Volume 7 Supplement 1

16th International Symposium on HIV and Emerging Infectious Diseases

Open Access

Influence of water on the circulation of the West Nile Virus in horses in Southern France

  • Sophie Pradier1, 2Email author,
  • Alain Sandoz3,
  • Gaëtan Lefebvre3,
  • Annelise Tran4,
  • Sylvie Lecollinet5 and
  • Agnès Leblond2, 6
Retrovirology20107(Suppl 1):P186

DOI: 10.1186/1742-4690-7-S1-P186

Published: 11 May 2010

Background

West Nile Virus (WNV) affects humans and horses, potentially causing severe neurological manifestations. Recent outbreaks of West Nile fever in horses were reported in Camargue (2000, 2004), Var (2003) and Pyrénées Orientales (2006). The circulation of this virus is strongly influenced by environmental conditions. This study aimed at explaining the circulation of WNV in horses by quantifying water bodies around equine stables using Landsat images.

Methods

A total of 135 stables were selected in three French departments (Hérault, Gard, Bouches-du-Rhône) and 1161 horses were tested by serological analysis between 2007 and 2008.

15 Landsat images (August 2006 to August 2008) were classified into 3 classes: open water, flooded vegetation and other. Surface areas of the first two classes were calculated for buffers of 2 to 5 km around each stable and for each date.

Two multivariate analyses were conducted: GLMs to identify which environmental variables were involved in the viral circulation in stables and GRMs to identify the horse variables linked to WNV circulation after retrieving the effect of the environment.

Results

The best model distinguishing 46 negative stables (no positive horse considering an error threshold of 0.15) from 61 positive stables (at least 1 positive horse) used 2 km buffers and included mean area of flooded vegetation, total number of horses present in the stable, mean area of open water and X and Y geographic coordinate. The first two variables had a positive effect and the other three a negative effect. The model predicted correctly 73% of positive stables and 71% of negative ones. At the horse level, breed, activity and age were significant. See Figure 1.
Figure 1

Localisation of stables in the study zone (classification of a Landsat image of September 8 th 2007).

Discussion

These results can be used to target the surveillance of this human and equine disease in Southern France.

Authors’ Affiliations

(1)
Clinique équine ENVA
(2)
UR 346 Epidémiologie animale INRA
(3)
Tour du Valat
(4)
CIRAD Agirs
(5)
UMR 1161 Virologie INRA AFSSA ENVA
(6)
Clinéquine ENVL

Copyright

© Pradier et al; licensee BioMed Central Ltd. 2010

This article is published under license to BioMed Central Ltd.

Advertisement