Spatial analysis of gypsy moth populations in Sardinia using geostatistical and climate models

Abstract

Spatial fluctuations of the Sardinian population of the gypsy moth Lymantria dispar (L.) (Lepidoptera: Lymantriidae) were characterized using geostatistical and climate models. Data on gypsy moth egg mass abundance recorded at 282 permanent monitoring sites from 1980 to 2004 were incorporated in a geographic information system with the vegetational, geomorphological and pedological features of the sites. Statistical analyses revealed that the relative outbreak frequency was related to the predominant host tree, slope and elevation of the monitoring sites, whereas there was no correlation between outbreak frequency and exposure and soil type. By using bioclimatic modelling, probability maps of gypsy moth outbreaks were generated. The model identified a probability surface with climatic conditions favourable to gypsy moth outbreaks and thus potentially subject to defoliation. The maps included 92 sites where outbreaks never occurred, suggesting that the Sardinian climate may not be a determinant factor for gypsy moth outbreaks. The geostatistical method cokriging with outbreak frequency as a covariate was found to be the most suitable technique to estimate gypsy moth egg mass abundance. Semivariograms showed spatial correlation of egg mass abundance within the range 18.5–53 km. The results obtained were used to create regional gypsy moth distribution maps by cokriging, which demonstrated the outbreak foci and different infestation levels at each monitoring area. These results can help to delimit the treatment areas and develop rational gypsy moth management programmes.