Standard laboratory reference strains of Ae. aegypti and Ae. albopictus (FAO/IAEA, 2017, 2020) were used for all experiments. The Aedes strains were maintained following the “Guidelines for Routine Colony Maintenance of Aedes Mosquitoes” (FAO/IAEA, 2017) (Experiments 2.3 to 2.9). Aedes aegypti and Ae. albopictus strains originating from Brazil (Juazeiro) and Italy (Rimini) were transferred to the IPCL from the insectary of Biofabrica Moscamed, Juazeiro, Brazil, and from the Centro Agricoltura Ambiente, Bologna, Italy, in 2012 and 2018, respectively. These two institutions are IAEA collaborating centers for the development of mosquito SIT. These strains were used to assess factors that might affect mosquito flight ability (Experiments 2.3–2.9). Two recently colonized Ae. albopictus strains originating from Spain (Valencia, TRAGSA) and China (Guangzhou, Wolbaki) were used to assess flight ability at the IPCL (Experiment 2.9). In addition, a wAlbB strain of Wolbachia-infected Ae. aegypti (hereafter referred to as the “Singapore strain”) was independently maintained (Cheong Huat Tan, personal communication) at the Environmental Health Institute of the National Environment Agency, Singapore, to replicate two sets of standardization experiments (Experiments 2.6–2.7).

The rearing period had controlled conditions as follows: temperature of 28 ± 2°C, 80 ± 10% relative humidity (RH), and lighting of 14:10 h light: dark, including 1 h of dawn lighting and 1 h of dusk lighting for larval stages. Adults were separately maintained under 26 ± 2°C, 60 ± 10% RH, and 14:10 h light: dark, including 1 h dawn and 1 h dusk.

To perform the experiments, mosquitoes were reared following modified mass-rearing procedures developed at the IPCL (Mamai et al., 2019; Maïga et al., 2019; FAO/IAEA, 2020). Larval rearing started on Thursdays (day zero) when eggs were hatched and transferred to mass-rearing trays previously filled with 5 L of osmosis water on Fridays (day one). No larval feeding was performed during weekends and pupae were collected on day six after egg hatching. A 4% IAEA larval diet was provided daily: a 300 ml (0.66 mg/larva) meal on day one, a 300 ml (0.66 mg/larva) meal on day four, and a 200 ml (0.44 mg/larva) meal on day five. Larvae and pupae were separated on day six. Four liters of larval water were reused with an additional 300 ml for day six of larval rearing. Pupae were collected daily and sex-separated using mechanical and semiautomatic pupal sex sorters (John W. Hock Co., Gainesville, FL; Wolbaki, China).

For each experiment, pupae were aliquoted into 100 ml plastic cups (Medi-Inn, United Kingdom), each holding 110 male pupae and placed in cages (15 × 15 × 15 cm, BugDorm, BD4M1515, Taiwan) for emergence. About 100 (accounting for emergence and mortality rates) adults were maintained with access to a 10% sucrose solution until the day of the experiments.

To assess the sensitivity of the new FTD (version 2.0) (Experiment 2.9), three-to-four–day-old adult Aedes mosquitoes were exposed to cold stress conditions (4°C for 2 hours), which is known to significantly impact the flight ability of both Ae. aegypti and Ae. albopictus (Culbert et al., 2018). Males were allowed to recover for 2 h in the presence of a 10% sucrose solution prior to the test.

To assess the sensitivity of the new FTD (version 2.0) (Experiment 2.9), three-to-four–day-old adults were exposed to high-dose irradiation (100 Gy using an X-ray blood irradiator (Raycell MK2) (Gómez-Simuta et al., 2021) for Ae. albopictus and 150 Gy using the GammaCell 220 (Nordion Ltd., Kanata, Ontario, Canada) for Ae. aegypti), which is known to reduce the quality of mosquitoes (Culbert et al., 2018). Males were knocked down and held in a cold room at 4°C in compacted batches of 100/cm³ to simulate mass-transport conditions prior to irradiation. Males were allowed to recover for 2 h with a 10% sucrose solution prior to the flight ability test. Untreated (nonirradiated) male mosquitoes were kept in laboratory conditions with a 10% sucrose solution during both the cold stress and the irradiation treatments.

The test conditions for all experiments were 26 ± 2°C and 60 ± 10% RH, under a laboratory daylight regime (500–1000 lux), using untreated male mosquitoes (unless stated otherwise).

