|
|
||||||||
| ABSTRACT |
|
|
|---|
| INTRODUCTION |
|
|
|---|
The diurnal activity of adult Stegomyia with immature development in containers limits the availability of collection methods for large-scale sampling, which is a major drawback in epidemiological surveillance of arboviruses as well as the evaluation of the impact of control strategies and detection of allochthonous mosquito invasion into nonendemic regions.7 Development of new operational techniques for collection of adult Stegomyia females that can be used to monitor their densities would, therefore, be a valuable contribution to the prevention and control of arboviruses, such as dengue.7
Recently, Facchinelli and colleagues8 described a new sticky trap (ST) for collection of adult Stegomyia species, whose females lay eggs in small water containers. They reported that it was effective for collecting Ae. albopictus females and monitoring their population dynamics in Rome, Italy. The objective of this study was to evaluate the ability of the ST to collect adult Ae. aegypti and Ae. albopictus in a dengue-endemic region of Thailand compared with the CDC backpack aspirator (BA).7,9 Backpack aspirators are routinely used to collect Ae. aegypti10 and were already being used at the study area to investigate dengue epidemiology and evaluate vector intervention strategies.11,12
| MATERIALS AND METHODS |
|
|
|---|
Study design. Collections performed by sticky-trap (ST) and by modified CDC backpack aspirator Model 1412 (BA)10 were carried out from 4 November to 13 December 2005, for a total of 39 sampling days. All households in the study area were georeferenced before the trial and mapped using MapInfo® software. A total of 40 houses—16 in village 1 and 24 in village 2—was selected on a grid composed of 150 m x 150 m cells drawn on village maps. Each household was selected within a radius of 25 m from each grid crossing. As a consequence, selected households were located at least 100 m from each other (a distance assumed to be adequate to ensure minimal interaction between the 2 collecting methods used in contiguous households, based on the limited dispersal of Ae. aegypti15). Over the 40 selected houses, STs were placed in 20 of them (8 in village 1 and 12 in village 2) and BA was performed in the remaining 20 (8 in village 1 and 12 in village 2). Within each village, collecting methods were alternated in contiguous households.
Six STs were used per household: 3 indoors and 3 outdoors (120 in total). We considered indoor locations to be those spaces surrounded by at least 3 walls and covered by a roof. Indoor STs were located in dark places and room corners, near racks with clothes, and at least 2 m away from water containers. Outdoors, STs were placed within 3 m from the edge of the house roof, close to walls or vegetation, in shaded places, and at least 2 m far from water containers. Because Aedes collections in individual traps were low, for statistical analyses, the 3 STs, based on location (indoors versus outdoors), were considered a single sampling unit.
Sticky traps were serviced twice a week, replacing the adhesive sheets with new ones previously coated with glue with a spatula and filling the trap with fresh water collected from the Ping River. Mosquitoes caught on the adhesive sheets were counted and identified in the field with the help of a magnifying lens.
Backpack aspirator collections were carried out in the remaining 20 households without traps. As with the ST, BA collections were performed twice a week both indoors, in all accessible rooms, and outdoors on objects and vegetation within 3 m from the edge of the roof. The duration of the collection effort varied from 9 to 20 minutes, according to the size and the complexity of each premise. We considered as a single sampling unit to be each collection performed at the same premise and date, either indoors or outdoors. Captured mosquitoes were transported back to the Armed Forces Research Institute of Medical Sciences (AFRIMS) Entomology Laboratory in Kamphaeng Phet, killed by freezing, counted, and identified to species.
Aedes were separated by species and gender. Other Culicidae were identified to genus, and only Culex from ST collections was separated by gender. In only 4 of the 1,320 total STs serviced during the entire investigation, the number of Culex spp. was estimated as > 100, rather than counted.
Our sampling protocol was designed so that the human effort necessary to carry out catches with the 2 methods (i.e., 2 teams of 3 technicians, twice per week) was comparable. Therefore, results can be compared even though the 2 collecting approaches are intrinsically different. ST collects mosquitoes continuously for the entire time between trap inspection, whereas BA represents a "snapshot" of the adult population present at the specific time when the collection is made.
