Estimating Spatio-Temporal Dynamics of Aedes Albopictus Dispersal to Guide Control Interventions in Case of Exotic Arboviruses in Temperate Regions

Abstract

The increasing number of exotic arbovirus cases imported in Europe and the 2017 chikungunya outbreak in central/southern Italy highlight the urgency of evidence-based outbreak management plans to predict, prevent or interrupt spreading of these arboviruses to non-endemic countries in temperate regions. We here present the results of three mark-release-recapture experiments conducted in a peri-urban area of North-East Italy to estimate the spatio-temporal dynamics of the dispersal of Aedes albopictus females looking for oviposition sites. The Flight Range of 90% of the mosquito population (FR90) was found to exceed 200 m, consistently with data obtained from a previous study conducted in a highly urbanised area in Rome (Central Italy). Modelling results showed that dispersal can be so rapid that insecticide spraying within a 200m-radius around a potential infected case leaves >10% probability that a potentially infected mosquito escapes the treatment, even if this is carried out after only 2–3 days since the importation of a viremic case. These data provide evidence in favour of an update of guidelines for the control of exotic autochthonous arbovirus transmission in temperate areas and highlight the need of effective surveillance approaches and rapid response to contain the risks associated to imported viremic cases.

Full text

1
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
Estimating Spatio-Temporal
Dynamics of Aedes Albopictus
Dispersal to Guide Control
Interventions in Case of Exotic
Arboviruses in Temperate Regions
Francesca Marini
1,7, Beniamino Caputo1, Marco Pombi
1, Manuela Travaglio2,
Fabrizio Montarsi3, Andrea Drago4, Roberto Rosà5,6, Mattia Manica5 & Alessandra della Torre
1
The increasing number of exotic arbovirus cases imported in Europe and the 2017 chikungunya
outbreak in central/southern Italy highlight the urgency of evidence-based outbreak management plans
to predict, prevent or interrupt spreading of these arboviruses to non-endemic countries in temperate
regions. We here present the results of three mark-release-recapture experiments conducted in a peri-
urban area of North-East Italy to estimate the spatio-temporal dynamics of the dispersal of Aedes
albopictus females looking for oviposition sites. The Flight Range of 90% of the mosquito population
(FR90) was found to exceed 200 m, consistently with data obtained from a previous study conducted in
a highly urbanised area in Rome (Central Italy). Modelling results showed that dispersal can be so rapid
that insecticide spraying within a 200m-radius around a potential infected case leaves >10% probability
that a potentially infected mosquito escapes the treatment, even if this is carried out after only 2–3 days
since the importation of a viremic case. These data provide evidence in favour of an update of guidelines
for the control of exotic autochthonous arbovirus transmission in temperate areas and highlight the
need of eective surveillance approaches and rapid response to contain the risks associated to imported
viremic cases.
Knowledge about the spatio-temporal dynamics of dispersal of adult mosquito vectors of diseases is instrumental
for understanding mosquito-borne disease transmission dynamics and for eectively calibrating control inter-
vention in order to prevent or interrupt pathogen transmission. Active dispersal of adult mosquitoes is triggered
by the need to nd mates, sugar sources, resting sites and, in the case of females, hosts for blood-meals and
oviposition sites and is highly aected by intrinsic species-specic ight capability, as well as by ecological (e.g.
abundance and location of sugar sources, hosts, resting sites and oviposition sites) and climatic (e.g. temperature,
rainfall, light intensity, wind speed and direction at ground level) conditions1. Reliable estimates of active mos-
quito dispersal and ight ranges are thus very hard to obtain. is is even more challenging in the case of exo-
phagic and exophilic species which are particularly dicult to recapture in the frame of Mark–Release–Recapture
(MRR) studies1.
For the above reasons, little is known about the ight range of Aedes albopictus, the Asian Tiger Mosquito,
the species which in the last 30 years has invaded all continents (except Antarctica) thanks to the passive trans-
portation of its eggs mostly inside used tires and lucky bamboos, and has stably colonized not only tropical, but
1Dipartimento di Sanità Pubblica e Malattie Infettive, “Sapienza” Università di Roma, Piazzale Aldo Moro 5, 00185,
Rome, Italy. 2Dipartimento di Biologia, Università di Padova, Viale G. Colombo 3, 35121, Padua, Italy. 3Istituto
Zooprolattico Sperimentale delle Venezie, Viale dell’Università 10, 35020, Legnaro, (PD), Italy. 4ENTOSTUDIO
srl, Viale del Lavoro 66, 35020, Ponte San Nicolò, (PD), Italy. 5Department of Biodiversity and Molecular Ecology,
Research and Innovation Centre, Fondazione Edmund Mach, Via E. Mach 1, 38010, San Michele all’Adige, Italy.
