Seasonal changes in the larvel populations of Aedes aegypti in two biotopes in Dar es Salaam, Tanzania

Bull. Org. mond. Sante' 11972, 47, 245-255 Bull. Wld Hith Org. J Seasonal changes in the larval populations of Aedes aegypti in two biotopes in Dar es Salaam, Tanzania MILAN TRPIS' The seasonal dynamics of larval populations of Aedes aegypti was studied in two different biotopes in Dar es Salaam, Tanzania. The first biotope was located on the Msasani peninsula on the coast 6 km north of Dar es Salaam, where A. aegypti breeds exclusively in coral rock holes. The population dynamics was studied during both the rainy and the dry season. Seasonal changes in the density of A. aegypti larvae depend primarily on variation in rainfall. The population of larvae dropped to zero only for a short time during the driest period while the adult population was maintained at a low level. The second biotope was in an automobile dump in a Dar es Salaam suburb, where A. aegypti breeds in artificial containers such as tires, automobile parts, tins, coconut shells, and snail shells. The greater part of the A. aegypti population of this biotope is maintained in the egg stage during the dry season. It serves as a focal point for breeding during the dry season: with the coming of the rains, the population expands into the surrounding residential areas. More than 70 % of the larval population developed in tires, 20 % in tins, 5 % in coconut shells, and 1 % in snail shells. Whereas yellow fever is largely endemic in West Africa (Chambon et al., 1967) and in Ethiopia (Serie, 1968), it has occurred only sporadically in East Africa, in Uganda (Haddow, 1965) and Kenya (Lumsden, 1954; Garnham et al., 1946). In Uganda, Aedes africanus and A. simpsoni were the vectors of yellow fever in a double cycle of transmission among monkeys, and between monkeys and man (Mahaffy et al., 1942; Haddow et al., 1947; Gillett, 1951, 1955). An epidemic of dengue-like fever affected some of the southern provinces of Tanganyika during 1952 and 1953 (Robinson, 1955; Lumsden, 1955) and A. aegypti occurred in very high densities in the huts of the inhabitants and was therefore thought to be the vector. Subsequently Ross (1956) reported the isolation of chikungunya virus from human cases and from A. aegypti in the area. Considering the large number of papers published on different aspects of biology of A. aegypti, sur- prisingly little is known about the seasonal distribu- tion of larvae. Lumsden (1955) made an entomo- 1 Entomologist/Ecologist, WHO East Africa Aedes Research Unit, Dar es Salaam, Tanzania. logical survey in the Newala district of south-east Tanzania, after the outbreak of dengue-like epi- demics in 1952-53, but it was not possible to make a systematic study of the ecology of A. aegypti larvae. Other reports dealing with A. aegypti larvae in Tanzania mention only the geographical distribu- tion ofmosquitos (Harris, 1942; van Someren, 1968), while Teesdale (1956) collected valuable data on the biology of adults in the coastal area of Kenya. Yellow fever has not been reported from Tan- zania, but A. aegypti and other Stegomyia mosquitos that are known vectors of yellow fever are very com- mon in Tanzania. One of the main objectives of the WHO East Africa Aedes Research Unit is to study the distribution and seasonal abundance of mosquitos suspected of being vectors of yellow fever virus. The effectiveness of large-scale chemical control programmes depends largely on the proper timing of insecticide application. Information on seasonal variations, especially in the larval population den- sity in relation to the environment, is important for the selection of the correct methods and timing of chemical control. It is believed that in this region 2893 -245- M. TRPIS A. aegypti exhibits unique breeding habits not seen in other parts of the world. This paper reports: (1) the seasonal distribution of the rural and urban populations of A. aegypti breeding in coral rock holes on