Tests for the existence of genetic variability in the tendency of Anopheles culicifacies species B to rest in houses and to bite man

Rulletin of the World Health Organization, 60 (3): 427 - 432 (1982) Tests for the existence of genetic variability in the tendency of Anopheles culicifacies species B to rest in houses and to bite man P. RAWLINGS1 & C. F. CURTIS2 Experiments were carried out in SriLanka on the malaria vectorAnopheles culicifacies using the mark-release-recapture technique. Collections were made in cattle-baited huts fitted with exit traps, and in nightbiting catches on human subjects. The same individual mosquitos were caught biting cattle andman on different occasions and the numbers caught demonstrated an overall preference for the cattle. The mean interval between successive blood meals in thefield was estimated to be 2.3 days. Following blood-feeding on a cow in a hut, A. culicifacies werefound to rest in the hutfor I or 2 days, whereas A. subpictus and A. varuna generally exited on the night offeeding or on thefollo wing night. Thesame individualA. culicifacies were recorded resting in a hutfor I or2 days on different occasions. The apparent absence ofgenetic variability in host choice or indoor, resting behaviour is considered encouraging for the prospects of control of malaria transmitted by A. culicifacies. Mosquito species differ in their tendency to feed on human or animal blood and in the length of time they are likely to spend resting in houses. These character- istics have obvious relevance to the role of the species in the transmission of disease, and to its vulnerability to control by spraying of houses with residual insecti- cides. Evidence has recently been presented for ge- netic variability within Aedes species in their tendency to bite man or animals (10), and Gillies (5) has obtained some response to selection for different host preferences in Anopheles gambiae. Coluzzi et al. (2) have reported an association between certain chrom- osome inversion types and indoor and outdoor resting of Anopheles arabiensis, while on the basis of epidemiological data, Molineaux et al. (8) have postulated the existence of indoor and outdoor resting genetic varieties of A. gambiae s.l. Such ge- netic variability could be the basis for the evolution of behavioural resistance to house spraying. Curtis (3) ' Formerly Research Fellow, Department of Entomology, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, England. Present address: Tsetse Research Laboratory, Department of Veterinary Medicine, Langford House, Langford, Bristol BS18 7DU, England. 2 External Staff Member, Medical Research Council, and Honor- ary Senior Lecturer, Ross Institute, London School of Hygiene and Tropical Medicine, London, England. proposed that it should be possible to obtain evidence for or against variability of the above kinds by releasing differentially marked wild mosquitos caught in the two situations. Recapture of the mosquitos at the time of their next reproductive cycle would show whether or not individuals tend to be found only in the same situation in which they were marked. This paper reports experiments of this kind in Sri Lanka with the malaria vector Anopheles culicifacies, identified as species B by Green & Miles (6). A. culicifacies has long been recognized as having a tendency to bite cattle and to rest in houses. The experiments investigated the possible existence of a distinct man-biting (anthropophilic) variety and of genetic variation in duration of indoor resting. Curtis & Rawlings (4) found that about 67% of samples caught in cattle-baited huts in Sri Lanka soon after dawn were freshly blood-fed and the remainder were semi-gravid, i.e., had taken their last blood meal at least 22 h previously (12). This suggested that some mosquitos were spending the second day of their oviposition cycle indoors and others outdoors. As a background for the main study on resting be- haviour, this paper also reports studies on the length of the oviposition cycle and on interspecific variability in resting behaviour. 4192 -427 P. RAWLINGS & C. F. CURTIS MATERIALS AND METHODS Host choice Biting catches were made between 18h 30 and 21h 00, using aspirators and torches, from a cow tethered in a hut and from the bare legs of 5 men sit- ting in an adjacent house. Mosquitos caught biting the men were marked with yellow fluorescent dust and those biting the cow with magenta dust; all the mosquitos were released outside the hut at the end of the evening. On subsequent evenings, all mosquitos caught on each host were immobilized and checked for fluorescence. Unmarked mosquitos were allowed to revive, marked magenta or yellow as before, and released. This procedure was carried out on 6 succes- sive evenings, and on the 7th evening, all captured mosquitos were killed and identified. Length of oviposition cycle Freshly blood-fed female mosquitos were collected soon after dawn from a cow-baited hut (4), marked with magenta dust, and released back into the hut. The next morning, all mosquitos in the hut were collected and anaesthetized with carbon dioxide. The reproductive status of all marked females was assessed by external examination and they were then released. Unmarked freshly-fed females were marked with yellow dust and released. Collections were made for the next 4 days; no further marking was done but the reproductive status of all recaptured mosquitos was assessed and recorded in relation to the number of days since that mosquito had been marked as a freshly blood-fed individual. Interspecies variability in indoor resting behaviour A mud-walled hut measuring 1.9 x 1.4 x 1.8 m with a palm-thatch roof-a type typical of rural Sri Lan- kan houses-was constructed at a site at Angun Ara, Sri Lanka, an area known to contain large numbers of A. culicjfacies (4). The hut had 2 windows, to which were fitted conical window traps of 46 cm diameter, and a palm-leaf door covered by sacking. It was considered that the mosquitos would attempt to leave the hut mainly through the well-lit window traps, although exit through the open eaves or the interstices of the palm-leaf roof was also possible. For two nights, the hut was baited with a calf overnight and the eaves were left open to allow entry of mosquitos. In the morning all the mosquitos inside the hut and in the window traps were collected using a mouth aspirator. Samples of the anophelines from these collections were identified using a dissecting micro- scope. The mosquitos that had been caught resting inside the hut after the second night were released back into it. The eaves were blocked and the hut was not baited to prevent the entrance of other mosquitos. Collections were made from the window traps on the 2 subsequent mornings. Test for intraspecific genetic variability in indoor resting behaviour. An experiment was carried out to determine whether genetic variability might be responsible for different indoor resting tendencies. The procedure is diagrammed in Fig. 1. A hut with open eaves (hut A) was baited overnight with a cow, and freshly blood- fed female A. culicifacies were obtained the following morning. These were then released into hut B, which was fitted with window traps, and the eaves of which had been blocked by extending the mud walls up to the roof. The mosquitos were recaptured the next day and differentially marked according to whether they had remained resting in the hut or had exited into the window traps. Both types were released into hut A from which they could exit to oviposit and to which they could return to feed on the cow. Three complete cycles of events were observed as indicated in Fig. 1. RESULTS Host choice Table 1 shows that about 6 times as many anoph- elines were caught biting the cow as were caught on the five men. In the prevailing conditions of semi- darkness, it was not feasible to identify the species of HutA HutB Night Bbod fed Day I Colfetfrahly p Retase in hot Bfed femfebs ee.0, "*~ Collct from Colba from Niht exit traps, inide hut, merk oranpg rk white Day2,4,6 ReleaseinhutA Ix I MoIquito exit to Night onipoitand rtuom to refeed I Dey 3, 5, 7 Cfld fryes_ _ _ _ Refm in hut Bfad fenufes 4 Collect from exit Collect fram traps, identify inude hut, identitfyDay 4, 6, 8 ncptu m rk acapture, mtrk annuinder ornp rnsminder white WHO 81IJ26 Fig. 1. Outline procedure for study of genetic vari- ability in resting behaviour of A. culicifacies. Hut A was baited and had open eaves; hut B was not baited, had closed eaves and window traps. 428 BITING AND RESTING HABITS OF ANOPHELES CULICIFACIES Table 1. Number of anophelines caught biting cattle and human subjects on 7 successive nights; the first 4 columns refer to recaptures of marked mosquitos 1 st feed Species composition Total no. on 7th night 2nd feed Man Cow of anophelines (all captures) caught A. A. A. A. A. A. culicifacies subpictus culicifacies subpictus culicifacies subpictus Man 2 0 2 0 188 16 0 Cow 4 1 17 12 1150 89 27 these live mosquitos prior to release. However, on the seventh night, all captured mosquitos were killed and identified, and A. culicifacies was found to show approximately the same 6:1 preference for the cow (Table 1). The recaptured marked mosquitos were found to include both A. culicifacies and A. subpictus. Two A. culicifacies originally captured on the cow were recaptured biting man, and 4 found originally on the men were recaptured on the cow. This disproves the hypothesis that there are distinct anthropophilic and zoophilic populations of A. culicifacies in this area. A Fisher's exact test showed no significant associ- ation between choice of host of A. culicifacies at successive feeds (P = 0.23), but the numbers recap- tured were small, and may have failed to detect a slight association of this kind. Length of oviposition cycle Pooled recapture data from the two releases of freshly blood-fed female mosquitos are shown in Table 2. On day 1, a very small number (probably nullipars) were found to have re-fed but most of the recaptured mosquitos were semigravid. The largest number of freshly blood-fed females were recaptured on day 2, indicating that 2 days is the usual interval between blood feeds. However, a considerable number of freshly blood-fed mosquitos were recap- tured on day 3, indicating a gonotrophic cycle lasting Table 2. Number and reproductive status of female A. culicifacies recaptured on days 1 - 4 after a blood feed No. recaptured Reproductive No. status released Day 1 Day 2 Day 3 Day 4 Blood-fed 2443 4 192 78 76 Semi-gravid 50 2 12 4 Gravid 0 6 41 13 3 days. It can be assumed that most of the blood-fed mosquitos captured on day 4 had completed two cycles since their release. The mean length of the oviposition cycle was calcu- lated from the number of freshly blood-fed females caught on days 2 and 3 (192 and 78 respectively). This gave a value of 2.3 days for the average length of the oviposition cycle. Interspecies variability in indoor resting behaviour Almost all the A. culicifacies remained in the hut on the day after feeding, while most of the A. varuna exited into the traps, and other species were found about equally in the hut and the traps (Table 3). The same pattern of distribution of the species was found on day 2, when all mosquitos caught in the hut were returned to it. All the A. subpictus and minority species exited the next night. About half of the A. culicifacies caught in the exit traps entered them the night after feeding and the remainder on the following night. However, the traps accounted for only 24%o of those that were returned to the hut. The others are presumed to have left the hut through the palm-thatch roof or to have died. Test for intraspecific genetic variability in indoor resting behaviour Although the above study on resting behaviour showed virtually no A. culicifacies that were "exo- philic" in the sense of leaving a house immediately after feeding, it did indicate that individual mosquitos rested for different periods. This was found not to be a result of genetic variation, however, since individual mosquitos were found to rest for a different number of days in successive oviposition cycles. The contin- gency table (Table 4) yields a non-significant x2 value, indicating that there is no evidence for an association between an individual's resting behaviour in successive cycles, i.e., there is no evidence for the existence of "more endophilic" and "less endo- philic" genetic varieties. 429 P. RAWLINGS & C. F. CURTIS Table 3. Mosquitos found resting indoors and exiting into window traps after a blood meal on a cow in a mud hut with palm-thatch roof Percentage distribution of species Total Day' Location no. of No. A. culicifaciesanophelines examined A. A. Other caught Semi- varuna subpictus ano- Unfed Fed gravid Gravid phelinesb 1 Inside hutc 160 47 0 85.1 0 0 4.2 4.2 6.4 1 Exit trapsc 113 43 0 7.0 0 0 72.1 7.0 13.9 2 Inside hutd 263 47 0 76.5 12.7 0 0 8.5 2.1 2 Exit trapsc 78 78 1.3 1.3 1.3 0 53.8 34.6 7.7 3 Exit traps 37 37 10.8 0 56.7 0 0 27.0 5.4 4 Exit traps 26 26 3.9 0 0 96.1 0 0 0 On the nights preceding days 1 and 2, the hut was baited with a cow and the eaves were open. On the nights preceding days 3 and 4, the hut was unbaited and the eaves were closed. b Other species collected were A. vagus, A. tessalatus, A. annularis, and A. jamesi. ' Released away from the hut. d Returned to the hut. Table 4. Number of mosquitos recaptured inside hut and in window traps after successive blood feeds" Site Site of initial capture Total no of. of anophelines recapture Window trap Inside hut B caught Window trap 11 12 941 Inside hut B 13 7 384 a Contingency table, X2 = 0.64, P> 0.3. DISCUSSION Control of malaria transmitted by A. culicifacies should be easier to interrupt by residual spraying of houses than is the case with some other vector species (1). Virtually all A. culicifacies were found to rest on the walls for a day following feeding and some re- mained resting for two days (Table 3). Even those that exited after one day's rest may enter another house for a further day's rest (11). Although A. culicifacies may bite man or cattle and may have varying periods of indoor resting, no evi- dence was found that these characteristics were the properties of different genetic varieties. On the con- trary, the different forms of behaviour were demon- strated in the same individual on different occasions. This is encouraging for the prospects of a long-term solution to the problem of malaria transmitted by A. culicifacies. In contrast, in the Sudan savanna region of Africa, Molineaux & Gramiccia (9) reported the existence of a genetic variant of A. gam- biae s.l. that is known to bite man but does not rest in houses and can therefore survive in a sprayed area long enough to develop sporozoites. This presents an almost insuperable obstacle to malaria control by house spraying. The evidence for such genetic vari- ation in A. gambiae s.l. is still somewhat circumstan- tial and couild be investigated further by the methods outlined in the present paper. The fact that the same individuals have been shown to bite both man and cattle (Table 1) gives support to the concept of increasing the number of cattle near human habitations to "screen" human populations by diverting mosquitos to the cattle, which are their preferred hosts. Marginal farming, in which the wild animals are driven away but domestic animals are not introduced, may be particularly dangerous from the point of view of malaria. It is believed that anophelines in some parts of Asia have evolved behavioural resistance to insecticides by becoming more exophilic (7). However, in the case of A. culicifacies in Sri Lanka, our results on indoor rest- ing behaviour give no evidence for genetic variability on which natural selection could operate to produce such a behavioural resistance. 430 BITING AND RESTING HABITS OF ANOPHELES CULICIFACIES 431 ACKNOWLEDGEMENTS Dr A. N. A. Abeyesundere, Superintendent of the Sri Lankan Anti-Malaria Campaign, fully supported the above research and provided transport. Mr D. C. Goonatilaka and Mr C. Wickramage assisted in the collecting and running of the experiments. Professor G. Davidson, Dr G. B. White, and Mr J. Lines criticized the script. Financial support was provided by the British Medical Research Council. RESUME EPREUVES POUR RECHERCHER L'EXISTENCE D'UNE VARIABILITE GENETIQUE DANS LA TENDANCE D'ANOPHELES CULICIFACIES, ESPECE B, A SE REPOSER DANS LES HABITATIONS ET A PIQUER L'HOMME Des experiences r6cemment publi6es, ainsi que des obser- vations sur les moustiques vecteurs de maladie, ont revel une variation g6netique intrasp6cifique dans la prefrence trophique et la tendance A se reposer IA 'int6rieur des habitations. Des exp6riences de marquage-lAcher-recapture ont donc ete realis6es pour rechercher si une telle variation existe chez le vecteur du paludisme, Anopheles culicifacies, au Sri Lanka. Des captures nocturnes de moustiques piquant une vache et cinq hommes, suivies du marquage, du lAcher et de la recapture, ont revele des cas de moustiques individuels venant sur la meme espece d'h8te ou sur une espece d'hote diff6rente pour des repas sanguins successifs, sans qu'il y ait de signes d'une variabilite g6netique dans la pref6rence trophique. Le nombre total de moustiques captures a montre que A. culicifacies pref6rait se nourrir sur la vache que sur les hommes. La dur&e du cycle d'oviposition naturel (temps compris entre des repas sanguins