector R2 R3 CuA M3+4 CuA ia (i 962) Belkin et al. (i 979) Tanaka (i98o), Faran & Harrison 198 I) Linthicum Costa Subcosta (proximally) Radius (distally) R sector (proximally) Radial + 3 (distal part of stem) R2 R3 5 R4+ Media MI+2 terminology for the taxonomically important parts of the male genitalia are Comparative in Table II, while the system ofnumbering for pupal setae is illustrated in fig. 2. Harbach and shown Leaflets Coxite Cul Cu2 & Knight I98O Harbach 198o Harrison seta Apical seta Intermediate to as ventromesad (referred or setae) seta Phallosome (apical portion bearing Aedeagus Gonocoxite setae Accessory of veins that have functioned, since at least the Oligocene, as single, structural supports. Adopted here Harbach & Knight (198o) Vein Vein 2.2 R4+5 Media M1 M3+4 Cubitus Vein IA Comparative terminology for venation of wings. TABLE I: 3 and a comparison of their system with that Knight's system for the larval setae is shown in fig. employed here is set out in Table III. MALE TERMINALIA (GENITALIA) Gillies & De Meillon I968 Harpago Club Claspette club Lateral bristle Apical accessory hair Outer Inner accessory hair Phallosome leaflets) Leaflets Parabasal spines Equivalent terms used for parts of male genitalia. TABLE II:

8 and 12 on the metanotum (Mtn) are designated with the suffix CT as 0( Fig. 2. System ofnumbering ofpupal setae. Setae, 10, are setae to 9 on the cephalothorax. (From Harbach and Knight, 1980).

9 3. System of numberingof larval setae. A antenna. C cranium. P,M,T pro-, meso-, metathorax (from Harbach and Fig. Iotorol dorsal Knight, 1980)

10 Antenna Prothorax clypeal Outer clypeal Posterior mid, outer frontal Inner, Sutural Vertical hair Shaft Inner, median, outer shoulder dorsal External Propleural palmate Thoracic metapleural Long Palmate 2 Hair 6 ('[ateral abdominal) Hair hair Saddle Inner, outer caudal & Knight I98o Harbach 1962) (Belkin seta -C 6-, 7-C 5-, 8-C II-P IO-, I2-P 9-, 8-M Io-M 9-, 3-T Io-T 9-, to -VII -I - VII to to 6-V 6-I I-X distribution records, that were not included in the maps provided by Gillies and De Meillon New are given in the text, where they are cited in terms of latitude and longitude. In case the (1968), should wish to add these new degree-square records to the old maps by inking in the appropriate reader squares, we provide a blank copy of the map on the same scale as in the original publication, Gillies & De Meillon i968 Preclypeal clypeal Inner Head 3-C 4-C 9-C i-, -, 3-P Dorsolateral mesopleural Long Mesothorax Metathorax Abdomen Xth segment -, 3-X Equivalent terms used in larval chaetotaxy. TABLE III: DISTRIBUTION on which readings are for every 6-line of latitude and longitude (fig. 4).

11 Map of Africa used tbr species distributions showing lines of latitude and longitude at 6-intervals. Fig. 12 3O

12 Species referred to in the text are indicated by page numbers. ANOPHELES MEIGEN SUBGENUS series Myzorhynchus coustani Laveran, 1900 (complex) p. 64 caliginosu De Meillon, 1943 p. 65 namibiensis Coetzee, 1984 p. 66 paludis Theobald, 1900 p. 66 symesi Edwards, 1928 tenebrosus D6nitz, 1902 p. 66 ziemanni Grfinberg, 1902 p. 67 obscurus (Grfinberg), 1905 P. 67 series concolor Edwards, 1938 p. 67 smithii Theobald, 1905 caroni Adam, 1961 faini Leleup, 1952 hamoni Adam, 1962 CELLIA THEOBALD SUBGENUS series Neomyzomyia jebudensis Froud, 1944 p. 68 lovettae Evans, 1934 vanhoofi Wanson & Lebied, 1945 wilsoni Evans, 1934 buxtoni Service, 1958 cinctus (Newstead & Carter), p. 68 machardyi Edwards, 1930 maliensis Bailly-Choumara & Adam, 1959 natalensis (Hill & Haydon), 1907 p. 70 nili (Theobald), 1904 p. 70 somalicus Rivola & Holstein, 1957 vernus Gillies & De Meillon, 1968 section Rhodesiensis rhodesiensis rhodesiensis Theobald, 1901 rhodesiensis rupicolus* Lewis, 1937 p. 71 cameroni De Meillon & Evans, 1935 p. 71 Dr D. J. Lewis informs * that, when he originally named this species, the adjectival form rupicolus was intended, not the us form rupicola. The emendation of this name by Knight & Stone (1977) to rupicola was therefore incorrect. The substantive LIST OF SPECIES OF AFROTROPICAL ANOPHELES series Christya implexus (Theobald), 1903 p. 68 Smithii section section Ardensis ardensis (Theobald), 1905 p. 68 deemingi Service, 1970 p. 68 dureni dureni Edwards, 1938 dureni millecampsi Lips, 1960 p. 69 kingi Christophers, 1923 p. 69 vinckei De Meillon, 1942 emended spellirg was followed by Ward (1984).

13 lounibosi new species p. 73 rodhaini Leleup & Lips, 1950 ruarinus Edwards, 1940 p. 75 series Myzomyla barberellus Evans, 1932 bervoetsi D'Haenens, 1961 brunnipes (Theobald), 1910 p. 7.5 culicifacies Giles, 1901 (complex) p. 75 domicolus Edwards, 1916 dthali Patton, 1905 p. 76 flavicosta Edwards, 1911 p. 76 fontinalis Gillies & De Meillon, 1968 moucheti nigeriensis Evans, 1931 schwetzi Evans, 1934 p. 78 tchekedii De Meillon & Leeson, 1940 walravensi Edwards, 1930 funestus sub-group funestus Giles, 1900 p. 79 aruni Sobti, 1968 p. 86 parensis Gillies, 1962 vaneedeni new species p. 81 brucei Service, 1960 confusus Evans & Leeson, 1935 p. 86 fluviatilis James, 1902 p. 86 fuscivenosus Leeson, 1930 leesoni Evans, 1931 p. 87 rivulorum Leeson, 1935 p. 87 section Marshallii-hancocki marshallii complex marshallii (Theobald), 1903 p. 87 letabensis Lambert & Coetzee, 1982 P. 90 hughi Lambert & Coetzee, 1982 p. 91 austenii (Theobald), 1905 p. 93 berghei Vincke & Leleup, 1949 brohieri Edwards, 1929 p. 93 gibbinsi Evans, 1935 p. 93 hancocki Edwards, 1929 p. 93 hargreavesi Evans, 1927 p. 93 harperi Evans, 1936 p. 93 mortiauxi Edwards, 1938 mousinhoi De Meillon & Pereira, 1940 njombiensis Peters, 1955 seydeli Edwards, 1929 p. 94 upemba Lips, 1960 p. 94 species Unassigned azaniae Bailly-Choumara, 1960 p. 75 erythraeus Corradetti, 1939 ethiopicus new species p. 76 longipalpis (Theobald), p. 78 moucheti moucheti Evans, 1925 p. 78 Funestus section This is Edwards' original spelling, taken by us to be correct see note under rhodesiensis rupicolus. 10

14 section Wellcomei wellcomei wellcomei Theobald, 1904 p. 94 wellcomei ugandae Evans, 1934 p. 95 wellcomei ungujae White, 1975 p. 95 erepens Gillies, 1958 p. 95 distinctus (Newstead & Carter), 1911 p. 95 theileri Edwards, 1912 p. 95 section Demeilloni demeilloni Evaris, 1933 p. 96 carteri Evans & De Meillon, 1933 freetownensis Evans, 1925 garnhami Edwards, 1930 p 96 keniensis Evans, 1931 lloreti Gil Collado, 1935 sergentii sergentii (Theobald), 1907 sergentii macmahoni Evans, 1936 p. 96 series Pyretophorus christyi (Newstead & Carter), 1911 p. 96 daudi Coluzzi, 1958 p. 96 gambiae complex gambiae Giles, 1902 p. 105 arabiensis Patton, 1905 p. 105 bwambae White, 1985 p. 117 melas Theobald, 1903 p. 118 merus D6nitz, p. 119 series Paramyzomyla cinereus Theobald, 1901 turkhudi Liston, 1901 p. 120 azevedoi Ribeiro, 1969 p. 120 listeri De Meillon, 1931 p. 122 dancalicus Corradetti, 1939 maculipalpis Giles, 1902 p. 122 pretoriensis (Theobald), 1903 p. 122 rujtpes rufipes (Gough), 1910 p. 123 rujtpes broussesi Edwards, 1929 salbaii Maffi & Coluzzi, 1958 p. 123 series Cellla argenteolobatus (Gough), 1910 p. 123 brumpti Hamon & Rickenbach, 1955 cristipalpis Service, 1977 p.124 cydippis De Meillon, 1931 p.124 murphyi Gillies & De Meillon, 1968 p. 124 quadriannulatus (Theobald), p. 105 multicolor Cambouliu, 1902 p. 122 Neocellla series pharoensis Theobald, 1901 (complex) p.124 squamosus Theobald, 1901 (complex) p.126 swahilicus Gillies, 1964

15 diagnostic points, so that the key can be used in the absence of the 1968 volume. The characters. segments with laterally projecting tufts of scales Abdominal I Section segments II-VII... on Abdominal segments not so Hind tarsus with at least last 2 segments entirely pale... Section II Hind tarsus not so Hind tarsus 5 mainly or entirely dark, tarsus 4 white... Section III Hind tarsusnot so... IV Section Legs speckled, sometimes sparsel Legs not speckled Wing entirely dark or with pale spots confined to costa and vein Section V Wing not so... KEY TO THE FEMALES present is presented Gillies in and De Meillon (1968) with the addition of a The key essentially as described species. The main difference is the provision of figures to newly of number small illustrate for the first half of each couplet are on the left and for the second half on the right. Two illustrations ethiopicus and upemba, are not included owing to a lack of information on certain key species,

16 Wing without 6. pale spot on basal half of costa Section \I a with at least pale spot Wing basal half of costa 7 on Palps with apex dark Section \ with apex pale Palps 8. Palps with 4 pale bands Section VIII Wing with pale interruption 9. 3rd main dark area of vein 1, sometimes on with preceding pale fused area 3rd main dark area with no pale interruption 10 Wing with 2 pale spots 10. vein 5.1 Section X on with pale spot Wing vein 5.1 Section XI on Wing with abundant pale areas, costa with at least 4 pale spots Palps with 3 pale bands... 9 Section IX SECTION I. with laterally projecting tufts of abdominal scales. Mosquitoes Wing almost entirely dark, costa without any pale spots brumpti 1. 13

17 Hind tarsus 1-5 entirely dark argenteolobatus 2. tarsus 1-4, at least, with apical pale bands 3 Hind Hind tarsus tarsus 3 and 4 all white or narrowly dark basally, 5 Hind dark or at least basal halfdark 5 all sized species; abdominal scale tufts short and dark; Moderate or more of hind tarsus pale cristipalpis half Hind tarsus 5 and about apical halfof4 pale pharoensis 6. tarsus 5 all dark and 4 with much less than apical half pale 7 Hind and 2 with definite light and dark rings in addition to apical pale band cinctus Hind tarsus and 2 with pale bands at apices only 4 -'r Hind tarsus not so.6 Very large species, abdominal segments 1-7 with long lateral 5. of yellowish and dark scales; hind tarsus largely dark implexus tufts 14

18 small species; upper branch of 2nd vein largely pale swahilicus Very to moderate species; upper branch 2nd vein either Small entirely dark apart from apex or with a. Mosquitoes with hind tarsal segments 4 and 5 entirely white; abdomen without projecting tufts of scales. laterally Legs speckled 2 1. Legs not speckled 8 Hind tarsus 3-5 entirely pale 3 2. tarsus 3 dark at base 5 Hinds Palps with 3 pale bands, usually with some speckling; 3. vein with 2 accessory sector pale spots 1st accessory pale spot 4 sector Palps with 4 pale bands, unspeckled; 1st vein with at most few scattered pale scales only squamosus cydippis SECTION II. maculipalpis 15

19 Mid tarsus 2-4 entirely dark; vein dark at base, basal 4. of stem of vein 4 with small pale areas half tarsus broadly pale at apex; 1st vein with 2 Hind sector pale spots accessory Hind tarsus narrowly pale or dark at apex; Fore tarsus with 5-9 pale rings; stem of 4th vein largely pale machardyi 6. tarsus with 2.4 pale rings; stem of 4th vein largely dark 7 Fore and mid tarsus 2 and 3 pale-ringed apically; Fore fringe spot present on wing 6th maliensis Mid tarsus 2-4 with pale apices; vein pale at base, basal half of stem of vein 4 entirely pale deemingi pretoriensis 1st vein with accessory sector spot 6 (in part) natalensis and mid tarsus 2 and 3 dark apically; 6th fringe spot absent buxtoni Fore 16

20 Palps very shaggy and unbanded 8. with 1-4 or narrow pale bands irregular... Palps smooth with 3 pale bands, the 2 outer ones broad or rarely fused 14 Palps without pale bands; 9. pale spots at apex of no hind tibia of base of tarsus or caliginosus... Palps with 1-4 pale bands; apex of hind tibia broadly or narrowly pale 10 Pale fringe spot present opposite vein 5.2; base of 11. tarsus dark hind paludis... tarsus broadly pale... coustani (in part) Hind tarsus entirely dark basally 12. with only a few or scales there pale... (in part) tenebrosus tarsus broadly pale Hind base at Hind tarsus 3 entirely pale tarsus 3 dark at base Hind No pale fringe spot opposite vein 5.2; base of hind TI 17

21 segment... apical pale band on tarsus 2 one-fifteenth to one-eighth of segment ziemanni namibiensis 3rd main dark area (preapical dark spot) on vein without a pale interruption; 14. tarsus 1-3 without distinct apical pale bands rujpes (in part) fore pale bands Hind 3 entirely pale hancocki tarsus part) (in brohieri Hind tarsus 3 not so... Mosquitoes with hind tarsus 5 mainly or entirely dark, tarsus 4 white; abdominal Femora and tibiae with at most apical bands only. 2 Hind tibia with a pale streak on apex, 3-5 times as long as broad: apical pale band 13. hind tarsus 2 one-eighth to two-fifths length of on coustani (in part) Pale streak on tibia 1-3 times as long as broad; with a pale interruption, or with a short extension of 3rd main dark area on vein subcostal pale spot into the dark area on vein 1; fore tarsus 1-3 with apical the brohieri (in part) W. Afi'ica theileri mainly E. and S. Ali'ica SECTION III. without projecting tufts of scales. segments laterally tibiae speckled kingi... Femoraand 1. 18

22 without laterally projecting tufts of scales. segments Palps with 3 pale bands 2 1. Palps with 4 pale bands 4 3rd main dark area (preapical dark spot) of 1st vein without a pale interruption; 2. terga fairly heavily clothed with cream or yellowish scales, abdominal on segments 6 and 7 3 especially main dark area of vein with a pale interruption, sometimes fused with preceding 3rd spot; scaling on abdomen very scanty, confined to pale or rarely 7th terga 8th Base of costa with 2 pale spots; stem of vein 2 entirely pale; 3. tarsus with some speckling salbaii fore shaggy,; costa and st vein without usual main dark Palps symesi areas smooth; 2nd main dark area (median dark spot) of vein well defined and Palps with 2 pale interruptions rujpes (in part) SECTION IV. Mosquitoes with speckled legs, hind tarsus 4 and 5 not entirely pale; abdominal gambiae complex (in part) Base of costa with pale spot; stem of vein 2 extensively dark; fore tarsus not speckled dancalicus 19

23 All tarsi completely dark; wing without pale fringe spots posterior to 4. vein 3rd 1-4 with conspicious pale bands on at least the apices; Tarsi with pale fringe spots up to vein 5.2 or 6 5 wing 3rd main dark area (preapical dark spot) of vein with a pale interruption, sometimes 5. with preceding pale area gambiae complex (in part) fused tarsus 2 with about apical two-fifths to half white and Hind rest dark the tarsus 2 either with less than apical two-fifths white or else Hind marked with dark and pale bands 7 prominently fringe spots present opposite vein 6; hind tarsus 2-4 with conspicuous Pale and pale rings in addition to apical pale bands 8 dark vernus (in part) 3rd main dark area with no pale interruption 6 natalensis (in part) No pale fringe spot opposite vein 6; hind tarsus 2-4 with apical pale rings and 7. dark except for to 2 pale spots ardensis otherwise 2O

24 Fore tarsus mainly pale with 8. dark markings vinckei narrow tarsus mainly dark with Fore pale rings 9 narrow Scales on 8th tergum dense and distributed 9. whole segment, sometimes with a over scales on lateral borders of 7th tergum dureni dureni few V. Mosquitoes with wings entirely dark SECTION with pale spots confined to costa and vein l; or not speckled, hind tarsus 4 and 5 not entirely pale; abdominal segments without laterally legs tufts of scales. projecting Wings entirely dark or uni-colorous 2 1. with at least some areas of paler scales Wings costa or vein 1, on being sometimes inconspicuous 5 these and legs entirely dark... 3 Palps with 2 well marked pale bands; hind femur and tibia Palps narrowly pale at apex... concolor Large species, wing length or more ruarinus mm species, wing length 3.5 Small or less 4 mm Scales on 8th tergum scanty and confined to posterior margin dureni millecampsi 21

25 Very pale brown species with glossy mesonotum; semi-arid 4. only regions aspect dark brown, mesonotum not General cave-dwelling so; Head scales narrow, rodlike, yellowish throughout; 6. regions only semi-arid Light and dark areas on wings poorly contrasted; semi-arid 7. only regions Pale areas on wing very narrow, subcostal pale spot present on 8. only; cave-dwelling costa rhodesiensis rhodesiensis lounibosi with 3 pale bands, dark at apex Palps unbanded or banding indistinct 10 Palps rhodesiensis rupicolus (in part) caroni (in part) Palps with 2-3 pale bands, pale at apex 6 5. with or without pale bands, dark at apex 9 Palps dthali (in part) Head scales broader, white on vertex, dark laterally; all regions 7 rhodesiensis rupicolus (in part) Contrast between light and dark areas on wing well marked 8 rodhaini Pale areas on wing broader, subcostal pale spot on costa and vein smithii (in part) 22

26 Pale brown species with poorly contrasting light and dark 10. on wing; areas regions semi-arid and contrast of dark and light Colour variable; areas dwelling... caroni (in part) cave hamoni vanhoofi Head scales narrow, rod-like yellowish throughout azaniae (in part) 11. scales broader, white Head vertex, dark on Mosquitoes without a pale spot on basal half of costa; pale spots not confined to and vein 1; legs not speckled; hind tarsus 4 and 5 not entirely pale; abdomen without costa tufts of scales. projecting shaggy to near tip Palps smooth except at extreme base... 3 Palps Palps entirely dark; hind tarsus 3 and 4 dark 2. narrowly pale at or apices obscurus (in part)... laterally rhodesiensis rupicolus (in part)... SECTION VI. with pale scales forming Palps or less definite pale bands; more tarsus 3 and 4 narrowly hind broadly pale at apices or tenebrosus (in part) Palps with apex dark, sometimes only narrowly 3. 4 so with apex pale... 5 Palps 23

27 Stem of 4th vein largely pale, vein 5.1 with 2 pale spots or largely pale, 4. spots present opposite veins fringe entirely absent Costa entirely dark except for a few indistinct pale scales subapically; 5. with a broad apical pale band and otherwise dark except for a palps either with 4 bands or else with 3 bands, the subapical of these Palps much narrower than apical band 10 being apart from costa generally very pale, basal half of stems of Wing 2 and 4 entirely pale 8 veins areas on wing greater or about equal to pale areas, Dark half of stems ofveins 2 and 4 largely dark 9 basal Scales on distal half of palps (apart from main pale bands) pale brownish, 8. on.basal half dark brown; distal half of proboscis prominently those pale scaled on palps (apart from main pale bands) uniformly dark brown; Scales dark-scaled or pale scaling inconspicuous proboscis tchekedii Stem of vein 4-largely dark, vein 5.1 with narrow pale area, pale fringe spots smithii (in part) narrow sub-basal pale band daudi Outer half of costa with 1-3 well marked pale areas; palps not so 6 Palps with 3 pale bands, the subapical pale band broad and about 6. to apical band... 7 equal wellcomei wellcomei wellcomei ugandae 24

28 fringe spots present opposite all veins up to 5.2, stem of vein 5 broadly Pale near base keniensis (in part) dark Stem of vein 5 and vein 5.2 dark except for a narrow pale spot distal 10. fork, 5.1 with a single pale spot to Hind tarsus l-z[ with distinct apical pale bands; mesonotum clothed with 11. narrow scales and with no patch of scales above wing root very tarsus 1-4 entirely dark or with a few pale scales at apices of 1-3; Hind scales broad, a patch of broad scales present above mesonotal wing root 12 Mosquitoes with palps dark at the apex or without distinct apical pale band; at pale spot on basal half of costa, pale scales not confined to costa and vein legs not speckled, least tarsus 4 and 5 not entirely pale; abdomen without laterally projecting tufts of scales. hind Palps entirely dark or without distinct pale bands 2 1. with 3 pale bands 5 Palps No pale fringe spots posterior to vein 3, stem of vein 5 pale except at fork 9. sometimes narrowly near base and erepens.fuscivenosus (in part) Vein 5 with extensive pale areas, 5.1 with 2 pale spots 11 distinctus Mesonotal scales yellowish or bronze medianally and white elsewhere schwetzi (in part) 12. scales white throughout walravensi Mesonotal schwetzi (in part) SECTION VII. 25

29 Small pale brown species, pale patches 2. wings indistinct, basal quarter to half of costa on dark; head scales narrow and yellowish azaniae (in part entirely with well-contrasting light and dark areas, basal quarter of costa with Wings least at area, even if narrow; head scales not pale 3 so Costa with humeral pale spot, 3. subapical (preapical) pale spot on costa no 1st vein and without a humeral pale spot, subapical pale spot present Costa costa on 1st vein and Wings with pale fringe spots opposite vein 3 up to vein 5.2.jebudensis Wings with no pale fringe spots behind vein 3 Wings generally pale, contrast between pale and dark areas, apart from costa and vein poorly 5. anterior mesonotal scales scanty, not forming conspicuous tuft turkhudi defined; with well-contrasting light and dark areas; conspicuous anterior Wings scale tuft present mesonotal 6... obscurus (in part)... faini 26

30 main dark area (median dark spot) on vein with2 pale interruptions 7 2nd main dark area on vein with at most pale interruption 8 2nd Pale bands on palps very narrow, occupying apices of 2nd, 3rd and 4th segments and not 7. joints; vein 5.1 with a single pale spot wilsoni (in part) overlapping bands on palps variable in width, distal 2 bands overlapping joints; Pale with 2 pale spots rujtpes broussesi 5.1 Wing field apart from costa and vein costa 9 of and dark areas on wing about equally distributed, humeral and presector pale spots Pale present on costa 11 Basal fifth of vein entirely pale fifth of vein either dark or with a sub-basal pale patch not extending to Basal Wings scantily scaled, all wing scales very narrow.fontinalis 10. heavily scaled, upstanding scales moderately broad lovettae Wings predominantly dark, no pale spots on basal quarter base smithii (in part) 27

31 Basal pale band of palps about equal to or slightly narrower than median band, 11. overlapping base of 3rd segment cinereus (in part) broadly pale band on palps either much narrower than median band, scarcely overlapping Basal of 3rd segment, or both basal and median pale bands very narrow 12 base VIII. Mosquitoes with smooth, 4-banded palps, pale at apex, at least pale spot SECTION on half of costa; pale scales not contined to costa and vein 1; legs not speckled; hind tarsus 4 and 5 basal 3rd main dark area (preapical dark spot) of vein with 1. pale interruption 2 a main dark area without a pale interruption 3 3rd Abdominal terga clothed with yellowish scales; hind tarsus 1-4 with very 2. apical pale bands broad hind tarsi entirely dark or with a few pale scales at 2nd main dark area (median dark spot) of vein with 2 pale interruptions; 3. and mid femora with a subapical pale spot fore 2nd main dark area ofvein with only pale interruption; legs not so 4 Pale bands on palps broad, basal pale band overlapping base of 4. segment 3rd Pale bands on palps mostly narrow, basal band not overlapping base of3rd segment 5 Extreme base of costa pale multicolor north-east Africa only 12. base of costa dark listeri Extreme azevedoi southern Africa only not entirely pale; abdomen without laterally projecting tufts of scales. christyi Abdomen without such scales; apices of segments 1-3 schwetzi (in part) III wilsoni (in part) cinereus (in part) 28

32 No pale fringe spots posterior to vein 3, femora and tibiae 5. speckled inconspicuously Stem of vein 5 and vein 5.2 pale at, and adjacent to, the fork garnhami (in part) 6. of vein 5 dark 7 Fork Wing 4mm or less, flattened mesonotal scales not extending onto 7. demeilloni (Berg River form) scutellum 4.4ram or more; some flattened scales present on scutellum as Wing as mesonotum well or else this area entirely absent; costa with at least pale spot on basal half, pale scales not vein to costa and vein 1; palps 3-banded, pale at apex; legs not speckled, hind tarsus 4 and 5 not confined entirely pale; abdomen without laterally projecting tufts of scales. 2nd and 3rd main dark areas (median and preapical dark spots) 1. from vein absent Hind tarsus 5 entirely white, tarsus 4 white except for a broad median 2. band dark Vein 5.1 with pale spot, sometimes with a vestigial 2nd pale spot with 2 well developed pale spots 5 Vein vernus (in part) Pale fringe spots present opposite all main veins up to 5.2 or 6, femora and tibiae not speckled 6 carteri (in part) SECTION IX. Mosquitoes with a pale interruption on 3rd main dark area (preapical dark spot) of wellcomei ungujae 2nd and 3rd main dark areas present on vein 2 seydeli Hind tarsus 5 entirely dark 3 29

33 Subapical pale band on palps very narrow, confined to apex of 3rd 5. 6 segment Scutal fossae and lateral areas of mesonotum above wing root with scattered 8. abundant broadish scales 9 or Subapical pale band on palps about equal to or slightly narrower 9. apical band than Pale fringe spot present opposite vein 6; fore tarsus 1-4 with conspicuous basal 4. apical pale bands and mortiauxi No pale fringe spot opposite vein 6; fore tarsus 1-4 narrowly pale apically only berghei pale band on palps broad, overlapping apex of 3rd and Subapical of 4th segment 7 base Base of costa with 2 pale interruptions brunnipes Basal quarter of costa entirely dark walravensi (.in part) tarsus either all dark or with apical pale bands on segments and 2 only 8 Hind tarsus 1-4 with well marked apical pale bands 10 Hind Fossae and lateral areas ofmesonotum without scales harperi njombiensis pale band on palps much narrower than apical Subapical band walravensi (in part) 3O

34 Apical pale bands on hind tarsus 1-4 very broad, equalling at least twice the apical width 10. the segments austenii of Vein 3 largely dark or broadly dark at either end; mesonotal scales 11. narrow and golden very Mesonotal scales as in A hargreav,esi 12. scales as in B mousinhoi Mesonotal marshallii letabensis X. Mosquitoes with vein 5.1 with 2 pale spots; SECTION pale interruption on 3rd main dark no of vein 1, costa with at least area palps 3-banded, pale at apex; legs not speckled, hind tarsus 4 and 5 not entirely pale;, abdomen 1, laterally projecting tufts of scales. without and subapical pale bands on palps broad so that intervening dark band Apical either slightly broader, about equal to or narrower than either is Hind tarsus bands narrower, about equalling or less than apical width ofsegments 11 gibbinsi (in part) Vein 3 narrowly dark at ends; mesonotal scales various 12 Mesonotal scales as in C hughi Mesonotal scales as in D gibbinsi (in part) d pale spot on basal half, pale scales not confined to costa and vein band 2 pale and subapical pale bands narrower, intervening dark band much Apical broader than pale bands 16 31