The original FTD (Culbert et al., 2018) (version 1.0) was modified in several ways to measure its efficiency while improving handling processes. The original FTD consists of 40 transparent acrylic plastic (polymethyl methacrylate, PMMA) tubes (hereafter referred to as “individual internal tubes”) that were placed together within a larger PMMA tube (hereafter referred to as the “inner tube”). Gaps between individual internal tubes were filled with transparent silicone (OBI, Austria). When male mosquitoes escape from the flight device, they can be recollected in a larger cylindrical PMMA tube (hereafter referred to as the “containment box”), which is closed at the top end with mesh (Figure 1A and Supplementary Material S1). A fan (DC axial fan: 40 mm, Vapo: 12 V, airflow: 0.218 m³/min, acoustic noise: 20.6 dB, and rated speed: 6,000 rpm, Multicomp, United Kingdom) and a BG-Lure (Biogents, Regensburg, Germany) pellets holder made of PMMA is placed on top of the end mesh.

To operate, the inner tube was inserted and removed from the bottom of the containment box, which was secured with a mesh sleeve. Mosquitoes were first loaded into the FTD before the device was introduced through the bottom in the larger cylindrical tube because the side opening was too small to allow side introduction. In addition, after the test was completed, the device was taken to a cold room to knock down the mosquitoes prior to counting. These handling processes were found to be tedious for nonexperienced operators and may have increased the risk of damaging the FTD. Therefore, we designed a square containment box to replace the larger cylindrical tube. The containment part of the device is easily opened from the top, and the inner tube can be inserted or removed more easily through a larger opening on the front side of the containment box [Figure 1B, Figure 2 (version 1.1)].

In addition, another version of the FTD was developed. The new FTD (version 2.0) (Figure 1C) is made of eight individual internal tubes fitted in an inner tube of 4 cm in diameter. The tube holder (base) was resized to fit the inner tube better. The containment box and the tube length are all similar to version 1.1.

For all experiments, the following steps were followed to perform the flight test (Supplementary Materials S13, S14). Before any test, one to two pellets (10–30 mg) of BG-Lure (Biogents, Regensburg, Germany) were placed on top of the FTD. A fan was placed directly on the small BG decoy cap and switched on at its standard speed (3000 rpm unless stated otherwise). About 100 adult male mosquitoes (unless stated otherwise) were aspirated using a mouth aspirator and were blown into the FTD via a small 1 cm hole at the bottom of the device. Males were then confined within a small space (height × diameter: 1 × 7 cm) and flew upwards through one of the individual internal tubes and out into the large containment box. After 2 hours, the top of the inner tube was covered using a Petri dish (9 cm in diameter) to avoid further escapes, the fan was stopped, and the experiment was deemed as complete. The inner tube was then removed through the large opening of the containment box by tilting it slowly. When tilted halfway, the Petri dish covering the top of the tube was held to avoid losing it. To remove the nonescaped males from the inner tube, mosquitoes were blown through the bottom into a cage (15 × 15 × 15 cm, BugDorm, BD4M1515, Taiwan). The escaped males were removed using a mouth (or mechanical) aspirator from the containment box and transferred to another cage (15 × 15 × 15 cm, BugDorm, BD4M1515, Taiwan). All cages were taken into a freezer (−20°C) for 20–30 min and numbers were counted for each treatment.

To assess whether the fan speed could affect a mosquito’s capacity to escape, two fan speeds were tested. The FTD fan (Multicamp, United Kingdom) was set to either its highest speed (high, 6000 rpm) or at its standard speed (normal, 3000 rpm). Three FTDs (version 1.1) were tested with two-to-three–day-old male Ae. aegypti (Brazil strain) mosquitoes and four FTDs with four-to-five–day-old male Ae. albopictus (Rimini, strain) mosquitoes for each speed, respectively.

It was initially observed that as males were being loaded into the FTD, a large number of mosquitoes escaped immediately. Therefore, to ascertain a suitable length of time to perform the QC flight test, a range of times were tested—including 30, 60, 90, and 120 min. The test was repeated four times (with two, two, two, and three pseudoreplicates per repetition, respectively) with two-to-three–day-old male Ae. aegypti (Brazil strain) and twice (two pseudoreplicates each) with two-to-three–day-old male Ae. albopictus (Rimini, strain) for each duration.

