Study sites

The release trials were conducted in Morcote (45° 55′ 30″ N 8° 54′ 58″ E, 292 m a.s.l.), a small town in southern Switzerland (Fig. 1). A permit to release sterile male tiger mosquitoes in Morcote (BAFU-217.23–64633/7) was obtained from the Swiss Federal Office for the Environment in accordance with the Ordinance on the Handling of Organisms in the Environment (No. 814.911). Geographically, the main urban area of Morcote is divided into two parts: the eastern part, which includes the historic center of the village and a more residential area that borders the nearby town of Vico Morcote to the northeast, and the western part, which consists solely of residential areas, with a significant proportion (over 20%) of secondary residences [25]. The release area, approximately 45 hectares, covered the entire urban area and was therefore almost entirely isolated from other towns except for the northeastern side bordering Vico Morcote. The control site, of about 14 hectares, was set in Caslano (45° 58′ 16″ N 8° 52′ 59″ E, 278 m a.s.l., 50 ha), a small town about 5.5 km straight-line distance from Morcote, with similar geographical and housing characteristics (Fig. 1).

Fig. 1

Release (red) and control (green) areas in southern Switzerland. The boundaries of the study municipalities, Morcote and Caslano, are outlined in blue. On the right, the yellow star marks the release point for the 2022 and 2023 MRR, while the red points indicate the 75 release stations for the 2023 SIT trial. MRR: Mark release and recapture; SIT: Sterile insect technique

Both towns follow an IVM since 2000 [9, 13]. In 2022 and 2023, the integrated control measures consisted in the regular treatment of catch basins in public spaces with the larvicide VectoMax® FG (Valent BioSciences Corp., Libertyville, IL, USA), applied every 6 weeks [26]. Citizens are encouraged to remove temporary water containers from their properties and to cover or treat permanent ones with the larvicide VectoBac® G (Valent Biosciences Corp., Libertyville, IL, USA). During 2021, before starting the sterile male releases, no significant differences were found in the average number of eggs collected per ovitrap in the SIT and control area (Fig. S1) although the data include another invasive Aedes species (i.e., Ae. japonicus) in addition to Ae. albopictus.

Mass rearing of Aedes albopictus

Sterile Ae. albopictus males were bred at the Centro Agricoltura Ambiente “G. Nicoli” (CAA) production facility in Italy, approximately 250 km from the release site, using methods outlined in [27]. The mosquito colony used for the SIT trials was initiated from eggs collected in Ticino in 2019 (strain CH) and maintained under standard conditions of 28 ± 1 °C, 85% relative humidity, and a photoperiod of 14∶10 h (L∶D). Mosquito larvae were mass-reared following the IAEA protocols [28, 29].

Preliminary mark release and recapture (MRR) study

The release of sterile, marked Ae. albopictus males took place on August 2nd, 2022, at the peak of the tiger mosquito reproductive season in southern Switzerland. At the CAA production facility, male pupae (strain CH F11) aged 20–48 h were separated from female pupae (with a residual female presence below 1%, considered acceptable in areas non endemic for arboviruses [23, 30, 31]) using an automatic version of the Fay-Morlan glass plate separator (Guangzhou Wolbaki Biotech Co., Ltd, Guangzhou, China) [32, 33] and irradiated with 55 Gy in water using a Radgil2 X-Ray irradiator (Gilardoni Ltd., Mandello del Lario, Italy). After irradiation and emergence, 22,500 sterile males were cold shock-anaesthetized at 8 ± 1 °C for 30 min and marked with a fluorescent pink powder according to the Food and Agriculture Organization of the United Nations (FAO) and IAEA 2017 protocol [34]. The males were then transferred to 15 plastic cups (MONOUSO ENVALIA GROUP S.L., Valencia, Spain; 10 cm diameter, 5 cm height, 350 cc capacity; 1500 individuals per cup), which were stacked into a cardboard cylinder. The cylinder was placed inside a polystyrene container, and the temperature was maintained at 10–12 °C using phase change materials (PCM ice gel packs; Blu Ice, Green Ice, Dryce Srl, Milano, Italy).