successifs) etait en moyenne de 2,3 jours: la plupart des femelles recapturees revenaient se gorger A nouveau au bout de 2 jours, mais certaines revenaient apres 3 jours. L'6tude des moustiques attir6s par une vache servant d'appAt dans une hutte et captures dans des pieges de fenetre ou au repos a l'int6rieur de la hutte a montre que la plupart des A. varuna et A. sub- pictus cherchaient A s'echapper de la hutte la nuit meme du repas sanguin, alors que presque tous les A. culicifacies se reposaient A l'interieur de la hutte. Les moustiques r6cemment gorg6s, captures au repos A l'interieur de la hutte, y ont ete remis et le moment ou ils cherchaient A s'6chapper 6tait etudie A l'aide de pieges de sortie. La moitie des A. culicifacies ont ete ainsi captur6s la premiere nuit apres le repas sanguin et l'autre moitie la nuit suivante. Pour determiner si ce comportement de repos d'un ou de deux jours etait une caract6ristique individuelle des femelles adultes, on a lache dans une hutte des femelles gorg6es, et celles qui ont e captur&es dans les pieges de fenetre le soir suivant (repos de 1 jour) ont ete marqu6es d'une certaine couleur, alors que les femelles captur6es a l'int6rieur de la hutte (repos de 2 jours) 1'etaient d'une couleur diff6rente. On a relache les deux cat6gories et on les a laissees pondre dans la nature et revenir pour un nouveau repas sur une vache-appat. Les femelles gorg6es ont ete a nouveau lich6es dans la hutte et on a recueilli celles qui pOn6traient dans les pieges de fenetre et celles qui se reposaient dans la hutte; les endroits oii les recaptures 6taient faites ont ete enregistr6s. Les resultats n'ont pas montr6 de caractere constant dans le comportement individuel en ce qui concerne le repos lors de plusieurs cycles d'oviposition successifs, c'est-a-dire qu'il n'y a aucun argument en faveur de l'existence de variet6s plus ou moins endophiles. Ces r6sultats sont discutes du point de vue de la lutte antipaludique. II semble qu'il serait possible de o proteger >> des populations humaines en accroissant le nombre de bovins, car ces derniers sont les hotes animaux pr6fer6s deA. culicifacies. Il est probable que l'absence apparente de varietes comportementales g6n6- tiquement distinctes dans l'6spece A. culicifacies retardera l'evolution de la resistance comportementale et prolongera l'efficacite de la lutte au moyen de pulv6risations d'insecti- cide a l'int6rieur des habitations. REFERENCES 1. BRUCE-CHWATT, L. J. Essential malariology, London, Heinemann, 1980. 2. COLUZZI, M. ET AL. Behavioural divergences between mosquitoes with different inversion karyotypes in poly- morphic populations. Nature, 266: 832- 833 (1977). 3. CURTIS, C. F. Reviewer's comment. Tropical diseases bulletin, 76, Abstract 3045: 1131 - 1132 (1979). 4. CURTIS, C. F. & RAWLINGS, P. A preliminary study of dispersal and survival of Anopheles culicifacies in relation to the possibility of inhibiting the spread of insecticide resistance. Ecological entomology, 5: 11 - 17 (1980). 5. GILLIES, M. T. Selection for host preference in Anoph- eles gambiae. Nature, 203: 852- 854 (1964). 6. GREEN, C. A. & MILES, S. J. Chromosomal evidence for sibling species of the malaria vector Anopheles (Cellia) culicifacies Giles. Journal of tropical medicine and hygiene, 83: 75 - 78 (1980). 7. ISMAIL, 1. A. H. ET AL. Studies on malaria and res- ponses of Anopheles balabacensis balabacensis and Anopheles minimus to DDT residual spraying in Thai- land. Part I. Pre-spraying observations. Acta tropica, 31: 129- 164 (1974). 432 P. RAWLINGS & C. F. CURTIS 8. MOLINEAUX, L. ET AL. Assessment of insecticidal impact on the malaria mosquito's vectorial capacity from data on the man biting rate and age composition. Bulletin of the World Health Organization, 57: 265 - 274 (1979). 9. MOLINEAUX, L. & GRAMICCIA, G. The Garki project. World Health Organization, Geneva, 1980. 10. MUKWAYA, L. G. Genetic control of feeding prefer- ences in the mosquitoes Aedes (Stegomyia) simpsoni and aegypti. Physiological entomology, 2: 133 - 145 (1977). 11. REISEN, W. K. ET AL. Observations on the diet activity patterns of some Punjab mosquitoes (Diptera, Culi- cidae). Biologia, 22: 67 - 77 (1976). 12. SLOOF, R. & HERATH, P. R. J. Ovarian development and biting frequency in Anopheles culicifacies Giles in Sri Lanka. Tropical and geographical medicine, 32: 306-311 (1980).