Statistical analysis. Collection results were analyzed separately for Aedes and Culex. A sampling unit was considered positive when either 1 (or more) Aedes female or Culex specimen was collected during a single service. When 2 out of 3 STs in the sampling unit were nonoperative (i.e., without water or turned upside down by residents), we excluded the entire unit from analysis.
The distribution of mosquito counts was positively skewed. An attempt to normalize data with a log-transformation left the distribution not normal (test for normality based on skewness and kurtosis P < 0.001). We, therefore, compared the proportions of positive catches in the two collecting approaches carried out indoors and outdoors by
2 tests, 2-sided Fisher exact tests or Mantel–Haenszel
2 tests. For the outcome of interest, the relationship between mosquitoes per catch, method of capture, locality (outdoors, indoors), and village and household (considered as clusters) was also examined by regression analysis using a negative binomial model. The latter is considered more appropriate when data are over-dispersed (i.e., when observed variance is greater than the mean mosquito number). The degree of over-dispersion was measured as the variance mean ratio (VMR).
For descriptive purposes, data are presented as total number of catches, number of positive catches, and range (i.e., minimum and maximum number) of mosquitoes for each catch. All statistical analyses were performed using STATA 9 statistical analysis software (Stata Corp., College Station, TX).
| RESULTS |
|
|
|---|
A total of 19,007 adult mosquitoes were collected and identified, of which 10,761 were collected during 348 attempts in village 1 and 8,246 during 511 attempts in village 2, respectively. In both villages, Culex spp. represented the majority (~90%) of mosquitoes collected. The rest of the mosquitoes collected were Aedes spp., Armigeres spp., and Anopheles spp.
Sticky traps and BAs collected a total of 11,524 and 7,483 specimens, respectively. Table 1
shows the total number of ST and BA catches and the total number of mosquitoes caught, either indoors and outdoors, in the two villages, subdivided by species and gender for Aedes. Gender and species identification was successful for 98.4% of the Aedes specimens collected by ST and for 100% of those collected by BA.
|
|
Females.
Comparison of the proportion of positive catches obtained by the 2 methods, carried out adjusting for data collected in the 2 villages, showed that outdoors the frequency of positive ST catches was significantly higher than that of BA catches (70.8% versus 9.0%,
2 M-H = 167.9, P < 0.001), whereas indoors, no differences between the 2 methods were observed (52.7% and 51.9%,
2 M-H = 0.01, P = 0.94). This is confirmed by the regression model controlling for clustering of catches repeated in the same household and possible heterogeneity between villages, which shows that the observed difference between the 2 methods was not different indoors (β = 0.14, P = 0.45, 95% confidence interval (CI) from –0.499 to 0.796), whereas it was significantly different outdoors (β = –2.92, P < 0.001, 95% CI from –3.547 to –2.304).
The proportion of positive catches obtained by ST was significantly higher outdoors than indoors (
2 M-H = 15.24; P < 0.001), whereas the opposite (more indoors than outdoors) was observed with BA (
2 M-H = 90.59; P < 0.001).
Males.
The frequency of positive ST catches was significantly lower than that of BA catches both indoors and outdoors (indoors, 18.2% versus 52.4%,
2 M-H = 55.3, P < 0.001; outdoors, 3.2% versus 9.0%,
2 M-H = 6.17, P = 0.013). The proportion of positive catches was significantly higher indoors than outdoors for ST (
2 M-H = 25.22; P < 0.001) and BA (
2 M-H = 92.3; P < 0.001).
Females versus males.
The frequency of positive ST catches was significantly higher for females than for males, both indoors and outdoors (indoors, 52.7% versus 18.2%,
2 M-H = 209.7, P < 0.001; outdoors, 70.8% versus 3.3%,
2 M-H = 56.6, P < 0.001). There were no differences in BA catches (indoors, 51.9 versus 52.4%,
2 M-H = 0.01, P = 0.99%; outdoors, 9% versus 9%,
2 M-H = 0.03, P = 0.86).
Ae. albopictus.
