Risk-Based Mapping Tools for Surveillance and Control of the Invasive Mosquito Aedes albopictus in Switzerland

Abstract

Background: In Switzerland, Aedes albopictus is well established in Ticino, south of the Alps, where surveillance and control are implemented. The mosquito has also been observed in Swiss cities north of the Alps. Decision-making tools are urgently needed by the local authorities in order to optimize surveillance and control.

Methods: A regularized logistic regression was used to link the long-term dataset of Ae. albopictus occurrence in Ticino with socioenvironmental predictors. The probability of establishment of Ae. albopictus was extrapolated to Switzerland and more finely to the cities of Basel and Zurich.

Results: The model performed well, with an AUC of 0.86. Ten socio-environmental predictors were selected as informative, including the road-based distance in minutes of travel by car from the nearest cell established in the previous year. The risk maps showed high suitability for Ae. albopictus establishment in the Central Plateau, the area of Basel, and the lower Rhone Valley in the Canton of Valais.

Conclusions: The areas identified as suitable for Ae. albopictus establishment are consistent with the actual current findings of tiger mosquito. Our approach provides a useful tool to prompt authorities' intervention in the areas where there is higher risk of introduction and establishment of Ae. albopictus.

Keywords: Aedes albopictus; ecological niche model; environmental factors; ovitrap; regularized logistic regression; surveillance.

Figure A1

Heatmap of the Spearman’s rank correlation coefficient between predictors selected by the ensemble model.

Figure A2

Heatmap of the Spearman’s rank correlation coefficient between the selected predictors and the other predictors (only variables with absolute correlation coefficient larger than 0.2 for at least one selected predictor are included).

Figure 1

Occurrence of Ae. albopictus in Ticino, Switzerland, from 2005 to 2015. (a) The empty dots represent observed cells; (b) the solid dots represent cells with establishment of the mosquito. Refer to methods for definition of observed and established cells. Maps modified from https://map.geoadmin.ch/ (accessed on 24 February 2022).

Figure 2

Coefficients assigned by the ensemble models to the mostly used features (i.e., features used by at least half of the models).

Figure 3

Probabilities of establishment of Ae. albopictus in Switzerland. The map on top shows the average risk estimate over the years 2015–2018. The color gradient shows the probability of establishment from 0 (white) to 1 (red). The bottom map represents the uncertainty of the prediction (higher values representing more uncertain predictions). Dots show cells where the presence of Ae. albopictus was monitored in 2021, the color green representing cases where it was found absent from the cell, blue where it was present but not established, and red where it was established (data source: Swiss Mosquito Network and info fauna—CSCF). Maps modified from https://map.geoadmin.ch/ (accessed on 24 February 2022).

Figure 4

Probabilities of establishment of Ae. albopictus in Basel. The map on top shows the average risk estimate over the years 2015–2018. The color gradient shows the probability of establishment from 0 (white) to 1 (red). The bottom map represents the uncertainty of the prediction (higher values representing more uncertain predictions). Black dots show the positions where Ae. albopictus was established in 2019, which are used to compute the feature car distance to establishment (distance from the closest cell with establishment in the previous year). Maps modified from https://map.geoadmin.ch/ (accessed on 24 February 2022).

Figure 5

Probabilities of establishment of Ae. albopictus in Zurich. The map on top shows the average risk estimate over the years 2015–2018. The color gradient shows the probability of establishment from 0 (white) to 1 (red). The bottom map represents the uncertainty of the prediction (higher values representing more uncertain predictions). Black dots show the positions where Ae. albopictus was established in 2019, which are used to compute the feature car distance to establishment (distance from the closest cell with establishment in the previous year). Maps modified from https://map.geoadmin.ch/ (accessed on 24 February 2022).