6Center Agriculture Food Environment, University of Trento, 38010, San Michele all’Adige, Trento, Italy. 7Present
address: Biotechnology and Biological Control Agency (BBCA) Onlus, Via Angelo Signorelli 105, 00123, Rome, Italy.
Correspondence and requests for materials should be addressed to A.d. (email: ale.dellatorre@uniroma1.it)
Received: 21 August 2018
Accepted: 3 June 2019
Published: xx xx xxxx
OPEN
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
also temperate regions due to the capacity to overcome the cold months in a state of embryonic hibernation2.
ese capacities have turned the species into one of the most successful invasive animal species worldwide3 and
in a global public health threat due to its competence to transmit a large number of exotic arboviruses (such as
Chikungunya (CHIKV), Dengue (DENV),4 and Zika5). is is already testied by Ae. albopictus central role in
large CHIKV epidemics in Indian Ocean in 2006–076 and in the Caribbean and Americas since 20157 and in
several cases of autochthonous transmission of CHIKV and DENV in temperate countries8.
Few MRR studies have been carried out to investigate the capacity to disperse of both host-seeking gravid
Ae. albopictus females looking for a host and blood-fed females looking for oviposition sites. In the former case,
most marked host-seeking females were collected within 100 m from the release site (Hawaii9, Missouri10, La
Réunion Island11), while investigation carried out in Texas, on host-seeking females emerged from larvae marked
with stable isotopes, showed that approximately 79% of them were found within 250 m from their natal site and
all of them remained within a 1 km distance12. However, females released in the forest were shown to ight up
to 1 km to reach an urban area with high densities of possible hosts13. In the case of oviposition-associated dis-
persal, studies exploiting the release of Rb-labelled females and recapture of Rb-marked eggs reported presence
of Rb-marked eggs in ovitraps throughout the whole 160 m radius urban area sampled in Singapore14, while in
Brazil most Rb-marked eggs and uorescent dusted females were found within 100–200 m from a release site13,15.
In Florida, Davis et al.16 showed that self-marked gravid Ae. albopictus female dispersal increased with time, but
seemed to stabilize around 90 m from the initial marking site.
In 2008, we carried out 3 MRR experiments within a highly urbanised area in the city of Rome (central Italy)
to analyse dispersal of orescent dusted Ae. albopictus females released aer blood-meal and recaptured by sticky
traps (STs), and obtained ight ranges of 90% of the released females ranging between 168 and 236 m17. e rst
aim of the present study - carried out with the same design, but in a peri-urban area in north-east Italy - was
to understand whether the ight ranges assessed in Rome could represent a common feature of Ae. albopictus
female dispersal associated to search of an oviposition site in temperate regions. In addition, we exploited the
data to model the temporal dynamics of the dispersal. is parameter has been largely neglected in previous
studies, despite its high relevance for models estimating the dynamics of spread of autochthonous transmission
of exotic arboviruses in temperate regions, as well as for the calibration of the spatial/temporal scale of insec-
ticide treatments to be implemented to prevent or interrupt spread of arbovirus outbreaks. In fact, the recent
cases of autochthonous CHIKV and DENV transmission in Europe8 highlight the urgency and timeliness to
improve preparedness to predict, prevent or interrupt spreading of these and other exotic tropical arboviruses to
non-endemic northern countries.
Results
Mark-release-recapture experiments. A total of 356 marked Ae. albopictus females were collected
out of the 3,959 released ones (MRR1 = 101/1,149, MRR2 = 214/1,600; MRR3 = 41/1,210; Supplementary
TableS2). In addition, 16,938 wilds Ae. albopictus were collected: 79.3% females (MRR1 = 3,607; MRR2 = 7,065;
MRR3 = 2,754), 14.4% males (MRR1 = 669; MRR2 = 1,533; MRR3 = 234), while for the remaining 6.4%
(MRR1 = 250; MRR2 = 401; MRR3 = 425) was not possible to determine the gender, due to bad preservation of
carcasses glued on the STs. Some specimens, mostly females, of Culex sp., Culiseta sp.. and Aedes caspius were also
found in the STs but not counted.
e recapture rates at 11 days from release were 8.8% (MRR1), 12.9% (MRR2), 3.0% (MRR3) and signicantly
dierent among the three experiments (Supplementary TableS1; comparison between GLMs with or without
releases as covariate: GLM deviance = 96.5, df = 2, P < 0.0001). Recapture rates observed in MRR1 and MRR2
were higher than those observed in the study carried out with a similar experimental design in a smaller study
area (i.e. 250 m radius vs 500 m radius, 55 STs vs 96 STs) in Rome (i.e. the odds of recapturing a marked mosquito
was 0.45 lower in Rome with a condence interval of 0.39–0.59; Binomial GLM: deviance 39.1, p-value < 0.001).
is may be due to the larger sampling area and the higher number of STs used in the present study as opposed
to Rome and/or to an inherent bias in using STs as recapture method (i.e. the power of recapture is dependent
on the competition with other natural oviposition sites in the study area, usually quite abundant in urban areas).
