Geostatistics as predictive tools to estimate Ixodes ricinus (Acari: Ixodidae) habitat suitability in the western palearctic from AVHRR satellite imagery

Geostatistics (cokriging) were used to model the cross-correlated information between satellite-derived vegetation and climate variables and actual records of the tick Ixodes ricinus in the western Palearctic. The output was used to map the habitat suitability for I. ricinus on a continental scale. A database of collecting localities of I. ricinus was built up from a total of 812 records. This database has been cross-tabulated with satellite NOAA AVHRR pictures obtained from 1982 to 1994 over the Palearctic at 10 day intervals, with a resolution of 8 km per pixel. A cokriging system was generated to exploit satellite-derived data and to estimate the distribution of I. ricinus. Three vegetation (standard NDVI values) and four temperature variables output the habitat suitability prediction with a sensitivity of 0.98 and a specificity of 0.92. Results obtained with the model closely agreed with actual records of the tick, with 4 and 3% of false-positive and false-negative sites, respectively. Such statistical analysis can guide field work towards the correct interpretation of the distributional limits of ticks and may also be used to make predictions about the impact of global change on tick distributional ranges.