One-larva-per-container mosquito surveys conducted in Bangkok-Thomburi, Thailand, in 1969.

270 NOTE One-Larva-per-Container Mosquito Surveys Conducted in Bangkok-Thonburi, Thailand, in 1969* by ROBERT J. ToNN 1 & Y. H. BANG2 One-larva-per-container surveys have been con- ducted in Bangkok since 1966. Tonn et al. (1969) reported the results of surveys made during the three climatic seasons of 1967 at 14 sites in Bangkok-Thon- buri. In January and February 1969, the surveys were repeated by the WHO Aedes Research Unit at 13 of the original sites; in this report, the results of these surveys are compared with those obtained in the surveys that were made in December 1966 and in January and February 1967. Methods and materials Study sites. In the original surveys, an attempt was made to select a variety of sites so that the total re- sults would be representative of all Greater Bang- kok-Thonburi. Shop-house, slum, low-income, mid- dle-income, and high-income areas were chosen in northern, southern, and central areas, and a subur- ban garden area in Thonburi was also included. The areas were described by Tonn et al. (1969) and a summary is given in Table 1. Makkasan was not included in the present study since a pilot insecticide project was being carried out there. Collection methods. The 1-larva-per-container sur- vey method used in these surveys was described in de- tail by Sheppard et al. (1969). This method entails the collection of a single larva at random from each container in which there are larvae. The larvae were placed in collection bottles numbered to correspond with the principal breeding habitats (indoor water jars, miscellaneous indoor containers, ant traps, out- door water jars, miscellaneous outdoor containers, and natural habitats. All the larvae were found to be Aedes aegypti. The information gathered from the surveys in- cluded a premises index, a positive container index, * This study was supported by Public Health Service research grant No. CC 00261 from the Center for Disease Control, Atlanta, Georgia, USA. 1 Formerly Project Leader, WHO Aedes Research Unit, Bangkok, Thailand. Present address: Project Leader, WHO East Africa Aedes Research Unit, Dar es Salaam, Tanzania. 2 WHO East Africa Aedes Research Unit, Dar es Salaam, Tanzania. 27221 Table 1. List of study areas in Bangkok and Thonburi Area f Description Bangkok Sam Sen shop-house market near Chao Phraya Tha Tian shop-house-river port market on Chao Phraya River 22 July Circle shop-house, not near river or klong Soi Aree high-income, single-dwelling housing Soi 49-55 high-income, single-dwelling- housing Huay Kwang government housing project Klong Toei, Abattoir slums near Port-Authority office Soi King Phet low-income to slum housing Bang Pho northern low- to middle-income housing Krung Thep Bridge southern low- to middle-income housing Thonburi Bang Plat northern suburban residential area, including shop-houses Phran Nok Road shop-house and slum housing Along Me Klong urban low- to upper-middle-income housing and data on the kinds ofcontainer and the proportion of each kind present at the various sites. Since all the mosquitos collected were Ae. aegypti, the indices are an expression of the Ae. aegypti population. The results of the latest surveys were compared statis- tically with those of similar surveys completed dur- ing the cool season of 1966-67. In both series of surveys, 100 houses were sampled. Results A summary of the results of the two series of sur- veys is given in Table 2. The information is com- pared by means of a statistical analysis of data from MOSQUITO SURVEYS IN BANGKOK, THAILAND, 1969 tm> -X 0 .w * 0 .- a.o a) CD(0 a) N r-(0 a) (0 co m m0) am) '- N _- o o OD OD T- 0 ( 4 '- 0 t N- C') C' '4*N19 N lq C' C'NC'N (0 ( N '- 0) 0 0 C'* C W - (0 q c' N (0 t N (0 N X (0 N6iC' CN ni (0Nq C') C') N 0) N tD r)0~ - ( N a) 0 o, 0 O NO 0000003 N O (0rt N o--- - N 0 Niiso- O b a) _ cs X N _ o eD O N _ 4b _' _N~~~~~~ M w _- N LO N - 0O0CO N ONqo CO) - CO CD t o 8- 00 C ON q _ N _ CN N - c U1 oN oa0 ( NC (0 ) m 00 m N ' (0 N N N a LO co N0 a) It)0 00D '-6- No - C' C W(0_ O C') (0 0 c C') (0 a)O 0 '-0N0 (D 00 '-ND (0(0e °OQ.