In total, 149 females and 1 male Aedes albopictus were collected by ST and 5 females and 2 males by BA. The frequency of positive ST catches was significantly higher than that of BA catches both indoors (8.6% versus 1.9%) and outdoors (34.7% versus 0.5%) (Fishers exact P < 0.001). The proportion of female-positive catches obtained by ST was significantly higher outdoors than indoors (
2 = 54.7, P < 0.001).
Culex spp.
Table 3
shows the number of catches, the proportion of positive catches, the total number of specimens caught, and the minimum and maximum number of specimens/catch in the 2 villages. Sticky traps and BA collections resulted in total numbers of 10,535 and 6,635 Culex specimens, respectively. Gender determination was carried out only for specimens collected in the ST; females represented 90% of the sample.
|
2 = 0.31; P = 0.57), whereas outdoors the frequency of positive ST catches was significantly higher than that of BA catches (97.7% versus 55.8%,
2 = 43.15, P < 0.001). In village 2, the frequency of positive ST catches was significantly higher than BA catches indoors (97.7% versus 86.3%,
2 = 11.61, P < 0.001) and outdoors (96.2% versus 38.7%,
2 = 97.08, P < 0.001). Results from the regression model controlling for clustering of collections repeated in the same household and the heterogeneity between villages indicate that there were significant differences between the 2 methods outdoors (β = –1.574, P < 0.001, 95% CI from –2.312 to –0.837) but not indoors (β = 0.096, P = 0.793, 95% CI from –0.621 to 0.813).
| DISCUSSION |
|
|
|---|
In our comparison of results from ST catches with those obtained by BA, it is relevant to stress that the 2 approaches collect different fractions of the mosquito populations (i.e., mainly ovipositing females for ST and mainly resting males and females for BA). The ST collects mosquitoes continuously. The BA captures mosquitoes only for the few minutes that the collecting is carried out. Therefore, our comparison was based on overall effort required to carry out sampling activities (from implementation of field work to specimen counting and identification), which was comparable between the 2 approaches.
Taking these differences into consideration, our results show that the ST collects more female Aedes than BA, the percentage of positive catches being double in the case of Ae. aegypti and almost 20 times higher in the case of Ae. albopictus. A similar trend, although less marked, was also observed in the case of Culex collections. The only result in contrast with this trend concerns Ae. aegypti males, which were collected mostly by BA. This result is expected because the ST is targeted at ovipositing females.
When the results are analyzed separately for indoor and outdoor catches, it is clear that the larger number collected by the ST is due mainly to outdoor catches, where ST consistently shows a higher percentage of positive catches and higher numbers of Aedes females/catch when compared with BA. The same pattern is also observed for Culex and Armigeres spp. These results may reflect the fact that the 2 methods collect different fractions of the mosquito populations and may be also linked to intrinsic constrains in the use of BA outdoors. It is likely that mosquitoes are more difficult to locate outdoors and/or may be able to escape aspiration more easily outdoors than when confined indoors.
The comparison between ST catches carried out indoors versus outdoors shows that the percentage of Aedes-positive catches, as well as the overall number of females/catch, is higher outdoors; i.e., 57% and 80% positive outdoor ST catches for Ae. aegypti and Ae. albopictus females, respectively. The higher outdoor collection rate for Ae. albopictus, compared with that for Ae. aegypti, may reflect a different distribution of females of the 2 species and the greater exophily tendency of Ae. albopictus.4 The higher number of male Ae. aegypti collected indoors and the general observation that this species tends to rest indoors2,10 suggest that females are leaving the indoor environment to find oviposition sites outside, as assumed by Dibo and others18 and confirmed by Favaro and others19 in Brazil. Movement studies and assessment of gonotrophic stages of females collected in the ST, which we have proved to be feasible, are needed to prove that females leave houses when looking for oviposition sites in Thailand.
| CONCLUSIONS |
|
|
|---|
The ST has several operational advantages compared with the BA: i) it operates continuously and is, therefore, less biased by local circumstances during the short period of aspiration; ii) it is not biased by collector expertise; and iii) it does not require a power source. Variation in operator placement of STs can, however, have a significant impact on the results.