35 Apical pale bands on hind tarsus 4, and sometimes on tarsus 2 and 3 also, 2. onto bases of succeeding segments 3 extending 2nd main dark area (median dark spot) on vein with 2 pale interruptions; 3. of hind tarsus 4 and 5 sometimes broadly pale rufipes broussesi (in part) bases main dark area on vein with pale interruption; bases of hind tarsus 2nd and 5 at most narrowly pale 4 4 costa and vein much broader than subcostal pale spot domicolus on of costa with 2 pale interruptions; 3rd main dark Base equal to or narrower area than subcostal pale spot 5 of hind tarsus 3 and 4 dark or at most with a few Apices scales 7 pale of costa with or no pale interruption 8 Base of costa with 2 pale interruptions 9 Base Bases of hind tarsus 5 and ofother segments dark.6 Base of costa with pale interruption, 3rd main dark area (preapical dark spot) Pale fringe spot present opposite vein 6 lloreti No pale fringe spot opposite vein 6 barberellus of hind tarsus 1-3, and sometimes I-4, distinctly Apices 12 pale-banded 32

36 Mesonotal scales fairly broad, extending 9. whole scutum and over scutellum onto carteri (in part)... scales variable, but flattened scales confined Mesonotal at most anterior to thirds of scutum two Very small species, wing length 2.8mm 10. less brucei (in part) or or moderate species, wing length 2.9mm Small more 11 or Hind tarsus entirely dark; preaccessory dark spot 11. vein on absent usually.freetownensis... 2 narrowly but distinctly pale apically; pre-accessory and spot present dark demeilloni (in part)... 3rd main dark area (preapical dark spot) equal to 12. narrower or subapical pale spot than 3rd main dark area much broader than subapical pale spot 13 Mesonotal scales broadish and white, only slightly less dense 13. posterior on of scutum than anteriorly, and extending onto third scutellum... Vein 6 either with pale fringe spot 8. with pale scales at apex or vein of brucei (in part) Vein 6 without pale fringe spot and with no pale scales at itsapex riz,ulorum(in part) Hind tarsus.jlavicosta (in part).flavicosta (in part) scales Mesonotal posterior third of scutum scanty, narrow on yellowish-brown and

37 Moderate-sized species, wing more than 3.2mm keniensis (in part) 14. species, wing 3mm or less 15 Small Fore tarsus 4 dark or indistinctly p-ale at apex; pale fringe spot 15. vein 6 usually absent* opposite Large species; fork of vein 5 pale garnhami (in part) 16. or moderate-sized species; vein 5 dark at fork 17 Small Base of costa with or no pale interruption of costa with 2 pale interruptions demeilloni (in part) Base Small species, wing about mm rivulorum (in part) 18. or moderate-sized species, wing mm demeilloni (in part) Small pale spot on vein 5.1, no pale interruption on 3rd main dark area dark spot) of vein 1, costa with at least pale spot on basal half; palps with less than 4 (preapical pale at apex; legs not speckled, hind tarsus 4 and 5 not entirely pale; abdomen without bands, laterally projecting tufts of scales. moucheti Fore tarsus 4 with well marked apical pale band; 6th fringe spot present* ben,oetsi SECTION XI. Mosquitoes with Palps with apex pale and no other pale bands 2 1. with 3 pale bands 3 Palps *But see note under moucheti nigeriensis in Gillies and De Meillon, (1968). 34

38 Subapical pale spot on costa and vein about equal to or broader than 4th 2. dark area (apical dark spot), pale fringe spots present opposite main 3, 4.2, 5.1 and 5.2 nili Congo form veins pale spot narrower, usually much narrower, than 4th main dark Subapical Joints of hind tarsal segments narrowly or broadly enveloped in pale bands, 3. least tarsus 5 pale-ringed basally at nili somalicus Preaccessory dark spot on vein about twice as broad as pale spots on either 4. of it.fuscivenosus (in part) side Preaccessory dark spot absent or, Subapical pale band on palps broader than or slightly narrower than subapical 6. band anl 3rd main dark area (preapical dark spot) of costa and vein dark areas, no pale fringe spot present opposite vein 5.1 longipalpis Pale banding on hind tarsus narrow and apical only. 4 if present, narrower or only slightly broader than adjoining pale spots 5 Basal area of vein proximal to 1st main dark area (presector dark spot), pale with a broad dark spot culicifacies Basal area ofvein entirely pale 6 equal to or narrower than sub-costal pale spot aruni (in part) pale band on palps much narrower than subapical dark band, Subapical main dark area broader than subcostal pale spot 7 3rd 35

39 Moderate-sized species, wing more than 3.3 mm.demeilloni (in part) 7. species, wing 3.3 mm or less 8 Small Tip of vein 6 with a few pale scales, 6th.fringe spot sometimes present parensis (in part) 8. of vein 6 dark, no 6th fringe spot present.funestus group Tip (in part; mainly highlands) demeilloni (extreme southern Africa only) cameroni ethiopicus, berghei, mortiauxi, upemba, daudi, brumpti, and Cristipalpis. fuscivenosus, Inner clypeal hairs (setae 2-C) close together, separated from each other 1. a distance much less than between inner and outer by (2-C and 3-C) clypeals sergentii macmahoni clypeals separated from each other by a distance about equal to or Inner than between inner and outer clypeals 2 greater clypeals simple, frayed or lightly feathered, the fraying if present not mainly Inner to apical half 4 confined KEY TO THE FOURTH STAGE LARVAE The following species, known only as adults, are not included: caliginosus, vernus, deemingi, fontinalis, Section Outer clypeal hairs (setae 3-C) with 8 or more branches Section II Outer clypeals simple or with less than 8 branches 3 Inner clypeal hairs (setae 2-C) strongly branched in apical half Section III 36

40 Saddle hair simple or with 2-4 branches 5 Thorax and abdomen laterally and ventrally with 5. spicules, not numerous in belts arranged Sides of abdomen and thorax without spicules 6 Width of main abdominal plate 6. segment V equal to at least three-quarters on between bases ofpalmate hairs (setae 1-V) distance Both long mesopleural hairs (setae 9-M,10-M) simple, occasionally 7. split one 2-3 branches into 1, at least, ofthe longmesopleural hairs feathered orwithmore than3 branches 8 Saddle hair (seta l-x) with at least 5 branches Section IV Section V Section VI This plate not more than two-thirds distance between palmate hairs 7 Section VII 8. long metapleural hair (seta 9-T) simple, (seta 10-T) feathered Section VIII Both long metapleurals feathered Section IX 37

41 inner and outer clypeals (sub-genus ). between Leaflets of abdominal palmate hairs on segments III-VII (setae 1-III to 1-VII) 1. simple, without serrations Branching of outer clypeals (setae 3-C) extending 2. almost whole over of stem; posterior clypeals (setae 4-C) long, reaching length far as bases as clypeals simple or branched, branching not extending to basal Outer of stem; posterior clypeals very short, not reaching much more quarter shaft hair (seta l-a) with more or many more than 20 branches; Antennal clypeal hairs (setae 3-C) either short and simple or else long and with outer shaft hair with less than 20 branches; outer clypeals with Antennal or more branches coustani group 40 SECTION I. Larvae with inner clypeal hairs (setae 2-C) much closer together than distance implexus (in part) Palmate hairs on segments III-VII with well developed serrations 2 of inner clypeals (2-C) concolor than half-way to bases of inner clypeals 3 up to about 30 branches obscurus SECTION II. Larvae with outer clypeal hairs (setae 3-C) with 8 or more branches, inner clypeal (setae 2-C) widely separated. hairs hair (seta I-X) with more than 5 branches Saddle bervoetsi (in part) Saddle hair simple 2 38

42 Palmate hairs on abdominal segments I and II (setae 1-I, 1-II) undifferentiated; inner clypeal 2. (2-C) with delicate fraying argenteolobatus (in part) hairs Palmate hairs on segments II-VII (setae 1-II to 1-VII) with 20 or more leaflets; 3. hairs (setae l-p, 2-P) on basal tubercles that are partially shoulder or separated, inner hair -P) with branches swahilicus fused hairs averaging per specimen less than 20 leaflets; shoulder hairs on large fused Palmate tubercles, inner hair with branches of abdominal palmate hairs on segments III-VII Filaments 1-III to 1-VII) short and blunt (setae (in part) pharoensis squamosus murphyi Palmate hairs partly differentiated on segment I (1-I) and fully so on segment II (1-II); inner clypeals (2-C) with well marked branching 3 pharoensis (in part) Filaments of palmates longer and drawn out 39

43 hairs (setae 3-C) with less than 8 branches; inner clypeal hairs separated by a distance equal clypeal or greater than that between inner and outer hairs. to Both long mesopleural hairs (setae 9-M, 10-M) simple or occasionally split; 1. of abdominal palmate hairs (setae 1) long filaments drawn out 2 and long mesopleural feathered; palmate hairs with short blunt-tipped One Inner shoulder hairs (setae l-p) not especially flattened, mounted on 2. tubercles inconspicuous wilsoni lovettae of their stem 4 sides clypeals half or more length of inner clypeals; branches of inner clypeals Outer largely confined to inner aspect 5 SECTION III. Larvae with inner clypeal hairs (setae 2-C) strongly branched in apical half; outer filaments ru3qpes (in part) Inner shoulder hairs greatly flattened, mounted on well formed basal tubercles 3 Outer clypeal hairs (setae 3-C) very short, one quarter or less length of inner 3. (2-C); branches of inner clypeals arising from both clypeals 4O

44 Inner clypeal hairs (setae 2-C) with most secondary branches arising 4. apex near Inner clypeals with most secondary branches arising about half-way up the stem Outer clypeal hairs (setae 3-C) half to three-quarters length of inner 5. (setae 2-C), posterior clypeals (4-C) reaching not more than half-way to clypeals cinctus vinckei Shoulder hairs (setae l-p, 2-P) mounted on large basal tubercles which are widely 1. inner shoulder hair (1 -P) much flattened 2 separated; Outer clypeal hairs (setae 3-C) about half length of inner clypeals (2-C) kingi 2. clypeals much less than half length of inner clypeals 3 Outer dureni bases of inner clypeals (2-C) nili clypeals about as long as inner clypeals, posterior clypeals extending up to Outer beyond bases of inner clypeals or somalicus SECTION IV. Larvae with saddle hair (setae l-x) having at least 5 branches; outer clypeal hairs 3-C) with less than 8 branches; inner clypeal hairs (2-C) widely separated and not strongly (setae branched. hairs on basal tubercles which are either fused or narrowly separated, inner Shoulder hairs not especially flattened 6 shoulder 41

45 shoulder hairs (setae 2-C) with at least half of the branches flattened Inner leaf-like, the apices fimbriated 4 and and posterior clypeal hairs (setae 3-C, 4-C) less than one-fifth length of inner Outer (setae 2C) maliensis clypeals and posterior clypeals about one-third to two-fifths length of inner Outer elypeals Inner shoulder hairs (setae l-p) with one-third to half of the branches 5. some of them deeply forked bifid, The simple long metapleural hair (setae 10-T) flattened and lance-like, about two-thirds 6. of feathered hair (seta 9-T) brunnipes (in part) length Metapleural hairs not so 7 Inner shoulder hairs with no fimbriated branches although some may be bifid apically. 5 machardyi buxtoni Inner shoulder hairs with at most a few only ofthe branches bifid apically ardensis (in part) 42

46 2-3 accessory tergal plates present on most abdominal segments 8 7. accessory tergal plate present 11 Only Main tergal plate on 5th abdominal segment two-thirds or less distance between bases 8. palmate hairs (setae l-v) austenii (in part) of Posterior clypeals hairs (setae 4-C) less than quarter length of inner clypeals (setae 2-C); 10. metathoracic tergal plates present schwetzi (in part) no clypeals at least half length of inner clypeals; paired metathoracic Posterior present domicolus (in part) plates Outer clypeal hairs (setae 3-C) with 6 or more branches bervoetsi (in part) 11. clypeals simple or with a few fine branches 12 Outer Main tergal plate three-quarters or more distance between palmate hairs 9 Outerclypeal hairs (setae 3-C) simple brohieri (in part) Outer clypeals with 2-4 branches 10 43

47 Inner clypeal hairs (setae 2-C) short and abruptly 13. not more than quarter length of frontoclypeus; 1st pointed, only partially so abdominal palmate differentiated, 2nd palmate 1st on rest of abdomen 14 as Main tergal plate on abdominal segment V almost equal to distance 14. bases of palmate hairs (setae l-v) between (in part) brohieri (in part) seydeli Main tergal plate on abdominal segment V wider thandistance between bases 12. palmale hairs (setae l-v) and occupying one third to half of length of segment.flavicosta (in part) Main tergal plate not so 13 palmate (seta l-i) undifferentiated, 2nd palmate (1-II) abdominal gibbinsi Inner clypeals not so; Main tergal plate half to three-quarters distance between palmate hairs 15 Posterior clypeal hairs (setae 4-C) bifid or with well marked lateral 15. branches marshallii (in part) Poster clypeals simple 16 44

48 Posterior clypeal hairs (setae 4-C) not reaching beyond half distance to bases 16. inner clypeals (setae 2-C) seydeli (in part) of Abdominal palmate hairs (setae 1-I to 1-VII) with very short stumpy 17. about one-sixth length of the blade filaments, (in part) marshallii letabensis hair (l-x) with less than 5 branches; outer clypeal hairs (setae 3-C) simple; inner clypeal saddle (setae 2-C) widely separated and not strongly branched. hairs Main abdominal tergal plates evenly rounded laterally, on segment 1. equal to up to two-thirds distance between bases of palmate hairs V plates slightly turned up laterally, on segment V nearly equal Abdominal distance between bases of palmate hairs to distinctus wellcomei (in part) wellcomei ugandae wellcomei erepens theileri wellcomei (in part) wellcomei VI. Larvae with wide abdominal tergal plates, equalling on segment V three-quarters SECTION more the distance between bases ofpalmate hairs (setae 1-V); saddle hair (seta 1-X) with less than or branches; no coarse spicules on sides of thorax and abdomen. Outer clypeal hairs (setae 3-C with 5 than 8 branches; inner clypeal hairs (setae 2-C) wide apart and not strongly branched. less long mesopleural hair (seta9-m) feathered 2 1. longmesopleural hairs (9-M, 10-M) simple 4 Both Posterior clypeals reaching almost to bases of inner clypeals 17 hancocki (in part) Filaments of abdominal palmate hairs longer hughi (in part) hancocki SECTION V. Larvae with numerous spicules on sides and undersurface of thorax and abdomen; (seta l-v) 45

49 Inner shoulder hair (seta l-p) with about 5-12 branches, mounted on very small, 2. developed tubercle scarcely shoulder hair with about branches, mounted on a well developed Inner 3 tubercle Width of main abdominal tergal plate at most four-fifths distance 3. palmate hairs (setae l-v) between tergal plate four-fifths or greater than distance between palmate Main hairs sergentii macmahoni Depth of main abdominal tergal plate on segment V equal to about half 4. more depth of segment 5 or abdominal tergal plate on segment V more or less twice as wide as deep, Main with no completely detached accessory plates.funestus subgroup normally tergal plate 3 or more times as wide as deep, 1-3 accessory plates Main on nearly all segments 6 present Saddle hair simple, 2 accessory tergal plates 7 azaniae sergentii sergentii Main tergal plate much less than half depth of segment 8 Saddle hair (seta I-X) with 3-4 branches; accessory tergal plate present on most segments.flavicosta (in part) 46

50 hair (seta 8-C) branched; ventral surface of abdomen without Sutural spicules hair simple; ventral surface of abdomen with belts of very fine Sutural spicules With 3 accessory tergal plates on most segments tergal plate barberellus (in part) accessory Outer and posterior clypeal hairs (setae 3-C, 4-C) very short with 2-3 branches, 11. more than quarter length of inner clypeals (setae 2-C). not leesoni confusus Saddle hair (seta I-X) with 2-4 branches 9 8. hairsimple 14 Saddle Inner clypeal hairs (setae 2-C) abruptly tapered, distal third fine 9. filamentous and tchekedii Inner clypeals not so 10 schwetzi Outer and posterior clypeals at least half of inner clypeals 12 47

51 Outer clypeal hairs (setae 3-C) simple 12. bifid 13 or elypeals with 3-4 branches domicolus (in part) Outer Paired metathoracic plates present; 1st abdominal segment usually with 13. accessory plates domicolus (in part) 2 Inner shoulder hair (seta l-p) greatly flattened, widely separated from 14. hair (seta 2-P); inner clypeal hairs (setae 2-C) median shoulder hair not so, bases of shoulder hairs close together; inner Inner simple 15 clypeals Main abdominal tergal plate 2--3 times as wide as deep; sutural hair C) simple or bifid (seta tergal plate 3--4 times as wide as deep; sutural hair branched Main base from metathoracic plates; no accessory plates on 1st abdominal No segment longipalpis frayed or with fine branches natalensis mousinhoi rivulorum 48

52 inner clypeal hairs (setae 2-C) widely separated, not strongly branched in apical half; outer spicules; hairs (setae 3-C) with 8 or fewer branches. clypeal Inner shoulder hairs (setae l-p) poorly developed, basal tubercles very small or 1. 2 non-existent shoulder hairs well developed, arising from well formed basal Inner 6 tubercles Basal spine of pleural hairs (setae 9-12) large, curved 2. sharply pointed and Inner shoulder hairs (setae l-p) minute, with 1-3 branches.jebudensis 3. shoulder hairs almost half size of median hair (2-P), with about 12 Inner Inner frontal hairs (setae 5-C) small, not reaching more than three-quarters 4. to base of posterior clypeal hairs (setae 4-C) 5 distance frontal hairs normally developed, reaching approximately to base of Inner clypeals posterior smithii caroni SECTION VII. Larvae with both long mesopleural hairs (setae 9-M and 10-M) simple; main plate on abdominal segment V not more than two-thirds distance between bases of palmate tergal (setae l-v); saddle hair (l-x) with less than 5 branches; thorax and abdomen without belts of hairs gambiae complex Basal spine ofpleural hairs not so 3 or more branches 4 faini 49

53 Inner frontal hairs (setae 5-C) reaching half to two-thirds distance to bases of 5. clypeals (setae 4-C), inner clypeals (setae 2-C) posterior frontals reaching two-thirds to three-quarters distance to bases of posterior Inner inner clypeals simple clypeals, Shoulder hairs (setae l-p, 2-P) arising from large, widely spaced tubercles, 6. hair (l-p) greatly flattened ardensis (in part) inner Saddle hair (seta I-X) with 3-4 branches 8 7. hair simple or bifid 13 Saddle minutely frayed hamoni vanhoofi Shoulder hairs not so 7 accessory tergal plates present on most abdominal segments 9 3 tergal plate only 10 accessory 5O

54 1st abdominal palmate hair (seta 1-I) undifferentiated, fully differentiated hairs 9. short spiky filaments austenii (in part) with abdominal palmate well differentiated, fully developed hairs with 1st filaments long The simple long metapleural hair (seta 10-T) flattened and lancelike, 10. about two-thirds length of feathered hair (seta 9-T) Metapleural hair not so 11 Bases of shoulder hairs (setae l-p, 2-P) fused barberellus (in part) 11. of shoulder hairs separated 12 Bases 1st abdominal palmate hair (seta 1-I) a branched hair, 2nd 12. (1-II) narrow and lanceolate palmate abdominal palmate partially developed, 2nd palmate fully 1st differentiated 3 accessory abdominal plates present on some segments at least plate only 16 accessory longipalpis (in part) brunnipes (in part) njornbiensis walravensi 51

55 Inner caudal hairs (setae 2-X)with straight branches culicifacies (in part) 14. caudal hairs with some branches, at least, hooked 15 Inner Shoulder hairs (setae l-p, 2-P) mounted on large fused tubercles; 15. with paired dorsal plaques thorax hairs mounted on small, well separated tubercles; no Shoulder plaques thoracic Basal spine ofmesopleural hairs (setae 9,M-12,M) large, curved and 16. pointed 17 sharply Outer clypeal hairs (setae 3-C) with 4-7 branches argenteolobatus (in part) 17. murphyi erythraeus dthali Basal spine of pleural hairs not so 18 Outer clypeals simple or with 2-3 branches (in part) cydippis (in part) argenteolobatus 52

56 Outer clypeal hairs (setae 3-C) with 3 or more short stiff 18. branches Outer clypeal hairs (setae 3-C) stiff, spine-like, not more than twofifths 19. length of inner clypeal hairs (setae 2-C) hargreavesi harperi VIII. Larvae with long metapleural hair (seta 9-T) simple, long mesopleural hair SECTION 10-M) feathered; tergal plates on abdominal segment V not more than two-thirds distance (seta bases of palmate hairs (setae l-v); saddle hair (l-x) with less than 5 branches; thorax and between without belts of spicules; inner clypeal hairs (setae 2-C) widely separated, not strongly abdomen Inner and median frontal hairs (setae 5-C, 6-C) simple or with up to 6 branches 2 1. and mid frontal hairs with 7-14 branches 3 Inner tergal plate; a group of spicules on basal third of antenna accessory longer than the rest markedly 3 accessory tergal plates; antenna curved, spicules uniform in With size moucheti moucheti Outer clypeals simple, occasionally bifid 19 moucheti nigeriensis Outer clypeals slender, about halflength ofinner clypeal hairs branched in apical half; outer clypeal hairs (setae 3-C) with 8 or fewer branches. lounibosi ruarinus 53

57 Antenna with a group of spicules on the basal third of inner border 3. longer than the rest markedly rhodesiensis rodhaini 3 accessory abdominal plates present on segments 1V-VII culicifacies (in part) 4. plate present, rarely 3 plates on segments VI-VII 5 accessory Inner shoulder hair (seta l-p) with 5-12 branches, mounted 5. a very on scarcely developed tubercle small, shoulder hair with about branches, mounted Inner well on tubercle 6 developed Antennae without such spicules. 4 azaniae (in part) Saddle hair (seta I-X) split into 2-3 branches demeilloni (in part) Saddle hair simple 7 Hair 6 on abdominal segments IV to VI (setae 6-IV tovi) with 10 or 7. terminal branches more.freetownensis Hair 6 on segments IV to VI with not more than 7 branches 8 54

58 Thoracic palmate hair either an ordinary branched hair or with 8. narrow leaflets 10 extremely Thoracic palmate with narrowlanceolate leaflets 9 Long mesopleural hair (seta 9-M) with 3-5 branches 9. cameroni (extreme southern Africa only) only Long mesopleural with 5-15 branches Posterior clypeal hairs (setae 4-C) about three-quarters length of 10. clypeals (setae 2-C) inner (in part) demeilloni (in part) carteri keniensis lloreti mesopleural hair (seta 9-M) feathered; tergal plates on abdominal segment V not more than long distance between bases of palmate hairs (setae l-v); saddle hair (l-x) with less than 5 two-thirds Fully developed abdominal palmate hair (seta 1) tassel-like, [. 2 undifferentiated Fully developed palmate hair fan-shaped, differentiated 4 carteri (in part) Posterior clypealsabout halflength of inner garnhami.ection IX. Larvae with both long metapleural hairs (setae 9-T, O-T) feathered and at least branches; thorax and abdomen without belts o" spicules; inner clypeal hairs (setae 2-C) widely not strongly branched in apical half; outer clypeal hairs (setae 3-C) with 8 or fewer separated, branches. 55

59 Shoulder hairs (setae l-p, 2-P) with very small bases, inner hair (l-p) 2. developed, simple or with 3-4 branches poorly hairs on well formed basal tubercles, inner hair with Shoulder branches Shoulder hairs (setae l-p, 2-P) mounted on large fused tubercles, 3. hair with 25 or more branches; pecten plate markedly inner hairs mounted on separated tubercles, inner hair small Shoulder 5-10 branches; pecten plate without spicules with Both long mesopleural hairs (setae 9-M, 10-M) feathered 5 4. long mesopleural hair simple or occasionally bifid 6 One Palmate hairs (setae 1-IV to VI) fully developed on segments IV 5. VI only, filaments short and stumpy to azevedoi spiculate laterally rnaculipalpis dancalicus turkhudi developed palmate hairs present on segments III to VII, Fully drawn out and finely pointed filaments cinereus 56

60 Inner shoulder hair (seta l-p) poorly developed, not more than half 6. of median hair and with less than 10 branches 7 length shoulder hair well developed, at least three-quarters length Inner hair and with more, usually many more, than9 branches 8 ofmedian Filaments of fully developed palmate hairs (setae 1) short and rather blunt-tipped; 7. Africa southern of typical palmate hair longer and finely pointed Filaments North Africa apically; Inner clypeal hairs (setae 2-C) minutely or conspicuously frayed 9 8. clypeals simple 11 Inner Outer clypeal hairs (setae 3-C) with 4-7 branches argenteolobatus (in part) 10. clypeals simple or with 2-3 branches cydippis (in part) Outer listeri multicolor spines of pleural hairs large, curved, sharply pointed 10 Basal spines of pleural hairs small and obtusely pointed rufipes (in part) Basal argenteolobatus (in part) 57

61 Shoulder hairs (setae l-p, 2-P) well separated, mounted on small 11. 2nd abdominal palmate hair (seta 1-II) undifferentiated, tubercles; 3rd palmate (1-III) partly so of shoulder hairs closely approximated or fused; 2nd abdominal Bases differentiated even if small, 3rd fully developed 12 palmate Inner shoulder hairs (setae l-p) placed on unusually large fused 12. basal spines of meso- and metapleural groups (9-, 10-M, 9-, 10-T) tubercles; shoulder hairs on medium-sized tubercles; basal spines of pleural Inner not so groups salbaii large and deeply pigmented christyi pretoriensis 58

62 comprehensive setal counts have been published for a few species, notably in the coustani Although marshallii species groups, our knowledge of the chaetotaxy of pupae generally has not advanced and not felt that an adequately illustrated key could be constructed at the present time. was pupae of the following species are still unknown at the present time: caliginosus, deemingi, The lounibosi, rodhaini, bervoetsi, eythraeus, ethiopicus,fontinalis,fuscivenosus, berghei, mortiauxi, upemba, vernus, brumpti and cristipalpis. daudi, A median keel present between the respiratory trumpets implexus l. keel absent 2 Median (in part) tenebrosus with 5 or more ribs 5 Trumpet coustani (in part) tenebrosus Seta 9 very short and blunt, 9-VII about one eighth length of 6. VIII (fig. 5, 1) 7 segment Seta 1-VII simple or bifid paludis (in part) 7. symesi Paddle fringe with hairs only or with slender spines which 8. and imperceptibly change to hairs (fig. 5, h) 9 gradually fringe with spines only or with spines and a more or less Paddle transition between spines and hairs (fig. 5, i) 40 abrupt Paddle fringe continued beyond the apical hair (l-p) fringe terminating at or before apical hair 24 Paddle border of the paddle with hairs to near base 11 Internal border bare or, if hairs present, they only reach about Internal half-way to the base 15 l-v, 1-VII a black tuft of more than 10 branches 13 Setae 1-VI, 1-VII simple or with 2-5 pale branches, not as a tuft; Setae KEY TO PUPAE much in recent years. The key presented here is therefore essentially the same 'preliminary' key very that compiled by Gillies and De Meillon (i 968), with few additions only. For the same reason it as Cephalothorax with rows of conspicuous hooks wellcomei section without such hooks 3 Cephalothorax trumpets with ribs (fig. 5,0) 4 Respiratory without ribs (fig. 5,p) 6 Trumpets Trumpet with 2-5 ribs paludis (in part) 4. Seta 5-VII a black tuft of about 12 branches tenebrosus (in part) Seta 5-VII with at most 6 branches ziemanni namibiensis Seta 9 longer and pointed 8 Seta 1-VII a dense pigmented tuft obscurus 9-V often strongly curved (fig. 5, k) 12 Seta 5-III with 8-9 branches somalicus Seta 5-III with branches nili (in part) 59

63 cinctus vinckei 9-V half or more length of following segment excluding Seta membrance 16 intersegmental 1-VI, 1-VII simple or trifid 20 Setae branched, pectinate or frayed 18 Setae 9-V curved at tip (fig. 5, m) ardiensis Seta 9-V to 9-VII straight 19 Setae 1-V simple to trifid; fringe with spines even if minute 21 Seta 1-V with 3-5 branches; fringe with hairs only natalensis (in part) Seta paddle hair (l-p) short, about one seventh length of Apical practically straight paddle, (in part) hancocki seydeli (in part) hancocki (in part) longipalpis pointed apical hair placed at the apex Paddle 5, d) ruarinus (fig. fringe consisting of hairs only 26 Paddle fringe consisting of spines changing to hairs 27 Paddle I-V, 1-VI about 1.25 to 1.5 times length or more of the Setae on which they are placed (fig. 5, c) 28 segment Setae 9-IV to 9-VII black nili (in part) Setae 9-IV to 9-VII pale in colour 14 Setae 1-IV, 1-V about the same as 5-IV, 5-V dureni These paler and less branched than 5-IV, 5-V 15. This seta less than half length of following segment 17 Seta 2-P simple longipalpis (in part) Seta 2-P branched leesoni Paddle fringe composed of hairs only natalensis (in part) Paddle fringe of hairs and spines, even if delicate moucheti (in part) 20. gibbinsi paddle hair longer, hooked or sinuous 22 Apical 2-P simple or trifid (fig. 5, c) 23 Seta Seta 2-P with 4-8 branches hughi letabensis lloreti (in part) Seta 1-III with 4-5 branches confusus Seta 1-III with 6-13 branches subgroup funestus brohieri 24. Paddle more or less rounded apically 25 Setae 1-V to 1-VII simple to trifid cameroni Setae 1-V to 1-VII with 5-9 branches wilsoni 27. These setae shorter, about as long as the segment or less 30 6O