>a) (0(0 C') CO 1D'0 a) N (0N0 COO') coN a ON 0 ( 0 0 ( C') (0 a <aC tD~ ND_NNc 9N OU 0 I F0W aw s~~o co 0 N co co w 0 w m 1- 0 Kt C1 Cu > 00MCa) N (0 N N (04 0 C N N4 6- ° : O O a)~~~~~~c a) ~~~~~~~~~~~~ Z C a a It0C - N (0 NO N c L6 C '- '- '-N N N LON co o a -~~~~~~~~~~ r O N NLO C LON C a) N NO LO 0o 00 + I4- > 04N ' I :!N 8 oS 0 o cDqqqwO ( I06 N (0 C am ( 0 N N (0-I o.-C0 ((01-- 0Oa a C c r-C' N ° C' °c C' N C' I 0° C a)C')uu-NO(0C')u- N0IIM CC'-- -N N C C C- ) C) CD) Co. a) M On o CN N M -t onM enM eD Cu- i 2 C 00 CN ) '- o0 0 00 C') 00 0 a) CO aN 000 m C C (0a z N ( N v0t 3 W) IN t lgt le le 'q CD 0 Ln _N O co r 00 It a) CL) C-a) a C) C-N (o N O~~~~~~~~-- ' CI N) '-NzC D(e xe ba IXIt _D_ 0 c r- co 0L N O0 C' ON4 (00 N1 I ' C')CL N- rN N1 a) N a) 00(0 , - N t cNC '- CY) '- N1 N1 C) C' u- N0 t C' co (0 (0 a) (00( 0 LO C- 1-)a) W(0a) 0 I) C 0 LO (0) IO a)(0 C) (0 N NI Oc'0 '! Ec NW0(0 -t N N %-- It (00 a7) N (0 + C')(0'§CD L(0 - 0COO (0 N T-(0LO N CONo a) mt C') N "* lq (0 Nl N- (0( NOr v a)~~~~~~~~~~~~~~~~~~~~L (0 a) (00 C' (0 N (0 N- a) (0W C a) (0(0 N ( 0(0(0(00((0N0 0 0r- c (0C 0 C' I(0(0 C') 0(0Co ') N e(0 CDN - E W/ Cu 2. c F 0 4) C CD Y Y2 0) 00 Cmm ac m0 0 ') IxY: X LO X ZV. CD (D : C Y >- m _- *- z so I0W 2 'aee mT0 E 0)D. 0 c - c E 0 co al co I I =00 0 .0 0 0 WCC DL o 0 C_ co C C C 'a us 05 : 271 CD 0 (n q- to (0 0 0 co C E 0 U) c0 0 -5 0 -c C tn ._ 0o -1~- -l-1 NOTES the 1966-67 and 1969 surveys conducted at each site and of the average data for all surveys carried out during each series of surveys. Premises index. The average proportion of houses with breeding populations of Ae. aegypti (premises index) was slightly lower in 1969 than in 1966-67 (64.8% against 76.6 %). Only two sites, Sam Sen market and Soi 49-55, showed a slight increase. The premises index was lower at all the other sites, with Tha Tian market (45% against 20%) and Soi Aree (92% against 69%) exhibiting the greatest reduction. Total number of containers holding water. Despite the premises index being lower in 1969 than in 1967, the actual number of potential larval habitats (water- of containers inspected. Among the 6 sub-divisions sampled, the greatest increase was in the total number of ant traps, 11 of the 13 sites showing an increase. Container index. In 1969, the average number of water-filled containers per house varied from 11.0 at Me Klong to 3.5 at Tha Tian market and the average number of Ae. aegypti-positive containers per house varied from 3.9 at Krung Thep bridge to 0.4 at Tha Tian. Of the categories listed in Table 3, water jars were the most important habitat (30% of a total of 4 800 jars contained larvae). There were rather more jars indoors (2 684) than outdoors (2 105), but a slightly higher percentage of those outdoors were positive. There were about half as many ant traps Table 3. Average number of different types of larval habitat per 100 premises 1966-67 season 1 1969 season Containers I ~~~~~~~~~~~~Percentage ___ __ners Average no. __% Average no. __% change Total no. of containers 585.7 100.00 601.6 100.00 + 2.71 Water jars: indoor 189.4 32.34 202.7 33.69 + 7.02 outdoor 210.5 35.94 151.4 25.17 - 28.08 total 399.9 68.28 354.1 58.86 - 11.45 Ant traps 150.5 25.70 198.2 32.95 + 31.69 Miscellaneous containers: indoor 29.1 4.97 39.3 6.53 + 35.05 outdoor 6.2 1.06 10.0 1.66 + 61.29 total 35.3 6.03 49.3 8.19 + 39.66 filled containers) increased during this period. Table 3 shows that the average number of such containers increased by 2.4% in the period between the first and second surveys. As can be seen in Table 2, this slight increase in the total number of water-filled containers was not statistically significant. The num- ber of ant traps increased by about 31 %, and that of water jars decreased by about 11 %. However, water jars still represent about 60% of the Ae. aegypti larval habitats. Miscellaneous containers increased