Can landscape metrics help determine the Culicoides imicola distribution in Italy?

• Main Authors: Carla Ippoliti, Marius Gilbert, Sabine Vanhuysse, Maria Goffredo, Giuseppe Satta, Eléonore Wolff, Annamaria Conte
• Format: Article
• Language: English
• Published: PAGEPress Publications 2013-11-01
• Series: Geospatial Health
• Subjects: landscape metrics, remote sensing, Culicoides imicola, bluetongue, Italy.
• Online Access: http://www.geospatialhealth.net/index.php/gh/article/view/72

<em>Culicoides imicola</em> is considered to be one of the main vectors of bluetongue disease in the Mediterranean Basin. However, local variations occur. For example, in Italy, <em>C. imicola</em> is a stable and abundant population in Sardinia and is widely distributed across the island, whilst in Tuscany on the Italian mainland, it ranges from low abundance in the west and coastal areas to absence in the eastern part of the region. Entomological surveillance data collected over 10 years were used to classify 52 sites as low to medium or high <em>C. imicola</em> abundance in Sardinia, and 59 sites as either positive or negative in Tuscany. The land cover was mapped from high-resolution remote sensing images using an object-based image analysis approach and a set of landscape metrics with 500 m buffers around each site. Multivariate analysis was used to test the statistical association of landscape metrics to <em>C. imicola</em> presence and abundance together with other eco-climatic and topographic variables. In Sardinia, 75% of the sites were correctly classified based on altitude alone and the inclusion of landscape- related variables did not improve the classification. In Tuscany, the mean annual temperature allowed classifying 70% of the positive/negative sites correctly. When landscape metrics was included in the multivariate model, an improvement up to 80% was obtained. The presence of riparian vegetation and water was found to be positively correlated with <em>C. imicola</em> presence, whilst forest (including the edge between the forest and cultivated areas) was found to be negatively related to the presence of <em>C. imicola</em>.