However, it is worth to note that the recapture rates obtained in both studies are generally higher than those
obtained in MRR studies on Ae. albopictus females dispersal aer blood-meal carried out with dierent recapture
approaches9,10,13. ese results conrm that STs are a valuable tool for MRR studies on Ae. albopictus ovipositing
or resting females (see also Marini et al.17 for a detailed discussion on pros and cons of using STs in the context
of MRR studies).
Notably, more than 70% of recaptured marked females were collected within 150 m from release site (Figs1B–
D, 2A; Supplementary TableS2) and during the rst 6 days of sampling (Fig.2B; Supplementary TableS2).
Table1 summarizes the mean distance travelled (MDT), maximum observed distance travelled (maxODT)
and ight range of 90% (FR90) and 50% (FR50) of marked females in each MRR experiment. Cumulative MDTs
were 110 m (days 2–11), 77 m (days 2–16) and 68 m (days 2–16) in MRR1, MRR2 and MRR3, respectively.
Daily MDTs were statistically dierent between the rst two MRRs and MRR3 (Mann Whitney test MRR1 vs
MRR3: U = 4, P = 0.03; MRR2 vs MRR3: U = 1, P = 0.01), but not between MRR1 and MRR2 (Mann Whitney
test U = 27.5, P = 0.19). e maxODTs were 463 m (day-4), 433 (day-5) and 375 (day-6) in MRR1, MRR2 and
MRR3, respectively. e ight ranges were always ≤50 m for FR50 and ≤252 m for FR90 (Table1) and statis-
tically dierent among the three MRR experiments (i.e. the interaction eect between log10(annulus median
distance + 1) and release was signicant as shown by a Likelihood ratio test P < 0.01, so the three lines were not
parallel between them: MRR1 vs MRR2: t = 4.98, P < 0.01; MRR1 vs MRR3: t = 4.08, P < 0.01; MRR2 vs MRR3:
t = −9.06, P < 0.01).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

3
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Flight ranges estimates were on average in agreement with results of Gamma GLM and ZAG models, as shown
in Fig.3, which shows model results for the 90% percentiles of the distance up to which mosquitoes are expected
to travel, a proportion chose in compliance to allow comparison with the estimated FR90. e 90% percentile
values reached about 250 m aer 2 (MRR1) and 14 (MRR2) days since release (this estimate was not reached
during the sampling period in MRR3). On average, the 90% percentile of the distance up to which mosquitoes are
expected to y was stable in MRR1, but it was predicted to increase with time in MRR2 (from <150 m at day-2
to ~250 m at day-14) and in MRR3 (from <100 m at day-2 to ~200 m at day-14), although with shorter estimates
and wider condence intervals due to the lower numbers of mosquito recaptured (Fig.3).
Result of the Bernoulli GLM model allows to estimate the distance up to which 90% of marked mosquito
is expected to be detected. is estimate was stable in MRR1 and MRR2, ranging on average between 170 and
300 m, while a shorter range (between 100 and 200 m) was estimated for MRR3 (Fig.3). Similar to ight range
estimates, also the 50% percentiles of the distance up to which mosquitoes are expected to y were computed for
all three models (see Supplementary Material and Supplementary Fig.S1).
Effect of fluorescent dye marking on mosquito survival. The survival of marked Ae. albopictus
females kept under semi-eld conditions for the duration of the experiments was shown not to be aected by the
uorescent dyes. In fact, no dierences in survival between marked (N = 56, 43 and 43) and unmarked (N = 49,
48 and 48) females were recorded in MRR1, MRR2 and MRR3, respectively (MRR1: log-rank test χ2 = 0.15,
df = 1, p = 0.70; MRR2: log-rank test χ2 = 0.26, df = 1, p = 0.61; MRR3: χ2 = 0.10, df = 1, p = 0.76).
Figure 1. (A) Study area in Piove di Sacco (north-east Italy), where 96 sticky traps (yellow dots) were located
into concentric annuli of 50 up to maximum of 500 m radius around the release point (red star). Map layout
generation were done using QGIS 2.18 (QGIS Development Team (2017). QGIS Geographic Information
System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org) free and open source soware
for geographic information system. e base layer “Regione del Veneto – L.R. n. 28/76 – Formazione della
Carta Tecnica Regionale” was obtained from the Veneto Region Geoportal (https://idt2.regione.veneto.it/
condizioni_utilizzo_geoportale/) distributed under Italian Open Data License 2.0 (IODL 2.0 http://www.dati.