64 9-IV less than half as long as 9-V, blunt 29 Setae 9-IV more than half as long as 9-V brucei Seta macmahoni sergentii hair of paddle (l-p) long, usually more than one third Apical rivulorum dthali of paddle 35 length hair shorter, at most one third length of paddle 31 Apical 1-V simple or bifid 34 Seta 1-V branched 32 Seta 1-VII simple or bifid 33 Seta 1-VII with about 4-7 branches smithii section Seta 1-VII longer than segment VIII; apical paddle hair one Seta to one fifth length of paddle, straight sixth 1-VII shorter than segment VIII; apical paddle hair more Seta one fifth length of paddle, hooked than Seta 2-P simple to trifid at the apex 9-V to 9-VII hair-like, bifid, pectinate or showing Setae branching some jebudensis and (excluding wilsoni) 9-V simple, spine-like as usual 36 Setae 1-V to 1-VI shorter than the following segment, 1-VI with 2-6 Setae (in part) domicolus (in part) njombiensis walravensi 1-VI simple, as long as following segment 37 Seta 5-V, 5-VI with 4 or more branches; subapical paddle seta Setae variable (2-P) 5-V, 5-VI simple or up to 4-branched, subapical paddle seta Setae to trifid simple austenii barberellus (in part) domicolus demeilloni freetownensis keniensis Seta 9-IV pale, only about one sixth as long as 9-V azaniae Seta 9-IV dark, one third to two fifths as long as 9-V (in part) pretoriensis rufipes harperi garnhami 34. maculipalpis salbaii marshallii Seta (2-P) branched 35. dancalicus branches Seta 1-V branched, shorter than following segment schwetzi Seta 1-V simple to trifid, as long as segment Seta 9-IV reduced, half or less length of 9-V (in part) pretoriensis (in part) rhodesiensis Seta 9-IV longer, half or more length of 9-V culicifacies rhodesiensis (in part) 61

65 fringe of spines only Paddle fringe of hairs and spines Paddle fringe of strong, blunt spines Paddle 42 Paddle fringe of fine, pointed spines... azevedoi... fringe with hairs continued beyond apical hair 44 Paddle fringe not continued beyond apical hair 51 Paddle 1-VI, 1-VII in form of a tuft 45 Setae 1-VI, 1-VII not so Setae VI, 5-VII coarse and black; spines Setae outer edge of paddle on and blunt coarse... 1-VI, 1-VII and 5-VI, 5-VII in the form of frayed Setae pectinate or hairs These setae not so 48 buxtoni machardyi paddle hair (l-p) short, less than quarter length of paddle, Apical strongly hooked. not segment hargreavesi (in part)... pharoensis Setae 1-VI, 1-VII only two-thirds length of segment Meatus more than half length of trumpet much less than half length of trumpet argenteolobatus (in part) Meatus flavicosta Seta 1-IV branched fringe with small spines succeeded by hairs with hooked tips Paddle 5, i)... listeri (fig. (in part) cinereus multicolor with distal border straight (fig. 5, n) brunnipes Paddle with distal border more rounded Paddle l-v, 1-VI more or less half Setae long as following segment 55 as 1-V and 1-VI, or at least 1-VI, about Setae long or longer as than segment Seta 1-VII simple, 5-VII with 3-4 branches turkhudi Seta 1-VII with 6-8 branches, 5-VII with many branches concolor 45. nili Congo form moucheti (in part) These setae delicate, pale; spines on paddle finely pointed 46. Seta 5-VII with short stiff branches Seta 5-VII with fine lateral frayings kingi maliensis 48. paddle hair more than quarter length of paddle, hooked 50 Apical 1-VI, 1-VII three-quarters to about equal to length of Setae 49. squamosus swahilicus gambiae complex (in part) (in part) lloreti Seta 1-IV simple or bifid Paddle fringe with larger spines succeeded by straight hairs 62

66 1-VII branched 56 Seta 1-VII simple mousinhoi Seta Apical paddle hair long and curved 55. jebudensis Apical paddle hair (l-p) short and straight njombiensis (in part) cinereus (in part) Setae 1-V to 1-VII longer than following segment Setae 1-V to 1-VII not normally longer than following segment 58 Setae 1-VI, 1-VII bifid, occasionally trifid hargreavesi (in part) Setae 1-VI, 1-VII simple, occasionally split on one side 59 less than half total length of 59. Meatus of trumpet short, (in part) argenteolobatus tchekedii trumpet christyi complex (in part) gambiae Meatus longer, more or less half length of trumpet 63

67 setal counts for the pupae and larvae of South African populations of coustani, tenebrosus Complete ziemanni have been published by Coetzee and Newberry (1980). Computer analyses showed and pupae and larvae of the 3 species could be separated with a high degree of accuracy. If that was limited to 2 characters, about 90% of both stages could be identified correctly. In examination the antennal shaft hair (seta 1-A) in the larvae of tenebrosus was found to be more heavily particular, than in the other species, i.e. tenebrosus branches, coustani 4-9, ziemanni It was branched found that the number of ribs on the pupal trumpets was a useful character. Mean counts for also 3 species were, coustani 12.2, tenebrosus 7.0, ziemanni The authors concluded that these results the the fact that the 3 taxa came from independent gene pools and thus reinforced the decision reflected Gillies and De Meillon (1968) to elevate these forms to specific status. by abdomen. In cleared larvae or pelts ofthe coustani group patches of spicules can be seen on the larval border of segments II to VI or VII. In most species they are very few in number on medio-anterior I and easily overlooked, but in paludis they number forming a well defined patch segment in namibiensis approximately 50. and and Newberry found that the differences in the eggs described by earlier workers (see Coetzee and De Meillon) were not applicable to South African populations. Thus, the exochorion in Gillies has the same polygonal markings as in coustani and tenebrosus, while the characteristic closed ziemanni in coustani does not appear to be constant, even within the same egg batch. deck complex coustani tenebrosus namibiensis paludis* shaft Antennal A) hair I8 lo-8 5-II I4 of spicules No. ventral on of surface Abdominal +/-5 I on pupal IO--I 4 8- SUPPLEMENTARY ACCOUNT OF SPECIES Subgenus ANOPHELES Meigen MYZORHYNCHUS series ANOPHELES COUSTANI SPECIES GROUP Another useful character is the distribution of belts of fine spicules on the ventral surface of the These differences are summarised in Table IV. No. of ribs No. of branches trumpet segment O symesi ziemanni O--IO o, or 3-5 *based on small numbers only TABLE IV: Summary of larval and pupal differences within the coustani group. 64

68 coustan/laveran complex studies by Coetzee (1983) on South African populations of coustani Cytogenetic s. have revealed s. existence of 2 cryptic species, separable by the presence of the fixed inversion on the X a F1 hybrids of the 2 species had undeveloped testes in the males and asynapses in the chromosome. of the larvae. Adults and larvae of the 2 species, designated sp. A and sp. B, could not chromosomes separated reliably by morphological characters. In the pupae, however, the branching ofcertain be was found to be useful. In sp. A the sum of the branches ofsetae 6-I, 9oi and 6-II was 12 or less, setae in sp. B it was 13 or more. while many populations. In Senegambia this morph appears to be the only form of the species present in received from Dr M Cornet and from our own collections). (material BIOLOGY: In the highlands ofkenya G. Van Someren (1976) recorded males swarming ADULT dusk at levels of up to 5.5 m above the ground, sometimes out of sight. From a pool of females at in the same area, Rift Valley fever virus has been isolated (Anon. 1983), collected finding also a from southern Africa (McIntosh 1975). Unfortunately it is not known which species reported was in the above reports. involved sp. A. South Africa, Transvaal, from DISTRIBUTION: number of localities in the lowveld. This a Sp. B. South Africa, Transvaal, from 2 localities near Johannesburg in the highveld. coustani group. It is possible that some at least of these records refer to sp. B. species coustani s.l., Senegal; 13N, 16W (Cornet and Chateau, 1974). Gabon; 0S, 10E (Service, Angola; numerous localities (Ribeiro and Ramos, 1975). 1976). cahginosus De Meillon Specimens ofcoustani s.l. with hind tarsus 3 entirely white IDENTIFICATION: sporadically occur the possibility that it may be found to occur widely in tropical Africa. suggests s.l. has been recorded from higher altitudes and from further south than any other member of the NEW RECORDS: Angola; recent records by Ribeiro and Ramos (1975) extend the known distribution considerably (fig. 6). Fig. 6. Distribution of caliginosus. 65

69 Closely resembles ziemanni, from which it differs in the following characters: IDENTIFICATION: Wing, apical fringe spot reduced as in coustani, rarely extending beyond vein 2.2 (fig. 7). 1. Pupal seta 5-VI with 11ol 8 branches, compared with 5-12 in ziemanni. The branching of certain 2. setae tends to be greater than in ziemanni, but the overlap between the 2 species is too great for larval to be used for identification of individual specimens. However, the number of spicules this the on surface of segment I serves to separate the larvae (Table IV). The egg is ventral in ziemanni. as (1984) gives full setal counts for the larvae and pupae. Coetzee of chromosomes showed that namibiensis and ziemanni are apparently Examination rain forest in Cameroun. Biting was almost exclusively at ground level with in peak in the first part a the night from 18-22h00. A similar early peak was found by Cordellier and Geoffroy (1976) in of In the Congo Republic, Carnevale and Le Pont (1973) caught moderate numbers offemales night. houses at night by the use of a light-trap. inside RECORDS: Gabon; 0S, 10E (Service, 1976). Central African Republic; 4N, 18E NEW 1969). Republic of the Congo; 3S, 14E (Carnevale, 1974b). Rwanda; ls, 29E (Rickenbach, 1967). Ethiopia; 6N, 36E (O'Connor, 1967). (Vermylen, tenebrosus D6nitz under coustani species group. See BIOLOGY: In Tanzania, Mosha and Magayuka (1979) found one specimen out of 625 ADULT RECORDS: Gabon; this species was reported by Galliard (1932), but has not since been NEW from West Africa. It is interesting, therefore, that Service (1976) found it in light-trap recorded again in Gabon (0S, 10E). A specimen we examined appeared typical, but the possibility catches, sibling species should be borne in mind. Ethiopia; 6N, 37E (Rodhain, Boutonnier et al., 1977). of and Burundi; ls, 30E; 3S, 29E (Vermylen, 1967). Angola; 17S, 19E; 12S, 14E; Rwanda 17E (Ribeiro and Ramos, 1975). 10S, narnlbiens s Coetzee namibiensis Coetzee, Syst. Entomol. 9: TYPE LOCALITY: Kavango, Namibia. Fig. 7. narnibiensis. Wing. but that in the former there were floating inversions constantly present that homosequential were in ziemanni. absent LARVAL HABITAT: Not known. ADULT BIOLOGY: Adults were collected in a cattle enclosure. DISTRIBUTION: Only known from Kavango, Namibia. paludis Theobald See under coustani species group. IDENTIPICATION: ADULT BIOLOGY: Rickenbach et al. (1973) studied its biting habits at various levels up to 30m forest in the Central African Republic, with a secondary peak towards dawn. In Uganda, gallery and Ssenkubuge (1974) found biting activity to remain more or less Haddow throughout the even IDENTIFICATION: dissected to have infective larvae of DiroJTlaria imrnitis. 66

70 ziemanni Grfinberg See under coustani species group. IDENTIFICATION: by Gillies (1984) in Gambia showed that the tendency for ziemanni to avoid houses studies linked was the reluctance to penetrate through small apertures to reach the host. Thus, with than 5 times more many ziemanni entered through a 20 cm-square aperture as through a 20 x 1.8 cm slit, whereas as of the endophilic species gambiae s.1. and funestus passed through both entryways in equal females When the host was outdoors, studies showed that host-seeking females detected and numbers. towards a bait of2 calves from a distance of oriented (Gillies and Wilkes, 1972). The same m (1981) found that the tligh speed of females in authors wind-tunnel in the field was in the range of a m/s. In Angola, Ribeiro et al. (1964) and Ribeiro and Ramos (1975) reported total of 846 a negative for sporozoites. In Kenya, the human blood index in females caught in lighttraps dissections was 0.09, most of the feeds being on bovids. Biting activity in this area was mainly in the early BIOLOGY: Chandler and Highton (1975) found that ziemanni larvae appeared in rice LARVAL in Kenya when the rice reached a height of 75 cm and when filamentous algae and Salvinia fields RECORDS: Senegal; 14N, 16W; 14N, 17W (Cornet and Chateau, 1974). Central NEW Republic; 4N, 18E (Rickenbach, 1969). Rwanda and Burundi; ls, 30E; 2S, 30E African 18S, 20E; (Ribeiro and Ramos, 1975). The posr;bility exists that 19E; of these records could some to namibiensis. Dr H. Ribeiro (in litt.) states that in 40% of specimens the apical fringe spot refer was not extending beyond vein 2.2, as in that species. Such specimens reduced, commoner towards were south of Angola, but Dr Ribeiro notes that, in 2 ofthem, pupal seta 5-VI had only 3-4 branches. the Africa; Abundant in the Transvaal, 23S, 31 E; 24S, 31 E, but South in Natal, 27S, scarce (Coetzee and Newberry, 1980). 32E obscurus (Grfinberg) The antennal shaft hair (l-a) of the larva has more than 20 branches, thus IDENTIFICATION: (2-VI) suggested by Gillies andde Meillon (1968). The variety nowlini Evans antepalmate was by White (1980). synonymised BIOLOGY: Little is known about its feeding habits but Rickenbach et al. (1974) ADULT the capture of 2 females from vegetation in Cameroun which gave positive reactions to recorded tests for bovid and mixed primate/bovid bloodmeals. precipitin Gabon; 0S, 10E, (Service, 1976). Angola; 5S, 12E; 9S, 14E; 10S, 14E DISTRIBUTION: pattern. Angola; 9S, 19E; 9S, 20E (Ribeiro and Ramos, 1975). DISTRIBUTION: ADULT BIOLOGY: The exophilic habits of the coustani group are well known. Experimental of the night, practically ceasing after 01h00 (Chandler, Highton and Hill, 1975). In Uganda, part activity was almost exclusively at ground level (Haddow and Ssenkubuge, 1974). biting were well established. 1967). Angola; 5S, 12E; 7S, 12E; 7S, 15E; 7S, 18E; 8S 13E; 9oS, 16OE; (Vermylen, 9S, 18E; 10S, 14E; 10S, 17E; IIS, 17E; 12S, 14E; 12S, 15E; 15S, 12E; 17S, 9S,17E; it from the coustani species group. This is a better discriminating character that the distinguishing (Ribeiro and Ramos, 1975). Central African Republic; 4N, 18E (Cordellier and Geoffroy, 1976). ANOPHELES Series concolor Edwards From Vincke and Leleup (1949). Floats absent; deck occupying entire upper surface of egg Egg: and completely surrounded by a well developed frill. Chorion with strongly marked reticular 67

71 implexus Theobald BIOLOGY: Earlier work on the feeding behaviour of implexus in forests in Uganda had ADULT that it could be taken abundantly on human bait, 73% of this activity being in the daytime. shown findings were puzzling since, in the absence of catchers, mosquitoes would rarely encounter These hosts in the forest. Further detailed studies by McCrae et al. (1976) showed that the normal potential for the species were in fact cattle, that were herded at night in stalls several hundred metres hosts the forest. Females that had fed on cattle, as well as unfed females, could be caught resting outside tree-trunks in the daytime inside the forest, while hourly catches at night established that most of on population was leaving the forest at night to feed outside and only returning at dawn. From the on the association of Hydracarine larval mites on adults of implexus McCrae (1976) studies CELLIA Theobald Subgenus series NEOMYZOMYIA Angola to jebudensis. They noted that the specimens differed from the typical form by the northern clypeal hairs (setae 2-C) being only 2/3 3/4 the length of the posterior clypeals (4-C). This inner the forest near Yaounde, Cameroun, was on a platform at 23 m above the ground. Cordellier and in 1976) captured two females on human bait at ground level in gallery forest in the Central Geoffroy deemingi Service deemingi Service, Bull.ent.Res. 59: 647. x97o. CHRISTYA Series that they had little influence on adult survival. concluded DISTRIBUTION: Central African Republic; 4N, 17-18E; 5ON, 16OE; 8ON, 20OE; (Cordellier and Geoffroy, 1976). Rwanda and Burundi; S, 30E; 2S, 30E; 3S, 30E; 2S, 29E; 3S, 29E (Vermylen, 1967). Angola; 9S, 14E; ls, 14OE (Ribeiro and Ramos, 1975). South Africa; the record from the Karoo desert (Jupp et al., 1980) awaits contirmation. SMITHH section jebudensis Froud Machado et al. (198I) have attributed a series of 3rd stage larvae from IDENTIFICATION: of development of the posterior clypeals is unique amongst African, which suggests degree when more material becomes available, it will be found to represent an undescribed species. that, ARDENSIS section ardensis Theobald RECORDS: Angola; I2S, I4E (Ribeiro and Ramos, I975). NEW cinctus Newstead and Carter BIOLOGY: Rickenbach et al. (i 973) recorded that the only specimen caught biting man ADULT Republic, but remarked that adults of this species are normally caught resting on tree African trunks. NEW RECORD: Central African Republic; 4N, I8E (Cordellier and Geoffroy, I976). TYPE LOCALITY: Mambilla plateau, Nigeria. 68

72 Adult key section II. Described from a single female. Closely resembles IDENTIFICATION: from which it differs in the following characters: maliensis, 4"4.5 mm. Palps (fig. 8,a) Legs (fig. 8,b): All femora and tibiae with numerous pale spots, tarsus on all legs. o- 3 pale spots, mid tarsus -4 with pale apices. Wings (fig. 8,c): Generally paler, pale fringe with the fact that it is of particular importance, from the point of view of the transmission of stressed malaria, that the adults never leave the natural enclosure ofthe forest gallery and that in this rodent ldgi Christophers ollaeles RECORD: Ethiopia; 6N, 36E (O'Connor, 967). NEW Wing length: present opposite vein 5., vein pale at base, vein. with a pale spot, stem of vein 4 largely spot pale-scaled. and early stages not known. Fig. 8. deemingi, a-female palp. b-hind Leg. c-wing. natalensis, d-egg. BIOLOGY: The single specimen known was captured in a light-trap in gallery forest. ADULT Only known from the Mambilla plateau in Nigeria at a height of 65o metres DISTRIBUTION: feet). (5,500 lm-eti milleeamlsl Lips.aOlheles ADULT BIOLOGY: Yoeli and Bone (967) reviewed the behaviour of this subspecies. They environment they are exposed to a rather narrow temperature range of x6-xc. 69

73 natalensis Hill and Haydon Aopheles (fig. 8,d, material from Transvaal): With 2 very small, polar decks, each surrounded by Egg: a RECORDS: Ethiopia; 14N, 39E; 13N, 38E; 12N, 37E; 12N, 38E (O'Connor, 1967). NEW 3S, 30E; 4S, 30E (Vermylen, 1967). Angola; 7S, 15E; 8S, 15E; 10S, 17E; 12S, Burundi; 11 S, 14E (Ribeiro and Ramos, 1975). These new records extend the known distribution to 15E; north in the Ethiopian Highlands and Mso show that it is widespread in Angola. The status of the larvae from West Africa is still uncertain. natalensis-like (1980) included the name natalensis form multicinctus in the list of Afrotropical, White nil/theobald complex progress has been made in recent years in unravelling the problem of the nili complex, and it is Little that combined morphological and cytotaxonomic studies clear required. Even after the are of somalicus, there still seem to be at least 3 entities within the complex, viz. (1) dark- recognition man-biting populations known to be an important secondary vector in West Africa and winged, Ethiopia. (2) Dark-winged and apparently zoophilic populations, widespread if localised western much of the Afrotropical Region. (3) Pale-winged populations, associated with malaria over in the Congo basin in Zaire and referred to by Gillies and De Meillon (1968) transmission 'A. nili as form' Recent cytogenetic studies by Miles et al. (1984) have shown Congo evidence of no differences between populations of a presumed zoophilic population from Namibia chromosomal a man-biting population from the Congo Republic. However, and are continuing to treat nili as we the name covered more than one species and refer to it, therefore, if the nili complex. as BIOLOGY: Detailed studies of the bionomics ofnili* have been carried out in ADULT hamlet on a forest/savanna border region of the Republic ofthe Congo (Carnevale 1974a, 1974b, Carnevale the of up to 100/man/night during the dry season. Numbers fell to half this figure in the wet catches season, presumably through flushing out of larvae. Biting activity was later than previously season, with high levels continuing up to 04h00 in the dry season. recorded in other studies, the residual population resting indoors by day As found to be but a small was in biting catches ranged from % at different seasons; the index from light-trap rates was consistently higher ( %). Dissection of the ovaries of biting and freshly catches females showed the presence of uncontracted or partly contracted ovariolar engcrged in the sacs of specimens; only 6% at night and 16% in morning catches had fully contracted majority It was sacs. that most females fed again the same night after oviposition, although it concluded possible that was hours passed before re-feeding. Maturation of the eggs at ambient temperatures took 35-40h. some oviposition and the subsequent feed it was concluded that there must be between delay of24h or so a becoming gravid, before oviposition. Precipitin tests on gorged females caught indoors after or via window traps were 84% positive for man, 2.4% mixed leaving plus domestic animals and man animals. Despite the presence of a small number of cattle and sheep outside the houses at 13.6% no fed females resting outdoors by day could be found. Sporozoites night only found in 1/1981 were dissections, despite the close contact with man and moderately high parous rates. Its role gland a as secondary vector was thereby confirmed. Carnevale kindly informs *Dr. that the wing markings of this population were of the dark type known from West Africa us elsewhere. and developed frill; floats with about 28 float ridges, extending along whole length of egg; poorly with faint polygonal markings. chorion although Gillies and De Meillon (1968) had synonymised it with natalensis. and Boreham 1978, Carnevale et al. 1977, Carnevale, Bosseno and Zoulani 1978, Carnevale and Zoulani 1975). This species was the dominant biting man throughout the year, with (10%) of those entering to bite at night. The site of main biting activity followed the proportion of the inhabitants, outdoors during the early part of the night, indoors later on. Parous movements From the recapture of previously marked and released females it was shown that an interval of at least 3 days elapsed between successive feeds (mean temp. 26.2C); since there was little delay 70

74 (1975) noted that the population reached a minimum at the time before midnight. This became later as the season wore on, continuing up to 03h00, and occurred and after the rainy season most biting took place later still. The authors made the interesting during that, on clear nights, the presence of moonlight inhibited biting in woodland when the observation were bare of leaves but not at other seasons. In the same country, studies with window traps trees that only 24% of the feeding population remained in experimental huts next morning and showed in rooms with verandah traps (Coz, 1971 ). In the western lowlands of Ethiopia, Krafsur (197.0a) 17% recorded that peak populations were associated with river level, not local rainfall. Night-biting also indoors were 6.6 times greater than spray-catches in the morning. The sporozoite rate in 619 catches was 1.3%, compared with 2.4% in gambiae s.1. and 1.7% in funestus. Almost 1/3 ot" dissections in riverside villages were due to nili. inoculations RECORDS: Senegal; 13N, 16W (Cornet and Chateau, 1974). Ethiopia; 8N, 38E NEW 1967). Rwanda and Burundi; 2S, 39E (Vermylen, 1967). Angola; (dark-winged (O'Conner, rhodesiensis rupicolus Lewis RECORDS: Algeria; Holstein et al. (x97i) records its presence in the Oasis oftassili (23N, NEW in the extreme south-east of the country. oe), cameroni De Meillon and Evans cameroni De Meillon & Evans, Ann.trop. Med.Parasit. 29: cameroni, Coetzee & Du Toit, Mosquito Svstem. : Adult key section XI, larval key sections VI, VIII. Description mainly from IDENTIFICATION: and Du Toit (x979). Coetzee Head and palps (fig. 9,b): as in rhodesiensis. Pharynx: (fig. 9,c) as in rhodesiensis but the teeth with. spicules which are finely drawn out. Mesonotum: clothed in fine hairs with a few hair-like terminal extending back from the anterior promontory to a greater distance than in rhodesiensis. Legs: scales with a few pale scales at the apices of the hind tarsal segments. Wings: (fig. 9,a) costa and xst dark with large pale areas; apical pale spot extending to the 3rd vein; wing field with well marked vein spots present at the bases of the fork cells and near the cross-veins; apex of vein 5-x pale, pale extending onto the fringe; lower branch of vein 5.2 without pale area. sometimes genitalia: (fig. 9,d,e) gonocoxites with a few scales laterally and ventrally; harpago slightly Male lateral club expanded and bent forwards, apical seta longer than club, outer accessory seta pointed, the single inner accessory seta of equal length and more than half the length of the apical seta; and Seta I-III with 4-9 branches, I-IV 3-7 and I-V to I-VII with x- 3 branches. Seta 5-III to 5-VII (x-p). respectively 6- x, 5"9, 3-8, 3-7 and x- 7 branches. Seta 9-VII about o.6 length ofsegrnent, on IV with In Burkina Faso, Gayral et al. of the early rains. In the early part of the dry season, when the nights were cooler, most biting lls, 17OE; 17S, 19E; (pale-winged form) 9S, 15E; lls, 13E (Ribeiro and Ramos, form) 1975). TYPE LOCALITY: Fish Hoek, Cape Town. aedeagus with about four pairs of leaflets, some of which are finely serrated. Pupa: Outer paddle fringe (fig. 9,t") composed of fine hairs only, not extending beyond apical hair about o. 3 of segment, on II and III small and stubby. spine-like, Head darkly pigmented; antennae almost black with spicules of equal length sparsely Larva: distributed over whole antenna. Clypeus: inner (2-C) and outer clypeals (3-C) simple, bases widely separated, outer slightly shorter than inner. Shoulder hairs: inner and median hairs (x-p, 2-P) well mounted on large fused tubercles (in rhodesiensis the bases are usually narrowly developed, Mesopleural hairs: long hair (9-M) with 3-5 branches, (x o-m) simple. Palmate hairs: separated). x-ii to x-vii) (fig. 9,h): fully developed leaflets with well developed shoulders and with short, (setae stout filaments. Tergalplates: (fig. 9,g) o.6-o. 7 (rarely o.8) times as wide as distance between palmate (x-v). (occasionally 2) accessory tergal plates present on V-VII. Saddle: hairs (x-x) simple. hairs Not known. Egg: 71

75 9..-In. cameroni, a-a'ing, b-female palp. c-pharynx, d-harpago, e-phallosome leaflets, f-pupal paddle, g-larval Fig. segment h-leaflet ot" palmate hair (1-\'). \'. d e 79