the Msasani penin- sula and in man-made containers in an automobile dump in the city of Dar es Salaam; (2) estimates of the larval populations and of the mortality of A. aegypti; (3) the relationship between the distri- bution of rainfall and of larval and adult popula- tions of A. aegypti in two different biotopes; and (4) the importance of different types of artificial containers in the breeding of A. aegypti. DESCRIPTION OF THE BIOTOPES The climate of the Dar es Salaam area is generally hot and humid with small seasonal and daily varia- tions in temperature. The mean daily temperature is about 26°C, the mean seasonal range is 4°C, and the mean daily range is about 8°C. The coolest days and nights are in July and August. The highest temperature ever recorded was 35.2'C in February 1962 and the lowest temperature recorded was 12.8°C in August 1955. From May to the end of September, Dar es Salaam is in the south-east monsoon season. After a long passage over the sea this air stream becomes very moist and, by comparison with land tempera- tures, relatively cool. In October, the monsoon moves southwards over the Dar es Salaam area and from October to December there is a short rainy season. From late December to the beginning of March the north-east monsoon blows; it is warm, and gives only a little rain. This is the hottest season of the year as the sun is nearly overhead. From March to May the monsoon moves back north- wards and the main rainy season begins. There are thus pronounced seasonal variations in rainfall and winds, although the total rainfall is less than would be expected for an equatorial coast in East Africa, averaging about 1 100 mm a year. The variation in total annual rainfall may be as high as 50% above or below the average. The Msasani peninsula and Bongoyo Island are located about 6 km north of Dar es Salaam and have been fully described by Trpis et al. (1971). Msasani village is located on the south-western side of the peninsula and a new residential area is under construction on the south-eastern side. The Buguruni biotope is located in the southern part of Dar es Salaam and has an area of approxi- mately 1 ha; it contains wrecked automobiles, discarded tires, automobile parts, tins, coconut shells, and snail shells and is surrounded by African-type houses. There is a well in the middle of the dump that furnishes drinking-water for the surrounding houses. The dump area is shaded by mango and palm trees (these, which have been introduced, and occur in the following order of abundance: Mangi- fera indica, Anacardium occidentale, Cocos nucifera, Pithecellobium dulce); there are no shrubs. The spe- cies of herb and sedge that occur in the surrounding area are found in the following order of abundance: Acalpha indica, Commelina africana, Assystasia gan- getica, Cassia occidentalis, Cyperus sp. METHODS Msasani biotope Thirty rock holes of different sizes and in loca- tions were chosen for weekly sampling. Water with larvae was transferred with a pipette to a plastic dish. All the larvae collected were taken to the laboratory for identification of the species and the instar and for a complete count. The water was replaced in the rock holes. The number of holes sampled weekly was reduced to 15 in July 1969, the other 15 being sampled monthly in order to com- pare the results of weekly and monthly sampling. During the 6-month period of study it was found that only certain holes were able to hold water. Some of the holes were porous and they either did not hold water at all or they held it only for a short time (a few hours). The quadrat method was used to study the distri- bution of the holes able to hold water. The average number of rock holes with water was 34.8/ha dur- ing the rainy season in April 1970 and 9.0/ha during the dry season in July 1970. The data for rainfall were obtained from the closest meteorological station, which is on the coast about 6 km south of the study area. A rain gauge was set up in the