gov.it/iodl/2.0/). (B–D) Distribution of recaptured Aedes albopictus females in the sticky-traps during the three
mark-release-recapture experiments (MRR1, MRR2 and MRR3).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

4
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Oviposition pattern under semi-field conditions. A positive correlation was observed in MRR2
(Spearman’s rank correlation: r = 0.92; t = 5.38, df = 5, p < 0.01) between the frequency of recaptured Ae. albopictus
females and the oviposition pattern of blood-fed females simultaneously to the released ones and kept under
semi-natural conditions for the entire length of the experiments (Fig.4). No correlation was observed in MRR3
(Spearman’s rank correlation: r = 0.42; t = 1.05, df = 5, p = 0.34), when a strong thunderstorm and heavy rainfall
occurred few hours aer the release, likely aecting more the released females than those kept under semi-natural
(and “sheltered”) conditions. Consistently with the temperature drop aer MRR3 (Supplementary TableS3), the
oviposition dynamics of marked females in MRR2 and MRR3 showed a 2-day shi in the peak of oviposition
activity.
Discussion
e rst aim of the present work was to study active dispersal of blood-fed Ae. albopictus females in a peri-urban
habitats in order to conrm or dispute that evidence obtained in a similar study carried out in a highly urbanized
study area in Rome17 can be generalised and represent a reliable estimate of Ae. albopictus female dispersal asso-
ciated to search of an oviposition site in other temperate regions.
e results obtained showed that 90% of marked Ae. albopictus blood-fed females were recaptured within
250 m from the release site (FR90 = 177–252 m), a range consistent with estimates from the results obtained in
Rome (FR90 = 168–23617). Moreover, the comparison between the oviposition patterns of females kept under
semi-eld conditions and the recapture rates in MRR2 supports the hypothesis that the observed dispersal was
mostly triggered by the need of Ae. albopictus females released at the blood-fed stage to nd a suitable oviposition
site aer blood-meal digestion (Fig.4). erefore, 250 m may represent a consistent and possibly generalizable
estimate of FR90 of blood-fed Ae. albopictus females in Italy, and possibly in temperate regions, as this value did
not signicantly dier in relation to the ecology (peri-urban/rural vs urban) or the size (500 m vs 250 m radius) of
Figure 2. Cumulative proportion of recaptured Aedes albopictus females on total of recaptured across the study
area (A) and during the rst 11 days aer release (B), in the three mark-release-recapture experiments (MRR1,
MRR2 and MRR3).
Days aer
release MRR1 MRR2 MRR3
Daily MDT (Daily maxODT)
2 75 (73) 95 (114) — —
3 80 (287) 46 (275) — —
4 171 (463) 72 (388) 58 (71)
5 104 (436) 97 (433) 35 (71)
6 106 (333) 72 (131) 78 (375)
7–11 107 (352) 121 (333) 25 (168)
12–16 — — 132 (433) 25 (192)
FR90 252 209 177
FR50 50 31 21
Table 1. Mean distance travelled (MDT), maximum observed distance travelled (maxODT) and ight ranges of
90% (FR90) and 50% (FR50) of recaptured Aedes albopictus females in three mark-release-recapture experiments
(MRR1, MRR2 and MRR3). All parameters are expressed in meters.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

5
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
the two study areas tested. However, it is important to highlight that the results obtained are based on mosquitoes
reared under optimal conditions, while adult tness and behaviour in nature are certainly aected by several
features such as competition for trophic resources at the larval stage and climatic conditions. For instance, Mori18
reported that Ae. albopictus emerging from larvae reared under crowded conditions dispersed further than those
reared in uncrowded conditions. Also, increased skip oviposition was reported in adults exposed to low-quality/
crowded versus high-quality/uncrowded larval habitats16. ese observations imply that gravid females will
move greater distances when larval habitat quality decreases and competition for resources increases. On the
other hand, extreme climatic conditions negatively aect survival and dispersal. In fact, compared to results from
MRR1 and MRR2 (which were carried out under optimal climatic conditions), results from MRR3 (which was
aected by heavy rainfall and a severe drop in temperature) show a much lower recapture rate, a delayed oviposi-
tion dynamics between released females and females kept in sheltered cages, and lower-distance dispersal.
e second aim of the study was to model the temporal dynamics of the dispersal, a parameter to our knowl-
edge previously analysed only in a study carried out in Florida showing that the dispersal of gravid Ae. albopictus
females increased with time, but seemed to stabilize around 90 m from the initial marking site16. Our analyses of
the temporal patterns of dispersal revealed that the distance travelled by blood-fed females is longer and does not
always increase over time. In fact, results from MRR1 showed that the potential exists for a very rapid dispersal,
as it is predicted that about 90% of marked mosquitoes have own within 250 m in 72 hours since release and
the remaining 10% have own even longer (Fig.3). Both in Padua and in Rome single females were found at the
limits of the study areas only 4 days since the release (maxODT Padua = 464 m; maxODT Rome = 290 m). In
particular, Bernoulli GLM model showed that the detection probability was not aected by the day of the release
in MRR1 and MRR2, suggesting that some individual mosquitoes could rapidly disperse in very few days aer
release (see Supplementary TableS2, day 4).