76 Only known from the extreme southern tip of the continent. First collected at DISTRIBUTION: Hoek near Cape Town in the nineteen-thirties, it remained unknown for more than 4 years Fish it was rediscovered breeding in the Palmiet river some 5 kilometres away. until lounibosi new species* rabaiensis, Lounibos, Cah. ORSTOM Ent.med. I7:28 unavailable nomen nudum (attrib. White) LOCALITY: Kombeni River, Rabai, Kilifi District, Kenya (3 55'S, 39 35'E). TYPE material on which this species is based consists of one (apparently) reared adult and a series of The collected from Pandanus axils by Dr L P Lounibos and Dr R Subra. It is closest to rhodesiensis, larvae, was subsequently found breeding in ground pools in the same locality by Dr White. It differs which the latter species in the following characters: from the clypeal hairs (2-C, 3-C) is almost entirely absent. Owing to the condition of the larvae of setal counts were not possible. Table V sets out the results ofcomparative counts offraying complete It will be seen that the inner frontal hair (5-C) shows the greatest reduction. In 2 larvae rhodesiensis. hairs were simple, thus approaching the condition seen in the rock pool-breeding species, these Both shoulder hairs are greatly reduced in lounibosi, but it may be noted that the reduction ruarinus. much greater in the case of the inner hair (I-P) than the outer hair (2-P). In rhodesiensis the former is has more branches than the latter, while in lounibosi this relationship is reversed. The clypeal hair are usually frayed in rhodesiensis and in some specimens the outer clypeal (3-C) is bifid. hairs The palmate hairs (I-I to I-VII) in the larvae are stumpy and without filaments, whereas in 3. the filaments are well developed and finely tapered (fig. o, e, f). rhodesiensis and eggs unknown. pupa BIOLOGY: Nothing known. ADULT Larval setae hairs 2-P shoulder of o-m Branching No. Range Mean S.E. examined 8 o o o.29"* o. 5 o ** o ** o ** o x 7.7** o xo 8.8** o o7 o o o-8 3. xo o x3 xx.o x xo.o 0.55 o o o.75 o.66 o.8 LARVAL HABITAT: Sunlit or shaded rock pools in river beds, in association with rhodesiensis. ADULT BIOLOGY: Nothing known. Female Wing. The contrast between light and dark areas reduced.. The larva shows a great reduction in the branching of numerous setae. This is most marked in the 2. hairs (setae 5-C, 6-C, 7-C) and shoulder hairs (I-P, 2-P), (fig. io, a-d). Furthermore, fraying frontal of the clypeal hairs and the branching of 7 other hairs in lounibosi and the local population of lounibosi rhodesiensis No. examined Range S.E, Mean on 2-C Fraying hairs 3-C clypeal of 5-C Branching hairs 6-C frontal -7 x4.o 9 x3-2o x C of -P Branching o.65 long pleural hairs o-t Difference significant at the % level TABLE V: Comparison of branching of larval setae in lounibosi and rhodesiensis from the Kombeni River, Rabai. his account of the ecology of mosquitoes breeding in Pandanus axils, Lounibos referred to the larvae he found *In as rabaiensis White, in press. In fact, no description of the taxon had been published by G B White and rabaiensis is therefore a.-in. nomen nudum. We are accordingly describing the species as new and are dedicating it to Dr Lounibos. 73

77 t,'. 10.,-In. Iounibosi a-frontal hairs, c-shoulder hairs, c-palmate hair (1-\"..,.,4n. rhodesiensis. I)-Fmntal tairs, d-stnlcicr Fi, l'-ix'atlct of palmate hair (1-\'). hairs, 74

78 leaf axils of immature Pandanus at levels within the reach of a person standing on the forest floor. the was not possible to sample from higher crowns of the Pandanus plants, which were up to om or It above the ground. The quantity of water per leaf axil was considerable, averaging oml. more occurs in the dense and thorny understorey on the lower slopes of a steep embankment. In Pandanus uninviting habitat and protected by the vicious spines of Pandanus leaves, the species has up to this eluded systematic study. It was collected on occasions by Lounibos (once as a single larva), now once by Dr R Subra. However, careful searches by Dr G B White at the same season in and years, as well as by Dr A W R McRae and other collectors, failed to provide any further succeeding In the dry season there may be ground pools, created by the drying up of the Kombeni specimens. below the hillside where Pandanus grow. Typical rhodesiensis were found by Dr White in one of river, shaded pools in July 98. This species has been reported from the area previously by Van these et al. 955). Someren finding of a leaf axil-breeding is of exceptional interest, especially since the tree The niche in tropical Africa does not seem to have been exploited by an anopheline. The association hole with Pandanus has been reported from Thailand in the case ofsint0n0ides Ho, (Harrison of Scanlon, x975) but not apparently from elsewhere.* and On the basis of his own investigation of the biotope, Dr White (in litt.) has raised the question Note. whether the presence of larvae in Pandanus may not have been fortuitous. He suggests the possi- of that lounibosi is not a distinct axil-breeding species but simply rhodesiensis that, in the absence of bility pools, temporarily colonises Pandanus. In his view, the reduction in branching ofhairs could ground phenotypic and ascribable to the peculiar features of the habitat. He cites the analogous but be phenomenon of the 'hairiness factor' from natural container habitats that has been shown opposite cause a striking increase in the branching of larval hairs in certain culicine genera (Mattingly, to While accepting that environmental influences could have some effect on the larvae, our own 1975). is that the differences between the two foa'ns are too great to be attributed to this factor alone view that lounibosi should be regarded as specifically distinct. and ruarinus Edwards RECORDS: Angola; ls, 14E, 16S, 13E (Ribeiro and Ramos, 1975). The report by NEW series ly.ornyla azanlae Bailly-Choumara RECORD: Djibouti; 12N, 42E (Rodhain, Carteron et al., 1977). NEW brunnipe. Theobald Anophele. RECORDS: Gambia; 13N, 15W (Gillies and Wilkes, unpublished report). Angola; 12S, NEW (Ribeiro and Ramos, 1975). 17E culicifac es Giles As described by Beidas and Gillies (1980) from material from Oman, the egg does not appear Egg: differ from that ofthe type form in India, i.e. deck narrow, completely surrounded by frill which is to between the floats; chorion unornamented, floats with float ridges. continuous to be a complex of at least 3 cryptic species, termed sp. A, B and C (Green and Miles 1980, shown et al. 1983). Species A has recently been identified from material collected on the border Subbarao LARVAL HABITAT: (Data from Lounibos, 979, and White, in litt.): Larvae were collected in whole question is one of great interest and warrants further investigation. The DISTRIBUTION: Only known from the type locality on the coast of Kenya. et al. (1967) of this species from Tanzania, in the vicinity of Lake Victoria, requires Suenaga confirmation. Note. The morphological taxon culicifacies in India, Pakistan, Thailand and Sri Lanka has been the United Arab Emirates and Oman (Akoh et al. 1984) which makes it seem likely that between on Socotra is a peripheral form of sp. A. culicifacies hackeri in Malaya, * species normally found in split bamboos, is occasionally found at ground level in the old leaf bases of a Nipah palms (Reid, 1968). 75

79 dthali Patton BIOLOGY: Although regarded as of no medical importance in Africa (Gillies and De ADULT 1968), it has been implicated in southern Iran as a vector of malaria, 5/339 gland Meillon, having been positive for sporozoites (Manoocherie et al. 1972). This finding is puzzling, dissections the low degree of contact of the species with man. The human blood index from precipitin given in the Iranian studies was only in the range 1-20%. In the same country, Carmichael et al. (1968) tests only 1/476 blood meals positive for man and recorded the human biting rate as very low. The found that it might act as a vector on the Red Sea coast of the Sudan was suggested by possibility and Wernsdorfer (1967). Wernsdorfer RECORDS: Ethiopia: 16N, 38E; 15N, 36E; 13N, 37E; lln, 37E; 9N, 40E; 9N, NEW (O'Connor, 1967). Kenya: 4N, 36E (Rodhain, 1971). The last record extends the known 42E of the species southwards into Turkhana Province of Kenya, to the east of Lake distribution Turkhana. ethiopicus new species LOCALITY: Gambela, Illubabor Province, Ethiopia. The Type has been deposited in the TYPE Museum (Natural History). British single specimen, now lacking hind legs, was caught by Dr E F Krafsur, resting, engorged, A a house. inside IDENTIFICATION: length: 3 mm. Wing Head: Scales on vertex fairly broad, white. Palps: (fig. 11,a) smooth with 3 narrow pale Female the apical band not extending to apex of 5th segment. Pharynx: With pairs of rods and bands, 2 pale scales at or of tarsus 1. Wing: (fig. 11,b) Predominantly dark-scaled, main dark areas on costa and vein apex developed, no pale interruption in 3rd main dark area (preapical dark spot). normally In the absence ofhind legs the definition of the species is incomplete and it cannot be keyed out Note. species with which it could be confused would be rujqpes s.l. From all forms of this species it only by the absence of the second pale spot on the 2nd main dark area (median dark spot) ofvein 1. differs the other hand, if either or both these hind tarsal segments were not entirely pale, ethiopicus On run down to Section VII of the adult key, i.e. those with palps dark at the apex. In this Section would would key out to cinereus and other species in Paramyzomyia series. It differs from all these ethiopicus the presence of a pale fringe spot opposite vein 6 and of the vestiture ofbroad, white scales on the in flavicosta Edwards It has been known for a long time that jtavicosta has a curiously disjunct IDENTIFICATION: being widely distributed in the Northern Savanna zone while also occurring in distribution, Recently, Smith et al. (1977) have reported this species to be common in the northern Madagascar. while Ribeiro and Ramos (1975) recorded it from one locality in Angola. The southern Transvaal, form is paler than some northern forms (fig. 11,c) but, given the well known variability of African (fig. 11,d, material from Kaduna, northern Nigeria, coll. C D Ramsdale). Deck wide, frill Egg: developed, not continuous between floats; float ridges-; chorion without obvious poorly markings. BIOLOGY: In Angola, in the vegetated margins of a well and along the banks of slow- LARVAL moving rivers. cones, without roots. Mesonotum: Fairly thickly clothed with broad, white scales. Legs: Fore and mid legs dark except for a small patch of pale scales at apex of each tibia and with in the usual way. But if it were assumed that either hind tarsus 4 and 5, or 4 alone, were pale, the mesonotum. DISTRIBUTION: Only known from the type locality. species, the material must be regarded as typical offlavicosta. In the absence of cytogenetic this both forms have to be regarded as conspecific. studies 76

80 d Fig In. ethiopicus, a-female palp. I)-Wing../la'icosta. c-wing (from northern Transvaal). d-egg (Nigeria). 77

81 In the northern Transvaal, the species was not uncommon biting man outdoors. RECORDS: Gambia; 13N, 14W. Benin; 6N, 3E (Brenguesetal. 1969). Central African NEW 8N, 20E (Rickenbach, 1969). Angola; 10S, 17E. South Africa; 22S, 30E. Republic; longipalpls Theobald White (1972b) has recorded typical longipalpis adults from the eastern and IDENTIFICATION: sides of the Ethiopian Highlands at altitudes of m ( it), thus extending western known distribution of the species very considerably to the north. This finding also confirms the records of longipalpis from Ethiopia, which had been called in question by Gillies and De earlier (1968). Ribeiro and Ramos (1975) comment that larvae from Angola may have saddle Meillon (l-s) with 5 branches. hairs BIOLOGY: In Ethiopia, adults were found in rock crevices and hollow trees, both ADULT by rock hyraxes, (White, 1972b). inhabited RECORDS: Ethiopia; 8N, 34E. Rwanda and Burundi; 3S, 30E (Vermylen, 1967). NEW 8S, 36E (Freyvogel and Kihaule, 1968). Angola; 9S, 14E; 10S, 14E; 12S, 14E; Tanzania; 13E; 15S, 13E. 13S, moucheti Evans BIOLOGY: Service et al. (1977) recorded a sporozoite rate of 3.8% in Gabon. In the ADULT region of Uganda, Onori and Bentheim (1969) failed to find any positive glands in 1008 Busoga RECORDS: Central African Republic; ztn, 18E (Rickenbach, 1969), Gabon; 0S, 10E NEW 1976). Rwanda and Burundi; 2S, 30E (Vermylen, 1967). (Service, schwetzi Evans RECORDS: Angola: 12S, 15E; 14S, 21E (Ribeiro and Ramos, 1975). NEW term "funestus sub-group" was introduced by Gillies and De Meillon (1968) for the 3 species, The s.s., aruni Sobti and parensis Gillies, none of them being separable morphologically in the funestus in the funestus group have confirmed the phylogenetic closeness of these species. rearrangements of his findings (extracted from his fig. 7) are shown in Table VI. funestus and vaneedeni Some homosequential for their chromosomal banding arrangements which introduces an element of are into the separation of species based purely on inversion rearrangements. caution, (vaneedeni) arunz? parensis No. of fixed inversions on 3 chromosomal arms with shared funestus shared with not funestus ADULT BIOLOGY: dissections. FUNESTUS section ANOPHELES FUNESTUS SUB-GROUP stages. To them must now be added vaneedeni (see below), a species described by De Meillon et early (1977) under the name aruni?. The penetrating studies by Green (1982) on chromosomal al. IO confusu$ rivulorum fuscivenosus leesoni o TABLE VI" Chromosomal rearrangements in the funestus group, from Green (982) 78

82 the assumption that inversion rearrangements are a better guide to the phylogeny ot'a group On morphological characters alone, these results show that confusus and the funestus sub-group than by itseltl In fact, Green (1982) showed that it was more closely related to the very similar stands speciesfluviatilis and to culicifacies and should, therefore, not be classitied as Oriental member ot" a funestus group. the (1977) also showed the presence of a sex-limited esterase in the male accessory glands ot" Green and aruni? in southern Africa. funestus has been known for a good tnany years that an exophilic member of thefunestus sub-group It The existence of this form was noted by Gillies and De Meillon (1968), and after a.house-spraying. study an account of it was given by De Meillon et al. (1977). The authors showed that it detailed be distinguished fromfunestus by the greater pallor of the wings and the broader pale bands on could palps. These characters served to identify all except occasional specimens. Because ot" its the the male of aruni a well marked pale patch is constantly present at the base of the palpal club. in (in De Meillon et al. 1977) stated that he had evidence from a study of the polytene Green and spermatogenesis in hybrids that a sample of the insects they called aruni? and chromosomes are distinct species. Further studies of chromosome arrangements by Green and Hunt funestus and Green (1982) confirmed the distinctness of these two species from other members ot'the (1980) group, i.e. parensis, fuscivenosus, confusus, rivulorum and leesoni. Only aruni s.s. and brucei were funestus included in this comprehensive study. not the light of these latter findings we have reexamined the question of the relationship ot'aruni In the Transvaal "arun/" with particular reference to the male palps. It is well known that the and or absence of a pale patch at the base of the club in members of thefunestus group is a useful presence character. According to Gillies and De Meillon (1968) a basal pale patch is never discriminating in funestus, rivulorum, brucei and (apparently)fuscivenosus. In confusus it is present in about 50 present of specimens, and in parensis in up to 90 percent. In leesoni and aruni it is tbund in all percent The results of the present study have shown the existence of constant differences between specimens. males of aruni s.s. the results of crosses between the two species still remain unknown, a tbrmal description ot'the and African species now seems appropriate (see below). South funestus Giles BIOLOGY: Little recent work has been published on the biology of the early stages of LARVAL replacement, in this case of funestus by rivulorum, is mentioned by Aboul-Nasr (1970) as species in south west Uganda when the former species was eliminated by house-spraying. occurring BIOLOGY OF ANOPHELES FUNESTUS ADULT and fertilization. In Burkina Faso, about 3/4 of females were fertilized before the first feed Mating other records of dawn swarming by, and in view of the subsequent demonstration that no gambiae (q.v.) the antennal fibrillae become erect at each dusk but remain so for only 2h in by previous workers, these authors tbund a very low incidence of recognizable mating recorded in nulliparous females. plugs form one phyletic line, rivulorum and fuscivenosus a second line while, surprisingly enough, leesoni occurs in the northern Transvaal in a region from whichfunestus s.s. has been almost eliminated by resemblance to the equatorial species aruni Sobti, De Meillon et al. referred to the South Afi'ican form as "arum?" At the same time they pointed out that the two species were probably distinct since s.s. and the Transvaal "aruni", so that, although chromosomal arrangements in aruni The association of this vector with rice fields, formerly observed in Nyanza Province, funestus. now apparently no longer occurs (Chandler and Highton, 1975). The phenomenon of Kenya, Supporting details were not given. and Coz, 1973a). The authors suggest that some fertilization may occur at the subsequent (Brengues citing Harper (1944) as having observed males offunestus swarming at this hour. We know of dawn, afterwards, the validity of the original field observations at dawn must be placed in doubt. As 79

83 In Gambia, Gillies and Wilkes (1976) foundfunestus to be flying over open savanna evenly at heights up to 2m above the ground, their numbers declining steeply above that. rather when dispersing on host-seeking flights, the movements of females were not obstructed by However, 6m-high netting fence (Gillies and Wilkes, 1978). a selection. In conformity, with previous findings in East Africa, the human blood index has been Host to be fairly consistently high, 0.99 in the Nyanza Province of Kenya (Pull and Grab, 1974), found indoors and 0.94 outdoors (Chandler, Boreham et al., 1975), and in Tanzania, indoors 0.98 the rather lower figure of 0.24 outdoors (White et al., 1972). and entering. In northern Nigeria, indoor biting catches were 1.5 times greater than outdoor House (Molineaux and Gramiccia, 1980). Entry showed a small peak around nightfall but mainly catches after midnight (Rishikesh and Rosen, 1976). In Burkina Faso, a quarter to a third of occurred left the houses after feeding (Coz 1971, Brengues and Coz 1973a), while half of the females left 24h later. remainder cycle. In Burkina Faso, Brengues et al. (1968) found more than 80% ofbiting to take place after Biting The use of light-traps in Kenya (Chandler, Highton and Hill 1976, Chandler 1977), midnight. that considerable flight activity near houses took place in the early' part of the night revealed biting activity was, as usual, more evident after midnight. Perhaps the same feature was although observed outdoors in gallery forest in the Central African Republic, where Cordellier and being (1976) recorded a first peak at around sunset although the major part of biting activity Geoffroy place in the hours leading up to dawn. No such pattern was apparent in Liberia, either in forest took savanna villages, and only the usual late biting activity was observed (Kuhlow and Zielke, 1978). or outdoors in villages had largely fed on man. Brady showed that, from fed:gravid ratios in resting and outdoors, the total numbers feeding on man could be calculated. An extreme example of houses during the second day of the gonotrophic cycle was recorded by Brengues et al. (1969) in exophily Benin, where gravid females were largely absent from houses. In Kisumu, Kenya, grain stores south found to be a more productive source of females resting outside houses than pit shelters (Clarke were al., 1980). et cycle. In Burkina Faso, Brengues et al. (1968) and Brengues and Coz (1973a) recorded the Gonotrophic of the first cycle as 4-5 days, in later cycles as 2-3 days. The first blood meal usually tbllowed length days after emergence. 63% of females went through a pregravid stage, i.e. they required a two blood meal to mature the first egg batch. After oviposition, about two thirds of timales tied second the same night. again XIII). Brengues et al. (1968), Brengues and Coz (1973a) recorded values of the daily survival rate Longevity. of in Burkina Faso. In the same country, Brun (1973) obtained an annual mean value (p) p=0.84 from use of the ratio of immediate:delayed sporozcfite rates. Critical studies by Krafur of Garrett-Jones (1977) in Tanzania, using the rate of uptake of infection with Wu. bancrofti, gave a and of 16.3% daily mortality (p=0.835). This finding agrees well with earlier work in the same area value Polovodova's technique. In Ethiopia, using the more conventional (if less reliable) parous using Krafsur (1970b) calculated daily mortality rates to be 9-12%. The same author (1970b) noted rate, existence of a deficit of nullipars in the house-resting sample used for age analysis. the In Burkina Faso, Coz (1966) found 17.6% of females infected with Mermithidae. A follow- Parasites. study by Poinar (1977) described the life cycle and morphology ofthe parasite. Brun (1973), also up Burkina Faso, made the interesting observation that parasitization by Mermithids in adult in was commonest in house-catches near breeding sites and rare at distances greater than funestus 200m. to disease. The following series of gland dissections for sporozoites have been recorded: 5.6% Relation in Burkina Faso (Coz, 1973a), 1.6% in northern Tanzania (White, Magayuka and positive Flight pattern. Outdoor resting. Brady (1974) in Ghana and Brun (1973) in Burkina Faso found that td females Brun (1973) recorded the time of oviposition by females caught in the wild when freshly Oviposition. and set up in cages. Nearly 60% of females laid eggs in the first 2 hours after sunset (see Table fed Boreham, 1972), monthly rates ranging from % in central Tanzania (Freyvogel and Kihaule, 80

84 values of 1-1.5% in a long series of dissections in Busoga Province of Uganda (Onori and monthly 1969), and 2.3% in Bwamba County in the same country (White, 1973). In The Gambia, Benthein, Gillies and Gubbins (unpublished report) found only 0.1% infected in over (1973a) found 0.5% of females carrying infective larvae of Wu. bancrofti in Burkina Faso. Coz et al. (1968) and Brengues (1975) concluded that in the West African savanna its potential Brengues In a savanna region of northern Liberia, where infective larvae were found in 1.4% and important. n two villages, Kuhlow and Zielke (1978) came to the same conclusion, as did Maasch (1973) 0.4% in the coastal region of the same country. In three coastal districts oftanzania, 2.4%, 1.4% working 0.6% respectively harboured infective Wu.bancrofti larvae (Mosha et al. 1974, Magayuka 1974, and 1974). In villages on the coast of Kenya, wherefunestus was the main vector, 1% of females Bushrod infective larvae (Wijers and Kiilu, 1977). Heavy infections were thought to diminish flight carried range. isolations from funestus are listed in Table VII. Virus Ethiopia Western Uganda Bunyamwera Tataguine Orunga Akabane Gerrniston near virus Tataguine B Hepatitis vaneedenl new species aruni? De Meillon et al. Mosquito Vews 37: & Moore (i 972) Lee (I 979) Lee et al. 1976) Ota et al. (1976) Tomori & Robin (i976) Metselaar et al. (i977) Johnson in recognition of his contributions towards clarifying the status of this perplexing species. Tzaneen, LOCALITY: Pusela, Tzaneen, South Africa. The type has been deposited in the British TYPE (Natural History), and paratypes in the South African Institute for Medical Research. Museum a) Separation from aruni Sobti. IDENTIFICATION: females can generally be separated from aruni by the darker wings and particularly, by the The of the pale bands on the palps. These two characters are expressed as the "wing-tip ratio" breadth the "palpal band ratio", and the method of estimation is shown in fig. 12, a, b. When these are and against each other, as in fig. 13, the area of overlap between the two species covers a small plotted other members of the funestus group may still be possible. The male genitalia and early stages with as in other members of the funestus sub-group. The male palps differ as follows: are monthly rates ii'om 0-3.4% in an extensive series ot'dissections in western Ethiopia (Krafsur, 1967), a mean of 2.3% in Angola (Ribeiro and Ramos, 1975). In two areas ofsouth western Uganda 1971), unusually high rates of 5.1 and 6.2% were recorded (Kafuko et al., 1969), contrasting with the the 000 dissections. as a vector of filariasis equalled that ofgambiae, but that usually being less abundant it was less Jos plateau, Nigeria Kenya do. (i972) Prince et al. East Africa antigen TABLE VII: Virus isolations from funestus. species is named for Gideon van Eeden of the National Institute for Tropical Diseases, The only of their respective ranges of variation. Palps in which theratio ofthe two distal pale proportion to the intervening dark bands is less than 2 exclude a diagnosis ofaruni, although confusion bands 81

85 12. Method of estimation of: a-wing-tip ratio and b-palpal band ratio. Male palpal club in c- aruni, d- vaneedeni (5-10%), Fig. vaneedeni (90-95%). e- -I- 82

86 vaneed i band ratio Palpal Fig. 13. Segregation of vaneedeni and aruni by plotting wing-spot ratio against female palpal band ratio. In 5/53 males examined, the palps were as infunestus s.s., i.e. without a pale patch ofscales at the 1. of the club, opposite the joint between the 3rd and 4th palpal segments (fig. 12,e). In the base 90% a narrow band of pale scales was present (fig. 12,d). This band appears to be formed remaining the pale tips of 1-2 overlapping rows of scales. Occasional specimens are also seen in which by pale-tipped scales are present. In aruni, the pale patch is constantly present and consists of scattered broad band, some 3-4 scales across, the most proximal scales often being pale to near their bases a The subapical patch is usually smaller or about equal in size to the apical patch. Furthermore, in 2. minority of specimens of vaneedeni in which a basal pale patch of scales is present, the distance the the basal and subapical pale patches (B in fig. 12,d) is greater than the width of the between patch (A in fig. 12,d). In aruni, the ratio A/B is occasionally as low as 1, but it is normally subapical x aruni \ x x x (fig. 12,c). much greater than this (range in 21 specimens). 83

87 14. Segregation ot" populations ofvaneedeni andfunestus by plotting wing-spot ratio against female palpal band ratio (from Fig. Meillon et al. 1977). De Separation from funestus s.s. b) morphological grounds for separating these two species are less clear-cut. De Meillon et al. The be identified by this combination of characters. However, our own studies on populations of not from other areas show that the degree of overlap may be quite considerable (fig. 15). funestus from West African coastal regions in particular may have appreciably paler wings than Specimens from the Transvaal, in conformity with the findings ofservice (1960) that coastal populations those Nigeria had generally paler wings than those from dryer regions to the north. However, since it in well be that vaneedeni will be found to have amuch more restricted distribution thanfunestus, the may will be mainly confined to separation ofsouthern or eastern populations ofthe2 species. De problem et al. pointed out a number of other characters by which the 2 taxa differed, although these Meillon not be used for identifying individuals. Thus, they found that a pre-sector pale spot on the could was present in all except 3% of vaneedeni, compared with the 79% offunestus in which it was costa A pale spot in the middle of vein 3 was present in 99% of vaneedeni compared with 50% of lacking. They also showed that pale spots on the tarsal joints, which are never seen infunestus, are funestus. vaneedeni funestus 1.0 band Palpal ratio showed that populations of the 2 species in the Transvaal coul.d be distinguished by plotting (1977) ratios against palpal band ratios (fig. 14) and that only rare specimens of vaneedeni could wing-spot sometimes conspicuous in vaneedeni. 84

88 oo O" band ratio Palpal 15. Segregation of individual females of vaneedeni and funestus by wing-spot ratio and female palpal band ratio. Dotted Fig. of the specific distinctness of vaneedeni must rest on the results of crosses with Confirmation Green and Hunt (1980) demonstrated that, although the chromosomes in the 2 species are funestus. only sterile hybrids result from crosses. homosequential, sum up, the identification of vaneedeni by morphology rests on the demonstration of To or families offunestus-like individuals, in which: populations of males have a small pale patch at the base of the palpal club % In the females, when wing-spot ratios are plotted against palpal band ratios (as defined above), 2. percentage of specimens whose measurements fall within the areas enclosed by values of 1.0 and the for both ratios are as shown in Table VIII below. 1.5 Larva and Egg as in funestus. Pupa, apparently different from that offunestus. Not BIOLOGY: Essentially an outdoor-biting species, frequently caught biting man outside ADULT in the early hours of the night (De Meillon et al. 1977, Smith et al. 1977). In the daytime, houses can be found resting in pit-shelters, sometimes in quite large numbers, in association with females and other species. The great majority of specimens were found by precipitin tests to have fed leesoni cattle (Smith et al., loc.cit.). Tests on laboratory bred material fed on gametocyte carriers showed on susceptibility tofalciparum malaria (De Meillon et al., loc.cit.). The involvement of this species in full transmission in the Transvaal, in the virtual absence of the main vectors, remains malaria 2-0 vaneedeni O funestus 1-0 "i " lines iv,dicate areas enclosed by values of respectively 1.0 and 1,5 for both ratios (see Table VIII). LARVAL HABITAT: uncertain. 85

89 of Value ratio per cent per cent the specimens were from southern Africa or from a wider area per cent arun/sobti In discussing the identification of this species Gillies and De Meillon (1968) -IDENTIFICATION: that "none of the characters used are completely diagnostic". However, as shown under remarked it is possible to separate all males from other members ofthefunestus group by the presence vaneedeni, a fairly broad patch of pale scales at the base of the palpal club, coupled with the fact that the size of the subapical patch is equal to or greater than the distance between this patch and the basal of The females can be separated from all other members of the group vaneedeni by plotting the wing-spot ratio against the palpal band ratio (see figs. 12 and 13). It except be seen that measurements for aruni fall completely outside the range recorded for funestus. will from examination of small numbers (shown in brackets) ofparensis (22), rivulorum (10), established (5) and leesoni (5) that wing-spot and palpal band ratios fall within the same range asfunestus. confusus confusus Ev.ans and Leeson RECORDS: Ethiopia; 7N, 36E (White, 1972b), thus confirming earlier records. Central NEW Republic; Rickenbach (1969) reported larvae from several sites that corresponded to African If confirmed, this would extend the known distribution considerably to the west. confusus. HABITAT: In Ethiopia at 1500m, breeding in hippopotamus footprints (White, LARVAL 1972b). fluvlatilis James Oriental species has been recorded from the Wadi Hadhramaut in South Yemen by This from North Yemen in the south- corner of the Arabian peninsula. Earlier records of this species from the Afrotropical part of west were considered doubtful by Mattingly and Knight.(1956). A similar viewpoint has been Arabia by Harrison (1980), who suggested that the larvae from Yemen might really belong to expressed member of the African funestus group. some regards the larvae, the problem resolves itself into differentiating between leesoni and As Apart from their closely similar chaetotaxy and almost identical abdominal tergal plates, fluviatilis. sole character that might serve to separate these two lies in the small, paired, sclerotised plaques the plates) on the thorax. We have compared larvae of leesoni from East Africa, fluviatilis from (notal and Saudi Arabia and Maffi's material from Yemen. While leesoni sometimes possesses or 2 Iraq plaques on the metathorax, the fluviatilis we have seen have 2 well developed metathoracic small just anterior to a line drawn between the metathoracic palmate hairs, together with a plaques, plaque anterior to these again at a point corresponding to the junction of the meso- and median segments. The material from Yemen is not now in a very good state of preservation, netathoracic in one specimen the thoracic plaques are identical with those onfluviatilis from Iraq. Although but funeslus (Transvaal) funestus (overall) vaneedeni I.O 5 5 o TABLE VIII: of females which fall within the area enclosed by plotting Percentage ratio against palpal band ratio for different values of these wing-spot ratios. Two sets of values are given for funestus, depending on whether This is illustrated in fig. 12,c where the ratio A/B is equal to or greater than 1. In all the other patch. species this ratio is less than 1. Measurements for other members of the funestus group have not been plotted, but it has been Colbourne and Smith (1964) and, more recently, by Maffi (1971 not conclusive, the finding supports the view that fluviatilis does indeed occur in Yemen. 86

90 the face of it there is nothing surprising in this, sincefluviatilis is widespread in arid regions in On Middle East and the eastern parts of Arabia. In fact, since no member of the funestus group the in similar habitats in north-eastern Africa, but only in the cool, well-watered highlands of occurs it seems to us, on ecological grounds, as much less likely that leesoni should occur in Yemen Ethiopia, fluviatilis. than leesoni Evans RECORDS: Central African Republic; 8N, oe (Rickenbach, x969). Angola; 9S, x4e NEW and Ramos, x975). (Ribeiro rivulorum Leeson RECORDS: Ethiopia; 5 N, 36 E (Rodhain, x97 I). Angola; xo os, x4 E (Ribeiro and NEW variability shown by "marshallii" has led in the past to the description of The number of varietal a subsequently treated as synonyms by later workers. Recently, considerable light has been forms, on the problem by the studies of Lambert (x979a x98x and of Lambert and Coetzee (x982), shed have shown that. the taxon marshallii constitutes a complex of at least 4 species. This conclusion who based on the recognition of cytogenetic differences. Certain combinations of morphological was were found to be associated with these differences. characters constituted at present, the complex consists of 3 named species, marshallii, letabensis and hughi, As a 4th species which, for lack of adequate material, was referred to by Lambert as sp. E. At the and time, two other forms that are currently treated as synonyms, pitchfordi andpseud0c0stalis, were same to be biologically distinct from all of the 4 cytogenetically defined species. However, considered this latter conclusion was based on statistical analysis of morphological variation alone, since and Coetzee left them in synonymy with marshallii pending further study. Lambert most important diagnostic characters, taken from Lambert and Coetzee, are The follows: as Palps: Apical dark band about equal in width to adjacent pale bands. Legs: most tarsal segments. narrowly pale apically. very (for full setal counts see Lambert and Coetzee): Paddle: lateral fringe consisting of spine-like Pupa: changing gradually to hairs, not extending beyond apical hair (seta x-p) (fig. x6,b). setae (Full setal counts given in Lambert and Coetzee): Clypeal hairs: posterior (seta 4-C) with 3-4 Larva: branches, rarely I-2. Shoulder hairs: (fig. x6,f): floats very close, with decks exposed at either end ofdorsal surface, surrounded by Egg: narrow frill. a BIOLOGY: commonly caught biting animals outdoors. Contact with man appears to be ADULT rare. the Central African Republic, attributed by Cordellier and Geoffroy (x976) to marshallii, from uncertain. remains Ramos, 975). MARSHALLII-HANCOCKI section ANOPHELES MARSHALLII COMPLEX marshallil Theobald marshallii (restricted). Lambert & Coetzee, Syst. Entomol. 7:325 x98. arising from large tubercles, contiguous at the bases, inner hair x-p) with x4-o branches, median (z-p) x-8 branches. Abdomen: hair 3 on ventral surface of (seta II (seta x3-ii large with 5-9 branches arising symmetrically from the stem (fig. I6,d). segment DISTRIBUTION: (fig. 7) A widespread, highland species from Ethiopia to the Transvaal, South A number of new records of marshallii s.l. have been reported from Rwanda and Burundi Africa. 967) and from Angola (Ribeiro and Ramos, 975). Since species determination could (Vermylen, be made at that time, no further details are given here. Similarly, the identity of the specimens not 87

91 16..-In. marshallii, b-pupal paddle, d-larval hair 13-II. f-egg. letabensis, a-pupal paddle, g-egg..4n, hughi, c-pupal Fig. e-hair 13-II. h-egg..4n, upembae, i-female palp (diagrammatic). paddle, 88

92 F/. 1]:.Distribution ofmarshallii s.s. Data from Lambert (1979b), Lambert and Coetzee (1982). 89

93 letabensis Lambert and Coetzee letabensis Lambert & Coetzee, Syst. Entomol. 7:328. x98. TYPE LOCALITY: Greystones, Transvaal, South Africa. The main diagnostic characters are as follows (summarized in Tables IX, X below)" o TABLE IX: Summary of characters of marshallii complex (data from Lambert & Coetzee. x98). 90

94 IO-II 2-III marshallii/ letabensis 6-IO -3 6-x 5 7-x o 7-x5 I I--2 3 x- absent Palps: Apical dark band very narrow, much less than width of pale bands. Legs: pale bands on. rather more prominent than in marshallii. tarsi Paddle: lateral fringe as in marshallii but more extensive, with more than 2o very fine hairs Pupa: apical hair (seta x-p) (fig. x6,a). Setae as in marshallii except that seta 3-II is simple or bifid. beyond Differs from marshallii as follows. Clypeal hairs: posterior hair (seta 4-C) simple. Shoulder hairs: Larva: hair (I-P) with 9-4 branches, mounted on a small tubercle which is separated from the basal inner of median hair (2-P) by a distance greater than its own width. tubercle (fig. I6,g) Floats well separated with 2 polar decks, i.e. generally as.figured by Gillies and De Egg: ( 968) for marshallii but with openings of decks slightly more rounded. Meillon BIOLOGY: Occasionally caught resting in houses or biting man outside. It seems that the ADULT of"marshallii" caught on human bait in the coastal lowlands ofnorth-east Tanzania, reported series Gillies (1964), were mainly of this species. Although no positive glands have been reported, its by with man is sufficiently close for it to be potentially a secondary vector. contact (fig. i8) Widespread from Ethiopia to Natal, though generally at lower DISTRIBUTION: branches, egg with 2 polar decks and well separated floats. 7-9 hughi Lambert and Coetzee hughi Lambert & Coetzee. Syst. Entomol. 7:39 x98. LOCALITY: Makonde, Transvaal, South Africa. TYPE main diagnostic characters are as follows: The Palps: Apical dark band about equal in width to adjacent pale bands. Legs: Pale bands on hind. more prominent than in either marshallii or letabensis. tarsi Paddle: lateral fringe as in letabensis but extending a shorter distance (8-20 very fine hairs) Pupa: I ,c). Setal counts differing from those ofmarshallii and letabensis at Differs from marshaltii in the following characters: Clypeal hairs: Posterior hair (seta 4-C) Larva: Shoulder hairs: Median hair (seta 2-P) with 8-13 branches. Abdomen: Hair 3-II small, with 8- simple. hughi Seta 7-CT x2-ct 9-I 2-II 3-II 8-II III 2-IV 5-IV x-v 6-V -2 in letabensis T X: Summary of setal differences in pupae of the marshallii complex. TABLE than marshallii. altitudes The attribution of specimens of "marshallii" from coastal regions of Tanzania to letabensis is Note. based on examination of material exhibiting the following characters: females with broad palpal bands, pupal paddle fringe extending well beyond apical hair (I-P), posterior clypeal hair in pale (4-C) simple, inner shoulder hair (I-P) with o- branches, abdominal hair 3-II large with larvae numerous sites beyond apical hair (seta I-P) (fig. (see Table X). branches, arising mainly from one side of the stem of the hair (fig. i6,e). 2 Floats well separated, a single deck continuous between the floats (fig. x6,h). Egg: 91

95 Fig. 18. Distribution of letabensis. Data from Lambert (1979b), Lambert and Coetzee (1982) and Gillies (1964) see text. 92

96 BIOLOGY: Caught mainly on animals outdoors. ADULT Only known from highland areas in Transvaal and Zimbabwe. DISTRIBUTION: Lambert (x 9 austenii Theobald Ribeiro and Ramos (x975) comment on the variability of the size of the IDENTIFICATION: In Angola, larvae were caught in clear water in a shallow well dug in a depression, (Ribeiro and Ramos, x975). grassed-over BIOLOGY: The same authors report the capture of a small number ofadults biting man ADULT night in Angola. at RECORDS: Angola; xos, x6e; x2s, x5e; x2s, I7E (Ribeiro and Ramos, x975). NEW brohieri Edwards Service (I 977 a) has recorded further specimens from the forest belt near Enugu, VARIATION: With the exception of the absence of a pale apical spot on the hind tibia, the variation Nigeria. falls within the range seen in brohieri from the savannah. Some doubt, however, still observed about the conspecificity of these forms. remains RECORDS: Gambia; Keneba, I3N, x6w (Gillies and Wilkes, I97o ). Despite intensive NEW season, suggesting that a temporary extension of its range had occurred. Gayral et al. ( 975) one confirmed that brohieri and hancocki occur sympatrically, thus providing further evidence ofthe have RECORDS: Rwanda and Burundi; 2S, 29E; 3S, 29E; 3S, 30E; 4S, 30E (Vermylen, NEW Rickenbach (1969) reported the finding of a larva resembling that ofgibbinsi in the Central 1967). Republic. African hancocki Edwards RECORDS: Liberia; 6ON, 8W (Kuhlow and Zielke, 1978). Gabon; 0S, 10E (Service, NEW Congo Republic; 3S, 14E (Carnevale, 1974b). 1976). hargreavesl Evans BIOLOGY: Haddow and Ssenkubuge (1974) point out that Gillies and De Meillon ADULT forest in Uganda, Haddow and Ssenkubuge recorded a sharp peak in biting activity in the Bwamba hour of the night. 2nd RECORDS: Gabon; 0S, 10E (Service, 1976). Central African Republic; 4N, 18E NEW 1969). (Rickenbach, harperi Evans Ribeiro and Ramos (1975) have confirmed that the larvae of Angolan IDENTIFICATION: x) reported the existence of another chromosomally distinct form from northern which he referred to tentatively as marshallii sp.e. He discussed the possibility that this might Natal, pitchfordi (Giles), a taxon currently treated as a synonym ofmarshallii, but rejected this conclusion be on the basis of statistical analysis of wing and palp characters. plates in the larvae although, from their key, it is evident that the width ofthe plates does not tergal 2/3 the distance between the palmate hairs (x-v). exceed LARVAL HABITAT: catching in this and other localities over a period of o years, the species appeared in catches for only distinctness of the two taxa. specific gibbinsi Evans BIOLOGY: Maasch (1973) recorded the presence of an infective filarial larva in a single ADULT in coastal Liberia. This appears to be the first record in hancocki. female misin te rpreted Kerr's 1933) findings on the biting cycle in N igeria. H is figures show (1968) rapid a to a well marked peak in the 3rd hour after sunset, followed by a steady decline till sunrise. In rise have normally developed lateral arms to the median plate of the spiracular apparatus specimens thus distinguishing them from material from the type locality, Kenya. These authors (scoop), 93

97 that Angolan forms could be regarded as subspecifically distinct. suggest RECORDS: Angola; 12S, 15E; 15S, 13E (Ribeiro and Ramos, 1975) NEW seydell Edwards The posterior clypeals (setae 4-C) of a larva from Tanzania are simple, IDENTIFICATION: and short, reaching to slightly less than half the distance to the bases of the inner clypeals (2- slender In this it resembles the larva of brohieri. C). RECORDS: Rwanda and Burundi; 3S, 29E; 3S, 30E (Vermylen, 1967). NEW upemba Lips upemba Lips. Riv. Parassit. 21: sp. indet. Mattingly. Parc National de l'upemba. I. Mission G.F. de Witte. 32(3): According to Mattingly, generally resembles hargreavesi, from which it differs IDENTIFICATION: the following characters: in Scutal integument dark throughout, lacking the central greyish area characteristic ofhargreavesi. 1. Scutal scales much smaller than in that species, those on posterior two-thirds ofmedian area dark 2. brown. golden Subapical pale band on palps twice as wide as apical band, which occupies whole of5th segment. 3. dimensions of the pale bands on the palps are quite unlike those of any other Afrotropical The and should enable recognition of this species even though we are ignorant o'f the markings the hind tarsal segments. of specimen was one of only 2 anopheline mosquitoes present in a collection made by the Mis- The G.F. de Witte to the Parc Nacional de l'upemba in Katanga, Zaire, in It was described sion not named by Mattingly who, however, recognised it as a distinct species. The name was subse- but applied by Lips, without any further details being given except for the citation of Mattingly'quently description, in an appendix to one of a series of papers on the anopheline fauna ofzaire. The of this name was overlooked by Gillies and De Meillon (1968). existence has not been possible to locate the type and only specimen of this species. It is no longer to be It section WELLCOMEI wellcomei Theobald and De Meillon (1968) recognised the 3 subspecies, w. wellcomei, w. ugandae and w. erepens. The Gillies named was a very dark form from an isolated area in Tanzania and Kenya. The pale form with last forms lacking this overlay of pale scales predominate (ugandae). However, material southwards, more recently from the western limits of the northern savanna in Senegal (Dr M Cornet) collected from adjacent parts of The Gambia lacks the flavescent scales and has a rather darker wing, and characters diagnostic of subsp, ugandae. This pattern of variation, in which there is a tendency both certain peripheral as well as southern populations to become darker, means that ugandae is a for less satisfactory taxon than appeared when Gillies and De Meillon redefined it. However, the rather ugandae has to be applied to Sene-gambian specimens even though this invalidates to a certain name the picture of distribution previously presented. extent upemba, Stone. Mosq.Syst.Newsl. 1:28. TYPE LOCALITY: 1760 m, Parc National de l'upemba, Zaire. Lusinga, from a single incomplete female lacking fore legs and last 2 hind tarsal segments. Known dark band rather indefinite, occupying slightly less than apical halfof4th segment (fig. Intervening 16,i). in the British Museum (Natural History), and Dr De Coninck kindly informs us that it was found deposited in the Muse Royal de l'afrique Centrale, Tervuren, Belgium. never flavescent palps and proboscis (wellcomei) is typical of the northern savanna while, from Uganda wellcomel wellcomel Theobald BIOLOGY: Krafsur (1971) recorded 47 females negative for sporozoites in western ADULT 94

98 Ethiopia. Brengues et al. RECORDS: Angola, 9S, 13E (Ribeiro and Ramos, 1975). Central African Republic; 8N, NEW (Rickenbach, 1969). 20E wehcornel ugandae Evans RECORDS: We have received a series of females from. Dr A W R McCrae, collected from NEW coastal region of Kenya, 70 km south of Mombasa (4S, 39E). This represents a considerable the to the north-east of the known distribution of the subspecies. To this should be added the extension from Senegal (13N, 16W) and Gambia (13N, 14oW); also Angola; 9S, 16E; 10S, 17E; records 21E; 15S, 19E (Ribeiro and Ramos, 1975). 14S, This very characteristic form, still only known from a single female., differs rather markedly Note. the others in possessing a pale sector spot on the costa and with vein almost entirely pale. The from distinctus Newstead and Carter BIOLOGY: Ribeiro et al. (1981) have recorded larval infections with Coelomomyces LARVAL africana. RECORDS: Angola; lls, 17E; 14S, 21E; 15S, 19E; 15S, 20E; 16S, 22E; 17S, NEW 17S, 19E; 17S, 20E (Ribeiro and Ramos, 1975). 18E; erepens Gillies marked differences from wellcomei s.s. represents a peripheral geographical form ot" wellcomei is no longer tenable. We are, erepens elevating it to species level. accordingly, theileri Edwards There is still some disagreement about the status of this species in the IDENTIFICATION: savanna. Gillies and De Meillon (1968) concluded that records from West Africa probably northern to brohieri. On the basis of captures of adults in Burkina Faso, Gayral etal. (1975) have retirred key, Section II, couplet 15). It follows that, to establish the presence of this species in areas (Adult brohieri is known to occur requires the rearing of wellcomei-like larvae and the demonstration where theileri-like adults emerge from them. As far as we are aware, this demonstration has yet to be that made. (1968) noted the capture of a single female infected with Wu.bancrofti. wellcornel ungujae White wellcomei ungujae White. Mosq.Syst. 7(4): wellcomei subsp, indet. Gillies. Proc.R.ent. Soc.Lond. A 33:11. TYPE LOCALITY: Kibonde Nzungu*, Zanzibar. difference is sufficient to raise doubts as to its conspecificity wellcomei erepens Gillies. Proc.R.ent. Soc.Lond. A 33:9. taxon was described, at the subspecies level, for populations occurring on the north-east This of the known distribution of wellcomei in a semi-arid region of Tanzania and Kenya. Despite border in wing markings and less differences in the egg and male it was provisionally retained by Gillies and De Meillon as a subspecies. With the discovery genitalia, wellcomei ugandae extends across to the East African coast, south of Mombasa, the view that that this view, although without producing new evidence. This seems to overlook the fact challenged the wing markings in brohieri are highly variable, so that dark specimens key out with theileri that classification of theileri and wellcomei in the same group, on the basis of the morphology of the The although not of the adults, has been shown by Green (1982) to be soundly based. In a study of larvae chromosome rearrangements he found that the 2 species had 11 inversions in common and 9 that were not shared. LARVAL BIOLOGY: Ribeiro et al. (1981) have reported the finding of specimens infected with Mermithidae. *Not Kibonde, Nzungu, as cited by White above. 95

99 RECORDS: Rwanda and Burundi; 3S, 30E (Vermylen, 1967). Kenya" 0S, 34E NEW in litt.); Angola; 17S, 18E; 17S, 19E (Ribeiro and Ramos, 1975). The records fi'om (Fontaine, and Burundi and from Kenya (Kisumu) extend the known distribution of the species R;,vanda to the north. A record from Illubabor (8N, 34E), in the western lowlands of Ethiopia, rather 1967), could possibly refer to brohieri. (O'Connor, section DEMEILLONI 29E; ls, 30E; 2S, 29E; 2S, 30E; 3S, 29E (Vermylen, 1967). Angola; 7S, 16E; 7S ls, 9S, 14E; lls, 14E; ]2S, 14E; 12S, 15E; 13S, 13E; 15S, 13E (Ribeiro and Ramos, 21E; 1975). garnhami Evans HABITAT: Recorded breeding in ground pools in total darkness in LARVAL cave on Mount a Kenya at 6000 ft (1800m), (Van Someren and Mutinga, 1971). The wings of the adults Elgon, were and in certain specimens the tarsi were entirely dark. The variable, was.occupied by bats and cave porcupines. 29E; 2S, 29E; 2S 30E (Vermylen, 1967). ls, sergentii macmahoni Evans series PYRETOPHORUS christyi Newstead and Carter TO DISEASE: Rift Valley fever virus has been isolated from RELATION pool of this species a near Nairobi, Kenya (Anon. 1983). collected 1967). (Vermylen, daudi Coluzzi gambiae s.s. the work of Ribbands and Muirhead-Thomson in 944 and 945 on melas in West Africa Following by Muirhead-Thomson 95 on salt-water "gambiae" in East Africa the distinctive nature of and forms of the complex has been generally recognised, although their evolutionary status salt-water not understood until later. Their conclusions were based on behavioural, taxonomic and cross- was mating studies. It fell to Paterson 962), using the principles ofpopulation genetics, to demonstrate specific nature of the East African form, to which the name merus D6nitz was shown to be the Meanwhile, as regards freshwater gambiae the familiar concept continued to prevail of appropriate. consisted of 2 "races", but it was only following the initial studies ofdavidson andjackson ( 962) it Paterson (x963) concluded that these authors were dealing with 2 sibling species, to which that Later a species breeding in heavily mineralised water, sp. D, originally documented by Haddow C. al. (947), was described from Uganda (Davidson and White 972, Hunt 972, Davidson and et demeilloni Evans RECORDS: Ethiopia; 6N, 37E (Rodhain, Boutonnier etal., 1977). Rwanda and Burundi" NEW NEW RECORDS: Ethiopia; 6N, 37E (Rodhain, Boutonnier et al., 1977). Rwanda and Burundi" NEW RECORD: D,jibouti; 12N, 42E (Rodhain, Carteron et al., 1977). NEW RECORDS: Rwanda and Burundi: IS, 29E; ls, 30E; 2S, 29E; 2S, 30E; 3S, 29E M. Coluzzi informs us (in litt.) that he has Prot: melanic mutants in a laboratory colony of seen from Tanzania, which closely resemble the material described under the name daudi. In his view, this taxon is a similar melanic muta'nt of some member of the gambiae complex. THE ANOPHELES GAMBIAE COMPLEX a single, opportunistic species, adapted to a variety ofenvironments. Holstein 952) suggested that Paterson et al. 963) added a third from southern Africa. These were designated as species A, B and Hunt 973). This brought the gambiae complex up to a total of 6 species. 96

100 these names was undertaken by Mattingly (I977) who proposed the names gambiae Giles, of Patton and quadriannulatus Theobald for sp. A, B and C respectively. This action has arabiensis general acceptance and is used here. More recently, sp. D, which is known to occur only in received single locality it, Uganda, has been named bwambae by White (985). One other name has been a forward in recent years quadriannulatus davidsoni Ribeiro et at. (1979) for the form in the Cape put Islands. This taxon, which was described and separated on external morphology only, has Verde from the four species noted above, the status is well established of the two salt-water species, Apart Theobald and merus D6nitz, occurring respedtively on the West coast and on the East coast as melas as in some inland areas of southern and eastern Africa (Paterson et al. 964, Mahon et al. 976, well and Cross x983). Coetzee differences between the five species gambiae, arabiensis, quadriannulatus, melas and merus. The inversion ofpolytene chromosomes in the ovarian nurse cells of half gravid females (Coluzzi, i968 discovery an adequate tool for the identification of all members of the complex (Green 1972, Hunt provided 1973). A more recent technique employed for identification is that of horizontal starch gel I972, (Mahon et al. 1976, Miles 1978, 979). Morphological characters for differentiation electrophoresis of limited value. remain Banding sequences on the X chromosome and autosomes in different karyotypes allow Phylogeny. to be made on the evolutionary relationships of the members of the complex. For inferences et al. 979). Coluzzi recognition. In I964 Paterson published the results of a study at Chirundu, Zambia, Specific-mate that the three freshwater species, gambiae, arabiensis and quadriannulatus, co-existed in nature showing hybridizing and were therefore reproductively isolated.* The production of sterile male without is known to characterise laboratory cross-matings between nearly all members of the hybrids and it was this sterility which lead to the discovery of the complex. Exceptions complex c. x quadriannulatus and ( bwambae x. quadriannulatus (Davidson and White are gambiae I974). Hybridization is readily induced in the laboratory but the occurrence of Davidson I972 gambiae and melas by Bryan (x979): Mahon et al. than o. (976) found an F2 hybrid backcross between readily between some species-pairs than between others. more are ignorant of the precopulatory mechanisms involved in maintaining positive assortative We has gone into the search for morphological characters that might enable identification to effort of individual specimens of the complex. Reviews have been published by Davidson et al. species the view of one of us (MC) the term "reproductive isolation" should be avoided as this subscribes to a relational concept *In species. Reproductive isolation is an incidental consequence of positive assortative mating induced by specific fertilization of The question offormal names was discussed by Paterson 968) and White (1975), and the fixing now been shown by Cambournac et al. (1982) to have polytene chromosomes typical of arabiensis. of the different members of the complex is now primarily based on differences in Separation patterns of polytene chromosomes. Initial studies by Coluzzi and Sabatini (967, i968a banding I969) on the chromosomes found in the salivary glands of fourth stage larvae showed fixed I968b discussion of this matter the reader is referred to Coluzzi and Sabatini (1969), White (i973) and in nature is a rare event. Individual gambiae/arabiensis hybrids have been recorded by hybrids Ramsdale and Leport (I967) White (97xa), White, Magayuka and Boreham (972), Coluzzi in Molineaux and Gramiccia (198o), between arabiensis and merus by Mahon et al. 1976) and between gambiae and arabiensis. Coluzzi and Kitzmiller ( 975) estimated the occurrence of hybrids as less % in natural populations. According to Okereke (198o), crosses are made in the laboratory but they presumably involve a spatial element. In the laboratory the time of onset of flight mating, at dusk, and hence of availability for mating, was found to vary slightly between member activity But only in the cases of merus was the difference great enough (2o min.) to be of possible species. ecologically (Jones et al., i974). significance differentiation. The ultrastructure of the pharynx of gambiae s.l. has been studied by Morphological et al. (198I). The work includes excellent scanning electron micrographs. A great deal of Coluzzi mechanisms (Paterson, 1985). 97

101 White (x974) Green (x97x) and Ribeiro (x98o). Two main difficulties (I967) to hamper seem in this direction. One is the unreliability of colony-reared material (Chauvet, Davidson progress Dejardin, I969), the other, geographical variation within the and species (Reid, x973). The same characters have been investigated in the female adult: following banding, Shown to be a useful character for the determination of bwambae (q.v.). Little Palpal can be attached to the presence of4-banded palps in freshwater gambiae (Davidson et al. 967). value Boreham 97). See also under melas (p. xs). Paterson et al. (x963) reported the presence of 4- and palps in quadriannulatus as well. banded ratios. Valid for the partial separation ofmelas from freshwater species (Gillies and Wilkes Palpal Bryan 98o) and for merus likewise (Bushrod, 98 x). 969, 968) showed differences in the number of between salt-water breeding and the freshwater breeding species but could not reliably sensilla the freshwater species from each other. separate Wing markings. Zahar et al. (x 97 o) used certain wing measurements but could only separate 19. Hind tarsal segments to show the differences in pale banding between Fig. gambiae, arabiensis and b- merus and a- in Southern Africa. quadriannulatus bandings (fig. 19). Shown by Coetzee et al. (1982) to enable separation of southern African Leg of the vector species gambiae and arabiensis from quadriannulatus and populations merus. Generally larger in gambiae Spermatheca. than in arabiensis or merus but not a reliable s.s. for individual specimens (Clarke 1971, White and Muniss 1972, Eyraud et al. 1973). character (x97x) showed that quadriannulatus was intermediate between gambiae and arabiensis which Green make the ratio of spermatheca diameter:wing length (Clarke, x97x unreliable where these would and Dejardin x968 Chauvet, Davidson and Dejardin x969, Zahar et al. x97o Green (Chauvet I, Reid x973, Ribeiro x98o). Good separation can be obtained between some populations of 97 and arabiensis (see fig. I8 in Davidson et al. x967, for example) but not between others. gambiae the whole, it seems that while characters that are locally valid may exist such On hind leg as Gillies and De Meillon 968) stated that in a part of the East African coastal lowlands, where only s.s. occurred, females with 4-banded palps were absent. However, later work has shown that gambiae palps are equally common in both gambiae and arabiensis (eg. White 97a, White, Magayuka such Antennal coeloconic sensilla. Ismail and Hammoud gambiae from arabiensis at one locality. a 0'5 occur together. species Various setal counts and the shape of the male genital lobe give partial separation of Pupa. and arabiensis (Reid 975a, b, Ribeiro x98o ). gambiae Numerous setal counts, especially branching of the inner shoulder hair (seta x-p) Larva. banding patterns, no morphological characters that can be universally applied have yet been found. 98

