Skip to main content
Download PDF

An easier life to come for mosquito researchers: field-testing across Italy supports VECTRACK system for automatic counting, identification and absolute density estimation of Aedes albopictus and Culex pipiens adults

Parasites & Vectors volume 17, Article number: 409 (2024) Cite this article

Abstract

Background

Disease-vector mosquito monitoring is an essential prerequisite to optimize control interventions and evidence-based risk predictions. However, conventional entomological monitoring methods are labor- and time-consuming and do not allow high temporal/spatial resolution. In 2022, a novel system coupling an optical sensor with machine learning technologies (VECTRACK) proved effective in counting and identifying Aedes albopictus and Culex pipiens adult females and males. Here, we carried out the first extensive field evaluation of the VECTRACK system to assess: (i) whether the catching capacity of a commercial BG-Mosquitaire trap (BGM) for adult mosquito equipped with VECTRACK (BGM + VECT) was affected by the sensor; (ii) the accuracy of the VECTRACK algorithm in correctly classifying the target mosquito species genus and sex; (iii) Ae. albopictus capture rate of BGM with or without VECTRACK.

Methods

The same experimental design was implemented in four areas in northern (Bergamo and Padua districts), central (Rome) and southern (Procida Island, Naples) Italy. In each area, three types of traps—one BGM, one BGM + VECT and the combination of four sticky traps (STs)—were rotated each 48 h in three different sites. Each sampling scheme was replicated three times/area. Collected mosquitoes were counted and identified by both the VECTRACK algorithm and operator-mediated morphological examination. The performance of the VECTRACK system was assessed by generalized linear mixed and linear regression models. Aedes albopictus capture rates of BGMs were calculated based on the known capture rate of ST.

Results

A total of 3829 mosquitoes (90.2% Ae. albopictus) were captured in 18 collection-days/trap/site. BGM and BGM + VECT showed a similar performance in collecting target mosquitoes. Results show high correlation between visual and automatic identification methods (Spearman Ae. albopictus: females = 0.97; males = 0.89; P < 0.0001) and low count errors. Moreover, the results allowed quantifying the heterogeneous effectiveness associated with different trap types in collecting Ae. albopictus and predicting estimates of its absolute density.

Conclusions

Obtained results strongly support the VECTRACK system as a powerful tool for mosquito monitoring and research, and its applicability over a range of ecological conditions, accounting for its high potential for continuous monitoring with minimal human effort.

Graphical Abstract

Background

Mosquitoes (Diptera: Culicidae) are vectors of several human pathogens, such as malaria parasites, and dengue (DENV), chikungunya (CHIKV), yellow fever, Zika and West Nile (WNV) viruses, causing diseases resulting in approximately 700,000 deaths/year worldwide [1,2,3]. Mosquito-borne diseases (MBDs) are becoming a major public health concern also in Europe, where an increasing number of autochthonous cases of arboviruses and a few outbreaks have occurred over the last two decades [4]. Culex pipiens, the native vector species, has been responsible for the endemic transmission of WNV in Europe since 2010, particularly in central and southeastern Europe, with the largest outbreak in 2018 [5,6,7]. In Italy, many WNV endemic cases, some even fatal, are reported each year, mainly in northern Italy [8, 9]. In addition, the risk of exotic arbovirus transmission by the invasive vector species Aedes albopictus is rising [10]. This was revealed by two large CHIKV outbreaks occurring in Italy in 2007 and 2017 [11,12,13] and by several DENV outbreaks in Croatia, France and Spain, including three outbreaks in Italy in 2020 and 2023, with 93 cases in three separate regions [14,15,16,17,18]. For these reasons, effective mosquito control is important to mitigate the impact of MBDs [19, 20].

Effective vector monitoring is an essential prerequisite to acquire data to optimize control strategies and interventions as well as for evidence-based risk predictions [21, 22]. However, conventional entomological monitoring methods are labor- and time-consuming and do not allow high temporal or spatial resolution, as they can hardly be implemented on a large scale and for an entire season [23]. These methods require qualified staff to check traps in the field at regular intervals, collect samples, make entomological identifications and analyze data. The time required for these different steps hinders large-scale routine and standardized data collections as well as the scaling up of timely application of control measures.

The exploitation of specific sensors coupled with machine learning (ML) technologies has the potential to address these major limitations and produce a significant step forward in mosquito surveillance and control. In 2022, a novel system named VECTRACK (Irideon SL. Barcelona, Spain; www.irideon.es), based on a specific optical sensor mounted on a commercial BG-Mosquitaire trap (Biogents AG, Regensburg, Germany) and combined with a supervised ML algorithm, was developed for automatic counting and identification of Aedes and Culex mosquitoes by genus and sex [24]. The optical sensor includes an emitter of a collimated beam of near-infrared light and a receiver panel, which face each other through a transparent circular tube defined as the sensing zone. While entering the traps, mosquitoes pass through the sensing zone, casting a fast-changing shadow upon the optical receiver because of the modulation of the light beam by their wing flaps. This signal is automatically detected and recorded by the sensor, along with the GPS coordinates, date, exact time, ambient temperature and relative humidity at the instant of each capture. González-Pérez et al. extracted five features from each wingbeat recording via an application of digital signal processing method and analyzed these by different ML algorithms. After training and validation phases, the highest accuracy achieved during the testing phase performed on laboratory-reared mosquitoes was 94.2% for the identification of Culex vs. Aedes mosquitoes and > 99% for sex classification in both mosquito genera. When tested for the first time in the field to our knowledge—in two sites in Barcellona province (Spain) with predominance of Cx. pipiens in sympatry with Ae. albopictus—the system discriminated the target mosquito genera from other non-target insects with a balanced accuracy of 95.5% and classified the genus and sex of those mosquitoes with a balanced accuracy of 88.8% [25].

In this article, we report the results of the first extensive field evaluation of the VECTRACK system carried out in four regions across Italy with predominance of Ae. albopictus in sympatry with Cx. pipiens and compare the relative capture rates associated with different trap types. This allowed us to (i) evaluate whether the equipment of a commercial BG-trap for adult mosquitoes with VECTRACK sensor affected the catching capacity of the trap, (ii) estimate BG-trap (either equipped or not with VECTRACK) capture rate for Ae. albopictus based on known capture rate of sticky traps and (iii) assess the accuracy of the VECTRACK algorithm in correctly counting and classifying field Aedes and Culex mosquitoes by genus and sex.

Methods

Experimental design

The experiments were carried out in four urban and suburban areas across Italy in summer 2023 during the months of peak mosquito activity from July to August in Bergamo (north, Lombardy Region; GPS coordinates: 45°42′11″N, 9°49′58″E) and Rome (center, Lazio Region; 41°54′12″N, 12°30′51″E); from August to September in Padua (northeast, Veneto Region; 45°20′55″N, 11°57′14″E) and Procida Island (south, Campania Region, Naples; 40°45′38″N, 14°01′10″E).

Three sites were selected in each study area, and three types of traps were used and rotated each 48 h during the experiment (Fig. 1): (i) one BG-Mosquitaire (BGM), a commercial suction fan trap from Biogents AG (Regensburg, Germany) equipped with BG-Sweetscent chemical attractant (Biogents AG); (ii) one BGM equipped as the former plus the VECTRACK sensor, directly placed on the entrance of the trap (hereafter, BGM + VECT); (iii) four sticky traps (STs, [26]) located at 50–100 m from one another (the choice to deploy four STs instead of one was due to their limited capture performance, and they were counted as a single trapping event). Sticky traps were included in the experimental design to provide estimates of BGM and BGM + VECT capture rates based on the known capture rate for ST [22, 27]. The distance between sites (> 100 m and < 400 m) within each area was chosen to limit biases due to possible competition between traps while avoiding prolonged interruption of captures while rotating the traps among sites. The specific location of the traps at each site was selected to provide shade, nearby vegetation, shelter from rain and wind, access to electrical power and protection from theft. The complete trap rotation was performed within 6 days, and three replicates/area were conducted in the following weeks. Collected mosquitoes were counted and morphologically classified [28] by trained entomologists. Hereafter, we refer to this quantity as the number of “morphologically identified mosquitoes” to differentiate it from the number of mosquitoes counted and identified by VECTRACK ML algorithm.

Fig. 1
figure 1

Experimental design of VECTRACK testing in Italy. Three types of traps were compared: BG-Mosquitaire (Biogents AG; BGM); BG-Mosquitaire equipped with the VECTRACK sensor (Irideon SL; BGM + VECT); four sticky traps (STs; [26]). Each trap was located in one site and moved to a close site (100–400 m) after 48 h following a predefined order of rotation. The complete trap rotation was performed in 6 days. Two additional replicates were conducted in the following weeks in each study area. The same experimental design was implemented in four regions (black dots in the map): Lombardy (by the University of Milan), Veneto (by the Istituto Zooprofilattico Sperimentale delle Venezie), Lazio (by the Sapienza University of Rome and Istituto Superiore di Sanità) and Campania (by the University of Naples Federico II)

Full size image

Statistical analyses

Analyses were carried out using R version 4.3.2, with the “glmmTMB” package [29].

Assessment of depletion effect

We checked potential depletion effects on the mosquito population caused by the continuous and repeated collections conducted in each site. To do this, we applied a generalized linear mixed model (GLMM) considering the total number of captured and morphologically identified mosquitoes as the response variable and the number of days from the first collection in the site, the trap type and their interaction as covariates. We considered as random effect the site nested within the region. The response variable was assumed to follow a negative binomial distribution, using a log link.

Assessment of changes in BG-Mosquitaire trapping performance due to VECTRACK application

We investigated whether equipping BGM with a VECTRACK sensor changed its trapping capability. To do this, we compared the total number of collected mosquitoes (as well as the total number of either Ae. albopictus or Cx. pipiens, i.e. the only two species currently identifiable by VECTRACK) morphologically identified in BGM and in BGM + VECT. At the species level, we also compared the estimated number of males and females separately. For each analysis, we applied a GLMM assuming a negative binomial distribution for the response variable with log link, the trap type as covariate and two crossed random effects: the sampling period and the site nested within region.

Assessment of VECTRACK performances in species and sex identification

We compared the number of morphologically identified mosquitoes collected by BGM + VECT with the number of mosquitoes identified by VECTRACK algorithm, considering the overall number of identified mosquitoes, the number of Ae. albopictus and Cx. pipiens specimens and the number of males and females separately identified for these two species. For Ae. albopictus and Cx. pipiens specimens and their sexes, we also provide the balance accuracy (BA) metric as defined by González-Pérez et al. (2024), where BA = Se + Sp/2, and Se and Sp indicate the sensitivity (the proportion of positive specimens correctly classified) and specificity (the proportion of negative specimens correctly classified), respectively. We computed the Spearman’s correlation coefficients corrected for ties, and we fitted a linear regression model considering the number of morphologically identified mosquitoes as the response variable and the number of mosquitoes identified by VECTRACK algorithm as the covariate.

Results

A total of 3829 mosquitoes were captured by all traps in the four field areas, corresponding to 18 collection-days/trap type/site (1998 by BGM, 1554 by BGM + VECT and 277 by STs). All collected mosquitoes were successfully morphologically identified by trained entomologists: Aedes albopictus was the predominant species (90.2%) followed by Cx. pipiens (8.8%); 36 other specimens were collected (23 individuals belonging to other Aedes species, 12 Culiseta longiareolata and one not-identified specimen; Table 1). Figure 2 shows box plots of Ae. albopictus and Cx. pipiens captures for each type of collection in each area.

Table 1 Total of Culicidae captured by three types of traps during 36 48-h-long trapping sessions in four areas across Italy and morphologically assigned to species and sex
Full size table
Fig. 2
figure 2

Box plots of total mosquito (upper panel), Aedes albopictus (central panel) and Culex pipiens (lower panel) captures for each type of collection overall and in each study area in Italy. Dark blue: BG-Mosquitaire (Biogents AG; BGM); light blue: BG-Mosquitaire equipped with the VECTRACK sensor (Irideon SL; BGM + VECT); purple: sticky traps ([26], STs; N = 4; considered as a single trap)

Full size image

No relevant evidence of a depletion effect due to repeated captures at the sites was detected (Additional file 1: Fig. S1, Table S1).

BGM trapping performance following VECTRACK sensor application and calculation of capture rate

According to GLMM results, the total number of mosquitoes collected by BGM + VECT was on average 32.3% lower (95% CI 2.1–53.1%, P: 0.038) than BGM (Table 2). Similar performances were estimated for the total number of Ae. albopictus adults captured (33.8%, 95% CI 1.9–55.4%; P: 0.040) and of Ae. albopictus males (35.1%, 95% CI 2.3–56.8%, P: 0.038). However, we did not find evidence against the hypothesis of comparable trapping performance between the BGM and BGM + VECT when considering Ae. albopictus females, Cx. pipiens adults, Cx. pipiens males and Cx. pipiens females (Table 2).

Table 2 Results of the generalized linear mixed models assessing difference in the number of captured mosquitoes between trap types
Full size table

Results showed that the group of four STs/site consistently captured fewer mosquitoes than each BGM, irrespectively of the species or sex (Table 2). Interestingly, the BGM and BGM + VECT showed a multiplicative factor of 14.5 (95% confidence interval: 9.1–23.0) and 10.2 (95% confidence interval: 6.3–16.6) higher captures of Ae. albopictus females compared to a single ST. The capture rate of a single ST for Ae. albopictus (defined as the ratio between the number collected adult females and the number of adult females present within the flight range of this mosquito species) was previously estimated though mark-release-recapture experiments to be on average 1.838*10−4 [22, 27]. Consequently, according to our results, the capture rates of BGM and BGM + VECT are expected to range from 1.678*10−3 to 4.219*10−3 and from 1.161*10−3 to 3.045*10−3, respectively.

VECTRACK algorithm performance in Aedes albopictus and Culex pipiens identification

To assess VECTRACK algorithm performance in the identification of mosquito species, we focused only on the collections made by the BGM + VECT and compared the number of Ae. albopictus and Cx. pipiens morphologically identified by an expert operator with the identification provided by VECTRACK algorithm. Notably, the number of specimens belonging to other Culicidae species (not targeted by VECTRACK algorithm) was negligible (N = 11).

Overall, VECTRACK algorithm counted 53 fewer Ae. albopictus and Cx. pipiens in the trap compared to an actual number of 1543 morphologically identified specimens (Table 3). This difference amounts to 3.4% unidentified Ae. albopictus and Cx. pipiens regarding the total captures. VECTRACK algorithm counted 5% (N = 69) Ae. albopictus less than the operator and overestimated the number of Cx. pipiens by 10.4% (N = 16). This amounts to a balance accuracy (as defined in González-Pérez et al.) of 99.8% and 99.4%, respectively.

Table 3 Mosquitoes identified by operator-mediated morphological examination (M) or by VECTRACK algorithm (VT)
Full size table

Overall, results show a very high correlation between the two identification methods applied across different sites and at different times for the total number of Ae. albopictus and Cx. pipiens (Spearman’s correlation coefficient, rs = 0.96, P < 0.0001). Results from the regression model (Table S2) estimated a 1.03 (95% CI 0.99–1.08) relationship between the overall number of adult mosquitoes identified by VECTRACK algorithm and by the expert entomologist, meaning that for every 100 Ae. albopictus and Cx. pipiens identified by VECTRACK algorithm, the expectation is to have captured on average 103 (SD: 8.5) Ae. albopictus and Cx. pipiens (Fig. 3).

Fig. 3
figure 3

Relationship between the number of captured target species, Aedes albopictus or Culex pipiens, as identified by VECTRACK algorithm and by operator-mediated morphological examination. Points correspond to the total number of collected adult mosquitoes (left panel), total number of collected Ae. albopictus adults (central panel) and total number of collected Cx. pipiens adults (right panel) by the sensor (x axis) and operator (y axis). The black line represents the mean predicted values obtained with the considered linear regression; the dashed lines represent the corresponding 95% prediction interval. The red diagonal line represents the ideal perfect identification performance

Full size image

Similar good performance was obtained for Ae. albopictus with a correlation of 0.96 (P < 0.0001) between the two identification methods. Overall, the number of morphologically identified Ae. albopictus was comparable to the number provided by VECTRACK algorithm (Chi-square test, χ2 = 0.409, df = 1, P = 0.52). Results from the linear model estimated a 1.05 (95% CI 1–1.1) relationship between the number of Ae. albopictus identified by VECTRACK and by the operator corresponding to an expectation of 105 (SD: 7.9) Ae. albopictus for every 100 Ae. albopictus identified by VECTRACK algorithm (Fig. 3).

The number of identified Cx. pipiens was comparable between operator and VECTRACK algorithm (Chi-square test, χ2 = 1.64, df = 1, P = 0.2). However, a less strong correlation was present (0.70, P < 0.0001) with VECTRACK algorithm overestimating the number of Cx. pipiens and showing high residual uncertainty regarding the number of captured specimens. For every 100 Cx. pipiens identified by VECTRACK algorithm, we expect to have captured on average 77 (SD: 5) Cx. pipiens (Fig. 3).

VECTRACK algorithm performance in Aedes albopictus and Culex pipiens female and male identification

Overall, VECTRACK algorithm provides high balance accuracy for females (99.1% and 98.8% for Ae. albopictus and Cx. pipiens, respectively), but a lower value for males was observed (62.6% and 62.8%). In particular, VECTRACK algorithm underestimated the number of Ae. albopictus females (– 22.9%) and overestimated the number of males (53.8%), while an opposite trend was observed for Cx. pipiens mosquitoes (23.3% overestimation of females and – 6.4% underestimation of males).

For Ae. albopictus, a very high correlation between estimates provided by operators and VECTRACK algorithm was found for both females (Spearman’s correlation coefficient, rs = 0.97; P < 0.0001) and males (rs = 0.89 P < 0.0001). However, a significant difference was found in the identification of sexes between the two identification methods (Chi-square test, χ2 = 66.1, df = 1, P < 0.001). The misidentification of Ae. albopictus females and overestimation of Ae. albopictus males was found to be systematic and consistent across sites.

The estimated relationship between Ae. albopictus females counts by VECTRACK algorithm and by operator-mediated morphological identification is 1.28 (95% CI 1.23–1.33) and was found to correct the estimate of VECTRACK algorithm providing an estimate of 128 (SD: 5.6) actual catches for every 100 females identified by VECTRACK algorithm (Fig. 4). The estimated relationship between Ae. albopictus male counts by VECTRACK algorithm and the operator is lower 0.66 (95% CI: 0.57–0.76).

Fig. 4
figure 4

Observed numbers (dots) of Aedes albopictus (left panel) and Culex pipiens females (right panel) counted by VECTRACK algorithm (x axis) and by operator-mediated morphological identification (y axis). The black line represents the mean predicted values by the linear regression and the dashed line the 95% prediction interval. The diagonal red line represents the ideal perfect identification performance

Full size image

We did not find evidence against the hypothesis of a similar identification performance of Cx. pipiens sexes between VECTRACK algorithm and by operator-mediated morphological examination (Chi-square test, χ2 = 1.17, df = 1, P = 0.28). However, the correlation between estimates provided by the two identification methods was found to be lower for both females (Spearman’s correlation coefficient, rs = 0.67 P < 0.0001) and males (rs = 0.37, P = 0.025).

Discussion

Previous studies showed the performance of the VECTRACK system in identifying Ae. albopictus and Cx. pipiens mosquitoes reared under laboratory conditions [24] as well as in field experiments carried out in two sites in Spain with prevalence of Cx. pipiens [25]. We extended the testing of the VECTRACK system to four experimental areas from northern to southern Italy where Ae. albopictus prevails and implemented an experimental design including rotation of BGM, BGM + VECT and sticky traps in each site. This allowed us to assess not only the performance of the VECTRACK algorithm in identifying the target mosquitoes under different eco-climatic conditions but also the capture capacity of the system compared to conventional BGM traps widely adopted for mosquito monitoring. We also included ST collections in the experimental design, which are known to yield fewer Ae. albopictus female catches than traps targeting host-seeking females but have relevant operation advantages (e.g. lower costs, no need of power supplies and protection from rain and thefts), to provide estimates of BGM and BGM + VECT capture rates based on the known capture rate for ST [22, 27].

The first question was whether the equipment of the BGM trap with the VECTRACK sensor could affect the capture capacity of the trap, possibly because of a lower aspiration strength or a higher point of entrance. Although the total number of mosquitoes collected by BGM + VECT was on average lower compared to BGM, the results obtained by GLMM did not highlight a significant reduction in the collections of Ae. albopictus females and Cx. pipiens females and males. Only Ae. albopictus male captures were significantly lower in BGM + VECT. Notably, the BGM + VECT system has a huge advantage compared to conventional adult traps by offering the possibility to be used continuously over time with very low maintenance, as operators receive the capture data directly on their PCs. Thus, even if the BGM + VECT system might collect significantly fewer mosquitoes than BG traps, this would not reduce its very high potential for field study and monitoring. However, results demonstrate the very good performance of BGM + VECT in mosquito collection and also suggest that the sensor could moderately affect the BG performance as the collected mosquito has to pass into a 20 cm-long tube before being trapped. This evidence could lead to further improvement of the BGM + VECT system by slightly increasing the aspiration strength. The estimates we provided for the capture rates associated with different traps would allow appropriate adjustments for estimating the mosquito abundance in sites of data collection.

Second, the results obtained allowed us to quantify the heterogeneous effectiveness associated with different trap types in collecting female mosquitoes and predict estimates of absolute mosquito density. The obtained estimates agree with the relative capture rate estimated for ST and BG traps recently obtained by simultaneously fitting mosquito data collected with different trap types [22]. Mark-release-recapture experiments conducted in Italy [27] suggested that collecting 10 Ae. albopictus adult females with a ST over a 24-h period corresponds to a local density of 4332 (95% CI 3272–5884) females per hectare. Based on this result, the estimates from the present experiment suggest that collecting 10 Ae. albopictus females in a BGM over a 24-h period may correspond to a local density of 299 (95% CI 189–475) of adult females per hectare, while 10 Ae. albopictus females collected in a BGM + VECT may correspond to a density of 423 (95% CI 261–686) adult females per hectare.

Third, the capacity of the VECTRACK algorithm to identify target mosquito species was tested. It should be highlighted here that the VECTRACK machine learning algorithm has been trained to identify Ae. albopictus, Aedes aegypti and Cx. pipiens so far [24] and that previous studies have reported the results of the identification performance at the genus level [24, 25]. Here, we report the results at the species level, as Ae. aegypti is not present in Italy and Cx. pipiens was the only species of genus Culex found in the collections. In other words, we did not include the 11 mosquito specimens belonging to other species in the analyses, but due to the small sampling size, the effect of this is negligible. Overall, the VECTRACK species identification showed > 99% balance accuracy, with a 5% underestimation of Ae. albopictus and 10.4% overestimation of Cx. pipiens. These values are consistent with results from Spain, showing a balance accuracy of ~ 95%, and 7.6% lower counts by VECTRACK algorithm compared to morphological identifications [25]. The correlation between the two identification methods applied across different sites was very high for Ae. albopictus (0.96) and lower for Cx. pipiens (0.70).

Moreover, the capacity of the VECTRACK algorithm to identify the sex of the target mosquito species was tested. Results show very high balance accuracy for females of both species (~ 99%), a value higher than those (87%–93%) calculated in the Spanish field experiments. However, balance accuracy for males (~ 62%) is lower than in previous trials (80–95%). The correlation between sex identification by operators and VECTRACK was high for both Ae. albopictus females and males (Spearman’s correlation coefficient, rs > 0.89), although a significant systematic misidentification of sexes was found in all sampling areas. However, the correlation was lower for Cx. pipiens females (Spearman’s correlation coefficient, rs = 0.67) and males (rs = 0.37).

Finally, the output of the linear model allows to predict the ranges of females and males of both species actually collected, based on the number of mosquitoes identified by VECTRACK algorithm. For instance, the model estimates that (with a 95% prediction interval) for every 100 Ae. albopictus or Cx. pipiens identified by VECTRACK algorithm, the actual catches are 89–120 or 67–86. These estimates are in the range of the variability observed by simply positioning the trap in a different nearby location or on a different day of the week.

Overall, the obtained results strongly support the VECTRACK system as a powerful tool for mosquito monitoring and study and its applicability over a range of ecological conditions. Moreover, Irideon company is further training the machine learning VECTRACK algorithm and has already released an updated version, which we preliminarily tested, showing improved sex identification for Ae. albopictus, with underestimation of the number of females reduced from – 22.9% to – 11.8%, and overestimation of the number of males reduced from 53.8% to 17.2% (data not shown). It is also important to highlight that the system is designed to work 24 h/day, 7 days/week with minimum maintenance and that this can lead to a significant increase of collected data while simultaneously reducing human effort and the associated costs. Indeed, at present, its use should be restricted to areas with limited presence of species different from Ae. albopictus and Cx. pipiens, such as most urban areas in Mediterranean Europe. At this stage, it is not possible to predict whether the algorithm performance would be reduced at high frequencies of other mosquito species which it has never been trained for. However, the VECTRACK ML algorithm is being trained to identify other species, which will eventually make its exploitation possible in different geographical regions or as a sentinel trap in potential points of entry.

In addition, the VECTRACK system has other invaluable features, as it collects data on mosquito capture, as well as temperature and humidity, in real time. This allows us to study features of mosquito bionomics virtually impossible to be studied with conventional monitoring tools, such as the daily activity patterns of the target mosquitoes, whose knowledge would be instrumental for optimized mosquito control interventions. The network of research groups collaborating within the IN-FACT project (Extended Partnership Initiative on Emerging Infectious Diseases; MUR project no. PE00000007) is taking advantage of this unprecedented opportunity to implement a large-scale study based on the VECTRACK system on Ae. albopictus and Cx. pipiens circadian rhythm patterns across different eco-climatic conditions from northern to southern Italy in 2024–2025.

Availability of data and materials

The dataset used and analyzed in the study is available from the corresponding author on reasonable request.

Abbreviations

DENV:

Dengue virus

CHIKV:

Chikungunya virus

WNV:

West Nile virus

MBDs:

Mosquito-borne diseases

BGM:

BG-Mosquitaire trap

BGM + VECT:

BG-Mosquitaire trap equipped with VECTRACK sensor

ST:

Sticky trap

References

  1. World Health Organization. World malaria report 2023. Geneva; 2023 Nov. Licence: CC BY-NC-SA 3.0 IGO. https://www.who.int/publications/i/item/9789240086173.

  2. World Health Organization. Disease outbreak news; dengue—global situation. 2023. https://www.who.int/emergencies/disease-outbreak-news/item/2023-DON498.

  3. World Health Organization. Yellow fever. Fact-sheets. 2023 May. https://www.who.int/news-room/fact-sheets/detail/yellow-fever.

  4. Lühken R, Brattig N, Becker N. Introduction of invasive mosquito species into Europe and prospects for arbovirus transmission and vector control in an era of globalization. Infect Dis Poverty. 2023;12:109. https://doi.org/10.1186/s40249-023-01167-.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Simonin Y. Circulation of West Nile virus and Usutu virus in Europe: overview and challenges. Viruses. 2024;16:599. https://doi.org/10.3390/v16040599.

    Article  PubMed  PubMed Central  Google Scholar 

  6. European Centre for Disease Prevention and Control. Surveillance, prevention and control of West Nile virus and Usutu virus Infections in the EU/EEA. https://www.ecdc.europa.eu/en/publications-data/surveillance-preventionand-control-west-nile-virus-and-usutu-virus-infections. Accessed 21 May 2024.

  7. Veo C, Della Ventura C, Moreno A, Rovida F, Percivalle E, Canziani S, et al. Evolutionary dynamics of the lineage 2 West Nile virus that caused the largest European epidemic: Italy 2011–2018. Viruses. 2019;11:814. https://doi.org/10.3390/v11090814.PMID:31484295;PMCID:PMC6784286.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. European Centre for Disease Prevention and Control. Historical data by year—West Nile virus seasonal surveillance. Stockholm: ECDC. https://www.ecdc.europa.eu/en/west-nile-fever/surveillance-and-disease-data/historical. Accessed 09 Apr 2024.

  9. European Centre for Disease Prevention and Control. Epidemiological update: West Nile virus transmission season in Europe, 2023. Stockholm: ECDC. https://www.ecdc.europa.eu/en/news-events/epidemiological-update-west-nile-virus-transmission-season-europe-2023-0. Accessed 09 Apr 2024

  10. European Centre for Disease Prevention and Control. Increasing risk of mosquito-borne diseases in EU/EEA following spread of Aedes species. 2023. https://www.ecdc.europa.eu/en/news-events/increasing-risk-mosquito-borne-diseases-eueea-following-spread-Aedes-species. Accessed 21 May 2024.

  11. Rezza G, Nicoletti L, Angelini R, Romi R, Finarelli AC, Panning M, et al. Infection with chikungunya virus in Italy: an outbreak in a temperate region. Lancet. 2007;370:1840–6. https://doi.org/10.1016/S0140-6736(07)61779-6.

    Article  CAS  PubMed  Google Scholar 

  12. Venturi G, Di Luca M, Fortuna C, Remoli ME, Riccardo F, Severini F, et al. Detection of a chikungunya outbreak in Central Italy, August to September 2017. Euro Surveill. 2017;22:17–00646. https://doi.org/10.2807/1560-7917.ES.2017.22.39.17-00646.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Caputo B, Russo G, Manica M, Vairo F, Poletti P, Guzzetta G, et al. A comparative analysis of the 2007 and 2017 Italian chikungunya outbreaks and implication for public health response. PLoS Negl Trop Dis. 2020;14:e0008159. https://doi.org/10.1371/journal.pntd.0008159.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Gjenero-Margan I, Aleraj B, Krajcar D, Lesnikar V, Klobučar A, Pem-Novosel I, et al. Autochthonous dengue fever in Croatia, august-september 2010. Euro Surveill. 2011;16:19805.

    Article  PubMed  Google Scholar 

  15. Succo T, Leparc-Goffart I, Ferré JB, Roiz D, Broche B, Maquart M, et al. Autochthonous dengue outbreak in Nîmes, South of France, July to September 2015. Euro Surveill. 2016. https://doi.org/10.2807/1560-7917.ES.2016.21.21.30240.

    Article  PubMed  Google Scholar 

  16. Lazzarini L, Barzon L, Foglia F, Manfrin V, Pacenti M, Pavan G, et al. First autochthonous dengue outbreak in Italy, August 2020. Eur Surveill. 2020;25:2001606. https://doi.org/10.2807/1560-7917.ES.2020.25.36.2001606.

    Article  Google Scholar 

  17. European Centre for Disease Prevention and Control (ECDC). Autochthonous vectorial transmission of dengue virus in mainland EU/EEA, 2010-present. Stockholm: ECDC. https://www.ecdc.europa.eu/en/all-topics-z/dengue/surveillance-and-disease-data/autochthonous-transmission-dengue-virus-eueea. Accessed: 09 Apr 2024.

  18. Barzon L, Gobbi F, Capelli G, Montarsi F, Martini S, Riccetti S, et al. Autochthonous dengue outbreak in Italy 2020: clinical, virological and entomological findings. J Travel Med. 2021;28:taab130. https://doi.org/10.1093/jtm/taab130.

    Article  PubMed  PubMed Central  Google Scholar 

  19. European Centre for Disease Prevention and Control. Guidelines for the surveillance of invasive mosquitoes in Europe. Stockholm: ECDC; 2012 Aug. https://doi.org/10.2900/61134.

  20. European Centre for Disease Prevention and Control. Guidelines for the surveillance of native mosquitoes in Europe. Stockholm: ECDC; 2014 Nov. https://doi.org/10.2900/37227.

  21. Caputo B, Manica M. Mosquito surveillance and disease outbreak risk models to inform mosquito-control operations in Europe. Curr Opin Insect Sci. 2020;39:101–8. https://doi.org/10.1016/j.cois.2020.03.009.

    Article  PubMed  Google Scholar 

  22. Zardini A, Menegale F, Gobbi A, Manica M, Guzzetta G, d’Andrea V, et al. Estimating the potential risk of transmission of arboviruses in the Americas and Europe: a modelling study. Lancet Planet Health. 2024;8:e30–40. https://doi.org/10.1016/S2542-5196(23)00252-8.

    Article  PubMed  Google Scholar 

  23. Maia LJ, Oliveira CH, Silva AB, Souza PAA, Müller NFD, Cardoso JDC, et al. Arbovirus surveillance in mosquitoes: historical methods, emerging technologies, and challenges ahead. Exp Biol Med. 2023;248:2072–82. https://doi.org/10.1177/15353702231209415.

    Article  CAS  Google Scholar 

  24. González-Pérez MI, Faulhaber B, Williams M, Brosa J, Aranda C, Pujol N, et al. A novel optical sensor system for the automatic classification of mosquitoes by genus and sex with high levels of accuracy. Parasit Vectors. 2022;15:190. https://doi.org/10.1186/s13071-022-05324-5.PMID:35668486;PMCID:PMC9169302.

    Article  PubMed  PubMed Central  Google Scholar 

  25. González-Pérez MI, Faulhaber B, Aranda C, Williams M, Villalonga P, Silva M, et al. Field evaluation of an automated mosquito surveillance system which classifies Aedes and Culex mosquitoes by genus and sex. Parasit Vectors. 2024;17:97. https://doi.org/10.1186/s13071-024-06177-w.PMID:38424626.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Facchinelli L, Valerio L, Pombi M, Reiter P, Costantini C, Della TA. Development of a novel sticky trap for container-breeding mosquitoes and evaluation of its sampling properties to monitor urban populations of Aedes albopictus. Med Vet Entomol. 2007;21:183–95. https://doi.org/10.1111/j.1365-2915.2007.00680.x.

    Article  CAS  PubMed  Google Scholar 

  27. Marini F, Caputo B, Pombi M, Tarsitani G, Della TA. Study of Aedes albopictus dispersal in Rome, Italy, using sticky traps in mark-release-recapture experiments. Med Vet Entomol. 2010;24:361–8. https://doi.org/10.1111/j.1365-2915.2010.00898.x.

    Article  CAS  PubMed  Google Scholar 

  28. Severini F, Toma L, Di Luca M, Romi R. Italian mosquitoes: general information and identification of adults (Diptera, Culicidae)/Le zanzare italiane: generalità e identificazione degli adulti (Diptera, Culicidae). Fragmenta Entomol. 2009;41:213–372. https://doi.org/10.13133/2284-4880/92.

    Article  Google Scholar 

  29. Brooks ME, Kristensen K, VAN Benthem KJ, Magnusson A, Berg CW, Nielsen A, et al. Glmmtmb balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 2017;9:378–400. https://doi.org/10.32614/RJ-2017-066.

    Article  Google Scholar 

Download references

Acknowledgements

The authors are grateful to Departments of Biology and Biotechnologies ‘Charles Darwin’ and Molecular Medicine at Sapienza University (Rome), Department of Agronomy, Food, Natural Resources, Animals and Environment (DAFNAE) of University of Padua (Prof. Mazzon and Dr. Di Bernardo) and Società Agricola Miriam, Norman Bergamelli and Nicoli Mario (Bergamo municipalities) for hosting the experiments; Mariagrazia Bodini, Raffaele Cozzolino, Tecla Mammalella, Mattia Foti, Federica Frasca, Chiara Fulgione, Chiara Gentile, Mattia Toni, Mariagrazia Sapsiano and Federico Zorzi for their help in field work activities; Annamaria Saponara for the linguistic revision of the text. The authors also thank Hugo Osorio (Public Health Entomologist at National Health Institute Lisbon, Portugal) for the fruitful discussion and sharing of unpublished data.

Funding

This study was entirely supported by Ministero dell’Università e della Ricerca (Italy), Piano Nazionale di Ripresa e Resilienza, and EU within the Extended Partnership Initiative on Emerging Infectious Diseases project number PE00000007 (One Health Basic and Translational Actions Addressing Unmet Needs on Emerging Infectious Diseases). M Micocci was supported by PON Research and Innovation 2014–2020 (D.M. 1061/2021), PhD grant DOT1326YBR-1.

Author information

Author notes

  1. Martina Micocci and Mattia Manica contributed equally to this work.

Authors and Affiliations

  1. Department of Public Health and Infectious Diseases, Sapienza University of Rome, Rome, Italy

    Martina Micocci, Beniamino Caputo & Alessandra della Torre

  2. Center for Health Emergencies, Fondazione Bruno Kessler, Trento, Italy

    Mattia Manica & Piero Poletti

  3. Department of Infectious Diseases, Istituto Superiore Di Sanità, Rome, Italy

    Ilaria Bernardini & Francesco Severini

  4. Department of Biosciences and Pediatric Clinical Research Center “Romeo Ed Enrica Invernizzi”, University of Milan, Milan, Italy

    Laura Soresinetti & Sara Epis

  5. Department of Biology, University of Naples Federico II, Naples, Italy

    Marianna Varone, Paola Di Lillo & Marco Salvemini

  6. Istituto Zooprofilattico Sperimentale Delle Venezie, Legnaro, Italy

    Fabrizio Montarsi

Authors
  1. Martina Micocci
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Mattia Manica
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Ilaria Bernardini
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Laura Soresinetti
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Marianna Varone
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Paola Di Lillo
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Beniamino Caputo
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Piero Poletti
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Francesco Severini
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Fabrizio Montarsi
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Sara Epis
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Marco Salvemini
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Alessandra della Torre
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

AdT, FM, FS, MMan, MS and SE conceived and designed the study. BC, FM, FS, IB, LS, MMic, MS, MV, PDL and SE conducted the fieldwork and performed the morphological identification of mosquito samples. MMan, MMic and PP analyzed the results. AdT, MMan, MMic and PP drafted the manuscript. All authors contributed to and approved the final version of the manuscript.

Corresponding author

Correspondence to Alessandra della Torre.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

13071_2024_6479_MOESM1_ESM.jpeg

Additional file 1: Fig. S1 Potential depletion of mosquitoes collected in each trap type after repeated sampling in each location. Each trap was analyzed separately because of the potentially different capture rate. Points are observed values; dashed line is a simple linear regression fit to help visualize a potential trend.

13071_2024_6479_MOESM2_ESM.docx

Additional file 2: Table S1 Results of the generalized linear mixed models assessing potential depletion after repeated mosquito collections. BG-Mosquitaire trap (BGM) considered as reference. Parameters values and their 95% confidence intervals are not exponentiated.

13071_2024_6479_MOESM3_ESM.docx

Additional file 3: Table S2 Estimated linear relationship between operator-based morphological identification and identification by VECTRACK algorithm of females and males of Aedes albopictus and Culex pipiens.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Micocci, M., Manica, M., Bernardini, I. et al. An easier life to come for mosquito researchers: field-testing across Italy supports VECTRACK system for automatic counting, identification and absolute density estimation of Aedes albopictus and Culex pipiens adults. Parasites Vectors 17, 409 (2024). https://doi.org/10.1186/s13071-024-06479-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13071-024-06479-z

Keywords

Parasites & Vectors

ISSN: 1756-3305

Contact us