Compared with other kinds of sampling devices targeted at trapping adult mosquitoes (e.g., the Prince trap, the CDC Wilton trap, and the BG-Sentinel trap), the ST used herein is cheap (about 15 USD for 1 ST equipped with a set of 4 sticky sheets), easy to manipulate, and does not need a power source. However, it should be stressed that, because the ST collects mainly gravid females, it cannot in principle be used to estimate human–vector contact. It is reasonable to assume, however, that gravid Ae. aegypti females must have taken a human blood meal because of their high degree of anthropophily.2 The sticky trap should be further evaluated for surveillance of Stegomyia mosquitoes and/or monitoring the impact of mosquito control strategies. When interpreting the results of such investigations, it will be important to consider that efficiency of the ST is likely to be affected by the availability of alternative, natural oviposition sites, a phenomenon that was demonstrated for ovitraps by Focks.20
The ST may also represent an attractive tool for ecological and epidemiological investigations. Specimens collected by the ST could be used i) to study dispersal and longevity—results from the preliminary MRR experiment carried out in Rome by Facchinelli and others8 confirm that specimens dusted with fluorescent powders can be easily detected by observing the sticky sheets under UV light; ii) to study resting and/or ovipositing behaviors with respect to the indoor versus outdoor environments (Valerio and others, unpublished data); iii) to analyze the origin of bloodmeals in specimens captured before completion of bloodmeal digestion (Valerio and others, unpublished data); iv) to assay for the presence of pathogens, e.g., dengue virus21,22; and v) to assay for insecticide resistance alleles.23 We propose that in areas where alternative breeding sites are scarce and/or reduced as a result of integrated source reduction control activities, it will interesting to test the potential of ST as a mass trapping device to reduce vector densities, similar to the study conducted with a lethal ovitrap in Thailand.24
Received October 19, 2007. Accepted for publication February 15, 2008.
Acknowledgments: The authors thank AMA-DISINFESTAZIONI s.r.l. for financial and logistic support and Prof. Gianfranco Tarsitani for support and encouragement. The technical staff of the AFRIMS Entomology Laboratory in Kamphaeng Phet, Thailand, is acknowledged for its great help in the field. The authors thank the residents of the Thai villages for allowing them to inspect their houses.
* Address correspondence to Alessandra della Torre, Dipartimento di Scienze di Sanità Pubblica, Sezione di Parassitologia, Università "La Sapienza," Piazzale Aldo Moro, 5, 00185 Rome, Italy. E-mail: ale.dellatorre{at}uniroma1.it ![]()
Authors addresses: Luca Facchinelli, Laura Valerio, and Alessandra della Torre, Dipartimento di Scienze di Sanità Pubblica, Sezione di Parassitologia, Università "La Sapienza", Piazzale Aldo Moro, 5, 00185 Rome, Italy, Tel: 39-06-49694268, Fax: 39-06-49694268, E-mails: e.facchinelli{at}ucdavis.edu, laura.valerio{at}uniroma1.it, and ale.dellatorre{at}uniroma1.it. Constantianus J. M. Koenraadt, Department of Entomology, Cornell University, Ithaca, NY 14850, E-mail: cjk48{at}cornell.edu. Caterina Fanello, London School of Hygiene and Tropical Medicine, Keppel St., WC1E 7HT London, UK, E-mail: Caterina.Fanello{at}lshtm.ac.uk. Udom Kijchalao and James W. Jones, Armed Forces Research Institute for Medical Sciences, Bangkok, Thailand, E-mails: udomk{at}afrims.org and james.jones{at}afrims.org. Thomas W. Scott, Department of Entomology, University of California, Davis, CA, E-mail: twscott{at}ucdavis.edu.
Reprint requests: Alessandra della Torre, Department of Public Health Sciences, University of Rome "La Sapienta," Piazzole Aldo Moro, 5, 00185 Rome, Italy, E-mail: ale.dellatorre{at}uniroma1.it.
| REFERENCES |
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |