Vectors collected in these areas were tested and two of their species

resulted infected by WNV: Culex pipiens and Ochlerotatus caspius.

Throughout European Culex modestus is also considered potential bridge

vector from birds to mammals.

In this study we described the spatial distribution of WNV vectors located

in Northwestern Italy. This territory is close to WNV foci area and could be

the door of WNV entrance in Piedmont.

WNV Spread

Study Area

Introduction

West Nile Virus (WNV) is a reemerging

zoonoses in Italy. The first outbreak

occured in Tuscany in 1998, close to

wetland area. No other cases were

reported during next ten years, but virus

introduction continued. In Autumn 2008

new foci in human and horses were

detected in Emilia Romagna region and

WNV spread to neighbouring areas.

Mosquitoes collection

Cluster analysis

Evaluating risk of introduction of WNV in Eastern Piedmont, Northern Italy

Luigi Bertolotti , Donal Bisanzio , Francesco Cerutti , Andrea Mosca , Luca Balbo , Gonzalo M. Vazquez-Prokopec , Uriel Kitron , Mario Giacobini

1,2 1,2 1,2,3

5

4 4

5,6 1,2

Remote sensing data

Corine of landuse 2001

Digital Terrain Model

MODIS 16 days NDVI

MODIS 8 days Land Surface

Temperature (LST)

(week) (week)

36 Mosquito CO Bited Traps, weekly collection from May to October,

2000-2006

29 Mosquito CO Bited Traps, weekly collection from May to October,

2007-2008

Traps were placed randomly and they are not clustered

Since 2000 control strategies against O. caspius were performed using spray

biological lavicidal, Bacillus thuringiensis var. israelensis (Bti)

Two types of control were performed: spraying all rice fields every year by

helicopter (TR1) or treating only high abundance spots (TR2)

Study area

Week

Generalized linear mixed model (GLMM)

Getis-Ord Gi(*d) was used to find local clusters

Analyses between traps were based on inverse of

their distances

C. pipiens showed several clusters in all years

C.modestus and O. caspius showed clusters in few

years

High abundance and several significant high spot of

adult mosquitoes was recorded in areas near rice

fields, north of study area.

Random effect:

Trap location

Months of year

Fixed effect:

NDVI

Weekly mean temperature

Weekly rainfall

DTM

Seasonality (sinusoidals function)

Different treatments

Spatial correlation function was implemented

Model was trained with 2000-2006 data

Validation was performed on 2007-2008 data

collection

Conclusion

1,439,059 mosquitoes of different species were

trapped in 47430 trap/nights (Tab. 1)

O. caspius, C. pipiens and C. modestus were the

most abundant species

O. caspius and C. pipiens showed peaks in late

summer

C. modestus did not show seasonality

A model was built for C. pipiens and O. caspius

It was difficult to make a model for C. modestus,

it have not link with variables studied

Presence of vectors confirms that study area could be a way for the spreading of WNV in Piedmont

Areas near rice fields have a high risk of future outbreak. In this part of territory there is the highest presence of vector, and it

is a good habitat for migratory birds

Model confirm that Bti control of O. caspius may be considered a good method to decrease its population

C. modestus did not show any seasonability and this species seems to be linked to agricultural activity

The model captures ecological features of different species and it could be used to plan surveillance

This study was funded by Lagrange Foundation: Donal Bisanzio. This study was funded by Regione Piemonte, Ricerca Sanitaria Finalizzata 2008 program: Luigi Bertolotti

Bibliography

Autorino, Gian Luca, Antonio Battisti, Vincent Deubel, Giancarlo Ferrari, Riccardo Forletta, Armando Giovannini, Rossella Lelli, Severine Murri, and Maria Teresa Scicluna. 2002. West Nile virus epidemic in horses, Tuscany

region, Italy. Emerg Infect Dis 8, no. 12: 1372-1378.

Calistri, P, a Giovannini, G Savini, F Monaco, L Bonfanti, C Ceolin, C Terregino, M Tamba, P Cordioli, and R Lelli. 2009. West Nile Virus Transmission in 2008 in North-Eastern Italy. Zoonoses and public health: 1-9

Gobbi, F, G Napoletano, C Piovesan, F Russo, A Angheben, A Rossanese, A M Cattelan, et al. 2009. Where is West Nile fever? Lessons learnt from recent human cases in northern Italy. Euro Surveill 14, no. 10.

Monaco, F, R Lelli, L Teodori, C Pinoni, A Di Gennaro, A Polci, P Calistri, and G Savini. 2009. Re-Emergence of West Nile Virus in Italy. Zoonoses and public health.

Bolker, B.M. et al., 2009. Generalized linear mixed models: a practical guide for ecology and evolution. Trends in ecology & evolution (Personal edition), 24(3), 127-35.

DTM

NDVILST

Corine Landuse 2001 Remote sensing images, regarding sample periods (2000-2008), were

collected

All data were processed to correct atmospheric disturbance

Collected data:

2

2

Tab.1: Annual number of mosquitoes collected in the area

Correlograms on weekly collections from 2000-2008

Model detected area where high number of mosquitoes

were collected

Model fited better for C. pipiens than for O. caspius, due

to their different homerange

Department of Animal Production, Epidemiology and Ecology, Faculty of Veterinary Medicine, University of Torino, ItalyMBC - Molecular Biotechnology Center, University of Torino, ItalyDepartment of Pathobiological Sciences, School of Veterinary Medicine, University of Wisconsin-Madison, USAIPLA, Institute for wood plants and environment, Torino, ItalyDeptartment of Environmental Studies, Emory University, Atlanta, GA, USAFogarty International Center, National Institutes of Health. Bethesda, MD 20892, USA

1

2

3

4

5

6

GLMM represent a class of fixed effects regression

models with an inclusion of random effects for several

types of dependent variables

Tab.2: Results of GLMM model for C. pipiens and O. caspius

WeekWeek

CO Bited Trap2