e obtained results are very relevant for the planning of control activities to prevent autochthonous transmis-
sion of exotic arboviruses in temperate areas, an occurrence which is likely to become recurrent, as testied by the
Chikungunya outbreak occurred in central Italy in summer 2017 with almost 500 human infected cases19, as well
as in France20 and Croatia21, and by the models predicting future patterns of transmission under climate change
scenarios22,23. e delayed application of peridomestic/perifocal space spray treatments with insecticides around
households where human arbovirus infection has been reported is a critical issue concerning their eectiveness in
reducing transmission risk. Despite that, guidelines from the Italian and French Ministries of Health24,25 recom-
mend insecticide treatments in a 200 m buer area around residence of conrmed infected patient independently
from the delay between notication and symptom onset. Based on our results and the model’s predictions over
time, however, the recommended 200 m buer does not appear to be sucient to target 90% of the mosquito
population which may have been in contact with the infected case in her/his household. First, the FR90 was found
to exceed 200 m in some of the MRR experiments carried both in Rome and in Padua. Second, modelling results
showed that dispersal can be so rapid that insecticide spraying within a 200 m radius around a potential infected
Figure 3. Distance up to which 90% of mosquitoes are expected to travel estimated by three models (Bernoulli
GLM, Gamma GLM, Zero Altered Gamma model). On the x-axis the days aer release, on the y-axis the
distance from the release site. Dots represent the mean distance value; vertical lines represent the 95%
condence intervals obtained by non-parametric bootstrap. e horizontal solid black lines represent the ight
ranges of 90% marked Aedes albopictus (FR90, calculated according to Lillie et al.29, White & Morris30, and
Morris et al.31). Each panel identies a mark-release-recapture experiment (MRR1, MRR2 and MRR3).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

6
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
case could fail to reduce below 10% the probability that a potentially infected mosquito escapes the treatment,
even if this is carried out aer 2–3 days since the arrival of a viremic case (i.e. the shortest plausible time-interval
for activation of mosquito control activities aer the report of suspect case of arbovirosis to health authorities).
In conclusion, the above results suggest that insecticide treatments aimed to prevent risk of autochthonous
outbreaks of exotic arboviruses in temperate regions may fail to reach a non-neglectable part of the potential
vector population even if implemented in a 200-m buer aer reports of an infected case to health authorities,
considering an average of 6 days delay between symptom onset and detection26. Overall, this reinforce the claim
that control activities carried out around the place of residence of infected cases are not sucient to control an
emerging arbovirus outbreak, not only due to the importance of outdoor activities in the spread of Aedes-borne
diseases in temperate countries27, but also due to operational constraints in intervening in an adequate area in
due time.
Methods
Study area. e study area (≈80 hectares) lies within the municipality of Piove di Sacco (Padua province,
north-east Italy) and is characterized by either peri-urban or rural habitats (Fig.1A). e peri-urban habitat is
mostly characterized by two-story residential buildings with small gardens and includes football and recreational
grounds, small stores, schools, a fuel station and a church. e rural habitat comprises large corn and alfalfa cul-
tivations, small vineyards, horse and sheep pastures, and irrigation canals. A highway delimits the northern and
southern extremes of the study area.
Mark-release-recapture experiments. ree MRR experiments were performed in August 3rd (MRR1)
and 24th (MRR2), and in September 9th (MRR3) 2009, with the agreement of Piove di Sacco municipality.
Temperature data during the experiments (Supplementary TableS3) were obtained from the ARPAV agrome-
teorological station of Legnaro (45°20′51N, 11°57′08E), located about 10 km northwest from the centre of the
study area (45°18′21″N, 12°01′49″E). Rainfall was absent/very low during all MRRs (≤0.2 mm), except for a few
days of moderate rain in MRR1 (i.e. 9.2 mm on day-7 and 3.8 mm on both day-10 and day-11 since release) and
a strong thunderstorm (≈170 mm of rain) occurred few hours aer release in MRR3 (Supplementary TableS3).
Mark-release-recapture experiments were carried out using the same approach described in Marini et al.17 in
Rome.
Figure 4. Frequencies of marked Aedes albopictus females collected (grey bar) in the second (MRR2, (A) and
third (MRR3, (B) mark-release-recapture experiment versus frequencies of ovipositing females kept in semi-
natural conditions (single ovipositions, black line) during the same interval.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

7
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Aedes albopictus females released were obtained from eggs collected by >20 ovitraps located in Padua prov-
ince. Eggs were le to hatch and larvae were reared in plastic basins (35 × 28 × 8 cm) at larval density ∼1 larva/
mL with cat pellets as food (Friskies® Adults). Pupae were transferred to 40 cm cubic cages, where adults emerged
and were maintained with 10% sugar solution until the day of release. Eggs, larvae and adults were always kept
outdoors in a site sheltered from direct sunlight and wind.
