An improved experimental hut design for the study of Aedes aegypti (Diptera: Culicidae) movement patterns in Thailand


Knowledge of the behavioral responses of mosquito vectors to chemical insecticides is of paramount importance to understanding the epidemiology of disease transmission and optimum strategies for vector control. Several studies have employed the use of experimental huts to evaluate changes in mosquito behavioral responses that are relevant to disease transmission (i.e., disruption of human-vector contact) when they are exposed to insecticides, as compared to chemical-free conditions, including measurement of time and density of house entry, exit, and indoor resting, with attention concentrated mainly on Anopheles species (Smith 1965, Roberts et al. 1984, Grieco et al. 2000, Pates and Curtis 2005). Relatively few attempts have been made, however, to describe the responses incited by insecticides on other mosquito species using experimental huts (Symes et al. 1942, Kennedy 1947, Brown 1964, Lal et al. 1965, Moore 1977, Suwannachote et al. 2009).

Our previous research describing the movement patterns of Ae. aegpyti to identify chemical modes of action demonstrated that a portable hut based on the design of Achee et al. (2005) served as a successful tool for such studies (Suwonkerd et al. 2006, Chareonviriyaphap et al. 2005, Grieco et al. 2007). Although these huts mimicked indigenous Thai homes and produced consistent and reliable results, structural adjustments were identified that could increase hut longevity and improve mosquito sampling efficiency. These modifications included: 1) a raised platform to prevent structural damage from termites and soil moisture; 2) cement ant traps placed underneath the raised platform to prevent predation on knock-down mosquitoes during chemical trials; 3) a walkway around the perimeter of the hut to facilitate mosquito removal from window and door traps and; 4) increased airflow between the ceiling and exterior roof to aide in indoor heat dissipation.

As part of a larger proof-of-concept research program evaluating a Push-Pull vector control strategy to reduce host-seeking Ae. aegypti from inside homes and the peridomestic environment using minimal chemical dose and treatment coverage of standard vector control compounds, additional experimental huts have been constructed based on these design modifications (Figure 1). This report describes these huts, interception traps, and baseline studies, that were generated without chemical intervention to determine if changes in hut design would negatively affect mosquito movement patterns as compared to previous findings.

Figure 1.

Experimental hut fitted with window and door interception traps for the evaluation of mosquito behavior patterns.

The dimensions of the modified huts are 4 m wide × 5 m long × 2.5 m high with three windows (0.9 m wide × 0.6 m high) and one door (1 m wide × 2.4 m high) onto which can be affixed removable window and door traps, respectively (Figure 2). Hut frames are made of iron pipe, untreated wood planks are used for walls and flooring, with the roof constructed of zinc panels. The dimensions of the window traps are 0.60 m long × 0.90 m wide × 0.60 m high (Figure 3A). Louvers made of 1.6 cm non-treated hard wood were placed over the front opening of each trap with a hinged mesh flap used to cover the bevel opening during removal of trapped mosquitoes. The door trap is separated into two equal portions with each measuring 0.73 m wide × 0.56 m high × 0.93 m long (Figure 3B). Louvers and mesh flaps are integrated into each portion as described for window traps. The frames of both upper and lower traps are fixed to the hut door but the lower door trap can be opened independently and easily disassembled during experimental trials to allow collectors to exit and enter the huts during host rotation periods. Both window and door traps have three collection portals through which collectors insert manual aspirators for removal of trapped mosquitoes (depicted in Figure 3B).

Figure 2.

Graphic display of the modified hut with dimensions (A) ventilation space; (B) window trap; (C) door trap.

Figure 3.

Window (A) and door (B) intercept trap design with dimensions.

To validate the modified experimental huts, entry and exit movement patterns of Ae. aegypti populations from Kanchanaburi Province, Thailand, were evaluated under chemical-free conditions on four consecutive days for entrance trials and four consecutive days for exit trials following previously described methodologies (Grieco et al. 2007). Briefly, each test population consisted of 100 five-to-seven-day-old, female mosquitoes marked with fluorescent powder (in separate colors) using the methods of Suwonkerd et al. (2006) and Tsuda et al. (2001). Test populations were released 30 min before sunrise (approx.05:30) and collections were made from traps at 20 min intervals from 06:00–18:00. For each entrance experiment, three mosquito test populations (n = 100/release population) were released at a marked point 10 m NW from the door of each corresponding hut. Interception traps were positioned inside the hut (to capture entering mosquitoes) and two people were indoors to generate host cues and serve as collectors. For exit trials, one test population (n=100) was released at the inside center of each hut and traps were fixed to exterior hut walls (to capture exiting mosquitoes) while three, 2-person teams removed mosquitoes from traps using the same sampling scheme described for entrance experiments (i.e., 06:00–18:00). One volunteer occupied each hut during exit trials to generate host cues and was positioned underneath an untreated bed net for bite protection.

Results of movement pattern studies of Ae. aegypti females into and out of the modified huts are given in Table 1. Among the three huts, there were no statistical differences in the total number of recaptured mosquitoes entering (ANOVA; F=1.979, df =2, 9, P= 0.118) or exiting (F=0.783, df =2, P= 0.574). Average time trends for hut entry and exit indicate similar peaks at 07:00 for both experiment al types with exit behavior continuing throughout the day (Figure 4). These patterns are similar to those described from previous findings of Ae. aegypti responses without chemical intervention during the same seasonal period (Suwonkerd et al. 2006, Grieco et al. 2007, Suwannachote et al. 2009), thereby indicating no negative effects from design modifications.

Table 1.  Entrance and exit movements of Aedes aegypti to three experimental huts in Kanchanaburi Province, Thailand.
ExperimentHutTotal releasedRecapture in traps (marked mosquitoes)Total recapture in trap1Total recapture inside hut2% KD3 on floorAvg4 indoor temp (ºC)Avg indoor humidity (%RH)
  1. 1Includes mosquitoes marked from previous days.

  2. 2Collected using a back-pack aspirator following last sampling period (18:00).

  3. 3KD= knockdown specimens removed from hut floor (n/p = not performed).

  4. 4Average 12-h reading from four replicate trials for a given hut.

Figure 4.

Average number of Aedes aegypti females collected from entrance and exit traps fitted to experimental huts during validation studies (without chemical intervention).

In summary, the experimental hut design reported here has shown to be successful in measuring expected movement patterns of Ae. aegypti populations under chemical-free conditions. These results validate the huts as a standard tool for evaluating mosquito exiting and entering behaviors as part of the development of a Push-Pull strategy within the larger research program.


We extend gratitude to the hut construction team for their time, effort, and design suggestions and thank the Armed Forces Development Command, Sai Yok District, Kanchanaburi Province, Thailand, for support of the research program by providing land to serve as the study site. Funding for this research is provided by the Bill and Melinda Gates Foundation (Grant #48513) and the Thailand Research Fund (RTA5280007).