in num- ber by about 40 %; the more numerous, smaller mis- cellaneous containers may reflect a greater degree of urbanization in Bangkok-Thonburi. Altogether, 8 of the 13 sites showed an increase in the total number (2 636) as there were water jars, but the percentage occupied by Ae. aegypti was much higher (about 40%). Out of a total of 8 021 containers, 7 425 (92.6%) were water jars or ant traps; similarly, of 2 631 containers positive for larvae, 2 470 (93.3 %) fell in the category of water jars or ant traps. Again, 33.3% of water jars and ant traps combined were positive for Ae. aegypti, i.e., virtually the same pro- portion as the container index for all types of recep- tacle (32.8 %). Statistical analysis. Table 2 shows the results of a paired comparison of the two surveys. To derive t values for assessing the significance of differences be- 272 MOSQUITO SURVEYS IN BANGKOK, THAILAND, 1969 tween the years, individual differences for each loca- lity were compared with the mean difference for all localities and the variances and standard errors were determined; the t value is the mean difference divi- ded by its standard error. This procedure, which is the correct one when surveys have been carried out locality by locality, indicates significant differences between the 1967 and 1969 surveys for 6 of the 10 quantitative measurements listed in Table 2. If unpaired comparisons are made, thereby adding the variation from one locality to another, these diffe- rences disappear. For example, the standard error (of the difference between the means) for the total number of positive water jars becomes 30.19 and thus the t value (for 12 degrees of freedom) is 1.53. This value is below significance at the 0.05 probability level. If an analysis of variance is performed on the total number of positive water jars, it is found that whereas between years the F1,12 is 9.66, which is just significant at the 0.01 level, the F12,12 between locali- ties is 7.30, significant at the 0.001 level. The ana- lysis therefore indicates that the differences in indices for the 13 sites (owing to environment) were greater than the total difference between the 1967 and the 1969 surveys. In other words, even though there were significant differences between the results of the two surveys, these differences were not as great as those between the individual areas surveyed. Discussion The 1-larva-per-container method developed for larval mosquito surveys provided an excellent com- parison of kinds of habitat occupied and amount of breeding between different survey areas, or the same area surveyed at different times. The two series of surveys described here were conducted to illustrate that such comparisons can be made. In the 2-year period between the two series of sur- veys, many changes occurred; entire sections of slum areas disappeared and others appeared, according to population movements. This trend applies also, though to a smaller extent, to the dwellings of the various income groups. However, the general fea- tures noted in the second survey were identical with those found in the first survey. For example, the Chinese shop-house areas of Tha Tian and 22 July Circle have fewer containers and, overall, fewer mos- quito breeding places than do other areas. This is a result of the nature of the markets, i.e., they consist, basically, of crowded concrete slabs on which the inhabitants do not have enough room to store large amounts of water. Another example is the high-in- come housing areas of Soi Aree and Soi 49-55. These areas are inhabited mainly by foreigners, and a sur- prising amount of mosquito breeding occurs there. The surveys reveal that the servants, who maintain water jars and ant traps in their own quarters, are largely responsible for this situation. Both the premises index and the number of positive containers decreased significantly in the period be- tween 1966-67 and 1969. At the same time, the number of water-filled containers found and exa- mined in 1969 was slightly, but not