In the morning of each release, 2–4 days-old females were fed with debrinated fresh bovine blood by a
hand-made membrane feeding system28. Blood-fed females were transferred into paper-cups closed by a net in
groups of 20, and marked by gently dropping orange uorescent dust (Day-Glo Color Corp., Cleveland, OH,
U.S.A.) into the cup. Marked females were then transferred back into the cages and provided with a 5% sugar
solution. In the evening, cages were brought by car to the release site in the centre of the study area inside paper
boxes. At 10 PM - when Ae. albopictus is known to be inactive2 - marked females were released by opening the
cages, which were le open all night long, in order to leave time to mosquitoes to y away. Alive and dead indi-
viduals found in the cages the following morning were counted and excluded from the total number of released
mosquitoes.
Collections of adult mosquitoes were performed using 96 sticky-traps (STs)29 located at ground level in shel-
tered positions and geo-referenced, using a global positioning system device (Garmin GPSMAP 60CSx). In order
to have the same collection-power at every distance from the release site, the study area was virtually subdivided
into 10 concentric annuli (from 50 to 500 m radius) around the release site, with the number of STs/annulus
increased from the centre to extremes of the sampling area to maintain the proportion ST/area to an average
density of ≈1 ST/8,200 m in each annulus (Fig.1A). All STs were activated by adding ≈500 ml of tap-water and
adhesive sheets approximately 12 h aer release (day-1); only the ST located on the release site was activated 24 h
later (day-2), to avoid collecting individuals which did not disperse immediately aer release. Hereaer, the 12 h
between release and STs activation are indicated as “day-0”. STs were monitored daily for the rst 6 days and at
days 11 (MRR1, MRR2 and MRR3) and -16 (MRR2 and MRR3) aer release, as described in Marini et al.17. All
collected mosquitoes were counted and morphologically subdivided by genus/species and gender under a dissect-
ing microscope; the presence of uorescent dust was checked under UV light.
Recapture rates were calculated for each MRR experiment as the proportion of the number of recaptured
marked mosquitoes over the total number of released ones. Recapture rates in the three MRRs experiments were
compared by a Bernoulli GLM, with the number of mosquitoes recaptured within the rst 11 days as response
variable and the MRR experiment as covariate. ese and the following statistical analyses were carried out using
the statistical soware R30.
To allow direct comparison of the results obtained in the present study and those previously obtained in
Rome, dispersal of the released mosquitoes was analysed as in Marini et al.17, i.e. estimating: (i) the Mean
Distance Travelled during the entire sampling period (MDT, a parameter not inherently biased by trap loca-
tion or size of study area31–33); (ii) the daily MDTs for the rst 6 days aer release; (iii) the maximum Observed
Distance Travelled (max ODT), corresponding to the linear distance from the release site to the farthest positive
ST (i.e. where at least one marked mosquito female was collected); and iv) the Flight Ranges (FRs) based on the
linear regression of the cumulative number of expected recaptures (ERs) from each annulus (x-axis) on the log10
(annulus median distance +1). FR50 and FR90 values were calculated from the equation of regression line as 50%
and 90% of the largest ER value, respectively31–33. Additionally, the linear regression of the cumulative number
of expected recaptures (ERs) was also computed considering the log10 (annulus median distance +1), the release
and their interaction to test the dierence between ight ranges in each release (by comparing the slopes of the
three regression lines).
To provide estimates of the distance up to which marked Ae. albopictus females are expected to be found con-
ditional to the time of release, dispersal data were further analysed by means of the following three models: (1) a
GLM with Bernoulli distribution and logit link to model the probability to detect a marked mosquito condition-
ally to days and distance from release, and therefore to estimate the probability to detect a marked mosquito at
a given distance. e response variable in the model is the presence/absence of marked mosquitoes into the STs,
while covariates are the distances of STs from the release site and the days aer release; (2) a GLM with Gamma
distribution and log link (hereaer Gamma GLM) to model the distance travelled by marked mosquitoes con-
ditional to the days aer release and to compute the percentile of the distance travelled by recaptured marked
mosquitoes. e response variable in this model is the distance from the release site at which marked mosquitoes
were captured; in compliance with the Gamma distribution constrain, only strictly positive distances are consid-
ered. Covariate are the days aer release; (3) a Zero Altered Gamma model (hereaer ZAG34), which accounts
for the recaptured mosquitoes into the ST located on the release site (zero distance) and allows to estimate the
percentile of the distance at which mosquitoes are expected to be found, weighted for the probability that marked
mosquitoes collected actually moved from the release site. erefore, in this model the distance travelled is mod-
elled conditional to the days aer release by means of a two-step model. First, a Bernoulli model with a binary
response variable (i.e. value of 0 for mosquitoes recaptured on the release site and of 1 for mosquitoes recaptured
in another ST) is applied to evaluate the probability that a marked mosquito moves away from the release site.