Densities of 25, 50, 75, and 100 males were tested to see whether the number of males loaded into an FTD would affect the output of the escape rate. The test was performed with four FTDs for each density with two-to-three–day-old male Ae. aegypti (Brazil strain) mosquitoes. For Ae. albopictus (Rimini strain) mosquitoes, the density effect was assessed twice (four and two pseudoreplicates) with two-to-three–day-old males for each density.

To assess whether the color of the FTD could affect escape rates, two types of inner tubes (where the tube ends with transparent and pink silicon) were used (Supplementary Material S15). For each tube type (transparent and pink), the presence/absence of fan and lure was tested for Ae. aegypti (Brazil strain) and Ae. albopictus (Italy strain): one with the addition of BG-Lure only (pink/transparent + lure), one with fan only (pink/transparent + fan), one with fan and lure together (pink/transparent + fan + lure), and one FTD without a fan or lure (pink/transparent). The test was performed twice (four and two pseudoreplicates/treatment) with two-to-three–day-old male Ae. aegypti and with two-to-three–day-old male Ae. albopictus mosquitoes, respectively.

The same experiment was carried out with five-to-six–day-old male Aedes aegypti in Singapore (Singapore strain) using only the transparent FTDs (version 1.1). The escape rates from two FTDs with fan and lure (transparent + fan + lure) were compared to three FTDs of each of the following: with lure only (transparent + lure), with fan only (transparent + fan), and without fan and lure (transparent).

For the FTD to be tested without a fan (absence of a working fan), a fan was still placed on top of the device but without turning it on.

The effect of age on male flight ability was assessed for different age groups ranging from less than 1 day up to 8 days old for Ae. aegypti (three replicates/age group) and from less than 1 day up to 10 days for Ae. albopictus. In addition, two age groups (12–13 and 16–17 days) were assessed for the latter. Four replicates were performed for each Ae. albopictus age group.

The same experiment was repeated at the Environmental Health Institute, National Environment Agency, Singapore. Male Ae. aegypti (Singapore strain), aged between two to six days, were tested. Four age groups, including two to three, three to four, four to five, and five to SIX days, were used to assess and compare mosquitoes’ ability to fly using the 40-tube transparent FTD (version 1.1). The flight test was performed in total 21, nine, nine, and 27 times for the two-to-three-, three-to-four–, four-to-five–, and five-to-six–day-old mosquitoes, respectively.

To assess whether different strains of Ae. albopictus exhibit different flight ability scores using the 40-tube FTD (version 1.1), three strains from different origins, Italy (Rimini), Spain (Valencia), and China (Guangzhou), were evaluated. The flight test was repeated twice with three pseudoreplicates per repetition per strain with three-to-four–day-old males. Male mosquitoes were randomly selected from each of the six pseudoreplicates per strain (Rimini: 82; Valencia: 84; and Guangzhou: 79 mosquitoes in total), and their right wings were dissected and measured as a proxy for strain adult size (Nasci, 1990).

To assess whether a reduction of the number of individual internal tubes within the flight test inner tube could affect the performance of the FTD (version 1.1), a FTD was made containing eight individual internal tubes fitted in an inner tube with a diameter of 4 cm (version 2.0). The eight-tube FTD was also tested with both mosquito species. In addition, the effects of cold stress (4°C chilling for 2 hours) and high-dose irradiation (100 and 150 Gy for Ae. albopictus and Ae. aegypti, respectively) (see Section 2.1 .) on the sensitivity of the new FTD (version 2.0) were evaluated.

The fan speed (“normal” 3000 rpm vs. “high” 6000 rpm) had no effect on mosquitoes’ escape rate through the FTD in both Ae. aegypti [high: 0.76 (0.71–0.81, 95%CI), normal: 0.70 (0.65–0.75, 95%CI), χ² = 1.17, df = 1, p = 0.28] and Ae. albopictus [high: 0.90 (0.87–0.93, 95%CI), normal: 0.91 (0.88–0.94, 95%CI), χ² = 0.2, df = 1, p = 0.65].

The duration of the flight ability test did not significantly differ between 30, 60, 90, and 120 min in both Ae. aegypti (χ² = 5.35, df = 3, p = 0.14; Figure 3A) and Ae. albopictus (χ² = 3.57, df = 3, p = 0.31; Figure 3B).

The densities of male mosquitoes within the FTD between 25 and 100 did not impact escape rates in Ae. aegypti (χ² = 6.3, df = 3, p = 0.09; Figure 4A) but had a significant impact in Ae. albopictus (χ² = 26.2, df = 3, p = 0.0001; Figure 4B). The Tukey test shows that a density of 25 male Ae. albopictus led to a greater mean escape rate than that of the densities 75 and 100 (p < 0.05).

