Observations on the swarming and mating behaviour of Anopheles culicifacies Giles in nature*

Observations on the swarming and mating behaviour of Anopheles culicifacies Giles in nature* W. K. REISEN 1 & M. ASLAMKHAN2 The mating and swarming behaviour of A. culicifacies Giles was investigated during December 1975 at a cattle shed near the village of Sattoki, Lahore District, Punjab, Pakistan. On average, swarming commenced 20.9 min before sunset (light intensity 1414.4 lx) and ended 21.0 min after sunset (5.4 Ix) with pairing restricted to the period from 6.1 min before (467.2 Ix) to 15.8 min after sunset (26.9 lx). The swarms were principally composed of males, with females entering only for mating. On average, copulation lasted 27.2 s and was completed in flight. Mostfemales (71.8 %) collected while mating had taken a partial blood meal either the previous evening or on the same evening as mating. Allfemales in the swarms were nulliparous and 82.6% had ovaries developed to at least Christophers' stage hIa. Adequate knowledge of the swarming and mating habits of mosquitos is of importance in colonization attempts and may have direct relevance in genetic control experiments in assuring the adequate mating competitiveness of the laboratory-reared, released males. The mating and swarming behaviour of A. culicifacies Giles has been described only for an outdoor insectary population (1). Since genetic con- trol experiments with A. culicifacies are planned, we decided to study this behaviour in nature to ensure that the insectary observations were applicable to natural populations, especially since colonization procedures typically alter reproductive behaviour. METHODS AND MATERIALS All observations were made close to the Ghulam Mohammad cattle shed near the village of Sattoki, Lahore District, Punjab, Pakistan, which has been described previously (2) and is shown diagrammati- cally in Fig. 1. During the time of observation, buffaloes and cattle were tethered in front of the feeding troughs (Fig. 1). The elephant grass between the seepage canal and the cattle shed was about 4 m tall and extended above the roof. Fires were often lighted in the compound but the smoke did not appear to interfere with A. culicifacies swarming. * From the Pakistan Medical Research Centre, 6 Bird- wood Road, Lahore, Pakistan. 1 Research Associate. 2 Associate Professor. Swarms were observed from start to finish against the lighted western sky and were collected, in part, by several sweeps of a net to determine their compo- sition. Pairing times (the time the pair was first observed until the partners separated) were recorded by means of a stopwatch and the direction of the mating flight was noted. Other mating pairs were collected individually with a net, placed in separate tubes, and returned to the laboratory where the trophic condition was scored using the WHO criteria (3) and the ovarian condition was graded using Christophers' stages (4). Parity was determined by --=--l7 MAIN CANAL--===-= - - TREE I~ TE% S TROUGH tTROUGH E-N RTROUGH-j LOW PILE -ttLOW EARTHEN LOW OF STRAW 5eMNTROUGH WOn®1TRO ¶ ~ (1 CUTTING QUARTERS ) CATTLE SHED f, 4 / I CATTLESHEDH SEEPAE CANAL- Fig. 1. A diagram of the Ghulam Mohammad cattle shed and its environs with the locations of A. culicifacies swarm formations numbered 1 to 6. 3501 - 155 BULL. WORLD HEALTH ORGAN., Vol. 54, 1976 156 W. K. REISEN & M. ASLAMKHAN Table 1. Time in relation to sunset (- before and + after sunset) and light intensity at which swarming and pairing began and ended; based on 9 observations Time (min) Light intensity (lx) mean±SEa range mean+SEa range Swarming starts -20.9+1.70 -11 to -26 1414.4+79.4 1076.4 to 1614.6 Pairing starts -6.1 +1.49 0 to -12 467.2±72.8 86.1 to 699.7 Sunset 17 h 03.8 17 h 01 to 17 h 08 Pairing ends +15.8±0.92 +11 to +21 26.9+ 5.3 7.5 to 54.8 Swarming ends +21.0+0.94 +15 to +24 5.4± 1.4 2.2 to 12.9 a SE = standard error of mean. the degree of coiling of the ovarian tracheoles (5) and insemination was determined by spermatheca dissections. Temperature, relative humidity, and light intensity were recorded at 5-min intervals between 16 h 00 and 18 h 00 on four evenings using a Taylor hygro- meter a and a Kahlsico photometer b accurate to 0.1 footcandles (-1.1 lx). Photometer readings were all made at ground level with the photosensitive cell lying parallel with the ground. RESULTS AND DISCUSSION Swarm description Swarming activity commenced on average 20.9 min before sunset (Table 1) with one or two males flying slowly in circles above a selected portion of the courtyard. Swarms usually formed above low