Second, a Gamma GLM is tted on distances from the release site at which marked mosquitoes were captured.
Covariate was the days aer release for both sub-models. In all the three models, condence intervals for the
estimated quantities are computed by non-parametric bootstrap applied to original data.
Eect of uorescent dye marking on mosquito survival. Before each MRR experiment, Ae. albopictus
blood-fed females were removed from the cages by manual aspiration before (“unmarked group”; N = 50) and
aer (“marked group”; N = 50) the marking procedure described above. Both groups were kept in 25 cm cubic
cages kept outdoors in sites sheltered from direct sunlight and wind and provided with water bowl for egg-laying
and 5% sugar solution ad libitum.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

8
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Number of dead mosquitoes was recorded daily for the entire duration of the MRR experiment. Mortality rates
in the two groups were analysed using the Kaplan-Meier method and comparing survival curves by Log-rank test.
Oviposition pattern under semi-field conditions. During MRR2 and MRR3, individual marked
blood-fed females (N = 48 and 24, respectively) reared and handled as described above, were kept in plastic cups
closed with mosquito net and lined with lter paper, kept in sheltered sites outdoors. Oviposition activity was
monitored daily for the entire length of the MRR experiments, by counting the number of eggs deposited on the
lter paper in each cup. e correlation between the number of marked mosquito females collected in MRR2 and
MRR3 and the number of the ovipositing females during the same intervals was assessed by Pearson’s correlation.
Data Availability
All data generated or analysed during this study are included in this published article (and its Supplementary
Information Files).
References
1. Ser vice, M. W. In Mosquito Ecology: Field sampling methods 652–751 (1993).
2. Hawley, W. A. e biology of Aedes albopictus. J. Am. Mosq. Control Assoc. 1, 1–39 (1988).
3. GISD -Global Invasive Species Database. Available at, http://www.iucngisd.org/gisd/search.php. (Accessed: 21st July 2018).
4. Paupy, C., Delatte, H., Bagny, L., Corbel, V. & Fontenille, D. Aedes albopictus, an arbovirus vector: From the darness to the light.
Microbes Infect. 11, 1177–1185 (2009).
5. Epelboin, Y., Talaga, S., Epelboin, L. & Dusfour, I. Zia virus: An updated review of competent or naturally infected mosquitoes.
PLoS Negl. Trop. Dis. 11, e0005933 (2017).
6. Pialoux, G., Gaüzère, B.-A., Jauréguiberry, S. & Strobel, M. Chiungunya, an epidemic arbovirosis. Lancet Infect. Dis. 7, 319–327
(2007).
7. Zeller, H., Van Bortel, W. & Sudre, B. Chiungunya: Its History in Africa and Asia and Its Spread to New egions in 2013–2014.
J. Infect. Dis. 214, S436–S440 (2016).
8. Gossner, C. M., Ducheyne, E. & Schaner, F. Increased ris for autochthonous vector-borne infections transmitted by Aedes
albopictus in continental Europe. Eurosurveillance 23 (2018).
9. Bonnet, D. D. & Worcester, D. J. e dispersal of Aedes albopictus in the territory of Hawaii. Am. J. Trop. Med. 26, 465–76 (1946).
10. Niebylsi, M. L. & Craig, G. B. Dispersal and survival of Aedes albopictus at a scrap tire yard in Missouri. J. Am. Mosq. Control Assoc.
10, 339–43 (1994).
11. Lacroix, ., Delatte, H., Hue, T. & eiter, P. Dispersal and survival of male and female Aedes albopictus (Diptera: Culicidae) on
éunion Island. J. Med. Entomol. 46, 1117–24 (2009).
12. Medeiros, M. C. I., Boothe, E. C., oar, E. B. & Hamer, G. L. Dispersal of male and female Culex quinquefasciatus and Aedes
albopictus mosquitoes using stable isotope enrichment. PLoS Negl. Trop. Dis. 11, e0005347 (2017).
13. Maciel-de-Freitas, ., Neto, . B., Gonçalves, J. M., Codeço, C. T. & Lourenço-de-Oliveira, . Movement of dengue vectors between
the human modied environment and an urban forest in io de Janeiro. J. Med. Entomol. 43, 1112–20 (2006).
14. Liew, C. & Curtis, C. F. Horizontal and vertical dispersal of dengue vector mosquitoes, Aedes aegypti and Aedes albopictus, in
Singapore. Med. Vet. Entomol. 18, 351–60 (2004).
15. Honório, N. A. et al. Dispersal of Aedes aegypti and Aedes albopictus (Diptera: Culicidae) in an urban endemic dengue area in the
State of io de Janeiro, Brazil. Mem. Inst. Oswaldo Cruz 98, 191–8 (2003).