significantly, higher. Of the various types of container, the in- crease was in the average number of ant traps (47.7), while there was an almost equally large de- crease (45.8) in the average number of water jars. This means that the majority of new containers in the areas surveyed in Bangkok were small miscellaneous containers. It must be remembered that changes in the premises index (which can result from the presence of a single positive container in formerly negative premises) and changes in the number of potential larval habitats are unlikely to have much bearing on the amount of mosquito breeding or the size of the mosquito popu- lation. Breeding within the ecosystem is determined, in part, by the number of containers actually serving as larval habitats and by the potential of each positive container to produce adult mosquitos. For example, a positive water jar produces, as a rule, more adults than a positive tin can, especially over a long period. Thus tin cans, broken jars, etc., that hold water temporarily will increase the container count of an area, but may cause a decrease in the proportion of positive containers, i.e., the container index. At the same time, this change may have little influence on the actual mosquito population within the survey area. Other studies made by theWHO Aedes Research Unit do not substantiate the occurrence of a general reduction in the Ae. aegypti population in Bangkok in the 2 years between the surveys. Therefore, much emphasis should not be placed on the premises index or on the number of positive containers when popu- lations are compared, unless the type of container and its larval productive capacity are known. There are a number of factors that could account for mosquito population changes in the survey areas. The ecology of specific premises, and to some extent that of the entire area, changes each time premises are occupied by a new resident. New containers re- place old ones and different concepts of cleanliness may influence the number of miscellaneous con- tainers in the environment. The same houses in a 273 274 NOTES survey area were not necessarily visited by the same collector during each survey, and differences occur both in the type of premises and in the effi- ciency of the collector. Finally, the 1969 survey was conducted during the cool dry season, a climatic period when a reduction of Ae. aegypti activity is most likely to occur, whereas the 1966-67 survey covered all three seasons. Although these factors cannot lead to the conclusion that there was a real reduction in the numbers of Ae. aegypti from 1966 to 1969, they do illustrate that conditions playing no actual role in determining the size of a mosquito po- pulation are capable of producing a significant diffe- rence in survey results. This should not, however, reflect adversely on the value of 1-larva-per-contai- ner surveys as a means of assessing Ae. aegypti populations in a specific area, even when the surveys are continued over a long period. The difference between the results of the two sur- veys was less significant than the differences in indi- vidual areas during each survey. Significant diffe- rences between the two surveys were apparent in the premises index, the total number of positive con- tainers, and the numbers of ant traps and positive water jars. H6wever, these differences were smaller than those between a shop-house and a slum area examined during the same survey. Thus surveys can perhaps be used to measure small changes in an ecosystem over a period of years but any such meas- urements must take into consideration factors of human population such as changes in economic sta- tus and demands for mosquito control within the area, as well as climatic and other non-human en- vironmental factors that might influence the num- bers of mosquitos. REFERENCES Sheppard, P. M., Macdonald, W. W. & Tonn, R. J. (1969) Bull. Wid Hlth Org., 40, 467468 Tonn, R. J., Sheppard, P. M., Macdonald, W. W. & Bang, Y. H. (1969) Bull. Wid Hlth Org., 40, 819-829