The polystyrene box was placed in a cardboard box and transported by car for 3 h to the receiving laboratory. Upon arrival, the marked sterile males were immediately transferred to the designated release station in Morcote (Fig. 1) and released at 18:00 local time. The sterility level of released males was checked through a residual fertility test at the receiving laboratory (Method S1) and their quality was evaluated by assessing the mean mortality rate after transportation, based on the number of males that did not fly out of a sub-sample of four containers within 1 h at the release station (males that were not yet awake were counted as dead).

From August 3rd to 9th, human landing collections (HLC) were conducted every 24 h at 32 evenly distributed points within a 200-m radius of the release station (Method S2, Fig. S2). The mosquitoes collected were brought to the laboratory, euthanized, and examined using a stereomicroscope. The species, sex, and presence of marked sterile males was recorded. Several parameters were estimated based on the data recorded, including male recapture rate (ϴ), male daily survival rate (s), average life expectancy, and mean dispersal distance [35,36,37].

Small-scale SIT field trial

In 2023, the releases of sterile males were conducted throughout the entire mosquito season, from early May to the end of September, at 75 fixed stations covering the entire release area (Fig. 1). Based on the results of the 2022 MRR, the release stations were evenly distributed at distances of less than 90 m from one another (range: 50–80 m). Approximately 2000 sterile males were released per station each week (3000 sterile males per hectare), for a total of about 150,000 males released each week. The releases were carried out manually by two operators, who moved between the stations by car. On average, the release process took about 3 h. In the first 8 and last 2 weeks of the season, releases were carried out once a week (Table S1). During the hottest part of the season (from July to mid-September), sterile males were released twice a week to maintain a higher sterile-to-wild ratio and ensure their continuous activity throughout the week. In this case, instead of a single weekly release of 150,000 males, two releases of 75,000 males each were carried out. In total, approximately 3,300,000 males were released over 22 weeks in the release area.

Production and transportation of sterile males

At CAA, male pupae aged 2–48 h (CH F13-F16) were irradiated with 40 Gy in water using a Radgil2 X-Ray irradiator (Gilardoni Ltd., Mandello del Lario, Italy). The males were packaged similarly to the 2022 protocol, with 2000 individuals per cup for the first four releases, and 1000 individuals per cup for all subsequent releases (Table S1). At each release station, the cups were placed in suitable locations where microclimatic conditions were favorable for their awakening and takeoff and where the ants would not reach.

For most of the study period (until week 26), males were shipped via courier, with transportation time to the receiving laboratory averaging approximately 18 h. In week 27, we transitioned to direct transportation by car or train, reducing delivery time to approximately 3 h (Table S1). Sterile males were 2–4 days old at the time of release and were marked only for MRR events.

Residual fertility of the sterile males was assessed three times during the release season (May 3, June 27, and August 8, 2023) at the receiving laboratory, using the same method as in 2022 (see methods in Additional file 1). Mortality following transportation was assessed at each release by selecting five randomly chosen release stations. The residual presence of females in the cups was assessed when MRRs were carried out.

MRR studies

Three MRR tests were conducted in 2023: one at the beginning of the season (June 13th, 10,000 males), one in the middle (July 11th, 20,000 males), and one at the end of the season (August 22nd, 20,000 males). After emergence, the sterile males were powdered-marked following the same protocol as in 2022, packed in stacked plastic cups (1000 males per cup), and shipped by courier to the receiving laboratory. The marked males were released at the same location as in 2022 (Fig. 1) and recaptured at the same 32 spots as in 2022 (Fig. S2), but this time with adult traps (BG-Pro, Biogents, Regensburg, Germany, sentinel mode, baited with CO2) instead of HLC. Several studies have shown that HLC and adult trap catches are comparable [38]. Adult traps offer the advantage of covering the entire day and do not depend on operator attractiveness, as is the case with HLC. The traps were checked daily for a week or until marked males were no longer captured. The same parameters as in 2022 were estimated.

Entomological monitoring to assess the effectiveness of SIT application

The population of Ae. albopictus was monitored at both the release and control sites through three entomological parameters: the total number of Ae. albopictus eggs laid, the percentage of Ae. albopictus eggs that hatched, and the number of Ae. albopictus adult females. Eggs were collected using oviposition traps (ovitraps), which consisted of black 1.5-L plastic jars (Ramona Ø13/H12, Luwasa® Interhydro AG, Allmendingen, Switzerland) with a top-border efflux hole. Each jar contained 1.2 L of tap water and a wooden slat (20 × 2.5 × 0.5 cm) that served as a substrate for female Aedes mosquitoes to lay their eggs [9, 10]. Forty-five and 14 ovitraps (one ovitrap per hectare) were set in the release and control site, respectively (Fig. S3). The wooden slats were collected and replaced with new ones once a week from mid-April until mid-October. In the laboratory, the eggs of Ae. albopictus were differentiated from other mosquito species and counted using established methods [39].

Egg hatching rates were assessed every 2 weeks. After counting the Ae. albopictus eggs, the wooden slats were subjected to a hatching protocol [34, 37]. The number of hatched eggs was counted, and egg hatching rate was calculated as the number of hatched eggs divided by the total number of eggs on the slats.

Adult mosquitoes were collected using CO2-baited BG-Pro traps (sentinel mode) evenly distributed (one trap every two hectares) across the release (25 traps) and control (8 traps) sites. The traps were set every 2 weeks, starting from the beginning of May until the end of September, 24 h after each release, and were checked after 40 to 48 h. Aedes albopictus mosquitoes were identified and counted by sex.

Statistical analyses

A generalized additive mixed-effects model (GAMM) was used to compare mosquito population parameters of the release and control sites across the season. Seasonal trends in these parameters were highly non-linear, with a peak in summer. To accurately capture these patterns, smoothers were applied within the GAMM framework. The analysis for the total number of Ae. albopictus eggs laid, the percentage of Ae. albopictus eggs that hatched, and the number of Ae. albopictus adult females is detailed in Supplementary file 2, Supplementary file 3, and Supplementary file 4, respectively. Briefly, the response variable was modelled using a generalised additive model (GAM) with a negative binomial (total number of Ae. albopictus eggs laid, number of Ae. albopictus adult females) or quasi binomial (percentage of hatched Ae. albopictus eggs) family to address overdispersion. The model included a smooth effect for “sampling date”, represented as the day of the year (yday, a numeric variable), which interacts with the variable municipality. The municipality variable is categorical and consists of two levels: Morcote (release site) and Caslano (control site); its effect was modelled as a fixed effect. To account for variations in trap exposure duration we incorporated Activation.time (a count variable ranging from 6 to 16 days) as an offset (we used the log of Activation.time to make the assumption that doubling the activation time would lead to proportional doubling in the number of eggs) in the model. We controlled for the non-independence of observations by including trap ID (unique.ID, a categorical variable) as a random effect. Model complexity was evaluated, and the best-fitting model was selected using a Chi-square test and Akaike information criterion (AIC) and Bayesian information criterion (BIC) criteria. All statistical analyses were conducted using R (version 4.4.2) [40]. The significance level was set at 5%.

A spatial GAM was used to test whether geographical location played a role in determining the three entomological parameters. In particular, this model estimated a two-dimensional spatial abundance surface to better understand the treatment’s effectiveness and how its impact diminishes toward the edges of the treated area. Two separate models were fitted to analyse the response variable, using a GAM with a negative binomial family to account for overdispersion. One model was fitted for Morcote (release site), and the other for Caslano (control site). Both models included a smooth effect for the “sampling date”, represented as day of the year (yday), and a combined smooth effect for geographic coordinates to capture spatial variability within each site. The best-fitting models were selected by evaluating AIC and BIC criteria. All statistical analyses were conducted using R (version 4.4.2) [40]. The significance level was set at 5%.

Public awareness

The IVM plan, followed in both release and control area, includes already several activities for communication to citizens: the distribution of leaflets on tiger mosquito and how to fight it to all households; dedicated web pages (www.supsi.ch/go/zanzare and www.ti.ch/zanzare); a service of personal counselling by phone or by email; specific spots “Let’s take away the water” and the video “Let’s fight the tiger mosquito together” broadcasted on national television during the breeding season and viewable on many municipal websites; active communication with the media, especially at the beginning of the breeding season and at the end of July, before the seasonal peak; and punctual educational activities in schools and other contexts. To raise public awareness about the SIT trial, the media were invited to the release of sterile males, both in 2022 and 2023, and several interviews were broadcast both in the press and on television the following days. A public presentation was conducted in June 2023 in the municipality of Morcote to inform the community and solicit their support.