Spatial patterns of Scaphoideus titanus (Hemiptera: Cicadellidae): a geostatistical and neural network approach

Abstract

The spatial distribution of the leafhopper Scaphoideus titanus Ball, the vector of the pathogen Flavescence dorée of grapevine, was studied in the Asti Province (Piedmont) Italy. Field sampling of adults was carried out using yellow sticky traps both in vineyards subjected to different pest management regimes, and in woods containing American grapevines. The spatial correlation of S. titanus captures was studied using geostatistical analyses. An artificial neural network (ANN) was designed to operate as a spatial predictor driven by external factors (elevation, slope, height above channel, agricultural communities, perimeter-to-area ratio, potential solar radiation, and pest management) for estimating the population density of the leafhopper. The captures were spatially related up to 210 m: the variogram fitting was significant, but resulted in low R ² values. The ANN achieved a significant generalization of the infestation levels of S. titanus, permitting prediction maps based upon simulated pest management scenarios to be obtained. The most important factors affecting S. titanus population density were pest management, and secondarily agricultural communities.

Keywords:

• leafhopper
• spatial correlation
• variogram
• artificial neural network
• prediction map

Acknowledgments

We are grateful to Edoardo Sala and Matteo Gerbi, of ‘Laurea di I livello in Viticoltura ed Enologia', for help in field sampling. This research was conducted within a project supported by ‘Distretto dei Vini Langhe, Roero e Monferrato’ and ‘Unione Collinare Monferrato Valleversa’.