biotope in January 1970 and since then rainfall has been measured daily. Comparison of the rainfall figures from two sources for an 8-month period in 1970 revealed only negli- gible differences. Buguruni biotope A. aegypti larvae were found in discarded tires, tins, coconut shells, and snail shells in the Buguruni biotope. A complete count of the different types of breeding place was carried out before the study 246 LARVAL POPULATIONS OF A. AEGYPTI IN TANZANIA began in April 1969, in order to calculate the larval productivity of A. aegypti in the whole automobile dump, and this was repeated in August 1969. As the opening round the inside of the tires is very narrow and it is impossible to extract all the larvae, 9 tires were chosen and marked for regular sampling. A 15-mm hole was made in the middle of each tire, and the tires were stood up with the holes uppermost. At the time of sampling, the tires were turned upside down and all the water with the larvae flowed into a pan. The larvae were than col- lected with a pipette and the water replaced after the tire had been returned to its original position. The tires were arranged in groups of three according to diameter (50, 75, or 100 cm) and were sampled weekly. Nine tins were also marked and grouped accord- ing to size (250, 2 000, or 4 000 ml) and were sam- pled at weekly intervals, as were 9 coconut shells (250-300 ml) and 30 Achatina fulica snail shells (60-150 ml). The number of larvae, their developmental stage, and the state of the breeding site were recorded weekly. The population of A. aegypti larvae per hectare was calculated from the mean number of larvae per container and the percentage of con- tainers with water. For the April-July period the total number of containers in the dump was used in the calculations. For the August-September period the number of containers found in the biotope in August 1969 was used. The rainfall data were obtained from the meteorological station at Dar es Salaam airport, which is about 5 km from the Buguruni biotope. The figures for 1970 obtained from the airport and those obtained from our own rain gauge showed considerable differences for some days during the 8-month period for which a comparison was possible. RESULTS Msasani biotope The proportion of the rock holes filled with water is shown in Table 1. During the 6-month period of study, only twice (5 August and 30 September) were all the rock holes found to be dry and usually 61-90% of the holes contained some water. Some of the rock holes that contained water were always found to be positive for A. aegypti larvae (Table 1). The daily rainfall, seasonal changes in larval density per hectare, and seasonal changes in the population of female A. aegypti as indicated by the 4-hour biting catches are all shown in Fig. 1. The numbers of adults are expressed in numbers caught per man-hour and the numbers of larvae per hectare are shown on a logarithmic scale. The density of A. aegypti larvae was high at the beginning of the study in April because of the heavy rains in March and especially heavy rainfall in April. The larval populations declined in May from 588 to 177 larvae per hectare. The next peak appeared in late May after more than 20 mm of rain. The population -dropped towards the end of June and remained at a very low level until the beginning of August. The small peak on 1 July (56 larvae/ha) was the result of a small shower in the last week of June after a 17-day dry period. After 2 months of very low rainfall (June-July) the 28 mm of rain between 6 August and 9 August produced an increase from zero to 5 290 larvae/ha. Larval density was high until the end of August and then slowly dropped to zero towards the end of September (Fig. 1). The adult population of females indicated by the 4-hour biting catches (3 meln in an orange tent) follows in general the same pattern as the curve for the larval population (Fig. 1). It may be seen that the peak of the larval population on 22 April