102 taxonomy. Following the earlier studies of Micks et al. (1966), enzyme variation in members Chemical the gambiae complex has been reported. Mahon et al. (I 976) showed that in southern Africa of s.s., arabiensis, quadriannulatus and merus could all be separated by plotting distribution gambiae for 16 allozymes. The gene frequencies of diagnostic allozymes in gambiae frequencies and s.s. may vary geographically so these frequencies should be checked in each area against arabiensis identified specimens before total reliance is placed on electrophoresis for chromosomally Separation of the other species may be more straightforward, and Miles 1978, 1979) identification. able to define quadriannulatus, bwambae, melas and merus in terms of species-specific isozyme was hydrocarbons could be utilised in the same way (Carlson and Service I979, 198o cuticular and Service i983). Hamilton The practical importance of correct identification of wild material cannot be Cytogenetics. enough. Green 1981 cites a classic case ofhow confusion can arise when a mixture of emphasized or more cryptic species occur together in nature and are not treated as separate entities but as a two population. When WHO tests for dieldrin resistance were done on wild samples from three single in Zimbabwe, confusing re.sults were obtained with 98% survivalrate at one site, o.8% at localities second site and 43% at the third. However, when the specimens were identified cytogenetically, the of the dead mosquitoes were shown to be quadriannulatus which is not a vector of Plasmodium. most of the survivors were identified as arabiensis and were therefore resistant to dieldrin and a Some campaign with BHC (which exhibits cross resistance with dieldrin) by health authorities spraying have had no effect on disease control. would results of cytogenetic studies lie outside the scope of the present work, and the reader is The to Coluzzi et al. (I 979) for a full account. We restrict ourselves here to a summary of the referred aspects of the subject. main one of the most important aspects of work on African in recent years has been Perhaps the presence of fixed or floating inversions. Studies along these lines have been published by and (1973b), Coluzzi and Sabatini (1967, 1968a, 1968b, 1969), Curtis and Chalkley (1979), Akiyama and Hunt (1973), Davidson and White (1972), Green (1972, 1980) and Hunt (1972). Davidson et al. (1977) published the techniques for fluorescent banding of mitotic chromosomes. The Gatti objective is to attempt to correlate inversion frequencies with particular aspects of second or behaviour. The aim here is to increase our understanding of the interaction of the physiology environment and the biology of vectors and so enable us to interpret the impact of control physical with far greater precision than has been possible in the past. The importance of this in measures important example is provided by the detailed studies in Nigeria of Coluzzi et al. (1977, One carried out over a series of seasons which showed that certain inversion frequencies in both 1979) gambiae s.s. the south to the Sudan and Sahel savannas in the north. Within the same geographical zone in genetic polymorphism can be shown, according to the sampling procedure adopted. Thus, further karyotypes in arabiensis are correlated with biting man indoors and resting indoors. certain the same chromosome arrangements are those correlated with drier environments. This Moreover, that the population is non-random in its response to house-resting, and hence to demonstration to insecticide, is held to be the key to the failure of malaria control schemes in this zone exposure and Gramiccia, 1980). In the rather less extreme climate of Kaduna, Di Deco et al. (Molineaux showed that karyotypes in gambiae s.s. that were known to be associated with dry climates (1981) in frequency seasonally, being commoner in the dry season. In the Gambia, Bryan et al. varied found cytogenetic evidence thought to indicate the existence of two populations ofgambiae (1982a) and to combine these into a key. Oketch and White (x98i) give further details on the patterns of merus from other members of the complex. A different approach showed that the separation the collection of cytogenetic data on field populations and laboratory colonies of the gambiae These studies have had two main objectives. The first is the routine mapping of polytene complex. banding sequences and the characterisation of species in terms of banding patterns chromosome malaria control strategies can hardly be exaggerated, once the method.is adequately planning with data. supported and arabiensis change in a clinal fashion in samples taken successively from the forest belt 99

103 partial reproductive isolation from each other. It is not possible at the present time to showing assess significance of this finding, but it seems likely that they are dealing with two species. the other genetic studies have been published and the majority of these are concerned with Few relationships and genetic mapping (see Narang and Seawright, 1982). The genes for DDT linkage dieldrin resistance have been shown to be in linkage group II (Haridi, 1974) and just recently, and for dieldrin resistance more precisely on chromosome arm 3 (Hunt, 1987) (arm 2L in Coluzzi those Sabatini 1967, but see Green and Hunt 1980, for an explanation of the new arm designations). and differences within the complex. A summary of recent contributions to our knowledge ofadult Behavioural and larval biology in "freshwater gambiae" is given in the next section. Briefly, the behaviour between gambiae s.s. and arabiensis are as follo.ws. differences species are highly anthropophilic but, when alternative mammalian hosts are available, Both shows a much greater tendency to feed on animals. Since domestic animals in many areas arabiensis kept outside or in flimsy shelters at night, it is not surprising that arabiensis shows a greater are to feed and rest outdoors.. The effect of house-spraying on populations of this species is tendency reduced in comparison with gambiae s.s., e.g. in the Kisumu area of Kenya (Service et accordingly 1978). In regions where only the latter species occurs, house-spraying may be highly effective al., 1974). While both species are important vectors, the lesser degree ofman-mosquito contact (White, explains the much lower sporozoite rate frequently encountered in arabiensis compared partially gambiae s.s. (see Table XIV). White suggested that a further difference between the species may with in their respective survival rates. The main evidence for this comes from the markedly lower lie of natural populations of (mainly) arabiensis in an inland area oftanzania compared with longevity the evidence would need to come from an area where the two species shared the same however, environment. major differences in the larval biology of the two species have been discovered. However, No certain conditions arabiensis is said to proliferate faster than gambiae s.s. (White, Magayuka under and Boreham, 1972). Moreover in northern Nigeria, gambiae s.s. season while populations ofarabiensis build up gradually as the country dries up during the dry rainy (Di Deco et al., 1981). season (figs 20-23*). While gambiae s.s. and arabiensis co-exist widely over much of the Distribution. two trends are apparent. The first is the evident absence ofgambiae s.s. from all areas to continent, north-east of Uganda and the Kenya Highlands, whereas in West Africa its range extends well the the Sahel in Mauretania. The second is the general absence ofarabiensis from the forest belt and into itself in towns in southern Nigeria within the forest belt, but is evidently unable to extend establish there into the surrounding countryside (fig' 24). Both these anomalies would frdm to rule out seem simple climate-based explanations for the differences in distribution. To test for the existence of any effects Coz (1973c) maintained mixed laboratory coloniesofwest African strains ofgambiae s.s. such arabiensis and subjected them to alternative conditions of humidity over periods of up to 12 and A constant humidity of 80-90% R.H. in 3 experiments lead to the complete replacement of months. by gambiae. Exposure to seasonally low R.H., on the other hand, lead to conflicting results, arabiensis or neither species becoming dominant in the colonies. either quadriannulatus appears to have a very disjunct distribution although doubtless other populations in the equatorial zone remain to be discovered. White (1974) has pointed out localised tolerance of this species to relatively cool conditions on highland plateaux, in contrast to the the members of the group. other South African, Namibian and Botswanan distribution records of members of the gambiae complex given here differ *The the maps of Davidson and Lane (in White, 1985) in that only published data have been included. from a coastal population assumed to be almost exclusively gambiae s.s. We feel that to be conclusive, dominates the catches during the more humid areas of West Africa. Coluzzi et al. (1979) have shown that this species can nevertheless 100

104 20. Distribution of gambiae s.s. from Davidson and Lane (1981); additional records from Coz (1973b). Fig. 101

105 Distribution ofarabiotsis from Davidson and Lane (1981), with additional recordsfrom Coz (1973b), Green (1970), Fi21. and Subra (1982), Ribeiro and Ramos (1975), Coluzzi et al, (1979). The interrupted line incorporates the putative Mosha northern limit in the Nile Valley (Unpub. WHO reports) 102

106 22. Distribution ol'melas and merus. From Davidson and Lane (1981); additional records from Gillies and De Meilh:m Fig. Mosha and Subra (1982), Cross and Thcron (1983). (1968), 103

107 Fig. 23. Distribution ot'quadriannulatus and bwambae from Davidson and Lane (1981), White (1973). 104

108 key section IV, larval key section VII. Adult Van Someren (1969) described 4 albinoid females ofgambiae s.l. from Mombasa. VARIATION: degree of pale scaling was variable, but none reached the level ofalbinoism previously reported The Nigeria by Service (see Gillies and De Meillon, 1968). White and Davidson (1972) report the from of black pigment below the dorsal thoracic and abdominal integuments. McIver and peppering (1975) give an account of the sensilla occurring on the palps ofgambiae s.l. Hinton (1968) Siemicki surface structures of the egg as seen under the scanning electron microscope. figured garnbae Giles gambiae Giles, A Handbook of t/e C;nats or Mosquitoes, 2nd edition: gambiae group A, Davidson & Jackson, Bull. Wld Hlth Org. 27: gambiae species A, Paterson, S.Afr.J.Med.Sci. 28: gambiae (restricted), Mattingly, Mosq. Syst. 9: LOCALITY: Georgetown, The Gambia. TYPE arblensls Patton arabiensis Patton, at. Bombay nat.hist.soc. 14: costalis Edwards, Bull.ent.Res. 3: gambiae group B, Davidson & Jackson, Bull. Wld Hlth Org. 27: gambiae species B, Paterson, S.Afr.J.Med.Sci. 28: arabiensis Mattingly, Mosq.Syst. 9: quadriannulatus davidsoni Ribeiro, Ramos, Pires & Capela, Garcia de Orta.Ser.Zool. 8: quadrlannulatus (Theobald) Pyretophorus quadriannulatus Theobald, Union of S.A. Division of Veterinary Research, First Report, Pretoria, p244. Govt.Printer, costalis var. quadriannulatus, Evans, Mem.Lpool Sch.trop.Med. (new series), no. 3: gambiae, Evans, Mosquitoes of the Ethiopian Region. II. Brit.Mus. (Nat.Hist.): gambiae species C, Paterson, Paterson & Van Eeden, Med.Proc.(Med.Bydraes) 9: BIOLOGY OF FRESHWATER SPECIES OF THE ANOPHELES GAMBIAE ADULT COMPLEX rhythms. Detailed studies in the laboratory on the pattern of spontaneous flight activity in Circadian s.s. have been made by Jones et al. (1972a, 1972b), Jones (1973), Jones and Gubbins (1977, gambiae Thus, virgin females show intense activity at dusk at a time when mating takes place, while nature. but still unfed females are more active later in the night when host-seeking and blood- inseminated are the dominant activities. feeding activity was found to continue under conditions of constant darkness, there being slight Cyclical in the period of the cycle between different members of the gambiae complex; in gambiae differences the cycle was approximately 0.5h shorter than in arabiensis, and in merus about the same time or s.s. THE FRESHWATER SPECIES IDENTIFICATION: of partially melanic larvae ofgambiae s.l. in northern Nigeria, in which there was a general capture TYPE LOCALITY: Sheik Othman to Dthala, South Yemen quadriannulatus, Mattingly, Mosq. Syst. 9:325. TYPE LOCALITY: Onderstepoort, Trangvaal, South Africa. This work has shown that activity is bimodal with an initial concentrated peak soon after the 1978). from light to dark and with a secondary more sustained increase in activity during the transition and later periods of darkness. The initial peak is greatly reduced in inseminated females. middle laboratory findings on isolated females are held to underlie the pattern ofactivity observed in These longer (Jones et al., 1974). 105

109 Distribution ofarabiensis and gambiae in and around Benin City, Nigeria The proportion ofgambiae in each locality is Fig. in black. The stippled zones indicate the approximate limits of the urban and peri-urban areas (ti'om Coluzzi et al., shown 1979). 106

110 timing of pupation is controlled by an internal 'clock' set by the time of sunset (light-off) The larval development (Jones and Reiter, 1975). On the other hand, the timing of adult during light conditions (Reiter and Jones, 1975). altered andfertilization. Swarming and mating in the gambiae complex are preceded by erection of the Mating of the male antennae. In the laboratory, this occurs during the hour before light-off and fibrillae for 1-2h aft8rwards (Charlwood andjones, 1979). This means that detection offemales and remains 1980). The existence of markers in the field has not yet been reported. In the West African Jones, Brengues and Coz (1973a) found that mating occurs in less than 20% offemales before they savanna, their first feed. In the remainder it occurs 24h later. Mating plugs were present in 74% and 50% take fed nullipars, depending on the degree of development of the ovaries. They also suggested that of gambiae s.l. might mate at dawn following their first feed earlier in the night. For the many reasons given above, this would not seem to be possible. In the laboratory, gambiae s.l. physiological sexual activity was dependent on the presence of functional accessory glands in the male further 1972), which was also necessary tbr partial suppression of flight activity in recently fertilized (Bryan, (Jones and Gubbins, 1978). females pattern. In western Kenya, light-traps used under the eaves of a house suggested that flight Flight commences later on (Chandler, Highton and Hill, 1976). In West Africa, the height above activity ground at which females disperse across the open savanna has been studied in two areas in The the (Gillies and Wilkes 1976, Snow 1979). This work has shown that most females are flying Gambia to the ground with more than 50% and 80% in 2 different investigations being caught at a level close to lower the flight level, but later in the night flight level was always nearer the ground even tended wind speed was normally lower at this hour. Despite this tendency to stick close to the though it was found that a 6m-high fence was no barrier to the dispersal offemales and provided no ground, for a bait in the centre of a.fenced-off circle with a radius of 65m (Gillies and Wilkes, protection 1978). and outdoorfeeding. In Garki, northern Nigeria, human bait-catches ofgambiae s.l. were much Indoor same indoors and outdoors (Molineaux and Gramiccia, 1980). However, in the case ofarabiensis, the fact that chromosome inversion frequencies were different in the 2 populations (Coluzzi et.al., the is of interest in possibly suggesting genetic involvement in the mediation of this behaviour. 1979), in houses. After the departure of the day-resting population from houses at dusk in Movements Nigeria, Rishikesh and Rosen (1976) noted a high turnover of gambiaes.l., that were northern and leaving during the course ofthe night. All gonotrophic stages were involved. At Garki, entering of those leaving during the night were unfed or partly fed, and were thought to be leaving to most the search for blood (Molineaux and Gramiccia, 1980). It is known that populations in the continue African savanna often tend to show a marked tendency to fly out of.fiouses after feeding. In West Faso, for instance, Coz (1971) and Brengues and Coz (1973a) recorded more than 33% of Burkina s.s. leaving houses after feeding. Of those that stayed on through the following day, 40% left gambiae dusk. at laboratory experiments, Gerold (1977) showed that wild populations were polymorphic for In and departure from houses, females that took off soonest being also the first to be caught activation in exit traps. After contact with DDT, gambiae s.s. resting. Laboratorystudies have shown that gambiae s.l. selects a rough in preference to a Indoor surface for resting on, and that this choice is dependent on intact hind legs (Hansell, smooth-texture In houses in Burkina Faso, the majority of females (91% or 71%, depending on the type of 1970). were resting on the roofs (Brun, 1973). A deficit of gravid females resting in houses by construction) is dependent on pupation time, although some modification of this is possible under emergence is only possible during this short period. In a cage, males will swarm over a black marker on mating floor; in contrast to this, females will make brief single flights over the marker (Charlwood and the and arabiensis have been found to mate once only (Goma 1963, Bryan 1.968). The inhibition of ofgambiae s.l: in the early part of the night is followed by quiescence before the main biting activity of m or below. Negligible numbers were flying at 6m and above. In general, higher wind speeds remained flying twice as long as arabiensis. 107

111 but the actual level varies considerably, e.g. in Burkina Faso fed" gravid ratios in one d:.irnrmal, were 1.5" (Brengues and Coz, 1.973a), in another 3" (Brun, 1973), and in south Benin as much area on man are followed by resting indoors (Molineaux and Gramiccia, 1980). The converse taken where feeds on animals are followed by entry into houses for resting, is well known. In situation, Kisumu, Kenya, Joshi et al. (1973) produced evidence that this movement took place late in the or at dawn leading to the presence of large numbers of live, cattle-fed females in houses even night treated with insecticide. In northern Niger.ia, Coluzzi et al. (1979) showed that the frequency when certain chromosome inversion types in the house-resting population of arabiensis differed from of in outdoor resting or biting females. Since the same chromosome arrangements were also more that in savanna populations than in humid regions, Coluzzi postulated that the behavioural frequent involved in entering houses at night might be to higher temperature and hence greater response deficit of the indoor environment. saturation resting. In the absence of human sleepers or collections of animal hosts outside at night it is Outdoor to find few fed females in outdoor shelters, e.g. in Burkina Faso (Brengues and Coz, 1973a), usual most of the catch was of unfed, mainly newly emerged females. The finding by Brun (1973) in where same country of40% of females in artificial shelters placed in the centreofa village being freshly the is exceptional. Where a high degree of resting by engorged females in natural shelters, such as fed in river banks, was observed in the Sudan (Haridi 1972, Akiyama 1973a), the effect of DDTtreated holes surfaces in houses was held to be responsible. At Kisumu, Kenya, Service (1970c),Joshi et al. and Service et al. (1978) found that arabiensis was 2-3 times commoner in pit-shelters and (1975) stores than gambiae s.s., this effect being increased after house-spraying. Further studies by grain with 42% of arabiensis coming from structures outside sleeping huts. compared cycle. The well-known late-biting cycle of gambiae s.l. has been confirmed in a number of Biting that differences in biting cycle between member-species of the complex are not great. In suggests Province, Kenya, gambiae s.l. was active round the eaves of houses early in the night, as Nyanza by light-traps, although biting activity was mainly delayed until 3h after sunset and shown followed the normal pattern (Chandler, Highton and Hill, 1975, 1976, Chandler 1977). thereafter other species of the complex, biting by quadriannulatus in Ethiopia reaches a peak before Unlike and tails away to dawn (White, 1974). White points out that this is correlated with low midnight Carnevale et al. (1979) have proposed a revised system of classification of ovarian microscopy. defined in terms of the degree of filling (Degr6 de Remplissage or DR) of the follicle development, yolk granules. The system enables mathematical analysis ofthe growth of the follicle, and by its with they were able to show that, in gambiae s.s., the DR is highly correlated with the growth of the use and can, therefore, be reliably used in the study of factors influencing development of the follicle In the West African savanna there seems to be general agreement that the first gonotrophic ovaries. lasts 3-4 days, subsequent cycles (from blood meal to blood meal) 2-3 days (Brengues and Coz cycle Brun 1973 in Burkina Faso, Molineaux and Gramiccia, 1980 in northern Nigeria). Brengues 1973a, Coz concluded that the first blood meal is taken 24h after emergence, and that 42% of females and a second blood meal in order to mature the first egg batch. After oviposition, 2/3 offemales require again the same night (Brengues and Coz, loc.cit.), 95% of them according to Brun, or 'nearly feed according to Molineaux and Gramiccia (1980). In the more humid conditions ofdegraded forest all' the Congo Republic, generally similar results were obtained in a detailed study, including the use in mark/release methods (Carnevale et al., 1979). In this region, 70-75% of newly emerged females. of 2 blood meals in the course of the first cycle, which lasted 3-4 days. The 2 feeds were usually took on consecutive nights. After oviposition 88% of females returned to feed again the same night. taken 28:1 (Brengues et al., 1969). In northern Nigeria, it was estimated that about half the blood meals as Clarke et al. (1980) showed an even greater difference between the 2 species, only 6% ofgambiae, studies: Liberia, both savanna and forest (Kuhlow and Zielke, 1978), Central African Republic and Geoffroy, 1976), Burkina Faso, in a relict.patch of forest (Gayral et al., 1975), (Cordellier Nigeria (Molineaux and Gramiccia, 1980), Ethiopia (White, 1974). This uniformity northern temperatures in highland areas. night cycle. The process of oogenesis in gambiae s.l. was described by Fiil (1976) using electron Gonotrophic 108

112 duration of the interval between successive cycles was 2-4 days. Carnevale and Molinier Mean combined the 2 parameters interval between blood meal and oviposition and the (1980) to refeeding interval into a single biting rhythm (rhythm des piqures), which oviposition the proportion of a parous population feeding per day. In the case ofgambiae s.s. in the describes Congo Republic the value was ob oo o o o o o 6 d d d d d d 6 d Ob oo :o o. o d d d d d d 6 d d d d d d d d do TABLE XI: Human blood index (HBI) of engorged females of members of the gambiae" complex. 109

113 selection. Particular attention has been paid to the existence of differences in feeding behaviour Host gambiae s.s. and arabiensis. Most studies, summarised in Table XI, have tended to show that betwen human blood index (HBI) is lower in arabiensis. If cattle are present in any numbers, the high a of feeds by arabiensis are taken on the animal host. On the other hand, it is only in proportion samples, such as outdoor resting sites or in compounds with abundant cattle that evidence selective obtained of extensive animal feeding by gambiae s.s. (Service, 1970c). Conversely, where cattle is are came to calf-bait in relatively greater numbers than to human baits as compared with arabiensis s.s. In Madagascar, this difference in behaviour was even greater (Chauvet, 1969). In gambiae northern Nigeria, 31% of the catch on man of the gambiae complex was gambiae s.s. caught on donkeys, the rest being arabiensis (Molineaux and Gramiccia, 1980). Service those recorded the interesting finding of 2 females of gambiae s.s. that had fed (1970b) bats. on in the attraction of different persons has also been studied. By direct catches off the Variation of sleeping subjects Carnevale, Frezil et al. (1978) in the Congo Republic showed that gambiae legs reported by Boreham et al. of females were found to have fed on more than one subject in the course of obtaining 2.7% single a meal. Bryan and Smalley (1978), using ABO blood groups as markers, showed that small full sleeping under the same, inperfect bed-net received 1/7 of bites. Port et al. (1980), also children in The Gambia, demonstrated that size of individual host was the determining factor. working has been little recent experimental work in the laboratory on host-selection, but Aboul- There (1969a) showed that gambiae s.s. females, that were attracted to a human host or a calf on first Nasr into experimental chambers, fed at random on either host after being released for a second release Field studies by White, Magayuka and Boreham (1972) showed that there was no difference in time. confirming that individual females of both species do not. become conditioned to feed on one thus of host. Using a flight-tunnel, Omer (1979) found that, while arabiensis responded to changes in type of carbon dioxide by flying up-wind, the response to odour from a human hand had concentration a effect. In the field, Snow (1970) found that the removal of 95% of carbon dioxide from greater breath reduced the number of attacking gambiae s.s. and melas by 54%. expired zoophilic habits of quadriannulatus have been confirmed in Ethiopia by White et al. (1980) The shelters and outdoor resting sites, from 5-50:1 in mixed habitations, while in houses where animal people slept only arabiensis was ever found. However, Hunt and Mahon (1986) report that only has been collected in houses occupied solely by humans at three localities in quadriannulatus and one in South Africa. Sharp et al. (1984) report the occurrence ofquadriannulatus in Zimbabwe in northern Zululand, South Africa. Blood meal analysis on a sample of indoor resting gambiae huts from the same locality showed 96% to have fed on cattle. s.l. A limited number of studies on the age composition ofgambiae s.1. populations have been Longevity. mainly in terms ofthe proportion parous. These results are set out in Table XII. White (1974) made, discussed the important question of whethergambiae s.s. shows greater inherent longevity than has Brengues and Coz (1973b) found that mortality in cages was lower in gambiae s.s. apart from White's findings in Kisumu, evidence is scanty and it may be noted that high However, rates were estimated for arabiensis by Garrett-Jones and Shidrawi (1969) in Nigeria. In our survival the parous rate by itself cannot be relied upon to give accurate estimates of longevity, and as opinion above, the evidence provided by the application of Polovodova's method in Tanzania is not noted In Madagascar,. the mark/release/recapture experiments of Chauvet (1969) showed a conclusive. superiority in the survivorship ofmarked arabiensis over gambiae s.s. But, as the author clear-cut out himself, the exceptionally low values of the parous rate in both species suggested that pointed sampling bias may have been involved. Less degrees of bias have been reported, for example, some scarce, even arabiensis is closely associated with man as was found by Krafsur (1971) in Ethiopia. Similar results are obtained with alternative hosts. In Burkina Faso, Coz (1973b) recorded that but only 9% of s.s. fed on infants, children, adolescents and adults in the ratios of 1"2:2.5:3. Similar results were (1978) in Kisumu, using haptoglobins as markers. In their experiments incidence of sporozoites in cattle-fed and man-fed specimens of either gambiae s.s. or arabiensis, the and Boreham (1975). White et al. found that the ratio arabiensis: quadriannulatus approximated 1" in arabiensis. The parous rates at Kisumu shown in Table XII suggest that this was the case, and than in arabiensis. 110

114 Selfand Pant (1968) in Nigeria, who found that nullipars were much commoner in exit traps in by than resting indoors. houses d d d 6 d d V d d d d d d d TABLE XlI: Records of parous rates and estimated daily survivorship in freshwater gambiae. 111

115 only comprehensive published record of the use ofpolovodova's technique is that ofaboul- The (1969b) in Enugu, eastern Nigeria. His extensive results showed much lower survivalrates Nasr had been previously recorded in East Africa by Detinova and Gillies (1964) and Gillies and than (1965). In particular, he identified no females of gambiae s.l. (probably mainly gambiae s.s.) Wilkes than 5-parous. At the same time, he recorded exceptionally high monthly sporozoite rates older a year-round mean of 15.4%. It is difficult to reconcile these two conflicting findings. Clements with Paterson (1981) have reanalysed published data on the survivorship of a number of vectors. and conclude that the deficit of nullipars in the samples ofgambiae s.s. and arabiensis in Tanzania, They in both species increased regularly with increasing age. mortality Following the laboratory studies of Haddow and Ssenkubuge (1962) it has been Oviposition. to refer to the relationship between time and oviposition activity as the oviposition cycle, customary is characterised by the bulk of the eggs being laid in the first 3h of the night. This concept has which incorporated into the cyclical model of activity propounded byjones and Gubbins (1978). In been a detailed field study in Kenya on (mainly) gambiae s.s. McCrae (1983) has challenged this recent, meal, and he concluded that time of feeding and temperature were the main determinants of blood oviposition "cycle", not an endogenous rhythm. In contrast to McCrae's findings, it is the to note that Brun (1973) in Burkina Faso, also using wild-caught, freshly fed females that interesting set up for oviposition in the field, found a marked peak of oviposition in the early part of the were sites by (mainly) gambiae s.s. McCrae (t984) found that black targets were more oviposition than white ones. He also noted that, with white targets, eggs were laid when in attractive settled a whereas with dark surfaces they oviposited in flight. Turbid water from a natural breeding position, was more attractive than swamp or tap water. The presence of larvae, but not of pupae, was site Hour. till o XIII: Time of oviposition of 37 female gambiae s.l. and 46 funestus.collected when freshly fed in TABLE Faso, Brun (973). Mean temperature (max + min/2) C. Burkina arabiensis, by the findings ofomer and Cloudesley-Thomson (1968, 1970) in the Sudan. presumably searches showed that very small numbers of fed females could be found in a waterless Exhaustive throughout the 8 months-long dry season, resting in the thatch ofhouses and, to a lesser extent, area dry wells and animal burrows. All females were nulliparous and, in the early months, their in were all in an early stage of development. As the season progressed so did the growth of the ovaries so that by June and July only half-gravid to gravid females were caught. Thus, survival ovaries, was by a form of gonotrophic discordance which, apart from permitting blood-sucking for the achieved of nutrition and Water balance as well as for reproduction, also inhibited search flights for purposes sites. Since normal gonotrophicactivity was observed to continue right through the dry oviposition in villages along the river Nile, where breeding sites were constantly available, and since the season was equally arid in both, it is not easy to see what environmental factors induced climate discordance in neighbouring waterless areas. gonotrophic by Gillies and Wilkes, was more apparent than real. They also showed that the level of studied and pointed out that these laboratory studies took no account of the time of feeding. interpretation females that were caught when freshly fed in the wild and maintained at ambient Using he found that oviposition in cages showed a build-up from dusk to peak in the hour temperatures, In re-feeding experiments he showed that oviposition occurred 44-47h after the following.midnight. In other words, his field results matched those ofhaddow and Ssenkubuge in the laboratory, night. of females laying in the 2h after sunset (see Table XIII). In laboratory tests 64% choice of on repellent. -o6 8.o 8.o o 8.o x.4 gambiae funestus Aestivation. Considerable light has been shed on the problem of dry season survival ofgambiae s.l., 112

116 feeding. Laarman (1968) recorded the presence of glucose, thought to be derived from plant Plant juices, in wild-caught females of gambiae s.l. in Kenya. TABLE XIV: Sporozoite rates in the gambiae complex. Range of monthly values given in parentheses. 113