Greater number of male Ae. aegypti, the Brazilian strain, and Ae. albopictus, the Italian strain, escaped from the transparent tube more than those that escaped the pink-colored internal tube (p < 0.001, Table 1). There was a greater impact of the fan on the escape rate of Ae. aegypti, Brazilian strain (p = 0.03, Table 1), and Ae. aegypti, Singaporean strain (p < 0.001, Table 1). The presence of a lure did not enhance the escape rate of both Ae. aegypti strains (p > 0.05, Table 1), but there was a significant interaction between fan and lure on the escape rate of Ae. albopictus, Italian strain (p = 0.01, Table 1).

When considering the treatment of the FTD (i.e., with/without lure, with/without fan, and with/without lure and fan), a significant impact was found on the escape rate of Ae. aegypti, Brazilian strain (χ² = 14.8, df = 7, p = 0.03; Figure 5A), Ae. albopictus, Italian strain (χ² = 48.5, df = 7, p < 0.001, Figure 5B) and Ae. aegypti, Singaporean strain (χ² = 29.04, df = 3, p < 0.001; Figure 5C). Furthermore, a higher number of mosquitoes escaped from the reference FTD treatment where a transparent tube, a working fan, and lure were simultaneously used (Figure 5). There was a similar number of escapes from the pink-colored FTD in Ae. aegypti, Brazilian strain (Figure 5A), and Ae. albopictus, Italian strain (Figure 5B) regardless of the treatment. On the other hand, the difference in the escape rates between the pink-colored tube and the transparent tube tends to be greater in Ae. albopictus (Figure 5B) than that in Ae. aegypti (Figure 5A). Although more escapes were recorded from the FTD with fan and lure, they were not significantly different from those from the FTD with a working fan (Figure 5). The FTD baited with a lure or without lure and fan displayed lower escape rates than those of the reference FTD and the FTD with a working fan (Figure 5C).

Age had a significant effect on male flight ability in both Ae. aegypti (Brazil strain) (χ² = 62.7, df = 6, p < 0.001; Figure 6A) and Ae. albopictus (χ² = 1684.4, df = 11, p < 0.001; Figure 6B). The pairwise comparison of means showed that two-to-three–day-old male Ae. aegypti (reference age in Culbert et al., 2018) had higher flight ability than those younger than 2 days (p < 0.001) but lower flight ability than males older than 3 days (p < 0.001). Similarly, two-to-three–day-old and three-to-four–day-old male Ae. albopictus had significantly higher flight ability than those younger than 2 days (p < 0.001) but lower than males older than 4 days (p < 0.001). Flight ability of male Ae. aegypti and Ae. albopictus declined after the age of five to six and seven to eight days, respectively, leading to a significant decrease in 16-17–days-old male Ae. albopictus of (p < 0.001) as compared to seven-to-eight–day-old males. Similarly, age had a stronger impact on flight ability in male Ae. aegypti (Singapore strain) (χ² = 62.5, df = 3, p < 0.001; Figure 6C).

Similar escape rates of about 74 ± 20 (±95% CI) were observed when different strains of Aedes albopictus from different origins (Rimini in Italy; Valencia in Spain; Guangzhou in China) were tested using the 40-tube FTD (χ² = 0.68, df = 2, p = 0.70) (Figure 7).

Wing length of male mosquitoes varied significantly between Ae. albopictus strains from different origins (χ² = 57.03, df = 2, p < 0.001, Figure 7). Aedes albopictus (Guangzhou strain) exhibited a higher body size as compared to Valencia and Rimini strains (p < 0.01), whereas the other two strains have similar sizes (p = 0.77, Supplementary Figure S1).

When the new FTD with eight individual tubes (version 2.0) was compared to the 40-tube FTD (version 1.1), similar flight capacity was observed both in Ae. aegypti (χ² = 1.26, df = 1, p = 0.26; Figure 8A) and Ae. albopictus (χ² = 2.86, df = 1, p = 0.09; Figure 8B).

The eight-tube FTD (version 2.0) was as sensitive to stress conditions such as cold temperature and high irradiation doses in both Aedes mosquito species as the 40-tube FTD (version 1.1) (Ae. aegypti: χ² = 23.44, df = 2, p < 0.001; n = 6, Figure 8C; Ae. albopictus: χ² = 22.38, df = 2, p < 0.001, n = 6, Figure 8D).