projections such as a mound of dirt or pile of straw (swarms 1 to 5, Fig. 1), but on occasion they also formed at the corner of the cattle shed (swarm 6, Fig. 1). Swarms were consistently present at these locations, but a typical "marker" was not recog- nized. Swarms never formed above domestic animals or humans, and they formed only in the immediate vicinity of the cattle shed. Usually the swarms were fairly low, 1 to 4 m high, and spherical in shape. The swarms sometimes moved vertically or horizontally, but always returned to the area of formation. The individuals within swarms consistently "faced" a certain direction, usually north-west but occasionally north or west; this is contrary to the findings of a Taylor Instrument Corp., Ashville, NC, USA. b Kahl Scientific Instrument Corp., El Cajohn, CA, USA. Russell & Rao (1) whose insectary swarms always faced east. Swarming at Sattoki began at a much higher light intensity (1414.4 lx, Table 1) than that observed by Russell & Rao, which was "about 2.0 footcandles" ( 21.5 lx). At Sattoki, A. culicifacies females bite cattle earlier in the evening during December than they do in August (2), and perhaps swarming activity also shifts to earlier in the crepuscular period be- cause of the colder winter temperatures. In the evening, light intensity and temperature dropped rapidly while relative humidity increased markedly c 1 164S 10I00.T TEMPERATUREM 14 Hz 4~~~~~~~~ o a- r .0 HUMIDITY- 86hT0~~~~~~~~~~~~~~1 0.HMIIT 8 601630 165 70 175 13 on 8, 10, 12, and 17 December 1975. One footcandle _1 0.7639 Ix. SWARMING AND MATING OF A. CULICIFACIES (Fig. 2). Perhaps this rapid temperature and/or humidity change altered the normal response to light intensity. Swarming continued until late twilight (Table 1), ending 21.0 min after sunset with a mean duration of 41.0 min (range = 36-47 min), which was considerably longer than the 2 to 20 min observed by Russell & Rao (1). Swarm composition The composition of samples from 25 A. culicifa- cies swarms collected in six different locations (Fig. 1) around the cattle shed was 544 males and 14 females (ratio = 38.9: 1), indicating that these swarms were composed essentially of males with females entering the swarms to mate. Apparently attracted to similar micro-ecological conditions or perhaps " markers ", males of A. stephensi Liston were collected within five A. culicifacies swarms and males of Culex pipiens fatigans Weidemann from one, while females of A. annularis van der Wulp, A. nigerrimus Giles, and A. pulcherrimus Theobald were collected once each from A. culicifacies swarms. However, females of these species were not insem- inated. Mating Pairing began on average 6.1 min before sunset and continued until about 15.8 min after sunset, the average duration being 21.7 min (range: 15-31 min) (Table 1). After the swarm had completely formed, pairs could be detected flying, in copulation, away from the main body of the swarm. Occasionally, when a female apparently approached the swarm, a group of the males would dart in her direction. Presumably slight changes in the tone of male wing beats stimulated similar behaviour, since many of these "darts " did not result in successful pair formation. Mating was " tip-to-tip ", with the copu- lating pair leaving the swarm and slowly drifting towards the ground in a " wobbling-type " flight. It was not discerned whether the male or the female led the direction of the flight. The average copulation time for 76 pairs was 27.2 s (standard error of mean = 0.92 s). Copulation was normally completed before the pair reached the ground. Copulation time was considerably longer than the 15 s reported by Russell & Rao (1); however, they also observed that the pair usually separated when contacting the screen wall of the insectary or the enclosed vegeta- tion. The direction in which 52 mating pairs flew after leaving the swarm was found to be random with 26.9% flying north, 23.1% south, 19.2% east, and 30.8% west (x2 = 1.538, P>0.05). Two pairs were observed to remhin within the swarm and fly upwards. Of 46 mating pairs collected by net and dissected, 28.3% had not fed, 19.6% had recently fed, and 52.2% had fed the previous evening. In 82.6% the ovaries were developed to at least stage Ila and all females were nulliparous as indicated by the tight coiling of their ovarian tracheoles (5). Most blood meals were weak or partial feeds and in no instance were the ovaries developed beyond stage II. A single freshly-fed, replete female was observed to remain in copulation for 45 s, falling to the ground before the pair could successfully separate. Another replete female collected by net remained in copulation even after transfer to the collection tube and this pair seemed to experience considerable difficulty in sepa- rating. Apparently a replete midgut may be a mechanical hindrance to mating. Blood feeding by unmated A. culicifacies females was reported by Russell & Rao (1) and was suggested previously by Reisen et al. (6) who found unfertilized, freshly-fed females resting in cattle sheds and virgin females feeding on buffaloes. This initial blood meal was apparently used instead of a sugar meal to mature the ovaries to the resting stage II (5). Reisen et al. (6) also found evidence of multiple blood feeding in A. culicifacies with 2 blood meals required, at times, to complete ovariole development. Thus, a given female may take as many as three blood meals from emergence to initial oviposition. A. culicifacies is the primary vector of rural malaria throughout much of the Indo-Pakistan subcontinent (7), and thus this increased incidence of man-vector contact has con- siderable epidemiological significance by increasing vector efficiency. ACKNOWLEDGEMENTS The authors would especially like to thank Mr Aziz-ur-Rehman, Mr Ghulam Nabi, and Mr M. Bashir, whose careful observations and technical assistance added greatly to this study. This research was funded by Grant No. AI-10049 from the United States National Institute of Allergy and Infectious Diseases and by PL-480 Grant No. 08-325-20 from the United States Communicable Disease Center. The administrative assistance of the United States Agency for International Development in Pakistan is gratefully acknowledged. 157 158 W. K. REISEN & M. ASLAMKHAN RxTSUMJt OBSERVATIONS SUR LE COMPORTEMENT D ANOPHELES CULICIFACIES GILES EN CE QUI CONCERNE LE VOL NUPTIAL ET L'ACCOUPLEMENT DANS LA NATURE Au cours de decembre 1975, on a etudi6 le comporte- ment de A. culicifacies Giles en ce qui concerne l'accoup- lement et le vol nuptial dans un abri pour des bovins a proximite du village de Sattoki, district de Lahore, Punjab, Pakistan en vue de verifier si les comportements decrits par Russell & Rao (1942) pour une population d'insectarium pouvaient etre directement extrapoles A des populations sauvages. Le vol nuptial commengait environ 20,9 min (eclaire- ment 1414,4 lx) avant le coucher du soleil et se termi- nait 21,0 min apres (5,4 lx); il durait 41,0 min. L'accou- plement etait limite A la periode allant de 6,1 min avant le coucher du soleil (467,2 lx) A 15,8 min apres (26,9 lx), et durant 21,7 min. En general, la population sauvage de Sattoki semblait effectuer le vol nuptial et s'accoupler plus t6t au cours de la periode crepusculaire et pendant plus longtemps que la colonie d'insectarium de Russell & Rao. La copulation durait 27,2 s et s'accomplissait en vol. La plupart des femelles (71,8%) recueillies pen- dant l'accouplement avaient pris un repas de sang partiel, soit le soir prec6dent, soit le soir meme de l'accouple- ment. Bien qu'il ne paraisse pas obligatoire, ce repas de sang a et6 considere comme pris a la place du repas de sucre normal et comme necessaire pour foumir 1'6nergie indispensable au developpement de l'ovariole jusqu'au stade II de Christophers. On a estime que l'ingestion de ce repas de sang suppl6mentaire accroissait l'efficacite de A. culicifacies en tant que vecteur du paludisme, du fait de l'accroissement du nombre des contacts homme- vecteur. Toutes les femelles s'accouplant etaient nulli- pares et chez 82,6% les ovaires avaient atteint un deve- loppement correspondant au moins au stade IIa. REFERENCES 1. RUSSELL, P. F. & RAO, T. R. On the swarming, mating, and oviposition behavior of Anopheles culici- facies. Am. J. trop. Med., 22: 417427 (1942). 2. REISEN, W. K. ET AL. Observations on the diel activ- ity patterns of some Punjab mosquitoes (Diptera: Culicidae). Biologia, 22 (1976) (in press). 3. WORLD HEALTH ORGANIZATION. Manual on practical entomology in malaria. Part IL Methods and tech- niques. Geneva, 1975. 4. CHRISTOPHERS, S. R. The development of the egg follicle in anophelines. Paludism, 2: 73-89 (1911). 5. DETINOVA, T. S. Age-grouping methods in Diptera of medical importance. Geneva, World Health Organiza- tion, 1962 (Monograph Series, No. 47). 6. REISEN, W. K. ET AL. Observations on the resting habits and diel changes in the ovarian condition of some Punjab mosquitoes (Diptera: Culicidae). Bio- logia, 22 (1976) (in press). 7. BHATIA, M. L. & KRISHNAN, K. A. A. culicifacies Giles, 1901. In: Vectors of malaria in India, 2nd ed. Delhi, National Society of India for Malaria and Other Mosquito-borne Diseases, 1961, pp. 147-186.