117 to malaria. Values for the sporozoite rate recorded in different studies are shown in Table Relationship As noted by earlier authors (White 1974, Molineaux and Gramiccia 1980), infection rates XIV. tend to be higher in gambiae s.s. cattle are absent from villages this difference.may disappear. White (loc. cit.) suggested that Where additional factor may be that the longevity ofarabiensis in nature is less than that ofgambiae. The an of values of the sporozoite rate obtained in different studies is very wide. As White, Magayuka range Boreham (1972) pointed out, the greater availability of antimalarials in some districts than in and may depress the level of infectivity of the human population. Conversely, asjoshi et al. (1975) others the identification of chromosomes requires the dissection of females with ovaries at noted, st. III-IV, so that the youngest.age-group, those taking their first feed will be absent Christophers the sample. The sporozoite rate of the identified sample tends to be greater, therefore, than in from feeding population as a whole, as these authors' data showed. the No difference in the infectivity of gametocyte carriers to gambiae s.s. Djibouti, there has been some doubt as to whether the sporadic cases of malaria occurring In were imported or autochthonous. Carteron et al. (1978) concluded that they were transmitted there by arabiensis even though the recent finding of the species, by Rodhain, Carteron et al. (1977), locally only the second record of the complex from that country since the beginning of the century. was As in the case of malaria transmission, the relative importance ofgambiae s.s. and arabiensis Africa. as would primarily seem to reflect their differing degrees of contact with man. Brengues and vectors (1973b)showed that the uptake of microfilariae by the 2 species was generally similar but Coz in nature according to the size of blood meal. Brengues and Bain (1973) gave a detailed varied of the uptake ofmicrofilariae by gambiae s.s., while Brunhes and Brunhes (1973) described account histological changes in infected arbiensis. The final act of transference of infective larvae from the the transmission of filariasis in the West African savanna. In Burkina Faso, gambiae s.l. was reviewed for about half the transmission. Mean infectivity was 1.8% offemales and it contributed responsible estimated 56.1 infective bites p.a. Transmission by gambiae was mainly during the rainy season, an being responsible for dry season infections. A comparison of transmission in savanna and funestus villages of northern Liberia showed that it was much greater in the former area (Kuhlow and forest 1978). They recorded the infectivity ofgambiae s.s. as 1.2% and 1.6% in the savanna villages Zielke, 1.1% and 0.7% in the forest. In coastal Liberia seasonal infectivity in gambiae s.1. (including and ranged from % (Maasch, 1973). In western Ethiopia 1.24% ofarabiensis had infective melas (McConnell and Schmidt, 1973). Transmission in East Africa has been reviewed by White larvae and McMahon et al. (1981) and investigated experimentally by Crans (1973). White found (1971b) rates of 0.45% in an inland area of Tanzania, compared with 1.27% nearer the coast. infectivity et al. found rates in gambiae s.l. of % in spraycatches in houses in 4 coastal McMahon compared with % in night-biting.catches. The difference in the 2 samples villages reflected differences in age composition and possibly the greater tendency of merus to presumably Kenya (Wijers and Kiilu, 1977). coastal isolations from the gambiae complex are listed in Table XV. Apart from these, Prince et al. Virus recorded the isolation of hepatitis-b antigen from gambiae s.l. in East Africa. The role of (1972) transmission of this virus is still far from established, and it may be noted that, on mosquito grounds, Baylet et al. (1974) concluded that, in Senegal, gambiae s.l. was not epidemiological White (1974) makes the interesting point that the greater zoophilic tendencies ofarabiensis involved. than arabiensis in line with the greater zoophily of the latter species. and arabiensis was found by Chauvet et al. (1972) in Madagascar. On the other hand, Coz and Picq (1973) presented evidence that West African strains of gambiae s,s. were more susceptible than arabiensis. Other diseases. In recent years Bancroftian filariasis has been studied in detail in several regions of s.s. to the human host was described by Zielke (1975"). gambiae et al. (1968), Lamontellerie (1972), Brengues et al. (1975, bis), Brengues (1975) have Brengues outside by day. A similar figure of 1.1% of gambiae s.1. with infective larvae was reported from rest may facilitate the transfer of viral zoonoses to man. He also speculates that the rapid spread of fever in East Africa during the great epidemic of 1959 was the result of the interplay O'nyong-nyong 3 anopheline vectors with varying degrees of zoophily. of 114

118 Ethiopia Western Kenya Western Sindbis Germiston s.l. Br6s et al. 969) gambiae s.lo Lee & Moore (x97) gambiae s.l. Rickenbach et al. (I976) gambiae Ota et al. 976) arabiensis It is only relatively recently that the role of Emesine bugs as predators of house-haunting Predators. has been suspected. No quantitative information is available as yet, but evidence of their mosquitoes on gambiae s.l. resting in houses has been obtained in Kenya (Service, 1973a) and northern preying (White, Boreham and Dolling, 1972). Hudson and Clarke (1971) recorded Salticid spiders Nigeria gambiae s.1. resting in houses in Tanzania, while Service also showed by precipitin tests capturing (unspecified) spiders were active in predation in houses. that The presence of fungal infection with Coelomomyces in gambiae s.l. has been noted in a Parasites. of regions. In Burkina Faso and Nigeria, there are records of4 species of fungi (Rodhai]a and number 1971, Service 1973b, White and Rosen 1973, Coz 1974, Rodhainand Brengues 1974). In Gayral Service (1977b), and in Angola, Ribeiro "et al. (1981) have recorded C. indica. Among Kenya, parasites, infections in both gambiae s.s. and arabiensis with species of Parathelohania have Protozoal observed in Kenya (Service 1977b) and Nigeria (White and Rosen 1973, Hazard and Anthony been Weiser and Prasertphon 1981). The latter authors also recorded the presence of Ambly0spora 1974, and ACosema salivaria. Canning and Hulls (1970) studied the biology ofafosema algerae, which coluzzii originally been recovered from a laboratory colony ofgambiae s.l. in Tanzania. Service (1977b) had and Boreham (1972) found metace.rcarial cysts in 2 specimens of gambiae s.s. presence of Gregarines has often been recorded e.g. White (1971c), White, Magayuka and The (1972), who noticed their greater frequency in gambiae s.s. than in arabiensis. When Boreham BIOLOGY OF FRESHWATER SPECIES OF THE ANOPHELES GAMBIAE LARVAL COMPLEX ofgambiae s.l. breeding sites in Accra, Ghana, were reported by Chinery (1970). In western Surveys Surtees et al. (1970) noted that bush clearance and the planting of rice led to a 7-fold increase Kenya, the density ofgambiae s.l. In the same area, Chandler and Highton (1975, 1976) recorded that the in muddy pools created by the transplanting of rice resulted in prolific breeding ofgambiae s.l., open first adults appearing in traps 10 days after planting. After 14 weeks the number of adults the decreased. Rice over lm tall appeared to inhibit breeding. Apart from this factor, water rapidly and predation were considered to be important. In Madagascar, Subra et al. (1975) found depth breeding in water-storage tanks, provided they were open and the water muddy. arabiensis studies seem to show that there are not major differences between gambiae s.s. and arabiensis All their colonization of transient rain pools or of more permanent sources such as rice fields and the in of seasonal swamps. Both species may occur together in the same pools in Nigera (Service, edges In the same country White and Rosen (1973) found gambiae s.s. by itself in 56% of pools, both 1970b). in 42% and only one pool out of 57 with arabiensis alone. In Kenya, although Service et al. species found little difference generally between the 2 species in choice of breeding sites, in one (1978) arabiensis predominated in cattle hoof-prints. At the same time, in Tanzania White, village and Breham (1972) found that populations of arabiensis increased ahead of gambiae s.s. Magayuka showers in the dry season. In the Transvaal we have found quadriannulatus larvae in semi- after pools in dry river beds, suggesting that their breeding requirements are not essential'ly permanent from the other freshwater species. different Senegal plateau, Nigeria Jos Chikungunya Ngando-group Cameroun forest Tataguine Tataguine gambiae s.l. Johnson et al. (I977) TABLE XV: Virus isolations from gambiae complex. Mermithidae in 6.9% ofgambiae s.l. larvae in rice fields in Kenya, while White, Magayuka noted in Tanzania. staff, their appearance in preparations of adult salivary glands needs to be pointed out to training confusion with sporozoites. avoid 115

119 Fig. 25. Age distribution curve for imnatures ofarabiensis. A in rice fields. B in small ponds and pools (from Service, 1977b). ll6

120 ( b, 1977b) made detailed studies oflarval mortality in arabiensis in small ponds Service pools and in rice fields near Kisumu, Kenya. Although the overall mortalty from egg-hatch to or emergence, at about 93%, was the same in both types of habitat, the pattern of mortality was adult different (fig. 25), the mortality in the first two instars in small bodies of water being much less very to the habitat more quickly than the predators and the overall mortality (egg-hatch to returned was then estimated to be only 83.5%. The most important aquatic predators were beetle adult) and Hemiptera, while adults of a number of predaceous Diptera were also involved. Service's larvae suggested that, unlike the aquatic predators, the files were mainly taking 3rd and 4th observations gambiae species D, Davidson and White, Trans.R.Soc.trop.Med.Hyg. 66: bwambae White, Syst Entornol. 10: Hot Springs, Bwamba County, Uganda. Buranga (from White, 1973). As noted by Haddow et al. (1951) the main IDENTIFICATION: the unusually broad apical pale band on the female pulp. The mean in arabiensis (fig 26a,b). When the lengths of these 2 bands are plotted graphically, separation of 0.89 except a small percentage of bwambae from other members of the complex is achieved. White all b 26. Female palps in a bwarnbae, b gambiae s.s. (semi-diagrammatic). Fig. chaetotaxonomic features of the pupa or larva appear to exist to enable identification of No specimens. The larval pecten is similar to that in freshwater species. The egg, likewise, individual that of the freshwater species. resembles BIOLOGY: This species became well known from the classical studies ofhaddow and his ADULT from the East African Virus Institute, Entebbe, on "gambiae" in Semliki Forest. These co-workers later studies have shown that it attacks human bait outdoors in large numbers, but where and settlements occur it enters houses to feed with equal readiness. Spray-catches in houses yield human numbers of freshly fed females, but the high fed:gravid ratio of 1:0.7 indicates that most moderate that have fed and rested indoors, leave at dusk the following evening (White, 1973). females, catches can be made of males and females resting outdoors, particularly on tree trunks Abundant were assumed to be important in the Semliki Forest, although their precise role was not hosts White showed that a domestic pig was attractive to bwambae, although perhaps less so established. were human baits. The biting cycle of bwambae formed the basis of the classical description by than (1945), Haddow and Ssenkubuge (1973) of the cycle for the gambiae complex as a whole. Haddow than in rice fields. Interestingly, treatment of rice fields with insecticide showed that mosquitoes larvae and were more important in small pools. instar WATER SPECIES MINERALISED bwarnbae White gambiae (in part), Haddow, Bull.ent.Res. 36: TYPE LOCALITY: characteristic of bwambae is ratio of apical pale band subapical dark band is , compared with 1.04 in gambiae s.s. and recorded a single specimen in which the subapical dark band was missing. buttresses in the forest. Fed females have rarely been found in these situations. Wild animal and 117

121 picture given is unusual, perhaps, for the biting activity recorded in the daytime by catchers The in the forest. But, as Haddow commented, this may be as much dependent stationed features on connected with the forest environment as on intrinsic characters. any TO DISEASE: In a short series of dissections of specimens identified to species, 0.7% RELATION HABITAT: Abundant breeding occurs in water derived from hot springs at LARVAL of C. The most favoured sites were exposed ground pools, where the mud had temperatures churned up and polluted by buffaloes and other game. Larvae were also found in water been under low vegetation. The water was highly mineralised, ph , total solids 3.4 concealed g/l, sulphate, chloride and bicarbonate being the most important constituents. Larvae appear to 15 highly dependent on this mineralised medium, and attemps to rear them in tap or distilled water be not been successful (Davidson and White, 1972). This dependence contrasts with the tolerance have by the salt-water members of the complex to fresh water in the laboratory. In the type shown breeding is intense and continuous throughout the year, leading to the constant presence of locality, adult populations in the Semliki Forest. large Only known from the small area of the Rift Valley, west of Ruwenzori, where DISTRIBUTION: geothermal conditions exist. appropriate ratio to separate melas females from gambiae s.l. was to take palpal value of 0.81 as the a point for the ratio (lengths of segments 4+5: length of segment 3). Females with discriminating ratios of 0.81 and above were treated as melas and all below as gambiae s.s. This enabled the correct In this area, 42% offemales had 4 pale palpal bands, whereas respectively. were seen in gambiae none $o$o BIOLOGY: (Studies in the Gambia). Mating and swarming. Swarms have been observed in ADULT clearing flanked by tall mangroves from about 15 min after sunset, continuing for about 20 min a and Jones, 1980). In the presence of an almost full moon swarming continued for (Charlwood some rain longer. Swarms could be induced to form over white markers placed 20 the ground, even on on the vertical distribution of females flying across open ground between Studies swamps mangrove present were taken at levels of up to a maximum height of9m. While the results showed populations most mosquitoes were flying near the ground, in one series of tests 5.6% of the catch that taken at was highest level. Most females were flying at levels at which the wind speed the less than their flight was estimated in the field to be 1.2 m/s (Snow, 1982). The finding that most mosquitoes speed, were upwind (Snow, 1976, 1977) is not surprising. What flying unexpected was that, although the was of old females 2-parous and over decreased with height, the reduction proportion less was than in nullipars, so that the age composition of the high-flying population marked older than was those flying at low levels (Snow and Wilkes, 1977). In in area where the nearest villages were an from the mangrove swamps by a distance of 2km, catches ofmosquitoes up to 22h00 separated near village were appreciably less than the the swamps and represented 15% of the total night's near at the former site as opposed 27% at the latter. This difference presumably reflected the catch time between the 2 sites (Snow, 1982). dispersal bwambae were found infected with sporozoites compared with 7.2% in gambiae of in the same area s.s. 1973). (White, SALT-WATER SPECIES naelas Theobald In The Gambia, Bryan (1980) found that the IDENTIFICATION: effective way of using the most identification of 96% of melas and 92% of gambiae s.s., at the cost of 4% and 8% misidentifications natural markers could not be discerned. Entry of females into the swarm, followed by though was frequently observed. coupling, Flight pattern. villages were made by Snow (1975, 1982). Aerial samples of the very large 118

122 in houses. From the light-traps in houses Odetoyinbo (1969) concluded that there was a Movements deal of movement of unfed females from house to house. The reason for this seemed to be that, great to the intensity of mosquito attack, most of the human population slept under nets, so that owing and accessible hosts were difficult to find. exposed to melas, Wilkes (1968) reported no special difficulty with this species. He confirmed applied report of the relatively short span of life of this species in The Gambia, no individuals older Giglioli's 5-parous being found. than TO DISEASE: Bryan (1979, 1983) found a sporozoite rate of 0.32% in nearly 5000 RELATION dissections, although 3.5% of freshwater gambiae were infected in the same area. Owing to the gland human blood index observed, Bryan et al. (1982b), Bryan (1983), considered that, except when low in very large numbers, melas played negligible role in malaria transmission. present HABITAT: Ribeiro and Ramos (1975) showed that, in Angola, as elsewhere along the LARVAL of West Africa, the habitats of gambiae s.s. and arabiensis were well segregated, in terms of coast from those of melas. salinity, merus D6nitz BIOLOGY: Theron (1978) observed large swarms of males assumed to be composed of ADULT along the margins of a saline lake in Zululand, forming up from sunset under the canopy merus, by lakeside trees. In southern Mozambique, Clarke (1969) found that, inthe cool season, the formed between feeding and oviposition was extended to at least 4 days. interval TO DISEASE: In a coastal village in Tanzania, Bushrod (1981) showed that merus RELATION acting as a vector of bancroftian filariasis and was probably more important in this role than was s.s. In a Kenyan village Mosha and Petrarca (1983) found infection rates for malaria gambiae of 3.3% and filarial larvae of 1.1% in the cytologically identified merus sample. This is sporozoites higher than Bushrod's (1981) rate and conflicts with their conclusion that merus is a relatively even poor vector of malaria and filariasis. Their conclusions were evidently based on results from a village where the sample size of merus was too small for infection rates to be established. second HABITAT: Mosha and Mutero (1982) studied the biology ofmerus in a large brackish LARVAL on the coast of Kenya. They showed that adult populations remained low as long as the pond pond full and the salinity low. As the pond dried up and salinity rose to the equivalent of 50% sea- was population density peaked, only to fall again as the pond continued to dry and salinity rose to water, in excess of 100% sea-water. Laboratory tests showed that survival of larvae was high in levels and in concentrations of up to 60% sea-water but declined steeply at higher salinities. freshwater concluded that, in the, field optimal conditions are around 40-50% sea-water, and that poor They by predators that are normally less tolerant of salinity.. limited Although essentially a coastal species (e.g. Mosha and Subra, 1982), it is known DISTRIBUTION: Longevity. In contrast to Giglioli's (1965) conclusion that Polovodova's technique could not be survival, and perhaps avoidance by ovipositing females, leads to their disappearance from larval above 100% sea-water. Conversely, they suggest that survival in freshwater habitats may be pools thriving populations of merus can be found at considerable distances from the coast, usually in that with salt pans (Patersonet al 1964, Cross and Theron 1983, Muspratt and Henning association Coetzee and Cross (1983) studied one of these sites in the northern Transvaal, 300 km from 1983). sea, and showed that crosses between coastal merus and inland populations were fully the Recent findings (R. Marchand, in litt.) have shown that merus occurs at various inland compatible. in north-east Tanzania, up to 80 km from the coast. Mosha and Mutero (1982) suggest localities that reports of gambiae s.l. breeding in saline lakes in the East African Rift Valley in Tanzania and Kinoti, 1971), could be explained by the presence ofmerus. In this area (Lake Manyara) (Njogu were found in the equivalent of 1/3--2/3 sea-water. larvae.erie. PARAMYZOMYIA additional characteristic of this series is the absence of floats on the eggs. An 119

123 was formerly subdivided into two sections Paramyzomyia the basis of larval characters. on the discovery of azevedoi, which combines characters from both groups, i.e. usually with However, long mesopleurals (setae 9-M, 10-M) branched but with normal mouth brushes, both removes of the justification for this subdivision. We are, accordingly, treating Paramyzomyia much a as group. single turkhudi Liston RECORD: Djibouti; 12N, 42E (Rodhain, Carteron et al., 1977). NEW azevedoi Ribeiro Angola. Mosfimedes, Adult key section VII, larval key section IX. The adults IDENTIFICATION: not separable are listeri on external morphology. The pupae are instantly recognisable by the length of setae 9- from and 9-VII. Similarly, the undeveloped, tassel-like palmate hairs (setae 1-III to 1-VII) of the VI them not being easily discernible under the microscope, the longest about 25 We have of the aedeagus of specimens from the Karroo, described by De Meillon and Van Eeden compared (l-p) long, straight or hooked. Seta 1-II with 4-7 branches, 1-III 4-9, 1-IV 5-7, lov 5-6, 1-VI 1- hair and 1-VII simple. Seta 5 very similar on II-VII with 4-8 branches, gradually increasing in length 3 Seta 9-VI almost reaching apex of VII, 9-VII extending to posterior border of VIII. caudally. Branching of hairs generally reduced. Head: Heavily pigmented dark brown Larva: black.clypeal or (fig. 27,a): All simple, outer hair (3-C) about half as long hairs inner hair (2-C). Frontal hairs (5-C to as Simple or with a few short branches. Shoulder hairs (fig. 27,b): Inner and median hairs (l-p, 2-7-C): (3-T) 2-3 branched, 1-I and -II simple, 1-III 3-branched, 1-IV to 1-VII with 3-6 simple palmate or flattened branches. Tergal plates: Very small, on V about 0.4 times distance between scarcely (from Ribeiro, 1969) Without floats, deck occupying whole of dorsal surface, surrounded by Egg: a developed frill; chorion with large, irregular bosses, well the undersurface with a fine reticular on pattern. HABITAT: A species frequenting highly saline pools along the coast ofsouthowestern LARVAL or in suitable areas inland. As described by Ribeiro (1974), larvae abound in tidal pools, Africa canals, salt-pans and brackish wells in otherwise arid country. There is saline constant association a filamentous algae. While the salinity of the breeding sites with most commonly in the vicinity of was g/l, approximately pure sea-water, larvae found within extreme limits ranging from were g/l, equivalent to % sea-water. It has not been recorded whether development 13.5 always successfully completed in these exceptional levels, the highest recorded for any African was In lower salinities azevedoi was frequently associated with listeri, but Ribeiro notes that, in. restricted zone where the distribution of azevedoi overlaps with that of melas, the two species the separate sites. At the far southern end of its distribution, in the Upper Karroo, larvae occupied were in saline pools along a river-bed in association with much larger numbers of listeri, and the found as NaC1, was 14.1 g/l (De Meillon and Van Eeden, 1976). salinity, BIOLOGY: In Angola, adults were caught resting outdoors in holes in the ground and in ADULT azevedoi Ribeiro, Anais Esc.?v'ac.Saude pub. Med.trop 3" deaconi De Meillon & \;an Eeden, Mosq.Syst. 8:336. TYPE LOCALITY: larvae immediately distinguish it from listeri. Male genitalia: As in listeri but phallosome (aedeagus) with only 3-6 unserrated leaflets, the smallest under the name deaconi, with azevedoi from Angola and found them to be identical. Pupa: Outer paddle fringe with hooked spines changing gradually to straight, fine spines; apical poorly developed, widely separated basally and without basal tubercles. Mesopleural hairs: 9-M P) or feathered, 10-M feathered.palmate hairs (fig. 27,c): Undeveloped and tassel-like, thoracic single palmates (setae l-v). A single accessory tergal plate present. Saddle hair (l-x): Simple. cliffs as well as in vegetation and in houses. Females were also caught rocky human bait outdoors, on the range of hosts normally attacked has not been established. Gland dissections of 109 females but were negative for sporozoites. 120

124 Fig. 27. azevedoi, a Clypeal hairs, b Shoulder hairs, c Palmate hair (l-v). listeri, d Egg. 121

125 Fig. 28. Distribution of azevedoi. (fig. 28); The arid coastal belt of Angola from latitude ls down DISRTIBUTION: the to Cape Province, northern km inland from the mouth of the Orange River. some llst.rl De Meillon (fig. 27,d) Without floats, deck occupying whole ofdorsal surface, golden brown with 2 dark Egg: across it; well developed frill surrounding deck, chorion with fine bosses. bands HABITAT: In Angola, the species shows high tolerance LARVAL salinity, larvae having been to in water equivalent to found than 1.5 times sea-water. In saline situations it more often was with azevedoi (Ribeiro and Ramos 1975, De Meillon and Van Eeden 1976). associated 13E (Ribeiro and Ramos, 1975). 16S, multi(c)olor Cambouliu Anoph.l.t ls, 14E; 13S, 12E; 14oS, 12OE; 15oS, 12OE; 15os, 13OE; 16oS,12OE; Holstein et al (1971) point out that, in occasional specimens of hispaniola with cinereus by Dahl and White, 1978) from Saharan (synonymised one 10ng mesopleural oases, (9-M) may appear simple unless examined carefully under high hair They suggest that this power. account for some of the more southerly records ofmulticolor in Tassili. Another new, southerly may is from South Yemen (Kravchenko, 1979). record macullppls Giles RECORDS: Kenya; ls, 37E (Ouma, 1975). Angola; 7S, 12E; 9S, 18OE; 10os, 17OE; NEW 14E; 12S, 17E; 13S, 14E; 13S, 16E; 13S, 17E; 15S, 13E (Ribeiro and Ramos, 1975). 12S, pr.tod.tmis Theobald Anoph.l.t Service (1970a) recorded the capture of IDENTIFICATION: rather darker female in northern a in which the pale spots Nigeria, the fore and mid tarsi were much reduced, the apices of fore on tarsus 4 and 5 and of mid tarsus 3-5 were dark. In addition, the wing showed a dark fringe spot NEW RECORDS: Angola; IDENTIFICATION: NEOCEIJ.,IA serles 122

126 vein 1, the costa was pale at base, vein 3 almost all dark, apex ofvein 4 pale and vein 6 pale opposite base. at species of Parathelohania has been described from a larva from Nigeria (Hazard and Anthony, A 1974). RECORDS: Congo Republic; 3S, 14E (Carnevale, 1974b). Gabon; 0S, 10E NEW Angola; 10S, 14E, 12S, 14E; 13S, 12E, 13S, 13E; 13S, 14E; 14S, 12E; (Service,1976). 12E; 15S, 13E; 16S, 12E; 16S, 13E; 17S, 12E; 17S, 13E (Ribeiro and Ramos, 1975). 15S, ruflpes rufipes Gough, Coelomomyces. RECORDS: Senegal; 14N, 17W (Cornet and Chateau, 1974). Kenya; ls, 37E (Ouma, NEW the outline used was that of the fore leg, although the correct markings for the hind leg were error, shown. 1977). In 1968 we remarked on singular finding of a specimen of salbaii in a house in Niger. An even more remarkable the was the discovery in 1974 of a live female in London (!) in a cage offunestus originating occurrence houses in the Kisumu district of Kenya (0S, 34E). The sporadic appearance of adults over from a wide area suggests that the species may be commoner than suspected and that the larvae may such have been overlooked. For instance, we have seen specimens from Acholi District, sometimes tentatively assigned to prtoriensis and yet with shoulder hairs distinctly separated. Uganda, argenteolobatus Gough material collected by Dr C.A. Green in Harare, Zimbabwe, it appears that the name From covers more than one species. No cytogenetic data have been reported as yet, but the argenteolobatus variation observed in a preliminary study involves at least 3 characters: morphological As in fig. 29,d, from material from Harare, Zimbabwe, floats entirely lacking, frill B. upper surface usually open and partially divided into 2 or 3 decks. absent, Abdominal tergal scales. 2. Terga II-VII mainly clothed with bronzy scales. A. Terga II-VII black-scaled except for a few scattered white scales posteriorly. B. Wing fringe spots (posterior to vein 3). 3. Spots present opposite veins A. Spots present opposite veins 5.2 or 6 only. B.. specimens from Harare, collected and reared by Dr. Green in 1978, have the combination The characters 1B, 2B, 3A. of female in the B.M. (N.H.), apparently from Gough's type series from Onderstepoort, has the A 1?, 2A, 3A. characters type female ofpseudosquamosa Newstead and Carter from north-east Zambia and also in the The (N.H.), currently regarded as a synonym of argenteolobatus, has the characters 1?, 2A, 3B. B.M. from Mufindi, Tanzania, has the characters 1?, 2A, 3A. Material two egg-types are clearly those of different species, but until the scaling and wing pattern of The LARVAL BIOLOGY: Rodhain and Gayral (1971) reported a single specimen infected with Angola; 9S, 14E; 9S, 15E; 10S, 14E; 12S, 14E; 13S, 13E; 13S, 14E; 14S, 13E; 1975). 12E; 15S, 13E (Ribeiro and Ramos, 1975). 15S, salbaii Maffi and Coluzzi In Gillies and De Meillon, 1968, pl. 110b, the hind leg ofsalbaii is figured. In fact, owing to Note. an NEW RECORDS: Djibouti; 12N, 42E (Rodhain, Carteron et al., CELLIA series Egg. A. As figured in Gillies and De Meillon (their plate 111) from material fi'om Zambia, 1. small but well developed floats and a very broad deck. with females laying eggs with floats has been determined, the problem is unlikely to be solved. In the 123