was followed by a peak in adult population on 29 April. The decrease in the larval population between 22 April and 13 May was followed by a decrease in the adult population between 29 April and 26 May. The small increase in larval density from 13 May, which reached its peak on 26 May, was followed by an increase in the adult population from 26 May to 10 June. The decrease in the numbers of larvae in the period from 26 May to 23 June resulted in a decrease in the adult population between 10 and 23 June. The small peak on 1 July affected the numbers of adults on 9 July. On the occasions when the population of larvae reached zero (4 times dur- ing the period 24 June-5 August), the adult popula- tion did not drop to zero during the same period. The large increase in the larval population that occurred on 12 August, and the large larval popula- tions that were present until the end of the month, resulted in a considerable increase in the adult population during August and September. Although the larval density depends on rainfall, the number of larvae produced per millimetre of rainfall varies from season to season. During the rainy season the eggs deposited in rock holes hatch almost immediately after physiological aging. Dur- 247 248 M. TRPIS Table 1. Condition of the breeding sites and the seasonal distribution of the larval population of A. aegypti breeding in coral rock holes No. of Rock holes Rock holes positive Mean no. Estimated no. Estimate of larval Month Day rock holes with water for A. aegypti larvae of larvae per of larvae per productivity of sampled rock hole hectare ~~~~~~~~~~~theMsasanisampled ~No. %No. o l ae peninsula 15 30 12 40.0 7 23.3 46.3 1 611 350213 22 30 18 60.0 17 56.7 52.0 1 810 393 328 29 30 21 70.0 20 66.7 18.0 637 138 421 total 90 45 56.6 44 48.8 38.7 4058 881 962 6 30 25 83.3 22 73.3 16.9 588 127 831 13 30 22 73.3 16 53.3 5.1 177 38576 May 20 30 14 46.7 12 40.0 11.4 397 86 230 26 30 22 73.3 19 63.3 22.8 793 172 459 total 120 83 69.2 69 57.5 14.0 1 955 425 096 3 30 23 76.7 20 66.7 49.9 70 14 970 10 30 22 73.3 22 73.3 10.0 14 3 000 June 17 30 15 50.0 13 43.3 7.7 1 1 2 310 23 30 1 0 33.3 5 16.7 2.9 4 870 total 120 70 58.3 60 50.0 17.6 99 21150 1 15 4 26.7 4 26.7 39.7 56 11 910 9 15 8 53.3 6 40.0 9.1 13 2730 15 15 4 26.7 3 20.0 10.2 14 3 060 July 22 15 3 20.0 3 20.0 1.7 2 510 29 15 1 6.7 1 6.7 4.0 6 1 200 total 75 20 26.7 17 22.7 12.9 91 19 410 5 15 0 0.0 0 0.0 0.0 0 0 12 15 12 80.0 12 80.0 152.0 5 290 715 575 August 19 15 12 80.0 10 66.7 13.0 452 98 310 26 1 5 1 2 80.0 1 0 66.7 6.1 212 46110 total 60 36 60.0 32 53.3 42.8 5 954 859 995 2 15 9 60.0 7 46.7 5.9 53 11 527 9 15 10 66.7 3 20.0 1.4 13 2 827 16 15 7 46.7 5 33.3 31.6 284 61 770 Sept. 23 15 2 13.3 1 6.7 0.5 4 8 700 30 1 5 0 0.0 0 0.0 0.0 0 0 total 75 28 37.3 16 21.3 7.9 354 84 824 LARVAL POPULATIONS OF A. AEGYPTI IN TANZANIA 10 '0 '0 P0 10 o. '0 '010 205 _Ade -z X M ' " - 0 agypri .0 I A 0' a 102' 020*)'ii2 10.12 1 10 Slo1OA 2' IO i 1b2 02Apri~~~ ~IMa Jun Jul Auus Setme Oct. Apri I May June July August September Oct. 249 3.0 - COo 2.0 ,-' 1.5 c 0 0 0 a 05 -E -' CD 0,0 50 ao E E 30 - 20 c 10 CK 0 WHO 20654 Fig. 1. Seasonal distribution of rainfall and changes in the density of populations of A. aegypti adults and larvae in coral rock holes during the dry and the main rainy seasons, 1969. iQuIO 9000 8 000 P 000 6000 5000 4 000 3000 2 50C 2000 I 500 1 rooo~~~~~~~~~~~~1 0 4 t ~~~~Ii w b- o 4- 0 0 0 0) 0 0) 0 .0 E z 1 oC90 8070 60 SC 4 C a 30 2S 2C IC S 8 p a a E 4 -3 a 1 a~~~~~~~I 250 M. TRPIS ing the dry season the eggs accumulate and hatch only when flooded by the intermittent rains. During the period between 13 June and 5 August, when most of the holes were dry, the population of adults was still at a high level. The females laid their eggs on the walls of almost dry holes and eggs accumulated in the holes for almost 6 weeks. The 28 mm of rainfall between 6 August and 9 August caused a great increase in the larval population at the beginning of that