Analysis of Oak Tree Mortality to Predict Ambrosia Beetle Platypus quercivorus Movement

The dispersal and host selection processes of ambrosia and bark beetles determine which trees are attacked. Factors affecting each process must be understood to predict the attack probability. Here, we analyzed data on the attack status of Quercus crispula by Platypus quercivorus over a 10-year period (2004-2013) in a 16-ha plot located in a cool temperate forest of Japan. The distance from trees attacked in the previous year was calculated for each tree in the plot. In parallel, beetle dispersal probability was calculated by assuming that the beetle dispersal kernel follows a lognormal distribution. The calculated beetle dispersal probability was used as a candidate explanatory variable in the model to predict the probability of attack, together with tree dbh and local basal area of Q. crispula. The results confirmed a previous study showing that the attack probability increased with increasing dbh and basal area of Q. crispula within a 10-m radius. Of 100 patterns of prepared kernels, the kernel with low dispersal probability at a short distance (peaking around 300 m) was the best predictor. Our results may indicate that P. quercivorus individuals fly beyond suitable neighboring hosts when dispersing from the original host plant.