127 a series of adults and larvae reared from eggs from Dr. Green's Harare series has been meantime, in the B.M. (N.H.). deposited BIOLOGY: Ribeiro et al. (1981) report finding larvae infected with Coelomomyces LARVAL africana. RECORDS: Angola; 7S, 15E; lls, 17E; 12S, 15E; 12S, 16E; 12S 17E; 13S, 16E; NEW 17E; 17S, 19E; 17S, 20E (Ribeiro and Ramos, 1975). 13S, (C)rlst palp s Service cristipalpis Service, Ann.trop.Med.Parasit. 71: LOCALITY: Jos plateau, Nigeria. TYPE Adult key section I. IDENTIFICATION: Length: mm. Wing (fig. 29,b): shaggy with 4 irregular pale bands and scattered pale scales. Mesonotum: Dark Palps clothed throughout with fairly broad white scales which are absent from the seutal fossae, grey, the appearance of two dark spots. Legs: fore leg, femur mainly black, pale posteriorly; tarsus giving 4 and 5 all dark; hind leg as in fig. 29,c. Wings: as in fig. 29,a. Abdomen: Lateral basal angles of tarsi II-VII bearing tufts of black scales; terga clothed with bronzy scales, the sterna white- segments cerci dark-scaled. scaled, early stages not known. and Only known from the Jos plateau in Nigeria (9N, 8E) at an altitude of 1570 DISTRIBUTION: (5000 feet). metres cydlpp s De Meillon Material from Harare (collected by Dr. C.A. Green) is not exactly as figured by Gillies and De Egg: perhaps a better illustration of this egg. BIOLOGY: Ribeiro et. al. (1981) report finding larvae infected with Coelomomyces LARVAL africana. RECORDS: Rwanda and Burundi; 2S, 29E; 1-4S, 30E (Vermylen, 1967). Angola; 7S, NEW 8S, 15E; 10S, 15E; 10S, 17E; 12S, 15E; 12S, 17E; 13S, 14E; 13S, 16E (Ribeiro 15E; Ramos, 1975). and rnarphyi Gillies and De Meillon Anaphele W.F. Snow has worked (in collaboration with ourselves) in the type locality, Keneba, The Dr. for many seasons and yet only a single female of murphyi has been captured. Its abundance Gambia, that area has evidently declined over the past 20 years. in pharoensls Theobald complex Namibia, East and West Africa and Egypt, form a separate mating group and have including X-chromosome arrangements from populations from Zululand. It is concluded from this different pharoensis is a species complex of at least 2 sibling species. The implications of this discovery in that of vectorial role have yet to be explored. terms BIOLOGY: In the Kisumu district of Kenya, it was found that light-traps placed round ADULT eaves of houses caught adults on the wing during the early part ofthe night (Chandler, Highton the Hill, 1976, Chandler, 1977). However, since there was little biting activity at that time ofnight, and of it being later and slowly increasing up till shortly before dawn, this movement was most white with black patches, tarsus 2 basal 1/3 1/2 dark, tarsus 3 basal 1/4.z 1/3 dark, tarsi 4 and 5 all dark; mid femur with some patches of pale scales,tarsus white at apex, tarsi 2 and 3 basal 2/3 dark, ADULT BIOLOGY: Caught biting man at night. Meillon (after De Meillon, 1937). Fig. 29,e is It has been shown by Miles et al. (1983) that populations of pharoensis from most parts of Africa, not directed towards the host (Chandler, Highton and Hill, 197.5). In The Gambia, presumably and Wilkes 1976) showed that, when moving across open country towards villages, pharoensis Gillies 124

128 29. cristipalpis, Fig. Wing. b Female palp. c Hind leg. argenteolobatus, d Egg lharare). cydippis, e a Egg. 125

129 were mainly flying just above the ground, 64% of them in the lowest 60 cm. females has generally been observed, the human blood index in Kenya was found to be as low as 0.16, As majority of feeds being bovids (Chandler, Boreham et. al., 1975). In Tanzania, Mosha and the (1979) reported a single female out of42 dissected to have infective larvae ofbrugia sp. At Magayuka same time, the species was fully susceptible to experimental infection with Wu. bancrofti. Rift the fever virus has been isolated from a pool of mosquitoes collected near Nairobi, Kenya Valley 1983). (Anon., HABITAT: As in the case ofgambiae s.l., open muddy pools created by the transplanting LARVAL The output of adults fell to a low level once the rice had grown to a height of lm. 1975). RECORDS: Senegal; an, 17W; 13N, 16W (Cornet and Chateau, 1974). Ethiopia; 6N, NEW sclutaou Theobald complex Ollaeles cytogenetic studies have indicated that squamosus should be regarded as a species Unpublished (C.A. Green, pers. comm.). complex BIOLOGY: In The Gambia, females of squamosus were found to have similar flight habits ADULT White (1970). recorded the arrival offemales in some numbers to a light source more than Tanzania, km from the nearest squamosus breeding areas. He concluded that this was an instance of 10 flight, comparable to that known for pharoensis. migratory BIOLOGY: Ribeiro et. al. (1981 reported larvae infected with Coelomomyces africana and LARVAL indica as well as with Mermithidae. C. RECORDS: Senegal; 13N, 16W (Cornet and Chateau 1974). Central African Republic; NEW 15S, 13E; 17S, 18E; (Ribeiro and Ramos, 1975, coincident records of adults and larvae 12E; Yemen Arab Republic; Kuznetsov (1971) has provided the first records of squarnosus only). (including adults) from the Arabian Peninsula. of rice in fields in Kenya provided favourable conditions for breeding (Chandler and Highton, 5N, 36E (Rodhain, (Carteron et. al., 1977, Rodhain, 1971). Djibouti; 12N, 42E Carteron 37E; al. 1978). Central African Republic; 10N, 22E (Cordellier and Geoffroy, 1976). Rwanda and et. 1-2S, 30E; 3S, 29E (Vermylen, 1967). Angola; 9S, 13E; 10S, 15E; 15S, 20E; Burundi; 18E; 17S, 20E (Ribeiro and Ramos, 1975). 17S, to pharoensis, namely to be flying very close to the ground when flying across open country. In 20E (Rickenbach, 1969). Rwanda and Burundi, 2-3S, 29E; 1-3S, 30E (Vermylen, 1967). 8N, 3S, 35E (Njogu and Kinoti, 1971). Angola; 10S, 15E; 12S, 14E; 13S, 14E; 15S, Tanzania; 126

130 A. E. 1969a. Methods for determining the host preference for gambiae Giles ABOUL-NASR, Anophelinae). Bull. Soc. ent. Egypte 53: (Diptera:Culicidae A. E. 1969b. Determination of the physiological age in gambiae Giles, main ABOUL-NASR, of malaria in Eastern Nigeria (Diptera:Culicidae Anophelinae). Bull. Soc. ent. Egpte 53: vector ent. Egypte 53: Soc. J. 1973a. Exophily in gambiae species B in the Sudan. Trans. R. Soc. trop. Med. Hyg. AKIYAMA, 67:440. J. 1973b. Further isolations of translocations in gambiae species A. Trans. R. Soc. trop. AKIYAMA, Hyg. 67: Med. J. I., BEIDAS, M. F. & WHITE, G. B Cytotaxonomic evidence for the malaria vector species AKOH, of the culicifacies complex being endemic in Arabia. Trans. R. Soc. trop. Med. Hyg. 78:698. A Rift Valley fever virus. Weekly Epidemiological Rec. 58(13):97. ANON. R., DIEBOLT, G., LINHARD, J. & DIOP, S Frequence de l'antigene Australia etudide BAYLET, a l'activit6 anophelienne dans sept villages Seres au Sdndgal. Bull. Soc. Path. exot. comparativement 67: M. F. & GILLIES, M. T The egg of (Cellia) culicifacies adenensis Christophers. BEIDAS, Syst. 12: Mosq. J. N The mosquitoes of the South Pacific. Univ. Calitbrnia Press. 608pp. BELKIN, vector-borne tropical diseases. J. trop. Med. Hyg. 78: of P. F. L., CHANDLER, J. A. & JOLLY, J The incidence of mosquitoes tbeding on BOREHAM, and babies at Kisumu, Kenya. J. trop. Med. Hyg. 81: mothers J Calculation of anopheline man-biting densities from concurrent indoor/outdoor resting BRADY, Ann. trop. Med. Paras#. 68: samples. W. R House catches of adult gambiae species B in two areas of BRANSBY-WILLIAMS, E. A.fr. reed. J. 56: Zambia. J Ed.La filariose de Bancroft en Afrique de l'ouest..mem. ORSTOM No. 79, 299pp. BRENGUES, J. & BAIN, J. O Passage des microfilaires de l'estomac vers l'hdmocble du vecteur BRENGUES, les couples Wuchereria bancrofti dans gambiae A, W. bancrofti Aedes aegypti et Setaria -. A. aegypti. Cah. ORSTOM Ent. mid. 10: labiatopapillosa J. & COZ, J. 1973a. Quelques aspects fondamentaux de la biologie d' gambiae Giles BRENGUES, et d'funestus Giles, en zone de savane humide d'afrique de l'ouest. Cab. ORSTOM Ent. (Sp.A) 11: mid. J. & COZ, J. 1973b. Rdceptivit6 cdmpar6e des trois espbces du complexe gambiae Giles BRENGUES, en Afrique de l'ouest, vis-a-vis de Wuchereria bancrofti Cobbold. Cab. ORSTOM Ent. mid. 10:207 pr6sentes J., SUBRA, R. & BOUCHITE, B Etude parasitologique, clinique et entomologique sur. BRENGUES, filariose de Bancroft dans le sud du Dahomey et du Togo. Cah. ORSTOM Ent. md. 7: la J., SUBRA, R., MOUCHET, J. & NELSON, G. S La transmission de Wuchereria BRENGUES, Cobbold en Afrique occidentale. Bull. Wld Hlth Org. 38: bancrofti P., CAMICAS, J. L., CORNET, M., ROBIN, Y. & TAUFFLIEB, R Consid6rations sur BRES, des arboviroses au S6n6gal. Bull. Soc. Path. exot. 62: l'epid6miologie L-O Contribution l'6tude biologique et ecologique des vecteurs ma, jeurs de paludisme en BRUN, de l'ouest. Thise, De Inginieur, Fac. des Sciences, Rennes, Mem. ORSTOM. C 36(1), 223pp. Ati'ique M. J. & BRUNHES, J Etude histologique chez Mansonia uniformis Theobald et BRUNHES, gambiae B Giles, de l'evolution de Wuchereria bancrofti Cobbold et des interactions entre la filaire ces deux hotes. Cah. ORSTOM Ent. mid. 10: et J. H Results of consecutive matings of tmale gambiae species B with trtile and BRYAN, males. N'ature 218:486. sterile J. H Further studies on consecutive matings in the gambiae complex.,nature BRYAN, 239: J. H Observations on the member species of the gambiae complex in The Gambia, BRYAN, Ati'ica. Trans. R. Soc. trop. Med. I-Iyg. 73: West REFERENCES CITED ABOUL-NASR, A. E Entomological problems facing malaria eradication in Africa (Uganda). Bull. BOREHAM, P. F. L Some applications of bloodmeal identification in relation to the epidemiology 127

131 J. H Use of the palpal ratio and the number of pale bands on the palps in separating BRYAN, gambiae Giles s.s. and melas Theobald (Diptera:Culicidae). Mosq. Syst. 12: J. H gambiae and A. melas at Brefet, The Gambia, and their role in malaria BRYAN, Ann. trop. Med. Parasit. 77:1-12. transmission. J. H., DI DECO, M. A., PETRARCA, V. & COLUZZI, M. 1982a. Inversion polymorphism and BRYAN, speciation in gambiae s.str, in The Gambia, West Africa. Genetica 59: incipient J. H., PETRARCA, V., DI DECO, M. A. & COLUZZI, M. 1982b. Observazionisurnelas BRYAN, Gambia e zone limitrofe..larassitologia 23: nella J. H. & SMALLEY, M. E The use of ABO blood groups as markers tigr mosquito biting BRYAN, Trans. R. Soc. trop. Med. Hyg. 72: studies. F. M Annu. Rept. E. A. Inst. Mal. Vect.-borne Dis. Tan Dec. 1973: BUSHROD, F. M The gambiae Giles complex and Bancroftian filariasis transmission in BUSHROD, a coastal village. Ann. trop. Med. Parasit. 75: Tanzanian F. J. C., PETRARCA, V. & COLUZZI, M arabiensis in the Cape Verde CAMBOURNAC, Parassitologia 24: archipelago. E. U. & HULLS, R. H A microsporidian infection of gambiae Giles, ti'om CANNING, 17: Protozool. D. A. & SERVICE, M. W Differentiation between species ofthe Anophelegambiaecomplex CARLSON, by analysis of cuticular hydrocarbons. Ann. trop. Med. Parasit. 73: (Diptera:Culicidae) D. A. & SERVICE, M. W Identification of mosquitoes ot" gambiae species CARLSON, A and B by analysis of cuticular hydrocarbons. Science 207: complex A. G., MOTABAR, M., SUNDARAMAN, S., ROWHANI, F. & GOLESTAN, J CARMICHAEL, evaluation of OMS-33 (0-isopropoxy-phenyl methylcarbamate). Unpubl. document Entomological WHO/VBC/ P Epid6miologie du paludism humain en R6publique Populaire du Congo. Le CARNEVALE, gambiae dans la r6gion brazzavilloise. Cah. ORSTOM Ent. mid. 10: complexe P. 1974a. Comparaison de trois m6thodes de capture pour l'6chantillonage d'une population CARNEVALE, nili (Theobald) Cah. ORSTOM Ent. mid. 12: d' P. 1974b. Variations saisonnibres d'une population d' nili (Theobald) 1904 CARNEVALE. en Populaire du Congo. Cah. ORSTOM Ent. mid. 12: Rdpublique P. & BOREHAM, P. F. L Etudes des pr6f6rences trophiques d' nili Theo. CARNEVALE Cah. ORSTOM Ent. mid. 16: P., BOSSENO, M. F. & ZOULANI, A Etude du cycle gonotrophique d' ni!i CARNEVALE Cah. ORSTOM Ent. mid. 16: (Theo.) P., BOSSENO, M., MOLINIER, M., LANCIEN, J., LE PONT, F. & ZOULANI, A CARNEVALE. du cycle gonotrophique d' gambiae (Diptera:Culicidae) (Giles, 1902) Etude zone de tbrgt en d'afrique Centrale. Cah. ORST021,I Ent. mid. 17: d6grad6e P., FREZIL, J. L., BOSSENO, M. F., LE PONT, F. & LANCIEN, J Etude de CARNEVALE, 56: P. & MOLINIER, M Le cycle gonotrophique et le rhythme quotidien des piqfires CARNEVALE, gambiae (Giles), 1902 et nili (Theobald) Parassitologia 22: d' P., MOLINIER, M. & MOUCHET, J Relations math6matiques dans la d6velopment CARNEVALE, tbllicules ovariens d' nili (Theobald) Cah. ORSTOM Ent. mid. 15: des P. & LE PONT, F Epid6miologie du paludisme humain CARNEVALE, R6publique Populaire du en P. & ZOULANI, A Agressivite d' nili (Theobald) 1904 l'int6rieur eta CARNEVALE, des maisons. Cah. ORSTOM Ent. mid. 13: l'ext6rieur B., MORVAN, D. & RODHAIN, F Le probl6me de t'end6mie palustre dans la CARTERON, de Djibouti..Md. trop. 38: R6publique J The use of light traps to study flight activity and seasonal abundance of mosquitoes of CHANDLER, importance on the Kano Plain, Kenya. Internat. J. Environ. Studies 10: medical J., BOREHAM, P. F. L., HIGHTON, R. B. & HILL, M. N Feeding habits of CHANDLER, in the Kisumu area and their possible relationship to disease transmission. E. Aft. med. J. mosquitoes interpretation of its mode of transmission and notes on Nosema infections in mosquitoes. Tanzania, l'agressivit6 d' gambiae A en function de l'fige et du sexe des sujets humains. Bull. Wld Hlth,Org. II. Utilisation des pibges lumineux "CDC" comme moyen d'6chantillonage des populations Congo Cah. ORSTOM Ent. mid. 11: anoph61iennes. 52:

132 J. A. & HIGHTON, R. B The succession ofmosquito species (Diptera:Culicidae)in rice CHANDLER, in the Kisumu area of Kenya, and their possible control. Bull. ent. Res. 65: fields J, A. & HIGHTON, R. B The breeding of garnbiaegiles (Diptera:Culicidae) in CHANDLER, fields in the Kisumu area of Kenya.. med. Ent. 13: rice J. A., HIGHTON, R. B. & HILL, M. N Mosquitoes ofthe Kanc Plain, Kenya Part I. CHANDLERI, of indoor collections in irrigated and non-irrigated areas using human bait and light traps. J. med. Results 12: Ent. J. A., HIGHTON, R. B. & HILL, M. N Mosquitoes of the Kano Plain, Kenya. II. CHANDLER, of outdoor collections in irrigated areas using human and animal bait and light traps. J. med. Ent. Results 13: J. D. & JONES, M. D. R Mating behaviour in the mosquito gambiae s.l.i. CHARLWOOD, range and contact behaviour. Physiol. Entomol. 4: Close J. D. & JONES, M. D. R Mating in the mosquito gambiaes.l. II. Swarming CHARLWOOD, Physiol. Entomol. 5: behaviour. G Etudes en particulier au moyen de radioisotopes, sur l'6thologie et la physiologie CHAUVET, des espces A & B du complexe gambiae dans une zone de sympatrie 5. Madagascar. compar6es ORSTOM Ent. mid. 7: Cah. G.I DAVIDSON, G. & DEJARDIN, J Validit6 d'une rndthode ch6totaxique de CHAUVET, des larves des espbces A & B du complexe gambiae Giles/t Madagascar. Cah. ORSTOM distinction mid. 7: Ent. A et l'espbce B du complexe garnbiaeit Madagascar. Cah. ORSTOM Ent. mid. 6: l'esp6ce G., RAVAONJANAHARY, C. & DUVAL, J Receptivit6 compar6e / Plasrnodium CHAUVET, des esp6ces A et B du complexe gambiae Madagascar. Comptes Rendus Soc. Biol. 166:489- falciparum 491. W. A A survey of mosquito breeding in Accra, Ghana, during a two-year period oflarval CHINERY, control. II. The breeding of Culex (Culex) pipiensfatigans Wiedemann and (Cellia) gambiae mosquito in Accra. Ghana med. J.. 9: Giles J. L A note on blood.digestion and egg development in merus in Southern CLARKE, gambiae complex in northern Nigeria. Bull. Wld Hlth Org. 45: J. L., PRADHAN, G. D.,JOSHI, G. P. &FONT,AINE, R E Assessment.ofthe grainst0re CLARKE, at Kisumu, Kenya. J. med. Ent. 17: (Diptera:Culicidae) A. N,' & PATERSON, G. D The analysis of mortality and survival ratesin wild CLEMENTS, Syst. Entomol. 8: (Diptera:Culicidae). M A new species of () from Namibia (Diptera:Culicidae). Syst. Entonol. COETZEE, 9:1-8. M.' & CROSS, H Mating experiments with two populations ot" merus D6nitz COETZEE, J..ent. Soc. S. Afr.. 46: (Diptera:Culicidae). M. & DU TOIT, C. L. N. I979. A full description of the adult, pupa and larva of (Cellia) COETZEE, De Meillon and Evans, from the Cape Province, South Africa.,4osq. Syst. 11" cameroni M. & NEWBERRY, K The separation of the immatures of certain mosquitoes of the COETZEE, in Southern Africa using computer techniques, Mosq. Syst. 12: subgenus M., NEWBERRY, K. & DURAND, D A preliminary report on a morphological COETZEE, Mosq. Syst. 14: Ati'ica. M.J. & SMITH, S. A Problems of malaria in the Aden Protectorate (Report on a COLBOURNE, M Cromosomi politenici delle cellule nutrici ovariche nel complesso gambiae del genere COLUZZI, Parassiotologia 10: G. & DEJARDIN, J Caract6res ch6totaxiques de distinction entre larves (Stade IV)de CHAUVET, Unpu.b! document who/vbc/69-147, WHO/MAL/ Mozambique. J. L Potential use of the spermatheca in the separation of species A and B females ot'the CLARKE, as.an unbaited outdoor, shelter for.mosquit0es of the gambiae complex, and funstus of mosquitoes. J. appl. Ecol. 18: populations M Chromosomal and.cross-mating evidence for two species within (A) cous.tani C.OETZEE, distinguishing important malaria vectors in the gambiae Giles complex in southern character Unpubl. document WHO/MAL/442 70pp. visit). COLUZZI, M., DALLAI, R. & INSOM, E Morfologia e funzione delle armature cibariches helle zanzare. Parassitologia 23:

133 M. & KITZMILLER, J Anopheline mosquitoes. In Handbook ofgenetics Vol. 3. Ed. R.C. COLUZZI, Plenum Publ. Corp. pp King. Parassitologia. 9: complex. M. & SABATINI, A. COLUZZI, Parassitologia 10: complex. M. & SABATINI, A. 1968b. Cytogenetic observations on species C, COLUZZI, and melas of the merus gambiae complex. Cah. ORSTOM Ent. mid. 6: M. & SABATINI, A Cytogenetie observations on the salt-water species COLUZZI, and merus melas, of the gambiae complex. Parassiotologia 11" M., SABATINI, A., PETRARCA, V. & DI DECO, M. A Behavioural divergences COLUZZI, mosquitoes with different inversion karyotypes in polymorphic populations of the between complex. Nature 266: gambiae M., SABATINI, A., PETRARCA, V. & DI DECO, M. A Chromosomal ditfirentiation COLUZZI, adaptations to human environments in the gambiae complex. Trans. R. Soc. trop. Med. H.vg. and 73: R. & GEOFFROY, B Les moustiques de R6publique Centrati'icaine. Distribution, CORDELLIER, et ti'6quence des Culicin6s dans l'ouest du pays Les arbovirus isol6s. Travaux et documents de abondance No pp. I'ORSTOM J Contribution i l'6tude du parasitisme des adultes d'a.funestus par Gastromermis sp. (Mermithidae). COZ, Soc. Path. exot 59: Bull. J Etude comparative des tntres et des V6randas-piges, comme moyen de sortie pour COZ, Koumbia (Haute-Volta). Cah. ORSTOM Ent. mid. 9: moustiques, J. 1973a. Contribution la biologic du complexe A. gambiae Giles COZ, Ati'ique occidentale. Cab. en Ent. mid. 11" ORSTOM J. 1973b. Contribution l'6tude du complexe A. gamtnae. R6partition g6ographique et saisonnibre COZ, cn de l'ouest. Cah. ORSTOM Ent. mid. 11"3-31. Afrique J. 1973c. Les m6canismes d'isolement g6n6tique dans le complexe gambiae Giles. Cah. COZ, Ent. mid. 11: ORSTOM J Contribution l'6tude du parasitisme des Ouest-Ali'icain (Mermithidac ct COZ, Cah. ORSTOM Ent. mid. 11: Coelomomyces). J. & PICQ, j. j Etude en laboratoire de la receptivit6 a Laveraniafalcipara d' gambiae A COZ, gambiae B. Bull. Soc. Path. exot. 65: d' W. J Experimental intbction of gambiae and Culex pipiens fatigans with Wuchereria CRANS, in coastal East Ali'ica. 7. reed. Ent. to: bancrofii H. A. & THERON, D. L A net, distribution record of CROSS, D6nitz.,7. ent. Soc. S. Afr. merus 46:155. of the gambiae complex. Heredity 42: members C. & WHITE, G. B Culicidae. In Limnofauna Europaea. Ed. J. Illes, Gustav Fischer Verlag, DAHL, pp Stuttgart. G Genetic Control of Insect Pests. Academic Press, London Net, York. DAVIDSON, G. & HUNT, R. H The crossing and chromosome characteristics of DAVIDSON, new, sixth species a the gambiae complex. Parassitologia 15: in 27: G. & LANE, J. DAVIDSON, Distribution maps tbr the gambiae complex. Unpubl. report studies at the London School of Hygiene and Tropical Medicine Progress rpt. Mosquito 40:2-3. no. G., PATERSON, H. E., COLUZZI, M., MASON, G. F. & MICKS, D. W The DAVIDSON, gambiae complex. In Genetics of insect vectors of disease. Eds. J. W. Wright & R. Pal. Elsevier Publ. Amsterdam, pp Co., complex. Tram. R. Soc. trop. Med. Hyg. 66: gambiae MEILLON, B The eggs of some South Ati'ican anophelines. Publ. S. Afr. Inst. reed. Res. 7:305. DE M. & SABATINI, A Cytogenetic observations on sp. A and B of the gambiae COLUZZI, 1968a. Cytogenetic observations on species C of the gambiae CORNET, M. & CHATEAU, R Quelques donndes biologiques sur Aedes (Stegomyia) luteocephalus 1907, en zone de savane soudanienne dans l'ouest S6n6gal. Cah. ORSTOM Ent. mid. 12:97- (Newstead), 110. CURTIS, C. F. & CHALKLEY, J Lack of recombination between the X chromosomes of diflirent DAVIDSON, G. & JACKSON, C. E Incipent speciation in gambiae Giles. Bull. H,Td Hlth Org. DAVIDSON, G. & WHITE, G. B The crossing characteristics ol'a new, sixth species in the 130

134 MEILLON, B. & VAN EEDEN, G (Cellia) deaconi n.sp., ti'om South Ati'ica DE Mosq. Syst. 8:335:342. (Diptera:Culicidae). MEILLON, B.,VAN EEDEN, G.J., COETZEE, L., COETZEE, M., MEISWINKEL, R., DU TOIT, DE L. N. & HANSFORD, C. F Observations on C. species of the funestus subgroup, a a suspected exophilic vector of malaria parasites in northeastern Transvaal, South Ati'ica. 37: T. S. & GILLIES, M. T Observations on the determination of the age composition and DETINOVA, importance of populations of gambiae Giles and funestus Giles in epidemiological Bull. Wld Hlth Org. 30: Tanganyika. DECO, M. A., RISHIKESH, N., PETRARCA, V. & COLUZZI, M \'ariazioni stagionali in DI gambiae e arabiensis a Kaduna, Nigeria. Parassitologia 23: M., CARNEVALE, P. & COZ,J Morphologie compar6e des espbces Aet B du complexe A. EYRAUD, Mesures de la spermathbque. Cah. ORSTOM Ent. mid. 10: gambiae. M. E. & LINTHICUM, K. J A handbook of the Amazonian species of FARAN, (Diptera:Culicidae). Mosq. Syst. 13:1-81. (Nyssorhynchus) A Oogenesis in the malaria mosquito gambiae. Cell & Tissue Res. 167" FIIL, T. A. & KIHAULE, P. M Report on a limited Anopheline survey at Ilhkara, South- FREYVOGEL, Tanzania. Acta trop. 25: Eastern H Culicid6s du Gabon. III. Anoph6lin6s. Ann. Parasit. hum.'comp. 10:85. GALLIARD, identification of sibling species of the gambiae complex. Heredity 38: the P. L., PICHON, G. & HAMON, J Etudes 6cologiques GAYRAL, la faune culicidienne d'une sur tbresti6re en zone de savane africaine deuxibme partie. Annls Soc. ent. Fr. (N. S.) 11: relique J. L Evaluation of some parameters of house-leaving behaviour of gambiae s.l. Acta GEROLD, 45: Leidensia M. E. C..!965. The problem of age determination in A. melas Theo. 1903, by Polovodova's GIGLIOLI, ORSTOM Ent. mid. nos. 3 & 4: method.cab. M. T The role of secondary vectors of malaria in north-east Tanganyika. Trans. R. Soc. trop. GILLIES, Hyg. 58: Med. M. T Vector behaviour and bionomics. In Textbook ofmalaria Eds. W. H. Wernsdortier & GILLIES, A. McGregor. In press. I. M. T. & DE MEILLON, B The Anophelinae ofafrica, South ofthe Sahara. Publ. S. Aft. Inst. GILLIES, Res. no pp. Med. M. T. & WILKES, T. J A study of the age composition of populations ot'gambiae GILLIES, and funestus Giles in north-east Tanzania. Bull. ent. Res. 56: Giles M. T. & WILKES, T.J A comparison ofthe range ofattraction of animal baits and ot'carbon GILLIES, for some West African mosquitoes. Bull. ent. Res. 59: dioxide M. T. & WILKES, T. J The range of attraction of single baits for GILLIES, West African some Bull. ent. Res. 60: mosquitoes. M. T. & WILKES, T.J The range of attraction of animal baits and carbon dioxide tbr GILLIES, Studies in a freshwater area of West Africa. Bull. ent. Res. 61: mosquitoes. M. T. & WILKES, T. J The vertical distribution of some West African mosquitoes GILLIES, over open farmland in a freshwater area of The Gambia. Bull. ent. Res. 66:5-15. (Diptera:Culicidae) M. T. & WILKES, T. J The effect of high fences on the dispersal of GILLIES, West At'rican some (Diptera:Culicidae). Bull. ent. Res. 68: mosquitoes M. T. & WILKES, T.J Field experiments with a wind tunnel on the flight speed of GILLIES, some African mosquitoes (Diptera:Culicidae). Bull. ent. Res. 71: West L. K. H Tests for multiple insemination in gambiae Giles. Nature 197: GOMA, studies. Parassitologia 13: autecological C. A Cytological maps for the practical identification of females 0fthe three freshwater species GREEN, the gambiae complex. Ann. trop. Med. Parasit. 66: of C. A A sex-limited esterase in the accessory glands ofmales offunestus. Mosquito News GREEN, Moswito C. & SHIDRAWI, G. R Malaria vectorial capacity ot'a population ol' GARRETT-JONES, An exercise in epidemiological entomology. Bull. Wld Hlth Org. 40: gambiae. M., SANTINI, G., PIMPINELLI, S. & COLUZZI, M Fluorescence.banding techniques in GATTI, C. A Identifications ofmember species ofthe gambiae complex in the Zambezi valley. GREEN, Afr. J. Med. 16: Centr. GREEN, C. A The practical problem of identifying members of the gambiae complex in 37: