Background: Aedes albopictus is an aggressive invasive mosquito species that represents a serious health concern not only in tropical areas, but also in temperate regions due to its role as vector of arboviruses. Estimates of mosquito biting rates are essential to account for vector-human contact in models aimed to predict the risk of arbovirus autochthonous transmission and outbreaks, as well as nuisance thresholds useful for correct planning of mosquito control interventions. Methods targeting daytime and outdoor biting Ae. albopictus females (e.g., Human Landing Collection, HLC) are expensive and difficult to implement in large scale schemes. Instead, egg-collections by ovitraps are the most widely used routine approach for large-scale monitoring of the species. The aim of this work was to assess whether ovitrap data can be exploited to estimate numbers of adult biting Ae. albopictus females and whether the resulting relationship could be used to build risk models helpful for decision-makers in charge of planning of mosquito-control activities in infested areas.
Method: Ovitrap collections and HLCs were carried out in hot-spots of Ae. albopictus abundance in Rome (Italy) along a whole reproductive season. The relationship between the two sets of data was assessed by generalized least square analysis, taking into account meteorological parameters.
Result: The mean number of mosquito females/person collected by HLC in 15' (i.e., females/HLC) and the mean number of eggs/day were 18.9 ± 0.7 and 39.0 ± 2.0, respectively. The regression models found a significant positive relationship between the two sets of data and estimated an increase of one biting female/person every five additional eggs found in ovitraps. Both observed and fitted values indicated presence of adults in the absence of eggs in ovitraps. Notably, wide confidence intervals of estimates of biting females based on eggs were observed. The patterns of exotic arbovirus outbreak probability obtained by introducing these estimates in risk models were similar to those based on females/HLC (R0 > 1 in 86% and 40% of sampling dates for Chikungunya and Zika, respectively; R0 < 1 along the entire season for Dengue). Moreover, the model predicted that in this case-study scenario an R0 > 1 for Chikungunya is also to be expected when few/no eggs/day are collected by ovitraps.
Discussion: This work provides the first evidence of the possibility to predict mean number of adult biting Ae. albopictus females based on mean number of eggs and to compute the threshold of eggs/ovitrap associated to epidemiological risk of arbovirus transmission in the study area. Overall, however, the large confidence intervals in the model predictions represent a caveat regarding the reliability of monitoring schemes based exclusively on ovitrap collections to estimate numbers of biting females and plan control interventions.
Keywords: Arboviruses; Biting rate; Chikungunya; Dengue; Epidemiological modelling; Europe; Invasive mosquitoes; Ovitrap; R0; Zika.