month. It has been found that even 3 mm of rain is enough to flood and hatch some of the eggs that are laid near or in the bottom of the holes. Buguruni biotope Seasonal changes in population of larvae. During April and May, 299.1 and 205.4 mm of rain were recorded at Dar es Salaam airport. Populations of A. aegypti larvae were high when this study began on 14 April. Four peaks and four depressions of the population were recorded between April and July (Fig. 2). The first three peaks were probably Fig. 2. Seasonal distribution of rainfall and changes in the density of the population of A. aegypti adults and larvae in the automobile dump during the dry and the main rainy seasons, 1969. 20 000 15 000 A 10 000 4, 7000 A ° ooIO X IL \!\ g e 6000 2000 3000 -(larvae)- 300 2500 250 200 0 co 7500 _ _ 4, ~~~~~~~~~~~~~~~~~~~E l00 -aegypti 100 CD 800 aduts 9g 6.700 70 o 4,600 60 E 50n5 400 / 0 300 30 cot 250 25 200 20 .E 150 15 z 100oo 101020I bioI'1020 102 1 o20 1020ro ib20 I 21020 1 AprilI May June July August September October Nov. 70-- 70 60 - 60 so0 - 50 E 40 40 4 30 -30 20 20 10 1 20 ?20I,0b90'l ~ 11 0 1 10 0 J., I Li WH 20653 AprilI May June JulIy August September October Nov. LARVAL POPULATIONS OF A. AEGYPTI IN TANZANIA 251 caused by the high rainfall during April and May. There was not much rain during June and the large number of larvae was not related to the rainfall at the end of June and the beginning of July. However, it is probable that the rainfall in Buguruni was higher than that recorded at the airport. Probably the hatching in July of the eggs accumulated in June caused the increase of the larval population. Most of the containers were dry at the end of July and during August and the larval density in all breeding containers was zero or almost zero. The rainfall at the end of August and beginning of Sep- tember caused another sharp increase in the larval population. When the curve for the larval population is com- pared with that for the adult population (Fig. 2) it can be seen that the first peak of larval density on 28 April is reflected in the peak of adult density on 5 May, and that the next peak in the larval popula- tion on 19 May affected the adult density on 2 June. The build-up of the larval population between 2 June and 8 July caused the increase in the adult popula- tion between 16 June and 8 July. The decrease in the adult population between 8 July and 25 August is reflected in the drop in the larval population between 8 July and 15 September and the sudden increase in the number of larvae on 2 September had an effect on the build-up of the adult population between 15 September and the beginning of November. Breeding in different types of container The proportions of containers containing water in the breeding sites of the Buguruni biotope are summarized in Table 2. All the tires contained water from April to June but the proportion filled with water gradually decreased during July and the first week of August. All the tires were dry from 11 August until 25 August, but during September 22-67 % of them contained water. There were always some tins with water in them. From 3 % to 10% of the coconut shells held water between 14 April and 12 May. Snail shells contained water on 6 of the 24 sampling dates. The frequency distri- bution of the proportion of breeding containers with water is summarized in Table 3 for the 6-month period of study. The figures in Table 2 show that more than 70% of the larval population was produced in old tires; 20% developed in tins, 5% in coconut shells, and 1% in snail shells. Toxorhynchites brevipalpis larvae were found regu- larly in the Buguruni biotope, breeding mostly in tires and tins. The larvae of this species are known to be predators of A. aegypti larvae and undoubtedly affected the populations. Estimate of mortality of larvae. By comparing the numbers of first-instar larvae with the numbers of pupae (Beklemisev, 1949; Trpis, 1960), the average mortality of A. aegypti larvae