16. Davis, T. J., aufman, P. E., Tatem, A. J., Hogsette, J. A. & line, D. L. Development and Evaluation of an Attractive Self-Maring
Ovitrap to Measure Dispersal and Determine Sip Oviposition in Aedes albopictus (Diptera: Culicidae) Field Populations. J. Med.
Entomol. 53, 31–8 (2016).
17. Marini, F., Caputo, B., Pombi, M., Tarsitani, G. & Della Torre, A. Study of Aedes albopictus dispersal in ome, Italy, using sticy traps
in mar-release-recapture experiments. Med. Vet. Entomol. 24, 361–368 (2010).
18. Mori, A. Eects of larval density and nutrition on some attributes of immature and adult Aedes albopictus. Trop. Med. 21, 85–103
(1979).
19. Venturi, G. et al. Detection of a chiungunya outbrea in Central Italy, August to September 2017. Euro Surveill . 22 (2017).
20. Calba, C. et al. Preliminary report of an autochthonous chiungunya outbrea in France, July to September 2017. Eurosurveillance
22, 17–00647 (2017).
21. Lusic, B. et al. First case of imported chiungunya infection in Croatia, 2016. Int. Med. Case Rep. J. 10, 117–121 (2017).
22. Semenza, J. C. & Bettina, M. Climate change and infectious diseases in Europe. Lancet. 9(6), 365–75 (2009).
23. Semenza, J. C. & Su, J. E. Vector-borne diseases and climate change: a European perspective. FEMS Microbiology Letters. 365 (2)
(2018).
24. Italian Ministry of Health. Piano nazionale di sorveglianza e risposta alle arbovirosi trasmesse da zanzare invasive (Aedes sp.) con
particolare riferimento ai virus chiungunya, dengue e zia - 2018. Available from, http://www.salute.gov.it/portale/news/
p3_2_1_1_1.jsp?lingua=italiano&menu=notizie&p=dalministero&id=3374 Cited 13 August (2018).
25. Ministère des aaires sociales, de la santé et des droits des femmes & Ortmans, C. INSTUCTION N° DGS/I1/2015/125 du 16
avril mettant à jour le guide relatif aux modalités de mise en œuvre du plan anti-dissémination du chiungunya et de la dengue en
métropole. (2015).
26. Sochaci, T. et al. Imported chiungunya cases in an area newly colonised by Aedes albopictus: mathematical assessment of the best
public health strategy, https://doi.org/10.2807/1560-7917
27. Ajelli, M. et al. Host outdoor exposure variability aects the transmission and spread of Zia virus: Insights for epidemic control.
PLoS Negl Trop Dis 11(9), e0005851 (2017).
28. Hamidou M., et al. Guidelines for routine colony maintenance of aedes mosquito species version 1.0 Food and Agriculture
Organization of the United Nations International Atomic Energy Agency. Vienna Available, http://www-naweb.iaea.org/nafa/ipc/
public/guidelines-for-routine-colony-maintenance-of-Aedes-mosquito-species-v1.0.pdf (2017)
29. Facchinelli, L. et al . Development of a novel sticy trap for container-breeding mosquitoes and evaluation of its sampling properties
to monitor urban populations of Aedes albopictus. Med. Vet. Entomol. 21, 183–95 (2007).
30. Team, . D. C. : A language and environment for statistical computing. R Found. Stat. Comput. 2673 (2005).
31. Lillie, T. H., Marquardt, W. C. & Jones, . H. e ight range of Culicoides variipennis (Diptera: Ceratopogonidae). Can. Entomol.
113, 419–426 (1981).
32. White, D. J. & Morris, C. D. Bionomics of anthropophilic Simuliidae (Diptera) from the Adirondac Mountains of New Yor State,
USA 1. Adult dispersal and longevity. J. Med. Entomol. 22, 190–199 (1985).
33. Morris, C. D., Larson, V. L. & Lounibos, L. P. Measuring mosquito dispersal for control programs. J. Am. Mosq. Control Assoc. 7,
608–615 (1991).
34. Zuur, A. & Ieno, E. Beginner’s Guide to Zero-Inated Models with R. Highland Statistics’ Beginner’s Guide Series. (Highland Statistics
Ltd, 2016).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

9
SCIENTIFIC REPORTS | (2019) 9:10281 | https://doi.org/10.1038/s41598-019-46466-4
www.nature.com/scientificreports
www.nature.com/scientificreports/
Acknowledgements
We thank Luca Delucchi for providing the map used in Figure 1A.
Author Contributions
F.M., B.C., M.P. and A.d.T. designed the study. F.M., B.C., M.P., M.T., M.F. and A.D. carried out the eld studies.
M.M., F.M., B.C. and R.R. carried out the statistical analyses. F.M., B.C., M.M. and A.d.T. wrote the manuscript.
Additional Information
Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-46466-4.
Competing Interests: e authors declare no competing interests.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
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 Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted 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 license, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2019
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com