was estimated as 83.3% in the Buguruni biotope and as 31.7% in the Msasani biotope. DISCUSSION It is generally believed that A. aegypti originated in East Africa, and that they spread from there to other parts of the world (Dyar, 1928; Christophers, 1960). In the present investigations it was found that some populations of A. aegypti in the coastal regions of Tanzania breed in coral rock holes, a habit not found in populations of this species else- where. The elevated coral reefs are generally unin- habited and are distributed along the whole coast of East Africa and it seems that they play an impor- tant role in the breeding of A. aegypti in this area. The populations are thought to be wild and accord- ing to Mattingly's classification they belong to the subspecies formosus and would be labelled as grade F-G according to McClelland's classification of the abdominal colour patterns.1 It must be stressed that the coral island populations may breed there without any contact with man; however, as soon as man enters the biotope they will readily bite him: from an epidemiological point of view this behaviour widens their vectorial potential. During the dry periods of August the population ofA. aegypti larvae on the Msasani peninsula dropped to zero, but the adult population, although low in number, was not eliminated. During the hot dry season the greatest part of the rock hole population is maintained in the egg stage. The embryonated eggs of A. aegypti are able to withstand drought in rock holes for as long as 4 months. It is interesting to note that A. simpsoni, another vector of yellow fever, occurs together with A. aegypti in the coral rock holes (Trpis et al., 1971). The bit- ing behaviour of the rock hole population of A. simp- soni in relation to man is similar to that of A. aegypti. Another breeding site in the rock hole biotope is the shells of the terrestrial snail Achatina fulica dur- 'Hartberg, W. K. (1969) Genetical assessment of taxono- mic characters of Aedes aegypti L. in Tanzania (unpublished document WHO/VBC/69.152). r, O eD i r GO s- p. tC _- M )- 0le0 N _. M Wx 0 M 0 '- 0 W .* r- W W M M)M)M)M)0 W N M ,- a _- _ N 1- o0 rl- I t 10 0 i- 00i '- 9%0 % CNle 0) 0) 0) 0 0 t o leh ED W Nv* M 0 a% i- N 0 0 C M C 0 0N _t 0 - CI 0) 0) 0) 0 0 m I a o r- Co - - - M 0 M 0 WD * _ W a , N N , n 0 9 0 Cd 0 0Lo M 0 M Ma C 00000o M M M 0 0 0 Cl1 rl 0a 00 i- 0) o o X * o Cl 0 0 0 0 r' 00) 9 C ON Cl 0 C4 O 0 0 La 00E _- 06 6 Q6i 0) 9l% 0 Cl N4Cl Cl . 0 ID0 co Cl M 0 M 0 0 0 W co co 0) 0) N Cl 0Wd W 0 N09% 90; 065 0 ME CE) co) ME N 9% 0 0 0 09%ME9% CM 0 0 o 0 6 0 C cO 9%. 0l 0 co co co 0) 0 M 0 N1- N C4 00 0 W 252 M. TRPIS CE 0 o CE)CE) 1- $l co C4 Co co 0 0) 0) 9 0 0 0) CE 9 i O. 0 0 - 0Y) CE) CEf) CE) C') 0) CR) 0 LCd o 00 9% C 0 CE CE 0O0 M N Cl DCo0 0 0 0) co r.. 0 C D 88- 0) CE)Ir C4 0 o o o0 0) 0 Cd o. co W0) 0) 0 0- 0 0 co W.: o: LO r- W o r09 0 9 co0 co CdCll 0> N Cl 9% co CE CE) co CE CE CE) CE) OR 9%9% 9% 0O 0 NO Cl 00 0 0 M M NCl 0 CL Cq co 0) co CE 9 Cl NN Cl 0 W WO. 6 0 0Drl 0) 9%: le 9% Cl Cl Cl M 0 0) 0 C 0 F coM0 M Mn r- Nt 0 - Mn M 0)0 0) LO MMco le LO0) 0 0. 0 40 o *-C c 0 U) 0 C1 D Co la c ._ 0 'C C c c 0 C. CU w -C 0 co 0 c ._ 0 C4 CU U) 0 C c00 -o 0 C 0C.) CU I- *M >E- co snoE IL - 0. z z O0co 00 C >. 0~ CL 0 ._ * z D tA 0.- °-' E I0 C U-.:> CL 0 ,MQ M @0, z ° 2 CL co 0 0 0 9 0) CE) C' r- 0) N 0 9 W" M) I.: a: M Co co 0 9 Cl CE) 00i 0) 0 It C8 N 9 8 9 Cl Cl Cl 9 8 0o Cl N o. 09 coME 06 0 0r 1- Cl Cl 0) 9 o. cn 0 Cs) 00- 0 0 Cl Cl Cl 0 0 ~9o 1* o N. 0 aq ad r- * co 66 M Pl01 0 6 06 0 0 C 0 6 065 0 909 8 90i 900 0 Cl N Cl 0 90 0 0 6i 0 0 0 0) N 0 0 0 C) co Cd 0 Cl le N OD co. 09 0 ME 00 . 0) 8 0 M)M)9%Cl -i00 _ NCC 0 LARVAL POPULATIONS OF A. AEGYPTI IN TANZANIA o w o w XCE N f% N U0 0 % 0t0 N 0U CE) C4 0) 0) O0 0 N UM O O O _ CE C CE CE N 0 0(W) LO 0) Co - - N 0 as I - 0 00 _% U) U)0) CE) CE) 0 9o o o: o 6 00o CE o 0 CE)oo 00000 0-iC - o~~~o o o o o C C C N 9 0 9 0 9 0) 0D N 0) (D , Co -- C C0 0 U 0 0 A 0 9 9o 0 0 0 0 0 6 6 6 6 6 0 0 0 0 o~~~ o o o o C C CE CE N o 0 0 0 9 o.~o.o.o.o. o. o. o. o. 0 0 0 CEP) NW- C4 C14 N O CE __ N eq0 9 9 T- I- CO o 0 0 0 0 q q q 0 0 a o o o o C CE C c N 0 0 0 06i 0 0 0 4 w- v- LOCD) -NN 00 C#)T-C_ ) o) O 00 O0 0. O 0 9 9 9 0 ON o o o 0 o r- o C4 0 C) CE) ei v- 0 0 0 0 0 06 0 0 6 0 0 N4 CE 9 9o 0 o a o o o o o o 0 6 0 0 0 0 0; 0 0 0 0 0 0 0 0 0 0 6 0 46 0 0 0 6 0 0) 0) 0 u-N0toN 0 253 ) N 0 r-U) 0 N r-. N am 0 o w- N cII) CE)O 0 - o. o. ,- . Ns o C)_ 0 CIO) 0 CE) OCE) 0 w- 0 w- v- 0 CE) 0 CEi6 w- 0 0 0 0 0 -U- (0 'U- It- o00 o6 0 o C; _ 0 T- o _D 0 0 0 0 0 0 0 0 06 0 0 0 06 0 0 0 0 0 0 0 0i 0 0 0 0 0 0 c. 0 0 0 0 0 0 0 0 46 0 06 0 0 0 0 C0 0 0 0 0 0 0. 0 .0 CDtm 0 C. 0 0 CL 0cu CD m 0 (. 0 -c 0 0 co (A - CD a) n 0 (A C. 0 U_ 0 S co 0 0 ._ 0 -c 0 - 0 Co r-0t 0E = Co C 2 CNo0 E-i >002=Ew=- o CO 0 00 C CD 0- Co 6 02 6 uz E > c~o7.> m go - 0w. 60= 0,. co en3z 0- 6. E Zo 0 C) 0 2 c) C 0) co cr 0 CE) o)00) 0 0 06o 1% I-. 0 CE) co cn 0o o. o. q O O aD 0 0 co co 0 (C Co NCE NCEf) N CE) N4 0 0 0 0 0 0 0 v- CE)Cr- LU) 06 0 0 0 6 0 0 00 0 0 0 0 6 0 0 0 0 0 0 0 C; 0 0 0 Ic Q O <0 0 0U) U) (0 U) U) ) (0 0 0 ( 0 C LO UC) 0o 0 0) ) r- C-NN 0 0 046 0 CIO) CE) (V) 0 0 0 0 0 0 0 N1 N 6 0 0 0 0 0 0 0 0 0 ci 0 0 0 d 0 Table 3. Frequency of occurrence of breeding con- tainers with water in the Buguruni biotope during the 6-month period Percentage of No. of observations containers Coconut Snail with water Tires Tins shells shells 0 3 0 19 18 1-20 1 0 0 6 21-40 3 2 2 0 41-60 2 5 2 0 61-80 3 5 1 0 81-100 12 12 0 0 ing the 3-month period of the main rainy season (April-June). Eretmapodites quinquevittatus was also frequently found breeding in these shells. Most cities and some villages have one or more places where wrecked cars and tires are dumped. During the rainy season populations of A. aegypti are very large in this type of biotope, which may also contain tins, coconut shells, and snail shells. Large numbers of A. aegypti eggs accumulate in the tires during the dry season and the first rain of the wet season causes a large increase in the mosquito population. It is, however, very difficult to measure the absolute density of the population of A. aegypti in an automobile dump during the rainy season. The number and condition of the wrecked cars and parts preclude any accurate count and they all contribute significantly to breeding during the rainy season. Thus it must be borne in mind that the larval, and perhaps the adult, population productivity during April and May is probably higher than that shown in Table 2. Some of the 30 rock holes in the Msasani biotope were sampled weekly and some monthly; no sig- nificant differences were found. It is concluded that a generalized curve for the seasonal distribution of the A. aegypti population in the Dar es Salaam area would have three peaks: one in October-November, representing a very large but brief increase; one between March and June, representing in increase as large as the first one and twice as long; and an intermediate peak in January, representing a smaller and shorter population increase. The application of these results should be of assistance in predicting seasonal changes in the population of A. aegypti and thus help to make large- scale chemical control measures more effective. ACKNOWLEDGEMENTS It is a pleasure to express my appreciation for the help of the team of mosquito scouts, who so diligently per- formed their tasks. I am also indebted to Dr Brian J. Harris, Head, Botany Department, University of Dar es Salaam, for the identification of the plants in the Msasani and Buguruni biotopes and to Dr E. J. Gerberg, Acting Project Leader, for his comments on the manu- script. This study was supported jointly by Public Health Service research grant No. CC 00261 from the Center for Disease Control, Atlanta, Ga., USA, and the World Health Organization. R1PSUM t FLUCTUATIONS SAISONNIE-RES DES POPULATIONS LARVAIRES D'AEDES AEGYPTI DANS DEUX BIOTOPES A DAR ES-SALAM (TANZANIE) La dynamique des populations larvaires d'Aedes aegypti a ete etudiee dans deux biotopes a Dar es-Salam. Le premier biotope est represent6 par les anfractuo- sites de coraux dans la peninsule de Msasani oui de nombreux gites favorisent la multiplication de l'espece. Les densit6s larvaires y fluctuent principalement en fonction du rythme des precipitations. On estime que pendant la saison des pluies, en avril-mai, la population larvaire, par hectare et par mois, est en moyenne de 653 500 larves; en saison seche, de juin a septembre, elle tombe 'a 246 300 larves. Les densit6s maximales de larves et d'adultes sont observees au moment oii les pr6cipitations sont les plus abondantes. Les larves ne dis- paraissent completement qu'en periode de s&cheresse tres intense (en aofit) alors que les populations d'adultes se maintiennent a un niveau faible. Les aeufs d6pos6s sur les parois des anfractuosites s'accumulent pendant la saison seche et une pluie relativement faible peut 254 M. TRPIS LARVAL POPULATIONS OF A. AEGYPTI IN TANZANIA 255 entrainer un accroissement explosif du nombre des larves apres plusieurs semaines de s&cheresse. Par contre, des precipitations de meme importance, pendant la sai- son des pluies, n'ont que peu d'influence sur les densites larvaires. Le biotope de Buguruni, dans la banlieue de Dar es-Salam, est un cimetiere d'automobiles. A. aegypti y trouve des gites artificiels (pneus hors d'usage, boites de conserves, pieces d'automobiles) et des gites naturels (d6bris de noix de coco, coquilles de mollusques). La courbe des densites larvaires suit en general les variations du regime des pluies. Pendant la saison des pluies, la population larvaire, par hectare et par mois, est en moyenne de 23 200 larves; elle s'abaisse a 12 000 larves pendant la saison seche. Les larves disparaissent quasi completement a plusieurs reprises en aoOt et en sep- tembre, et les adultes sont pratiquement introuvables entre le debut d'ao(ut et la mi-septembre. La majeure partie des populations d'A. aegypti subsiste, au stade de l'ceuf, pendant la saison seche. Avec l'apparition des pluies, les moustiques pullulent. Les vieux pneus four- nissent plus de 70% des larves, les boites de conserves 20%, les noix de coco 5% et les coquilles de mol- lusques 1 %. Des larves de Toxorhynchites brevipalpis, qui exercent une action pr6datrice sur les larves d'A. aegypti, ont ete trouvees en grand nombre dans le biotope de Buguruni. Leur presence entraine une mortalit6 estimnee A 83% chez les larves d'A. aegypti, alors que dans le biotope de Msasani la mortalite larvaire n'est que de 32%.' REFERENCES Bekleniiev, V. N. (1949) Ucebnik Medicinskoj tntomo- logii, Moscow, Medgiz, pp. 490 Chambon, L. et al. (1967) Bull. WidHlth Org., 36, 113-150 Christophers, S. R. (1960) Aedes aegypti, the yellow fever mosquito: its life history, bionomics, and structure, London, Cambridge University Press Dyar, H. G. (1928) The mosquitoes of the Americas, Washington, Carnegie Institute Garmham, P. C. C. et al. (1946) Bull. ent. Res., 16,473-496 Gillett, J. D. (1951) Ann. trop. Med. Parasit., 45, 110 Gillett, J. D. (1955) Ann. trop. Med. Parasit., 49, 154 Haddow, A. J. (1965) Trans. roy. Soc. trop. Med. Hyg., 59, 436-458 Haddow, A. J. et al. (1947) Bull. ent. Res., 37, 301-330 Harris, W. V. (1942) Bull. ent. Res., 33, 181-193 Lumsden, W. H. R. (1954) Bull. Wld Hith Org., 11, 103 Lumsden, W. H. R. (1955) Trans. roy. Soc. trop. Med. Hyg., 49, 33-57 Mahaffy, A. F. et al. (1942) Trans. roy. Soc. trop. Med. Hyg., 36, 9-12 Mattingly, P. F. (1967) Ann. trop. Med. Parasit., 51, 392408 Robinson, M. C. (1955) Trans. roy. Soc. trop. Med. Hyg., 49, 28-32 Ross, R. W. (1956) J. Hyg. (Lond.), 54, 177-191 Serie, C. et al. (1968) Bull. Wld Hlth Org., 38, 873-877 Someren, E. C. C. van (1968) Bull. ent. Res., 57, 207-220 Teesdale, C. (1956) Bull. ent. Res., 56, 711-742 Trpis, M. (1960) Biol. Prdce, Sek. biol. a lek. vied, 6, no. 3, pp. 140 Trpis, M. et al. (1971) Bull. Wid Hlth Org., 45, 529-531